81699 2014 world development report Risk and Opportunity Managing Risk for Development A World Bank Group Corporate Flagship world development report 0 Risk and Opportunity Managing Risk for Development © 2013 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1234 16 15 14 13 This work is a product of the staff of The World Bank with external contributions. Note that The World Bank does not necessarily own each component of the content included in the work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http://creativecommons.org/licenses/ by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for com- mercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2013. World Development Report 2014: Risk and Opportunity—Managing Risk for Development. Washington, DC: World Bank. doi: 10.1596/978-0-8213-9903–3. License: Creative Commons Attribution CC BY 3.0 Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBN, eISBN, DOI, and ISSN: Softcover ISSN: 0163-5085 ISBN: 978-0-8213-9903-3 e-ISBN: 978-0-8213-9982-8 DOI: 10.1596/978-0-8213-9903-3 Hardcover ISSN: 0163-5085 ISBN: 978-0-8213-9964-4 DOI: 10.1596/978-0-8213-9964-4 Cover design: Heads of State Interior design: Debra Naylor The maps in the World Development Report 2014: Risk and Opportunity—Managing Risk for Development were produced by the Map Design Unit of the World Bank. The boundaries, colors, nominations and any other information shown on this map do not imply, on the part of the World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. Contents Foreword xiii Acknowledgments xv Overview Risk and Opportunity 1 Overview 3 Risk is a burden but also an opportunity 4 Risk management can be a powerful instrument for development 5 What does effective risk management entail? 10 Beyond the ideal: The obstacles to risk management 16 The way forward: A holistic approach to managing risk 18 The household 21 The community 23 The enterprise sector 25 The financial system 27 The macroeconomy 31 The international community 33 An institutional reform to mainstream risk management 36 In conclusion: Five principles of public action for better risk management 40 Some closing thoughts 42 Notes 43 References 45 Part 1 Fundamentals of risk management 51 1 Risk management can be a powerful instrument for development 53 Risk and opportunity 53 Why is risk management relevant for development? 55 What does risk management entail? 59 Risk management is cost-effective—yet not always feasible 70 Notes 73 References 74 Spotlight 1 Preparing for the unexpected: An integrated approach to disaster risk management in the Philippines and Colombia 76 v vi CONTENTS 2 Beyond the ideal: Obstacles to risk management and ways to overcome them 79 Missed opportunities for good risk management 79 Why aren’t people better at managing their own risk? 81 Obstacles beyond the control of individuals hamper their risk management 85 Why aren’t states better at filling in the gaps? 89 Putting it all together: A policy sequence 95 A way past the obstacles: Choosing policy priorities 98 Notes 101 References 102 Spotlight 2 Protecting the food consumption of the poor: The role of safety nets in Ethiopia and El Salvador 104 Part 2 The role of key social systems 107 3 Households are the first line of support to confront risk and pursue opportunity 109 Sharing good times and bad times 109 What risks do households face and how do they cope? 110 How do households prepare to manage risks, and what obstacles do they face? 115 How can government strengthen protection and foster better opportunities for households? 121 Putting it all together: Guidelines for policy implementation 130 Notes 133 References 134 Spotlight 3 Moving toward universal health coverage in Turkey and the Kyrgyz Republic 136 4 Cohesive and connected communities create resilience 139 Communities confront many risks 139 Communities as risk managers 140 Communities as providers of insurance 143 Communities as providers of protection 145 Cohesive and connected communities are more effective 148 Public policies to improve local risk management 152 Putting it all together: Policy principles and research priorities to foster resilient communities 159 Notes 161 References 162 Spotlight 4 Where criminal justice is not enough: Integrated urban crime and violence prevention in Brazil and South Africa 164 Contents vii 5 Fostering resilience and prosperity through a vibrant enterprise sector 167 Creating jobs and supporting innovation 167 The ways that the enterprise sector can help people confront risks 168 Flexibility and formality in the enterprise sector improve people’s resilience and prosperity 172 How can the government help enhance the flexibility and formality of the enterprise sector? 179 Putting it all together 185 Notes 187 References 188 Spotlight 5 Moving toward greater labor market flexibility: India’s uneven path 190 6 The role of the financial system in managing risk: More financial tools, fewer financial crises 193 The financial system can fulfill a socially beneficial function of risk management 193 For good risk management, people need a range of financial tools 194 Financial crises hurt people: How can they be prevented? 204 Resolving the tension between financial development and financial stability 211 A summary of policy recommendations 216 Notes 218 References 219 Spotlight 6 Building resilience to global economic shocks in the Czech Republic, Peru, and Kenya 222 7 Managing macroeconomic risk: Building stronger institutions for better policy outcomes 225 Increasing resilience and promoting opportunity through sound macroeconomic policy 225 Gearing macroeconomic policies toward aggregate stability 226 Generating sustainable fiscal resources to finance stabilization policies and long-term social programs 236 Putting it all together: What to avoid and what to do to improve risk management at the macroeconomic level 242 Notes 245 References 246 Spotlight 7 Diseases without borders: Managing the risk of pandemics 248 viii CONTENTS 8 The role of the international community: When risks exceed national capacity 251 Global problems call for global players 251 What circumstances call for action by the international community and why? 252 How does the international community enhance risk management? 256 How effective is the international community in resolving global risks? 262 Policy implications and takeaways 269 Notes 274 References 275 Focus on policy reform Mainstreaming risk management into the development agenda: Selected institutional reforms 278 Appendixes 287 Abbreviations and data notes 289 Background papers 291 Selected indicators 293 Index 325 Boxes 1 Five key insights on the process of risk management from the 1.6 Protection and insurance can provide complementary means World Development Report 2014 4 of managing risk 68 2 A risky world: Trends in risk across regions 9 1.7 Preparation for risk varies within and across regions and 3 How does preparation for risk vary across countries? 15 continents 69 4 Bringing the essentials of and obstacles to risk management 1.8 A “man-made” disaster: The Fukushima nuclear accident in together in policy design 17 Japan 71 5 Which systems for which risks? 20 2.1 The irrational, the uncertain, and the short-sighted: Some classic experiments reveal surprises about people’s 6 Should access to social insurance be tied to work behavior 83 status? 29 2.2 Strengthening the interaction between experts and policy 7 An independent fiscal council can help overcome procyclical makers to improve risk management 88 fiscal bias 34 2.3 Institutions to improve risk management: National risk 8 For certain global risks such as climate change, the assessments 91 international community should embrace incremental approaches that can lead to global solutions 37 2.4 Applying a robust decision-making methodology to deal with the risk of flooding in Ho Chi Minh City 96 9 Selected policy recommendations from the WDR 2014 38 2.5 Precaution in the face of risk: Striking a balance between costs and benefits with the precautionary principle 97 1.1 A risky world: Risks vary over time and across regions 54 3.1 Altruism, exchange, or social norms: What motivates family 1.2 While poverty has declined, many people around the world members to care for one another? 111 remain vulnerable to poverty 58 3.2 Improving service delivery by partnering with the private 1.3 When risk aversion becomes loss aversion: A view from sector 123 utility theory 61 3.3 Index-based insurance: The potential and the 1.4 Developing countries have increased their resilience over challenges 126 time 64 4.1 When sharing wealth is a religious mandate: The use of zakat 1.5 Leveraging new information and communication in Islamic communities 142 technologies for risk management 66 Contents ix 4.2 Refugees and internally displaced people: Moving from 6.6 Financial bailouts: “Too big to fail” versus moral managing displacement to facilitating opportunities 151 hazard 210 4.3 New communication technologies help communities manage 7.1 The new normal in the world economy: Heightened violence and local conflict 153 macroeconomic policy uncertainty in developed 4.4 Clean, green, and blue: Managing water and flooding in countries 227 Singapore 156 7.2 Relinquishing monetary policy flexibility: The ultimate 4.5 Using social marketing to increase access to sanitation in sacrifice? 231 rural Tanzania 158 7.3 Reserve requirement policy has substituted for 5.1 Worker, consumer, and environmental protection can also be monetary policy as a countercyclical tool in most profitable business 172 developing countries 232 5.2 Lessons from great depressions 175 7.4 Fiscal stimulus: The good, the bad, and the ugly 235 5.3 The complex relationship between globalization and 7.5 Managing commodity revenues in Chile: An example formality 178 of sound institution building and management of public resources 239 5.4 The resilience of the power sector during a period of civil unrest in Côte d’Ivoire was driven by protections for 7.6 Disclosing fiscal risks in Colombia: A path to greater contracts 181 transparency and credibility in risk management 241 5.5 Labor market flexibility alongside social support: Examples S7.1 An emergency response to a top global catastrophic from Germany and Denmark 183 risk 249 6.1 Better than cash: Electronic payments reduce risk and 8.1 Climate change and implications for development 255 costs 196 8.2 International support for disaster risk management 261 6.2 Housing finance can improve household resilience and 8.3 Regional risk-sharing solutions: Promoting financial opportunities 197 resilience to disaster risks 262 6.3 Innovative insurance mechanisms in Mongolia and 8.4 Global efforts to provide tools and databases for assessing Mexico 199 disaster risk 263 6.4 Private pension insurance to confront the risk of income loss 8.5 Well-managed risks can unleash development opportunities in old age 200 in fragile and conflict-affected states 264 6.5 Preparing for a banking crisis with crisis simulation 8.6 A tale of two disasters 272 exercises 209 Diagrams 1 The risk chain: The nature and extent of outcomes 2.2 An iterative process of decision making to prompt robust depend on shocks, exposure, internal conditions, and risk action in the face of uncertainty 95 management 12 2.3 A set of screens for assessing obstacles to risk management, 2 The interlinked components of risk management 13 and formulating policy responses 98 3 Key social and economic systems can contribute to risk 4.1 Three major types of violence and their spillovers 147 management in complementary ways 19 6.1 Combining financial tools improves preparation for losses of 4 Balancing the trade-offs in the institutional design of a different probability and severity 195 national risk board 39 8.1 The agents of the international community 252 1.1 The risk chain: The nature and extent of outcomes 8.2 Role of the international community 257 depend on shocks, exposure, internal conditions, and risk F1.1 Balancing the trade-offs in the institutional design of a management 63 national risk board 279 1.2 The interlinked components of risk management 65 2.1 Individuals, firms, and countries face many obstacles in managing risks 82 x CONTENTS Figures 1 Many people around the world are poor or live very close S3.1 Use of health care and out-of-pocket health expenditure to poverty; they are vulnerable to falling deeper into by income status, the Kyrgyz Republic, 2009 137 poverty when they are hit by negative shocks 5 4.1 People and their communities are affected by both 2 The benefits of risk management often outweigh the idiosyncratic and systemic shocks 141 costs 8 4.2 People respond to shocks on their own and by pooling 3 Education and health outcomes in developing countries risk with others, relying on informal credit and assistance are improving, but unevenly 22 to cope with shocks 143 4 People respond to shocks on their own and by pooling 4.3 Social capital helps reduce costly coping in Eastern risk with others 24 Europe and Central Asia 144 5 Self-employment is more prevalent in developing 4.4 Insecurity affects more than the victims 146 countries, especially in Sub-Saharan Africa and South 4.5 Membership in voluntary organizations varies widely Asia 26 around the world 149 6 Countries vary widely in the flexibility and formality of 4.6 Social exclusion based on ethnicity is not declining in their product and labor markets 28 many parts of the world 151 7 Financial inclusion in savings, credit, and insurance across 4.7 Who do you trust? Confidence in police, government, and developing countries at different income levels 30 religious leaders 155 8 Economic, financial, and social interconnectedness are 4.8 More people have cell phones than have toilets— on the rise 35 illustrating a failure to provide the most basic 1.1 Households in developing countries feel susceptible to protection 157 risk and are concerned by it 56 S4.1 Homicide rate in selected cities 164 1.2 Risk management tools can help people pursue 5.1 Wage employment rises with the level of economic opportunity 60 development 169 1.3 The benefits of risk management often outweigh the 5.2 The flexibility of the enterprise sector varies around the costs 71 world 173 3.1 Shocks to households vary considerably across 5.3 Countries with a higher regulatory burden experience countries 110 more severe recessions 174 3.2 Elderly people frequently live with other family 5.4 The formality of the enterprise sector varies around the members 112 world 176 3.3 Men who experienced violence in the household when 5.5 Typology of countries by the flexibility and formality of young are more likely to act violently as adults 114 their product and labor markets 180 3.4 Immunization rates have increased and infant mortality 6.1 The range of formal financial tools used by individuals has declined everywhere 115 varies by country and income 198 3.5 The risk of death from noncommunicable disease is 6.2 As financial systems deepen, they diversify their growing in all developing regions 116 institutional structure 198 3.6 Demand for preventive health care products falls steeply 6.3 The shares of formal and informal saving and borrowing as the price increases 117 change as countries develop 201 3.7 Low-income countries still lag in educational attainment, 6.4 Banks’ aggregate refinancing risk and foreign currency and some middle-income countries suffer from gaps in mismatches can increase systemic risk in developing quality 117 countries 205 3.8 Educational attainment is still uneven for boys and 6.5 Prudent financial inclusion can enhance financial stability girls from poor households, especially in lower-income but, if excessive, can weaken it 213 countries 118 S6.1 Bank nonperforming loans to total gross loans ratio, 3.9 Female labor force participation is still limited in some 2007–11 223 regions 121 7.1 Real economic activity is more volatile and more likely to 3.10 Social norms in many regions tolerate domestic decline sharply in developing countries 228 violence 124 7.2 Monetary policy has become countercyclical in some 3.11 Noncontributory pension programs have expanded developing countries 230 coverage in developing countries, especially for the 8.1 Damages from a disaster can exceed a country’s annual poorest 128 GDP in developing countries 253 Contents xi 8.2 A 2011 survey highlights divergent national interests and F1.1 Noncontributory pension programs have expanded coverage short-sighted views on climate change 267 in developing countries, especially for the poorest 282 8.3 Disaster-related donor assistance has focused more on coping F1.2 Increasing coverage will require higher levels of spending in than preparedness and prevention 268 countries with aging populations 283 Glossaries 1.1 Terms related to risk management 61 2.1 Economics terms used throughout the Report 81 Maps 1 Banking crises around the world, 1970–2011 31 7.1 Government consumption became countercyclical in 2 Government consumption became countercyclical in more than one-third of developing countries over the past more than one-third of developing countries over the past decade 234 decade 33 8.1 Developing countries in general have suffered higher 2.1 Crises repeat themselves in the absence of effective mortality associated with disasters 253 prevention 80 2.2 Deeply uncertain futures: Different climate models project very different changes for precipitation in Africa 94 Profile 1 The Gomez family: A modern tale of risk and resilience 7 Tables 1.1 Households in developing countries face many shocks 55 6.2 National financial sector strategy documents rarely 3.1 Poorer households are more likely to report using costly consider the trade-off between financial development and mechanisms to cope with shocks 113 stability 214 3.2 A systemic approach for policies to address multiple risks 6.3 Composition of financial stability committees in selected using multiple instruments 122 developing countries 215 3.3 Common features of programs to expand coverage of health 6.4 Policy priorities to improve the financial system’s role in risk insurance 127 management 217 3.4 Policy priorities to improve risk management at the 7.1 Policies to prevent or cope with fiscal risks 237 household level 130 7.2 Policy priorities to improve risk management at the 4.1 Policy priorities to improve risk management at the macroeconomic level 244 community level 159 8.1 Examples of global actions and factors underlying their 5.1 Policy priorities to improve the enterprise sector’s role in risk impact 258 management 186 8.2 Policy priorities to improve risk management at the S5.1 Studies of labor regulations in India 191 international community level 270 6.1 A taxonomy of macroprudential tools 208 F1.1 National policy measures limiting carbon emissions 284 Foreword In recent years, the world has suffered a multitude of crises. Financial and economic turmoil have disrupted the world economy through loss of income, jobs, and social stability. Intense natural disasters have devastated entire communities from Haiti to Japan, leaving a trail of fatalities and economic losses in their wake. Concerns about global warming have grown, as have fears about the spread of deadly contagious diseases. As I travel around the world, I hear the same concern: how can we become more resilient to such risks? The World Development Report 2014 (WDR 2014), Risk and Opportunity—Managing Risk for Development, helps provide answers to this pressing question. Another concern is the missed development opportunities that arise when necessary risks are not taken. Pursuing opportunities requires taking risks, but many people, especially the poor, are often reluctant to do so, because they fear the potential negative consequences. Failure to act can trap people in poverty, leaving them vulnerable to negative shocks and even less able to pursue opportunities that would otherwise improve their well-being. The inability to manage risk properly leads to crises and missed opportunities. This poses significant obstacles to attaining the World Bank Group’s two main goals: ending extreme poverty by the year 2030 and boosting shared prosperity of the bottom 40 percent of the population in developing countries. Managing risk effectively is, therefore, absolutely central to the World Bank’s mission. The WDR 2014 demonstrates that effective risk management can be a powerful instrument for development—it can save lives, avert economic shocks, and help people build better, more secure futures. This report calls for individuals and institutions to move from being “crisis fighters” to becoming “proactive and systematic risk managers.” There is substantial evidence that recognizing and preparing for risk can pay off abundantly. For instance, many developing countries displayed resilience in the face of the recent global financial crisis because they had previously reformed their macroeconomic, financial, and social policies. Protecting hard-won development gains by building resilience to risk is essential to achieving prosperity. That is true whether one is grappling with natural disasters, pandemics, financial crises, a wave of crime at the community level, or the severe illness of a household’s chief provider. Risk can never be completely eliminated. But people and institutions can build resilience to risk by applying a balanced approach that includes structural policy measures, community-based prevention, insurance, education, training, and effective regulation. Countries have learned how to manage risk in diverse settings, but, until now, research related to risk management in the developing world has not been synthesized into a single source that is easily accessible and well-referenced. This WDR aims to fill that gap. It serves as a valuable guide both for mainstreaming risk management into the development agenda, and for helping countries and communities strengthen their own risk management systems. The Report also offers important insight for changing the approach to risk in the Bank’s own operations. The World Bank Group is currently undergoing a transformation, which calls for shifting the institutional culture regarding risk from one of extreme risk aversion to one of informed risk taking. This year’s WDR cautions that the greatest risk may be taking no risk at all. I could not agree more. xiii xiv FOREWORD My hope is that the WDR 2014 will lead to risk management policies that allow us to minimize the danger of future crises and to seize every opportunity for development. Success on this front will help us build the world we all want: one free of poverty, with shared prosperity for all. Jim Yong Kim President The World Bank Group Acknowledgments . This Report was prepared by a team led by Norman Loayza, together with Inci Ötker-Robe. The other members of the core team were César Calderón, Stéphane Hallegatte, Rasmus Heltberg, Xubei Luo, Martin Melecky, Ana María Oviedo, and Kyla Wethli. Research analysts Sebastien Boreux, Kanako Goulding-Hotta, Rui Han, Harry Edmund Moroz, Anca Maria Podpiera, Jun Rentschler, Faiyaz Talukdar, and Tomoko Wada completed the team. Gilles Cols, Olga Jonas, Federica Ranghieri, and Anna Reva contributed to the Report’s spotlights. The Report was sponsored by the Development Economics Vice Presidency. Overall guidance for the prepara- tion of the Report was provided by Kaushik Basu, Senior Vice President and Chief Economist, and Asli Demirgüç- Kunt, Director of Research, Development Economics. The team benefited from continuous engagement with and advice from Martin Čihák, Quy-Toan Do, Mary Hallward-Driemeier, Aart Kraay, and Sergio Schmukler. An advisory panel comprising Laura Alfaro, Robert Barro, Thorsten Beck, Stefan Dercon, Ibrahim Elbadawi, Rohini Pande, Klaus Schmidt-Hebbel, Hyun Song Shin, and Jan Švejnar provided feedback and advice. The team also benefited from the advice of World Bank Chief Economists Augusto de la Torre, Shantayanan Devarajan, Marianne Fay, Ariel Fiszbein, Caroline Freund, Indermit Gill, Bert Hofman, Jeffrey Lewis, and Martín Rama. The team would like to acknowledge the generous support for the preparation of the Report by the Canadian International Development Agency, the Knowledge for Change Program, the Japan Policy and Human Resources Development Fund, and the World Bank Research Support Budget. The team also thanks the German Federal Ministry for Economic Cooperation and Development and the Deutsche Gesellschaft für Internationale Zusammenarbeit, which co-organized and hosted the WDR International Policy Workshop in Berlin, November 2012. Interagency consultations were held with the European Commission, the International Monetary Fund, the Organisation for Economic Co-operation and Development, several United Nations organizations, the World Economic Forum, and agencies for development cooperation in Denmark, Finland, France, Japan, the Netherlands, Norway, Spain, Sweden, Switzerland, and the United Kingdom. Valuable inputs were received from the World Bank Institute and all regional and anchor networks, as well as other parts of the World Bank Group, including the International Finance Corporation and the Multilateral Investment Guarantee Agency. Country consultations were held in Austria, Belgium, Brazil, Chile, Denmark, Finland, France, Germany, Indonesia, Japan, the Netherlands, Norway, Peru, Rwanda, Singapore, Spain, Sweden, Switzerland, the United Kingdom, and the United States. Most included academics, members of civil society, and public entities and governments. Consultations with researchers and academics were aided by ad hoc conferences organized by the Centre for the Study of African Economies, Oxford University, and the Center on Global Governance at the School of International and Public Affairs, Columbia University. The team also received valuable feedback at the African Economic Conference 2012, the Asia Development Forum 2013, and the Latin American and Caribbean Economic Association Conference 2012. The Report was skillfully edited by Nancy Morrison and Martha Gottron. Bruce Ross-Larson and Gerry Quinn provided additional editorial advice. The World Bank’s Publishing and Knowledge Division coordinated the design, typesetting, printing, and dissemination of the Report. Special thanks to Mary Fisk, Stephen McGroarty, Stephen Pazdan, Denise Bergeron, Andres Meneses, Shana Wagger, and Paschal Ssemaganda, as well as to the Translation and Interpretation Unit’s Bouchra Belfqih, Cecile Jannotin, and Michael Lamm. The Development Data Group contributed to the preparation of the Report’s statistical annex, coordinated by Timothy Herzog. The team also thanks Merrell Tuck-Primdahl, Vamsee Krishna Kanchi, and Swati P. Mishra for their guidance on communications strategy, and Vivian Hon for her coordinating role. Barbara Cunha, Birgit Hansl, and Manal Quota reviewed some of the foreign language translations of the Overview. The production and logistics of the Report were assisted by Brónagh Murphy, Mihaela Stangu, and Jason Victor, with contributions from Laverne Cook, Gracia Sorensen, and Tourya Tourougui. Ivar Cederholm, Elena xv xvi ACKNOWLEDGMENTS Chi-Lin Lee, and Jimmy Olazo coordinated resource mobilization. Irina Sergeeva and Sonia Joseph were in charge of resource management. Gytis Kanchas, Nacer Megherbi, and Jean-Pierre Djomalieu provided IT support. Background papers were provided by Joshua Aizenman, Phillip R. D. Anderson, Maximillian Ashwill, Emmanuelle Auriol, Ghassan Baliki, Thorsten Beck, Najy Benhassine, Nicholas Bloom, Julia K. Brown, Martin Brown, Daniel Buncic, Julio Cáceres-Delpiano, Sara Guerschanik Calvo, Olivier De Jonghe, Alejandro de la Fuente, Philippe de Vreyer, Mark A. Dutz, Maya Eden, Penelope D. Fidas, Roberto Foa, Rodrigo Fuentes, Garance Genicot, Gary Gereffi, Ejaz Ghani, Sudarshan Gooptu, Mikael Grinbaum, Federico H. Gutierrez, Ronald Inglehart, Susan T. Jackson, Olga B. Jonas, Jan Kellett, Ilan Kelman, Tariq Khokhar, Auguste T. Kouame, Aart Kraay, Sadaf Lakhani, Sylvie Lambert, Esperanza Lasagabaster, Ethan Ligon, Samuel Maimbo, William F. Maloney, Tom Mitchell, Ahmed Mushfiq Mobarak, Hernan J. Moscoso Boedo, Andrew Norton, Eduardo Ortiz- Juárez, Patti Petesch, Florence Pichon, Patrick Premand, Carlos Rodriguez Castelan, Natalia Salazar, Luis Servén, Francis J. Teal, Maarten van Aalst, Guillermo Vuletin, Koko Warner, Tetyana V. Zelenska, and Nong Zhu. Details of their contributions are listed at the end of the Report. For valuable contributions and advice, the team thanks Pablo Ariel Acosta, Tony Addison, Montek Ahluwalia, Ahmad Ahsan, David Aikman, Harold Alderman, Franklin Allen, Aquiles Almansi, Philippe Ambrosi, Goli Ameri, Walter J. Ammann, Dan Andrews, Paolo Avner, Edmar Bacha, Javier Baez, Hemant Baijal, Christopher Barrett, Scott Barrett, Kathleen Beegle, Tim Besley, Gordon Betcherman, Deepak Bhattasali, Indu Bhushan, Jörn Birkmann, Christiane Bögemann-Hagedorn, Uta Böllhoff, Patrick Bolton, Laura Elizabeth Boudreau, François Bourguignon, Carter Brandon, Juan José Bravo, Tilman Brück, Robin Burgess, Guillermo Calvo, Jack Campbell, Jason Cardosi, Michael R. Carter, Miguel Castilla, Michael Chaitkin, Marcos Chamon, Guang Zhe Chen, Maria Teresa Chimienti, Fredrick Christopher, Craig Churchill, Luis Abdón Cifuentes, Massimo Cirasino, Stijn Claessens, Daniel Clarke, Tito Cordella, Sarah E. Cornell, Gerardo Corrochano, Robert Cull, Julie Dana, Anis Dani, Jishnu Das, Joachim De Weerdt, Ximena Del Carpio, Jean-Jacques Dethier, Jacqueline Devine, Pierre Dubois, Patrice Dumas, Peter Ellehoj, Brooks Evans, Jessica Evans, Marcel Fafchamps, Paolo Falco, Shahrokh Fardoust, Thomas Feidieker, Wolfgang Fengler, James Fenske, Ana Margarida Fernandes, Adrián Fernández, Francisco Ferreira, Deon Filmer, Greg Fischer, James Foster, Marcel Fratzscher, Linda Freiner, Roberta Gatti, Francis Ghesquiere, Swati Ghosh, Antonino Giuffrida, David Gleicher, Markus Goldstein, George Graham, Margaret Grosh, Patricia Grossi, Mario Guadamillas, Conor Healy, Frank Heemskerk, Joachim Heidebrecht, Jesko Hentschel, Rafael Hernández, Matt Hobson, John Hoddinott, Niels Holm-Nielsen, Naomi Hossain, Andrew Hughes Hallett, Oh-Seok Hyun, Elena Ianchovichina, Ridzuan Ismail, Takatoshi Ito, Abhas K. Jha, Emmanuel Jimenez, Steen Jørgensen, Nidhi Kalra, Sujit Kapadia, Masayuki Karasawa, Corneille Karekezi, Supreet Kaur, Lauren Kelly, Igor Kheyfets, Beth King, Naohiro Kitano, Leora Klapper, Alzbeta Klein, Kalpana Kochhar, Kiyoshi Kodera, Friederike Koehler-Geib, Diane Koester, Robert Kopech, Anirudh Krishna, Jolanta Kryspin-Watson, Howard Kunreuther, Kiyoshi Kurokawa, Christoph Kurowski, Miguel Laric, Alexia Latortue, Sara Lazzaroni, Nick Lea, Daniel Lederman, Margaret Leighton, Robert Lempert, Sebastian Levine, Yue Li, Irina Likhacheva Sokolowski, Justin Yifu Lin, Kathy Lindert, Gladys Lopez, Augusto López Claros, Leonardo Lucchetti, Maria Ana Lugo, Olivier Mahul, Thomas Markussen, Will Martin, María Soledad Martínez Pería, Eric Maskin, Laura Mazal, J. Allister McGregor, Claire McGuire, Robin Mearns, Carlo Menon, Rekha Menon, Erwann Michel-Kerjan, Tim Midgley, Gary Milante, Suguru Miyazaki, Nuno Mota Pinto, Marialisa Motta, Joy Muller, Akira Murata, Lydia Ndirangu, Ha Nguyen, Giuseppe Nicoletti, Yosuke Nishii, Michel Noel, Alistair Nolan, Sharyn O’Halloran, Philip O’Keefe, Ory Okolloh, Michelle Ooi, Miguel Angel Ostos, Marcus C. Oxley, Robert Palacios, Pepi Patrón, Douglas Pearce, Brian Pinto, Russell Pittman, Jean-Philippe Platteau, Sandra Poncet, David Popp, Antonin Pottier, Prashant, John Primrose, Hnin Hnin Pyne, Ricardo Raineri, Anthony Randle, Martin Ravallion, Robert Reid, Ricardo Reis, Ortwin Renn, Changyong Rhee, Helena Ribe, Michelle Riboud, Jamele Rigolini, Dena Ringold, David Robalino, Jorge Luis Rodriguez Meza, Rafael Rofman, Jonathan Rothschild, Davinder Sandhu, Apurva Sanghi, Hans-Otto Sano, Yasuyuki Sawada, Stefano Scarpetta, Anita Schwarz, Paul Seabright, Junko Sekine, Amartya Sen, Rodrigo Serrano-Berthet, Shigeo Shimizu, Paul B. Siegel, Joana Silva, Emmanuel Skoufias, Marc Smitz, Irina Solyanik, Joseph Stiglitz, Adrian Stone, Stéphane Straub, Henriette Strothmann, Pablo Suarez, Kalanidhi Subbarao, Mark Sundberg, Olumide Taiwo, Tamanna Talukder, Kazushige Taniguchi, Finn Tarp, Gaiv Tata, Maria Hermínia Tavares de Almeida, Stoyan Tenev, Mehrnaz Teymourian, Erik Thorbecke, Klaus Tilmes, Carlos Tortola, Izabela Toth, Carolina Trivelli Ávila, Yvonne Tsikata, María Cristina Uehara, Tunc Tahsin Uyanik, Renos Vakis, Dominique Van De Walle, Ashutosh Varshney, Adrien Vogt-Schilb, Eiji Wakamatsu, Sophie Walker, Simon Walley, Christine Wallich, David Waskow, Masato Watanabe, Asbjorn H. Wee, Jonathan B. Wiener, Alys Willman, Lixin Colin Xu, Mohamed Mahdi Youssouf, and Asta Zviniene. The team also thanks the many others inside and outside the World Bank who provided comments. OVERVIEW Risk and opportunity Risk management can be a powerful instrument for development Managing risk for a life full of opportunities: a mother protects her child against malaria with a bed net in Ghana. FPO © Arne Hoel/World Bank OVERVIEW Risk and opportunity Risk management can be a powerful instrument for development The past 25 years have witnessed un- conducting it effectively, and how can precedented changes around the these obstacles be overcome? The WDR world—many of them for the bet- 2014’s value added resides in its em- ter. Across the continents, many phasis on managing risks in a pro- countries have embarked on a active, systematic, and integrated path of international integration, way. These characteristics under- economic reform, technological score the importance of forward- modernization, and democratic looking planning and preparation participation. Although challenges in a context of uncertainty. They and inequalities remain, economies also highlight the necessity to address that had been stagnant for decades are all relevant risks jointly, using all avail- growing, people whose families had suffered able tools and institutions. From a policy deprivation for generations are escaping poverty, and maker’s perspective, a proactive, systematic, and in- hundreds of millions are enjoying the benefits of im- tegrated approach to managing risks involves strik- proved living standards and scientific and cultural ing a proper balance between the contribution from sharing across nations. As the world changes, a host the state and the contribution from individuals, civil of opportunities arise constantly. With them, how- society, and the private sector, with the goal of en- ever, appear old and new risks, from the possibility of suring that these contributions are coordinated and job loss and disease to the potential for social unrest complementary. and environmental damage. If ignored, these risks The WDR 2014 argues that risk management can can turn into crises that reverse hard-won gains and be a powerful instrument for development—not endanger the social and economic reforms that pro- only by building people’s resilience and thus reducing duced these gains. The solution is not to reject change the effects of adverse events but also by allowing them in order to avoid risk but to prepare for the opportu- to take advantage of opportunities for improvement. nities and risks that change entails. Managing risks The WDR 2014 is not devoted to a detailed analysis responsibly and effectively has the potential to bring of specific risks. Its framework, however, can be im- about security and a means of progress for people in plemented to address particular, relevant sets of risks developing countries and beyond. in given regions and countries. Focusing on the pro- The World Development Report (WDR) 2014 cess of risk management allows the WDR 2014 to focuses on the process of risk management, address- consider the synergies, trade-offs, and priorities in- ing these questions: why is risk management impor- volved in addressing different risks in different con- tant for development, how should it be conducted, texts, with the single motivation of boosting develop- what obstacles prevent people and societies from ment (box 1). 3 4 WORLD DEVELOPMENT REPORT 2014 B OX 1 Five key insights on the process of risk management from the World Development Report 2014 1. Taking on risks is necessary to pursue opportunities 4. For risks beyond the means of individuals to handle for development. The risk of inaction may well be the alone, risk management requires shared action and worst option of all. responsibility at different levels of society, from the 2. To confront risk successfully, it is essential to shift from household to the international community. unplanned and ad hoc responses when crises occur to 5. Governments have a critical role in managing sys- proactive, systematic, and integrated risk management. temic risks, providing an enabling environment for 3. Identifying risks is not enough: the trade-offs and shared action and responsibility, and channeling obstacles to risk management must also be identi- direct support to vulnerable people. fied, prioritized, and addressed through private and public action. Source: WDR 2014 team. Risk is a burden but also an opportunity sumption by almost 10 percent and continued to be negatively affected three to five years later.1 Health Why worry about risk? In recent years, a multitude costs from high levels of crime and violence amount of crises have disrupted the world economy and to 0.3–5.0 percent of gross domestic product (GDP) have had substantial negative consequences on de- a year for countries in Latin America, without even velopment. Because of the 2008–09 global financial considering the impact of crime on lost output crisis, most economies around the world experi- stemming from reduced investment and labor par- enced sharp declines in growth rates, with ensuing ticipation.2 Loss of employment in countries as dif- loss of income and employment and setbacks in ef- ferent as Argentina, Bulgaria, and Guyana not only forts to reduce poverty. When food prices spiked in has lowered income and consumption but has also 2008, riots broke out in more than a dozen countries reduced people’s ability to find new work, worsened in Africa and Asia, reflecting people’s discontent and social cohesion, and in some cases increased domes- insecurity and causing widespread political un- tic violence.3 rest. The 2004 Asian tsunami, the 2010 earthquake Whether adverse consequences come from sys- in Haiti, and the 2011 multiple hazard disaster in temic or idiosyncratic risks, they may destroy lives, northeastern Japan—to name but a few—have left assets, trust, and social stability. And it is often the a trail of fatalities and economic losses that exem- poor who are hit the hardest. Despite impressive plify the increased frequency and intensity of natu- progress in reducing poverty in the past three de- ral disasters. Concerns about the impact of climate cades, a substantial proportion of people in devel- change worldwide are growing, and so are fears oping countries remain poor and are vulnerable to about the spreading of deadly contagious diseases falling into deeper poverty when they are struck by across borders. Indeed, the major economic crises negative shocks (figure 1). The mortality rate from and disasters that have occurred in recent years and illness and injury for adults under age 60 is two and those that may occur in the future underscore how a half times higher for men and four times higher vulnerable people, communities, and countries are for women in low-income countries than in high- to systemic risks, especially in developing nations. income countries, while the rate for children under Idiosyncratic risks, which are specific to individu- age five is almost twenty times higher.4 Mount- als or households, are no less important for people’s ing evidence shows that adverse shocks—above all, welfare. Losing a job or not finding one because of health and weather shocks and economic crises— inadequate skills, falling victim to disease or crime, play a major role in pushing households below the or suffering a family breakup from financial strain or poverty line and keeping them there.5 Moreover, forced migration can be overwhelming, particularly realizing that a negative shock can push them into for vulnerable families and individuals. Households destitution, bankruptcy, or crisis, poor people may in Ethiopia whose members experienced serious stick with technologies and livelihoods that appear illness, for example, were forced to cut their con- relatively safe but are also stagnant. Risk and opportunity 5 F I G U R E 1 Many people around the world are poor or live very close to poverty; they are vulnerable to falling deeper into poverty when they are hit by negative shocks More than 20 percent of the population in developing countries live on less than $1.25 a day, more than 50 percent on less than $2.50, and nearly 75 percent on less than $4.00. a. All developing countries, 2010 b. Developing countries by region, 2010 500 $1.25 per day 16 $1.25 per day % of population in each region 14 $2.50 per day $2.50 per day Total population, millions 400 12 300 10 8 200 6 4 100 2 0 0 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 US dollars US dollars Sub-Saharan Africa Middle East and North Africa South Asia Latin America and the Caribbean East Asia and Pacific Europe and Central Asia Source: WDR 2014 team based on data from World Bank PovcalNet (database). Note: $1.25 per day is a widely used measure of extreme poverty. However, $2.50 per day is considered a more relevant measure of extreme poverty for some regions, such as Latin America and the Caribbean. See Ferreira and others 2013. Yes, confronting risk, as the possibility of loss, is Risk management can be a powerful a burden—but it is also necessary to the pursuit of instrument for development opportunity. Risk and opportunity go hand in hand in most decisions and actions taken by countries, Whether risks are systemic or idiosyncratic, imposed enterprises, and families as they seek to improve or taken on voluntarily, development can occur only their fate. Indeed, risk taking is intrinsic to the pro- by successfully confronting risk and pursuing op- cess of development. Consider a few examples. Since portunity. Many crises and development losses are the 1990s, most developing countries have opened the result of mismanaged risks. No less important, their borders to seek international integration and many opportunities are missed because preparation higher economic growth, but in the process they for risk is insufficient and necessary risks are not have also increased their exposure to international taken—the “risk of inaction.” It is therefore essential shocks. Firms around the world have made invest- to shift from unplanned and ad hoc responses when ments to upgrade their technologies and increase crises occur to proactive, systematic, and integrated profitability, but the debt required to do so has risk management. As such, risk management can made them more vulnerable to changes in demand build the capacity to reduce the losses and improve and credit conditions. From Brazil to South Africa, the benefits that people may experience while con- millions of families have migrated to cities to seek ducting their lives and pursuing development op- better job opportunities and health and education portunities (drawing 1 and profile 1). services, where they have also become more exposed Risk management can save lives. Consider the to higher crime and benefit less from communal case of Bangladesh, where improved preparation for support. The motivation behind these actions is natural hazards has dramatically reduced loss of life the quest for improvement, but risk arises because from cyclones. In the past four decades, three major favorable outcomes are seldom guaranteed. cyclones of similar magnitude have hit Bangladesh. 6 WORLD DEVELOPMENT REPORT 2014 D R AW I N G 1 Risk management for everyone: A visual representation of key concepts To pursue opportunity, people must …not one risk, but many confront risk Job Job loss Disease Loss Opportunity Risk Crime Natural Natural Disasters disasters Financial Financial crises Crises ...and often burdened with obstacles risk burdened ...and often Sharing with with others canobstacles overcome to manage them to manage these them obstacles Job Disease Disease Job Job Community Family loss Loss Loss Natural Natural Natural Disasters disasters Disasters Crime Crime Banks Financial Financial Enterprises Financial crises Crises Crises Int‛l ,QW·O Government community Community ...through collective action and institutions Risk management can be a powerful tool for development Knowledge Protection Insurance Coping Drawing by Jason Victor for the WDR 2014. Risk and opportunity 7 PROFILE 1 The Gomez family: A modern tale of risk and resilience The Gomez family lives in a shantytown on the from school. Having two income earners (and a outskirts of Lima. Only a few years ago, the fam- willing grandmother) made the Gomez house- ily lived in a rural village in the Peruvian Andes, hold more resilient to whatever might happen. where they had a small farm. The region was And things did happen. Mario, the eldest prone to droughts, and they could never earn son, was injured in a traffic accident. There was enough income to escape poverty. Many of no car insurance, and the family had to bear the their neighbors had migrated to the city in the cost of Mario’s medical treatment. They could not 1980s, pushed by civil conflict in the countryside. have done it alone, and they didn’t have to. They The Gomez family refused to go for fear of los- relied on a public hospital, run and financed by ing their land and finding nothing better in the the state. Medical treatment there was of uneven city. The risk was too large. Peru was a different quality, but it provided basic services. The fam- place then: inflation and unemployment were ily had to spend some of their limited savings to rampant, and the threat of social unrest was ever supplement the hospital services and buy medi- present. cation, but all that was worth it because Mario In the 1990s, the macroeconomy was stabi- recovered. lized and the civil war ended. New opportunities The Gomezes had to dig into their assets once started to arise in urban and rural areas. At first, again, but this time for a very different purpose. these opportunities eluded the Gomez family. A Elena—the second daughter, whom everyone dam had been constructed near their village, but regarded as the brains in the family—came using its waters required the renovation of canals home one day and asked her parents if she could on their farm. They applied for a loan from a com- study English in the evenings. This was a good mercial bank but were denied, which came as no idea. Peru had recently signed several free trade surprise since it was their first time applying. Mr. agreements (one of them with the United States), and Mrs. Gomez came to believe that their chil- and exporting companies had started to grow, dren had no future in the village and decided to offering jobs to young, qualified people. English migrate to the city. This time, however, they did would be a big plus. not have to worry about losing their farm. They Some months before, however, her parents had been given a property title and were able to would have declined her initiative on the grounds sell the farm to a neighbor, who had the capital that it was not safe to be out at night. Police pro- to renew the canals. The money from the farm tection was scarce in the outskirts of the city, and would give the Gomezes a cushion as they took criminals took advantage of that. When a crime the momentous challenge of migration. wave eventually affected the Gomezes’ shanty- Lima, with just under 10 million inhabitants, town, the community put together neighbor- seemed like a huge and inhospitable place. That hood patrols (effective, although at times unduly is why they decided to move to the shantytown harsh). When Elena asked for English classes, the where many members of their village had relo- safety risk had been reduced, and she could go cated. There, they would find companionship, out to study in the evenings. As time passed, she cultural identity (all the festivals of their old and her family would be well prepared to benefit village were properly celebrated here), and, of from the period of stability and sustained growth course, help finding a job. Mr. Gomez found work that Peru was experiencing. on a construction site, but it was irregular, with Confronting risks and seizing opportunities frequent layoffs. Mrs. Gomez had to pitch in, and may have put the Gomez family on the path out she was fortunate to find work as a seamstress in of poverty, possibly forever. It was their work, ini- a textile enterprise. The grandmother helped out, tiative, and responsibility that made it possible, taking care of the children when they returned but they could not have done it alone. Source: WDR 2014 team. Note: A video of this fictional story is available in nine languages on the World Development Report 2014 website http://www.worldbank .org/wdr2014. 8 WORLD DEVELOPMENT REPORT 2014 FIGURE 2 The benefits of risk management often outweigh the costs 12 10 Benefit-cost ratios 8 6 4 2 Break- even point 0 Vaccinations Improved Early warning Nutritional Measures to reduce damage from: water and systems interventions Earthquakes Floods Tropical storms sanitation Source: Wethli 2013 for the WDR 2014. Note: The figure shows the median of benefit-cost ratios across a range of studies in each category (with a minimum of at least four esti- mates in each category). Above the dotted line, expected benefits exceed expected costs. The range of estimates within each category can be substantial, reflecting a diversity of intervention types and locations, and the sensitivity of estimates to variations in underlying assumptions. However, in almost all cases, even the 25th percentile of the ranges are above the break-even point. A cyclone in 1970 claimed over 300,000 lives, but to undertake new promising ventures. Some farmers one in 1991 claimed almost 140,000, and one in 2007 in Ethiopia, for instance, choose not to use fertilizer claimed about 4,000. Casualties have been greatly because they fear drought and other potential shocks reduced by a nationwide program to build shel- and thus prefer to retain savings as a cushion rather ters—from only 12 shelters in 1970 to over 2,500 in than investing in intermediate inputs.8 In contrast, 2007—along with improved forecasting capacity and farmers in Ghana and India have been more willing a relatively simple but effective system for warning to take on risk in search of higher yields—increas- the population.6 ing their investments in fertilizer, seeds, pesticides, Risk management can avert damages and prevent and other inputs—because they have rainfall insur- development setbacks. Countries as different as the ance.9 When aggregated, these gains can have much Czech Republic, Kenya, and Peru offer recent com- broader effects, contributing to improved productiv- pelling examples where macroeconomic preparation ity and growth for a country as a whole. has shielded the economy from the negative effects Crises and losses from mismanaged risks are of a global financial crisis. Having achieved lower fis- costly, but so are the measures required to better cal deficits, disciplined monetary policy, and lower prepare for risks. So, does preparation pay off? Ben- current account deficits, these countries experienced efit-cost analyses across a number of areas suggest a smaller decline in growth rates in the aftermath of that risk preparation is often beneficial in averting the 2008 international crisis than they did following costs, sometimes overwhelmingly so (figure 2). There the 1997 East Asian crisis. The same beneficial ef- seems to be a lot of truth in the old adage that “an fect of macroeconomic preparation seems to have ounce of prevention is worth a pound of cure.” For occurred in many other low- and middle-income example, a regimen of mineral supplements designed countries.7 to reduce malnutrition and its related health risks Risk management can unleash opportunity. Risk may yield benefits at least 15 times greater than the management tools—such as improved information, cost of the program.10 Similarly, improving weather crop insurance, and employment diversification— forecasting and public communication systems to can help people mitigate risk. The ability to mitigate provide earlier warning of natural disasters in devel- risk, in turn, can allow people, especially the poor, to oping countries could yield estimated benefits 4 to 36 overcome their aversion to risk and be more willing times greater than the cost.11 Risk and opportunity 9 B OX 2 A risky world: Trends in risk across regions The risks that people face have changed considerably over time, every region of the world. While Latin America, the Middle East and although this evolution has sometimes varied across regions. Risks North Africa, and Sub-Saharan Africa all have suffered significantly have eased in some areas—such as maternal health, where the mor- fewer years of recession in each decade since the 1980s, Organisa- tality rate has declined in all regions. Conversely the incidence of tion for Economic Co-operation and Development (OECD) countries crime has increased substantially in Latin America and Sub-Saharan have experienced more. Africa. Strikingly, the incidence of natural disasters has increased in a. Maternal mortality b. Homicides 25 Rate per 100,000 live births 800 1990 Rate per 100,000 people, 700 1981–1990 2000 20 1991–2000 600 annual average 2010 2001–2010 500 15 400 300 10 200 5 100 0 0 OECD EAP ECA LAC MENA SAR SSA OECD EAP ECA LAC MENA SAR SSA c. Incidence of natural disastersa d. Large recessionsb 4.5 0.4 4.0 1981–1990 1981–1990 Proportion of decade 1991–2000 1991–2000 3.5 0.3 Annual average 2001–2010 2001–2010 3.0 in recession 2.5 0.2 2.0 1.5 1.0 0.1 0.5 0 0 OECD EAP ECA LAC MENA SAR SSA OECD EAP ECA LAC MENA SAR SSA Source: WDR 2014 team based on data from World Bank World Development Indicators (database); EM-DAT OFDA/CRED International Disaster Database; United Nations Office on Drugs and Crime Homicide Statistics (database). Note: Figures show the simple average across countries in each region. OECD countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. a. Natural disasters include droughts, earthquakes, floods, and tropical storms. b. Large recessions are identified by following Barro and Ursúa 2012 and using as a threshold a 5 percent decline in GDP per capita growth from peak to trough. There were no large recessions in South Asia from 1991 to 2010. Comparing the cost-effectiveness of preparing is both unavoidable and necessary. For instance, a for risk with that of coping with its consequences family living in a violence-ridden community faces is one of the important trade-offs that must be as- safety, health, and property risks and must choose sessed. The choice between these actions depends how to allocate its limited budget to protect and in part on how the (certain) costs of preparing for insure against each of these risks. Likewise, a small risk compare to the (often uncertain) benefits of country prone to torrential rains and also exposed doing so.12 In addition, risk management requires to international financial shocks must decide how considering different risks and the relative need of much to spend in flood prevention infrastructure preparing for each of them (box 2). Given limited and how much to save to counteract the effects of resources, setting priorities and making choices financial volatility. 10 WORLD DEVELOPMENT REPORT 2014 When risks are taken on voluntarily in the pursuit prices their products will command in the market. of opportunity, another trade-off emerges: expected And governments decide the level of policy interest returns must be weighed against the potential losses rates and fiscal deficits in the presence of uncertain of a course of action. This trade-off is intensified external conditions, domestic productivity growth, when a higher return is possible only if more risk is and changes in financial markets. accepted. That is often the case with financial invest- ments, where a lower yield is characteristic of a more The analysis of choice under uncertainty in secure position, and higher yields with riskier posi- economics and public policy tions.13 A risk-return trade-off may also be perceived for certain development actions: for instance, public It is only natural, therefore, that the analysis of choice opinion and certain experts may link the pursuit of under uncertainty and scarce resources has been at higher economic growth with lower environmental the heart of economics and public policy for cen- protection or higher inequality.14 Although this and turies. The basic approach to decision under un- other risk-return trade-offs may not be present, risk certainty—introduced by Daniel Bernoulli in the management entails addressing them as a legitimate 1700s and modeled formally by John von Neumann possibility. and Oskar Morgenstern in 1944—is based on the Risk management involves not only considering notion that individuals optimize the expected “util- trade-offs but also taking synergies into account. ity” (or subjective perception of welfare) of possible These can make both preparation for and conse- outcomes.18 This expected utility approach relies on quences of risk less costly. They can also diminish individuals making rational choices, based on their risks and increase expected benefits. These “win- preferences for risk and their knowledge of potential win” situations are widespread and should be em- outcomes and respective probabilities. phasized—which is not to say that they are costless Notwithstanding its valuable insights, this ap- or always easy to implement. Investments in nutri- proach has been challenged on two important tion and preventive health, for example, make people grounds. The first is that individuals do not seem to more productive while reducing their vulnerability operate in a fully rational manner, possibly because to disease.15 Similarly, improvements in the busi- uncertainty makes the decision process so com- ness environment, such as streamlining regulations plicated that people prefer simple behavioral rules and improving access to credit, can induce the en- that evolve over time but are not always optimal. terprise sector to become more dynamic and grow The work of Maurice Allais in the 1950s and Daniel more quickly, while also making it more resilient to Kahneman and Amos Tversky in the 1970s focused negative shocks.16 At the macroeconomic level, dis- attention on the limitations and innate tendencies of ciplined monetary and fiscal policies—reflected in human behavior when confronting decisions under moderate inflation and sustainable public deficits— uncertainty.19 accelerate economic growth while reducing high vol- The second challenge to the basic expected utility atility in the face of external and domestic shocks.17 approach is that individuals do not make decisions in isolation but in groups, mainly because the poten- tial outcomes can be greatly affected by how people What does effective risk management act in coordination with others. The work of Duncan entail? Black in the 1940s and James Buchanan and Mancur As the ancient Greek philosopher Heraclitus wrote, Olson in the 1960s emphasized the shortcomings of the only thing constant is change. And with change and obstacles to collective action.20 Although origi- comes uncertainty. Faced with choices for bettering nally concerned with the state’s provision of public their lives, people make virtually every decision in the goods, the public choice approach extends to actions presence of uncertainty. Young people decide what to taken by any group, from households to communi- study or train for without knowing exactly what jobs ties of any size. The basic insight is how valuable and and wages will be available when they enter the labor at the same time elusive it is to coordinate collective market. Adults decide how much and how to save for action, especially in the face of uncertainty. retirement in the face of uncertain future income and A different strand of the economics literature is investment returns, health conditions, and life spans. also concerned with the collective action problem Farmers decide what to cultivate and what inputs to and offers critical principles to overcome them. use not knowing with certainty whether there will be In their pioneering work in the 1960s and 1970s, enough rain for their crops and what demand and Leonid Hurwicz, Roger Myerson, and Eric Maskin Risk and opportunity 11 studied the problem of mechanism design to may be positive (such as abundant rainfall or a wind- achieve efficiency in markets, organizations, and fall in terms of trade) or negative (illness or war). institutions. The critical insight here is that incen- They may affect small groups (such as a family or a tive constraints should be considered as important rural community) or large ones (a region or a coun- as resource constraints in understanding decision try). And they may occur suddenly (such as natu- making in the presence of uncertainty.21 This in- ral hazards or financial shocks) or gradually (such sight is vital when developing the best ways to coor- as demographic transitions, technological trends, dinate the collective action of any group, especially or environmental changes). Whether the outcomes under asymmetric information, diverging interests, from those shocks are positive or negative, large or and limited knowledge. It forces analysts and policy small, individualized or widespread, depends on the makers to see beyond aggregate resources and ques- interaction between shocks and the internal and ex- tion what informs and motivates the actions of peo- ternal conditions that characterize a social and eco- ple and organizations, including actions related to nomic system (such as a household, a community, managing risk. or a country). Importantly, the effect of shocks on people’s outcomes is also mediated by their actions to prepare for and confront risk. An analytical framework for risk management This interaction can be represented by a risk The insights derived from the economics of decision chain (diagram 1), which can be applied to different under uncertainty provide an analytical framework types of risks and contexts.22 For example, whether for risk management. The World Development Report someone becomes ill during a pandemic depends 2014 proposes that this framework consists of several on how contagious the virus is (the initial shock); interrelated steps: population density and living conditions in given areas (the external environment or exposure); peo- • Assessing the fundamental goals of and motiva- ple’s individual susceptibility (internal conditions, tions for risk management: that is, resilience in the such as their age or the strength of their immune face of adverse events and prosperity through the system); and the steps they take to prevent becom- pursuit of opportunities (discussed in the first two ing sick or contaminating others, such as frequently sections above). washing their hands or wearing a face mask (risk • Understanding the environment in which risks management). Similarly, whether an enterprise can and opportunities take place (referred to below as successfully take advantage of new technology and the risk chain). innovation depends on the characteristics of the technology (the initial shock); the infrastructure in • Considering what risk management entails: that the country, which may affect the enterprise’s access is, preparing for and coping with both adverse and to the technology (the external environment); how positive events (presented below under “The com- innovative the enterprise is (internal conditions); ponents of risk management”). and how much capital the enterprise has accumu- lated and how informed it is about the benefits and • Assessing the main obstacles that individuals and potential drawbacks of the new technology (risk societies face in managing risk, including con- management). straints on resources, information, and incentives In this context, risk is defined as the possibility of (discussed below in the section entitled “Beyond loss. Risk is not all bad, however, because taking risks the ideal”). is necessary to pursue opportunity. Opportunity is • Introducing the potential role of groups and col- defined as the possibility of gain, thus representing lective action at different levels of society to over- the upside of risk. People’s exposure to risk is deter- come the obstacles that people encounter in man- mined by their external environment. For example, aging risk (presented below in the section “The whether a house is exposed to the risk of coastal way forward”). flooding depends on its location. Vulnerability occurs when people are especially susceptible to losses from negative shocks because of a combination of large Understanding the environment in which risks exposure, weak internal conditions, and deficient and opportunities arise: The risk chain risk management. For example, a highly leveraged The world is constantly changing and generating financial institution that has taken very risky posi- shocks that affect individuals and societies. Shocks tions without counterbalancing hedges is vulnerable 12 WORLD DEVELOPMENT REPORT 2014 The risk chain: The nature and extent of outcomes depend on shocks, D I AG R A M 1 exposure, internal conditions, and risk management External environment Risk ks Shocks management anagement Outcomes Internal conditions Source: WDR 2014 team. Note: The feedback arrows in the risk chain diagram represent the potential for the outcomes of past shocks to affect exposure and internal conditions, as well as the propensity for future shocks. Similarly, the effectiveness of people’s risk management can significantly affect the nature of and propensity for future shocks. to an economic or financial shock. Likewise, a poor The components of risk management: household with few assets and volatile income may Preparation and coping be especially vulnerable to increased food prices. To achieve that goal, risk management needs to Risk management is the process of confronting combine the capacity to prepare for risk with the risks, preparing for them, and coping with their ef- ability to cope afterward—taking into account how fects. Resilience is characterized by the ability of peo- the up-front cost of preparation compares with its ple, societies, and countries to recover from negative probable benefit. Building on the seminal contribu- shocks, while retaining or improving their ability to tion from Isaac Ehrlich and Gary Becker, prepara- function. Much of the emerging literature on risk in tion should include a combination of three actions a development context emphasizes the important that can be taken in advance: gaining knowledge, role that risk management can play in increasing acquiring protection, and obtaining insurance.23 resilience to negative shocks. However, to increase Once a risk (or an opportunity) materializes, people prosperity and well-being, risk management also take action to deal with what has occurred through has an essential role in helping people and countries coping (diagram 2). A strong risk management strat- successfully manage positive shocks. Indeed, success- egy would include all four of these components: fully managing positive shocks is a critical part of knowledge, protection, insurance, and coping. They increasing people’s resilience to negative shocks over interact with each other, potentially improving each time. For example, a farmer’s ability to withstand a other’s quality. For instance, better knowledge can drought may be substantially influenced by how the lead to more efficient decisions regarding the alloca- yields from years of good rainfall were managed. tion of resources between insurance and protection. Thus the goal of risk management is to both decrease Likewise, better insurance and protection can make the losses and increase the benefits that people expe- coping less difficult and costly. Several obstacles, rience when they face and take on risk. however, often make this risk management strat- Risk and opportunity 13 D I AG R A M 2 The interlinked components of risk management Insurance To transfer resources across people and over time, from good to bad states of nature Knowledge To understand shocks, internal and Coping external conditions, and potential To recover from losses and outcomes, thus reducing uncertainty make the most of benefits Protection To reduce the probability and size of losses and increase those of benefits Preparation Coping Source: WDR 2014 team. egy difficult to achieve in practice, as is discussed in While knowledge of risks often has been lacking more detail below. in developing countries, it is increasing in several key areas, such as dealing with disease, economic Knowledge cycles, and natural hazards. And new technologies Obtaining knowledge and thus reducing the un- are greatly helping to improve knowledge of poten- certainties that people face when they confront risk tial shocks and inform responses to them. Farmers in and pursue opportunities is the first component of Ghana and 15 other African countries, for example, risk management. Knowledge entails more than just receive specific market information through their amassing information: while obtaining information mobile phones, which helps them improve their about possible events and their likelihoods is neces- response to changes in agricultural prices and de- sary, knowledge also involves using that informa- mand.24 Globalization and scientific advances have tion to assess exposure to those events and possible also improved understanding of many pathogens, outcomes and then deciding how to act. Knowledge including how they can be detected and diagnosed therefore contains elements of assessment and judg- rapidly to enable disease control. Improved tech- ment. Furthermore, people’s knowledge of risk de- nologies have also supported greater collaboration pends not only on the information they can access among scientists and policy makers, as well as en- but also on the quality of information that is pro- abling the media to inform people, even in remote vided by other social and economic systems. Indeed, parts of the world. public policy has an important role to play in im- proving the availability, transparency, and reliability Protection of information that may be relevant for risk prepara- Protection includes any actions that lower the prob- tion, including national account and labor statistics, ability and size of negative outcomes or increase the various market signals, and weather forecasts, among probability and size of positive outcomes. Develop- others. Moreover, the state can contribute by reduc- ing countries have made substantial improvements ing the uncertainty that can be created by erratic in some aspects of their risk protection in recent de- policies, protracted implementation of reforms, and cades. The percentage of people in low- and middle- frequent regulatory changes. income countries with access to improved sanitation, 14 WORLD DEVELOPMENT REPORT 2014 for instance, increased from 36 percent in 1990 to 56 formal insurance. For example, new devices for cars percent in 2010; meanwhile, the immunization rate can allow insurers to vary the insurance premiums for measles doubled from 41 percent to 83 percent they charge based on the quality of people’s driving.28 between 1985 and 2010.25 Improved sanitation and Together, knowledge, insurance, and protection increased vaccinations, alongside other preventive constitute preparation. The assets of households, health measures, have helped reduce infant and ma- communities and governments, as well as services ternal mortality rates. Similarly, following repeated provided by markets and the public sector, all influ- cycles of high inflation during the 1970s and 1980s, ence preparation for risk, which in turn affects out- many developing countries established sound fiscal comes. Overall, the extent of people’s preparation and monetary policy frameworks, which have helped for risk tends to be correlated with national income reduce the intensity and incidence of large recessions across countries. However, interesting variations (see box 2). Increased use of early warning systems within regions highlight the important role of policy has helped to protect populations exposed to natural in determining preparation for risk, over and above hazards, reducing fatalities when major events occur. access to resources (box 3). Insurance Coping To the extent that protection cannot completely The final component of risk management is coping, eliminate the possibility of negative outcomes, insur- which encompasses all actions that are taken once a ance can help cushion the blow from adverse shocks. risk (or, alternatively, an opportunity) has materi- Insurance includes any instruments that transfer alized. Coping, therefore, consists of deploying the resources across people or over time, from good to knowledge, protection, and insurance resources that bad states of nature. In certain cases, insurance for have been obtained during the preparation phase. particular risks is provided by specialized markets The relationship between coping and preparation in the financial system. However, because formal becomes very fluid when confronting an evolving insurance markets are often not widely available in risk. This includes updating relevant knowledge by developing countries, a larger burden is placed on monitoring and assessing emerging risks and then self-insurance, which is often pursued through rela- adapting and implementing any necessary and avail- tively costly and inefficient means, such as holding able responses. durable assets (like jewelry) that can be sold in the The choice of how much to prepare for risk has event of a shock. Large numbers of households also implications for the kind of coping that is needed, participate in informal, community-based risk shar- which, in turn, can contribute to vicious or virtuous ing, and microfinance and microinsurance programs circles in risk management. When effective prepara- are increasingly providing new instruments that help tion limits the damages from adverse shocks, coping people manage risk. Similarly, alongside traditional can be minimal—leaving more resources available safety nets, conditional cash transfers and other so- for further investments in risk management, reduc- cial insurance programs are a means for the state to ing vulnerability to future shocks, and so on. At the transfer resources to help the most vulnerable cope household level, for instance, having health insurance with adverse circumstances.26 can facilitate medical treatment and recovery, while There may be either synergies or trade-offs be- reducing out-of-pocket expenses, when a family tween insurance and protection as strategies to man- member falls ill or suffers an accident. At the mac- age risk. To the extent that having insurance reduces roeconomic level, evidence suggests that by reducing people’s incentives to prevent bad states from occur- losses from natural hazards, for example, preparation ring, insurance and protection act as substitutes for for risk may sustain and even accelerate economic each other. However, when the steps that people take growth.29 to attain protection facilitate or make it cheaper to In contrast, when preparation is limited or a insure against adverse outcomes, protection and in- shock is unexpectedly large, coping can be haphazard surance can complement each other.27 Being a non- and require costly measures—leaving few resources smoker, for instance, can make it easier and cheaper available for future risk management, worsening to obtain health insurance. Protection often must vulnerability to shocks, and weakening households’ be observable for insurance and protection to be ability to undertake new opportunities. For example, complements. While observability is already highly the loss of assets that occurs from natural disasters relevant for informal risk sharing in communities, in countries as different as Ethiopia and Hondu- technology may also make it increasingly relevant for ras—caused by direct damage from a hurricane or Risk and opportunity 15 B OX 3 How does preparation for risk vary across countries? Index of risk preparation across countries Least prepared quintile Most prepared quintile Missing data People’s preparation for risk at the country level includes actions by sanitation facilities, and an indicator of fiscal space based on gross and contributions from all social and economic groups and institu- public debt as a percentage of revenues (state support).a tions, including the state. An index of preparation for risk is charted This index shows that the extent of people’s preparation for risk on the map above. The index, developed for the World Development tends to be correlated with national income across countries, but Report 2014, comprises measures of assets and services across four only to a certain extent. People tend to be the most prepared in important categories—human capital, physical and financial assets, high-income countries (particularly in North America and western social support, and state support—that influence preparation for Europe), and least prepared in low-income countries (especially in risk. The component indicators for the index include: average years Africa), on average. However, substantial variation exists within of total schooling for the population aged 15 and over, and the regions. For example, Chile is reasonably well prepared for risk, immunization rate for measles (human capital); the proportion of while its neighbor to the east, Argentina, has only average risk prep- households with less than $1,000 in net assets, and an index of access aration despite having a similar level of income per capita. Likewise, to finance (physical and financial assets); the percent of the work- Ethiopia has better risk preparation than other countries in the force who contribute to a pension scheme, and the proportion of region with similar or relatively higher income per capita. This respondents stating that “in general, people can be trusted” (social underscores the importance of policies, over and above income support); and the percent of the population with access to improved level and access to resources, in determining preparation for risk. Source: Foa 2013 for the WDR 2014. Map number: IBRD 40097. a. Each indicator is rescaled to range between zero and one. The index, which is the average of the eight indicators, thus maintains the cardinal properties of the indicators, rather than simply being an average of rankings across the components. This approach follows in part the methodology used in the construction of the Worldwide Governance Indicators (see Kaufmann, Kraay, and Mastruzzi 2010). If necessary, each indicator is transformed so that an increase in its measure represents an improvement. drought, lack of insurance, and distressed sale of world in the midst of the 2008–09 crisis—including assets—has substantial short-term as well as long- bailouts of large financial firms, fiscal stimulus, and term effects: poor households can effectively become extended periods of monetary easing—helped calm trapped in poverty, making them more vulnerable markets in the short-run, these responses may have to future negative shocks and less able to undertake negative longer-term effects, including substantially new ventures for improvement.30 Similarly, while increased public debt and perverse incentives for fi- the coping responses by governments around the nancial institutions’ risk taking. 16 WORLD DEVELOPMENT REPORT 2014 Beyond the ideal: The obstacles to risk this information. Cognitive shortcomings are rel- management evant and pervasive obstacles to risk management in many circumstances, even in advanced countries. In If risk management can save lives, avert economic the United States, for example, a survey revealed that damages, and unleash opportunity—and, further- only 31 percent of homeowners in flood-prone areas more, if risk management is cost-effective and its were aware of the risk.31 The repercussions of extreme fundamentals are well understood—then, why aren’t instances of lack of information and knowledge— people and societies better at managing risk? Al- so-called “deep” uncertainty—are explored below. though the specific answer varies from case to case, it is always related to the obstacles and constraints Behavioral failures. Even if information exists, deci- facing individuals and societies, including lack of sion makers may be unable to turn knowledge into resources and information, cognitive and behav- actions and behaviors that prepare them for risk. In ioral failures, missing markets and public goods, and many cases, decision and policy makers seem to have social and economic externalities. This realization short memories regarding the origins of crises of leads to an important message. Identifying risks is various sorts. Systemic financial crises, for instance, not enough: the obstacles to risk management must are almost always preceded by unusually high credit also be identified, prioritized, and addressed through concentration and growth, and this process seems private and public action (box 4). to be well understood.32 Yet policy makers often do Consider the case of Mumbai. Its drainage system little to control credit booms. A false sense of security is more than 100 years old and barely capable of han- may underlie people’s inability to manage prepara- dling the annual monsoon rains. Reports and pro- tion for risk in normal times (by saving for a rainy posals have repeatedly spelled out how investments, day or completing disaster preparedness plans, for such as installing pumping stations and clearing instance). And a “paradox of protection” can arise: out debris, are needed to expand the capacity of the risk protection that suppresses losses for a long pe- storm drainage system. Yet with few exceptions, the riod creates a false sense of security, leading to de- proposals have not been acted upon. An exceptionally creased vigilance and risk awareness and potentially large monsoon hit the city in 2005, leading to more resulting in larger future losses.33 In many cases what than 400 deaths, extensive damage to buildings and might be perceived as irrational behavior may in fact infrastructure, and interruption of economic and fi- be the result of distorted incentives, incorrect or in- nancial activity. Afterward, a fact-finding committee sufficient knowledge, or particular social norms and made recommendations for overhauling the drainage cultural beliefs. system that were distressingly similar to those made in the 1990s. As of 2013, however, implementation is again lagging. As a result, India’s financial capital re- Obstacles beyond the control of individuals mains highly vulnerable to monsoon rains. hamper their risk management Missing markets and public goods. Markets in areas critical for effective risk management—credit, insur- Why aren’t people better at managing their ance, jobs—are weak or even missing in many de- own risk? veloping countries. So are public goods and services Lack of resources. Even when a risk management essential for risk management—economic and polit- strategy is cost-effective, individuals and groups may ical stability, law and order, and basic infrastructure. find it difficult to undertake because of large up- In fact, well-developed markets may be missing be- front costs and limited access to credit. Shortages of cause supportive public goods are flawed. If, for in- assets and finance, which are especially acute in poor stance, the justice system does not enforce contracts, and developing countries, can make the trade-offs it makes little sense to buy health, vehicular, or house inherent in risk management harder to handle. Gov- insurance, and no such market will exist.34 There are ernments may decide that, given their limited bud- many reasons why public goods are missing, but this get, current consumption spending is more pressing discussion considers only the most pertinent ones for than investments for disaster risk reduction. risk management. The first, already discussed, is lack of resources: the costly flood protections constructed Lack of information and cognitive failures. Relevant in the Netherlands, for example, are simply not fea- information may not exist or be available to decision sible for many similarly threatened developing coun- makers, or they may lack the ability to understand tries, like Bangladesh or Vietnam. The second reason Risk and opportunity 17 B OX 4 Bringing the essentials of and obstacles to risk management together in policy design Designing effective public policy must go beyond simply identify- ing—helping to identify critical gaps and revealing effective, low- ing potential risks to analyzing obstacles to risk management. cost interventions. Diagram a below presents a set of screens to assist in decision mak- a. A set of screens to aid risk management Risk Information Behavior Resource Policy Incentive assessment assessment assessment assessment assessment design How Are bad incentives leading to too Are Are cognitive Are What policies much risk much or too little risk taking? decision- and behavior resources should be are we makers ill biases impairing and access implemented? Because of Because of facing? informed? risk to resources market failures? government management? too limited? failures? This practical approach provides two important insights for the should aim for robust policies that may not be optimal in the most design of risk management policies: likely future, but that lead to acceptable outcomes in a large range of scenarios and that are easy to revise as new information becomes Be realistic. Simple risk management instruments should be pre- available. Starting with a strong foundation for risk management ferred when capacity is low. Policy makers should concentrate on requires a long-term perspective, creates the right incentives, and low-hanging fruit and win-win solutions. Soft measures that change minimizes the risk of unintended negative effects. It also helps incentives (such as improving zoning regulations for coastal areas) ensure that policies are flexible enough to be adjusted when new are preferable as a starting point to engineered measures (such information becomes available. (For more on both these insights, as dikes to prevent flooding). Furthermore, it is particularly cost- see the discussion entitled “Five principles of public action for bet- effective to strengthen the capabilities that are useful in managing ter risk management” at the end of this overview.) risks of different natures, such as the ability to complete large-scale evacuations (which can be useful for either a hurricane or a nuclear Thinking about both the fundamental components of and accident, for example). Realistic policy options should ensure that obstacles to risk management with these lessons in mind can help risk management avoids unintended negative policy consequences; identify which specific policies are most relevant in different con- provides the right incentives to build on everybody’s best capaci- texts. For example, countries with limited resources or weak institu- ties; and protects the most vulnerable, who are often least able to tional capacity should focus on policies that are foundational, while implement ideal but expensive solutions. countries that already have solid foundations for risk management in place can aim for more advanced policies. This framework is used Build a strong foundation for improved risk management over time. It throughout the World Development Report 2014 to organize and pri- often makes sense to create institutional arrangements when the oritize risk management policies across the four main components need for them is obvious, such as after a disaster event, and that of risk management (knowledge, protection, insurance, and coping) cannot be easily reversed once the memory of the event has disap- for different social and economic systems, from the household to peared. This institutional irreversibility should be combined with the international community. These are summarized in correspond- flexible implementation and continuous learning. Policy makers ing tables for each of these systems (diagram b). b. A framework for public policy priorities POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Protection Insurance Coping Source: WDR 2014 team. 18 WORLD DEVELOPMENT REPORT 2014 is related to the political economy of risk manage- rest of the world. Both negative and positive exter- ment. Governments may be reluctant to spend on nalities may complicate the process of risk manage- risk preparation because its costs are immediate and ment, making it less predictable and distorting its in- observable while its benefits, even if substantial, are centives. The solution is coordination and collective longer term and less visible. action, which can be difficult to obtain when there are wide differences in preferences, values, and expo- Government failures. Risk management can also be sures. For instance, externalities and collective action impaired by government failures stemming from failures may be why reaching a binding international capture by interest groups, corruption of govern- agreement on greenhouse gas emissions is proving ment officials, and distortionary policies. On policy so elusive. capture, enterprises and people who are negatively affected by certain risk management measures will Deep uncertainty and robust solutions naturally tend to oppose them and be vocal about it, while the people protected by these measures are “Deep uncertainty” is an obstacle to risk manage- often not aware of them (and therefore do not sup- ment that deserves special attention. Also known as port them), or lack the commensurate influence of Knightian uncertainty in economic circles,37 deep active lobbies. Powerful tobacco and asbestos lob- uncertainty refers to a situation for which even ex- bies, for instance, can block useful health regulations perts cannot agree on appropriate models to under- even in the presence of well-established scientific stand it, on the potential outcomes and probabilities evidence. On distortionary policies, sometimes even of its occurrence, and on how much importance well-intentioned measures can impair risk manage- should be given to it. Taking a broad perspective, the ment by distorting people’s incentives to manage difference between deep uncertainty and ordinary their own risk. An example is poorly designed post- uncertainty is a matter of degree, fluid, and evolv- disaster support that creates moral hazard and dis- ing. Building knowledge helps to reduce the degree courages risk management by individuals and firms. of uncertainty. The history of science is full of cases Similarly, overly generous safety nets or financial where deep uncertainty gradually became ordinary sector bailouts can undermine incentives for risk uncertainty, amenable to management and control. preparation. But while this happens, what should be done in the presence of “unknown unknowns”? Social and economic externalities. Risk management Under conditions of deep uncertainty, it is pref- actions undertaken by some people or countries erable to implement adaptive and robust policies may impose losses on others. For instance, overuse and actions that lead to acceptable outcomes in of antibiotics is creating ever more drug-resistant a large range of scenarios and that can be revised bacteria. Similarly, excessive exploitation of common when new information is available and when the natural resources such as oceans, forests, and the at- context changes.38 For monetary and financial mosphere—a phenomenon known in the literature policy, a promising practice is the use of stress test- as “the tragedy of the commons”—is leading to en- ing of banks and other financial institutions using vironmental degradation, climate change, and a fu- a broad range of situations, including forward- ture drop in economic growth.35 In a different realm, looking crisis scenarios.39 Above all, plans that are an expansion in the money supply to stimulate the designed for the most likely outcomes but that in- domestic economy in large advanced economies is crease the vulnerability to less likely events should creating destabilizing capital inflows to developing be avoided. For instance, dike systems built only for countries, as well as eroding the wealth of domestic standard rainstorms and tides can actually increase savers and taxpayers. Similarly, instituting trade bar- vulnerability by creating a false sense of security riers to protect domestic producers during economic and dramatically increasing the damages when a downturns imposes increased cost on trade partners flood does occur. and can lead to trade retaliation, possibly turning a downturn into a protracted world recession.36 Other The way forward: A holistic approach to risk management actions can generate benefits for managing risk people other than those bearing their cost, therefore creating incentives to “free ride.” That is the case, for Can individuals on their own overcome the obstacles instance, for countries that take costly measures to to risk management they face? Although individuals’ reduce greenhouse emissions, which can benefit the own efforts, initiative, and responsibility are essen- Risk and opportunity 19 tial for managing risk, their success will be limited without a supportive external environment. While Key social and economic systems can D I AG R A M 3 individuals on their own may be capable of dealing contribute to risk management in complementary ways with many risks, they are inherently ill-equipped to confront large shocks (such as the head of a house- hold falling ill), systemic shocks (such as a natural People’s risk management hazard or an international financial crisis), or mul- tiple shocks that occur either simultaneously or sequentially (for example, a drought followed by a food price shock and food insecurity). The state Civil society and the People can successfully confront risks that are be- private sector Social protection yond their means by sharing their risk management • Health, old age, and Households unemployment with others. They can pool their risk collectively • Family ties insurance through various overlapping social and economic • Assistance and relief Communities groupings (systems). Indeed, the need to manage risk Public goods • Collective action • Infrastructure and pursue opportunity collectively may often be a • Law and order Enterprise sector key reason why these groups or systems form in the • National defense • Jobs and income first place.40 These systems extend in size and com- Public policy Financial system plexity—from the household to the international • Macroeconomic • Insurance and credit management community. They have the potential to support • Regulatory framework people’s risk management in different yet comple- mentary ways (diagram 3). Their different scope may allow them to handle shocks and exposures that match their scale (box 5). International community • Resources, expertise, global rules, and coordination • The household is the primary instance of support, pooling resources, protecting its members—espe- Source: WDR 2014 team. cially the vulnerable—and allowing them to invest in their future. • Communities provide informal networks of insur- ance and protection, helping people deal with id- • The international community can offer expertise, iosyncratic risks and pooling resources to confront facilitate international policy coordination, and common risks. pool resources when risks exceed national capac- ity or cross national and generational boundaries. • Enterprises can help absorb shocks and exploit the opportunity side of risk, contributing to more These systems have mutual interactions, often stable employment, growing income, and greater complementing and sometimes substituting for each innovation and productivity. other’s risk management functions. For instance, various mechanisms of protection and insurance • The financial system can facilitate useful risk man- provided by communities, enterprises, the financial agement tools such as savings, insurance, and system, and the state can complement and improve credit, while managing its own risks responsibly. households’ self-protection and self-insurance. En- terprises rely on macroeconomic stability, public • The state has the scale and tools to manage sys- services, and financial products to remain dynamic temic risks at the national and regional levels, to and continue to provide income and employment provide an enabling environment for the other to people. The financial system can provide tools of systems to function, and to provide direct support insurance, saving, and credit only if enough house- to vulnerable people. These roles can be achieved holds and enterprises are able to participate in the through the provision of social protection (social system, and if the economy features a certain degree insurance and assistance), public goods (national of stability and predictability. Markets, in general, defense, infrastructure, law and order), and public can provide risk management tools and resources at policy (sound regulation, economic management). a growing scale if the necessary public services, such 20 WORLD DEVELOPMENT REPORT 2014 B OX 5 Which systems for which risks? Individuals face a multitude of risks, and various social and economic job loss of the head of the household or a burned-down house). The systems can help them manage risks that are beyond their means state must sometimes provide substitutes for these functions when alone. But which systems are most appropriate for which risks? Two markets are missing or not available to some. important principles provide a way to prioritize risk management Because systemic risks affect large groups of people, they can across systems: hardly be managed by individuals alone. Communities have an advantage in managing small systemic risks (such as local violence 1. The principle of subsidiarity suggests that risks should be handled or flooding) because of their proximity to the groups of people at the lowest level capable of handling them, to take advantage affected and their potential advantage in monitoring and resolving of the proximity to and greater knowledge of the agents most local tensions. The state also has an advantage in managing small directly affected by a risk, as well as the ability to monitor both systemic risks (such as moderate fluctuations in aggregate prices those agents and the risks that they face. or regional food shortages) because of its capacity to control the 2. The principle of comparative advantage suggests that risks should national macroeconomy and transfer resources between different be managed by the system that can handle them most effectively. parts of a country. Individuals and households are well placed to handle idiosyn- Because many agents within a country are severely affected cratic risks (such as minor injuries or income shortfalls) as long as the when large systemic shocks occur, such as economy-wide banking potential losses remain relatively small. They have an advantage in crises or natural disasters, the cross-support they can provide for managing these types of risk because of their proximity to the level one another is limited. In other words, it is difficult for the private at which the main impact occurs and because of their ability to sector alone to pool and insure for systemic risk. The state thus has a monitor conditions and efforts within the household. unique role in managing large systemic risks because it has the scale As the size of potential losses increases, the tools that individuals and tools to prepare at the national and regional levels. Support and have at their disposal can quickly be exhausted. The enterprise and coordination from the international community is needed when financial systems can thus provide effective tools and mechanisms large systemic risks cross national borders or overwhelm national (discussed in more detail in the sections below) for individuals to capacities. Spotlights in the WDR 2014 feature case studies of risk manage potential losses from large idiosyncratic shocks (such as the management by different support systems. Types of risk that can be managed by different systems and examples featured in the WDR 2014 spotlights Small Large idiosyncratic idiosyncratic risk risk Small systemic risk Large systemic risk System best placed to Individuals and The enterprise sector The community and The state and the manage risk households and financial system the state international community Health risks (Turkey and the Kyrgyz Food shortages Natural hazards (the Republic) (Ethiopia and El Philippines and Colombia) Spotlight examples Loss of employment and income (India) Salvador) Financial crises (the Czech Urban violence (Brazil Republic, Peru, and Kenya ) and South Africa) Pandemics (global) Source: WDR 2014 team. as the rule of law and a sound regulatory framework, more prevalent and the relative roles of the house- are in place and effective. The international commu- hold and the community are larger. For these coun- nity relies in part on responsible governments that tries, the international community may also play a are willing to cooperate to address global risks; in larger role through financial assistance and capac- turn, the international community can help govern- ity building. As countries advance—and informal ments and countries that lack resources and capacity mechanisms give way to formal ones—the relative for risk management. importance of the contributions from the enterprise The relative importance of these systems changes sector and the financial system grow. The potential with the level of development. In less advanced role of the state is larger in less developed countries, countries, and especially in fragile and conflict-af- but in these cases the state tends to suffer from more fected countries, informal mechanisms tend to be severe capacity and resource constraints. These limi- Risk and opportunity 21 tations call for a mutual, symbiotic relationship be- tional support to confront risk and pursue opportu- tween the state, civil society, the private sector, and nity. Extending Gary Becker’s metaphor in A Treatise the international community, as countries develop on the Family, households are “little factories” where (see below). goods and services of knowledge, protection, and insurance are produced, using both “intermediate inputs” obtained from the rest of society and the The state, civil society, and the private sector: pooled efforts and skills provided by family mem- Helping one another manage risk bers.43 How can the household contribute? None of the social and economic systems presented above works perfectly. Indeed, in certain cases they Protection and risk pooling for its members. Protection hinder rather than help people’s risk management. and insurance at the household level are particularly They have the potential, however, to become effective important for idiosyncratic risks and even more rele- support systems when their weaknesses are resolved. vant when market or social insurance is lacking. Pro- The state thus has an important potential role to play tection against adverse shocks is especially important by complementing and supporting the functions for the vulnerable within the household: the young, that households, communities, enterprises, and the the old, and the ill. For this purpose, families can financial system may serve. From this perspective, benefit from the resources that are available in soci- the state’s role goes beyond the narrow purpose of ety—all the more so if these resources are increasing correcting market failures and extends to address- and improving. Thus, for instance, higher incomes ing systemic risks, building institutions that enhance and better access to health services have increased each component of risk management, and providing immunization rates for measles to more than 70 direct support to vulnerable populations. percent in every region of the world, although Sub- It would be naïve, however, to ignore the fact that Saharan Africa still has much room for improvement the state often falls short in fulfilling its potential role. (figure 3a). Historically and throughout the world, examples of Moreover, sharing bad times (and good times) government failures are regrettably abundant.41 This occurs naturally in the household. Indeed, pooling is all too vividly evident in the case of fragile and risk within and across family generations has been conflict-affected countries. What to do then? Civil a basic form of insurance from time immemorial. society, the private sector, and the international com- The extended family plays an active role, especially munity can provide badly needed public goods and in developing countries. For instance, evidence from services—albeit imperfectly. Especially, but not only, Bangladesh, Ethiopia, India, Mali, and Mexico shows in democratic societies, they can also help improve that extended family members step in to help out in governance and the delivery of public services by a substantial way when their relatives fall ill.44 Simi- generating mechanisms to make the state responsive larly, evidence from several countries around the to the needs of the population and accountable for world indicates that family members who migrate its actions.42 assist their families through remittances when nega- The discussion that follows assesses the poten- tive shocks occur in their place of origin.45 tial contribution of each major system and suggests ways to improve their performance, individually and Allowing its members, especially the young ones, to in combination with other systems. The state’s po- make investments for the future. The role of house- tential contribution is presented in connection with holds extends well beyond protecting and insuring each system, reflecting its overarching role and al- members against negative events. Households invest lowing for an elaboration of specific recommenda- in the human capital and social skills of their mem- tions for public policy, as well as a discussion of their bers, especially the young, preparing future genera- rationale and trade-offs. tions to manage the risks and opportunities they will face. Schooling is one important example where progress has occurred in recent decades. The aver- The household age number of years of educational attainment has increased since 1960 in all regions—most substan- How can it foster resilience and prosperity? tially in regions that initially had the lowest attain- For most people, the household—defined as a group ment (figure 3b). However, the quality of education, of individuals related to one another by family ties— as measured by international exams in science, math, constitutes the main source of material and emo- and reading skills, is still lagging behind in many 22 WORLD DEVELOPMENT REPORT 2014 FIGURE 3 Education and health outcomes in developing countries are improving, but unevenly a. Measles immunization rate b. Educational attainment of 15–24 year-olds 100 10 % of children aged 12–23 months 9 90 Years of education 8 80 7 6 70 5 4 60 3 50 2 1990 1995 2000 2005 2010 1960 1970 1980 1990 2000 2010 OECD East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: WDR 2014 team based on data from World Bank World Development Indicators (database) (panel a) and Barro and Lee 2010 (panel b). Note: Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. low- and middle-income countries, without signs of expenditures, such as social insurance and private converging yet.46 health insurance, significantly reduces the incidence of catastrophic medical expenditures, especially for poor households.47 Given the fundamental impor- What characteristics improve the household’s tance of health for everything else people do, there contribution to risk management? is indeed great need for health insurance and much Households are small but complex units. The mo- room for improvement: only 17 percent of adults in tivations of their members can range from altruism developing countries report having contributed to to self-interest, the intrahousehold relationships can health insurance, and this share is as low as 2 percent be based on common goals or relative bargaining in some low-income countries.48 power, and the household’s connections to society can be fluid or remote. These characteristics can have Fairness within the household. One would like to great influence on how well the household func- think of households as nurturing, cohesive units. tions as a first line of support to confront risk and All too often, however, abuse and discrimination opportunity. occur within the family, making it a source of, rather than a solution to, risk. Compelling evidence shows Access and participation. Communities, labor and fi- that women’s economic and social empowerment nancial markets, and public institutions provide the can strongly influence whether the allocation of re- “intermediate inputs” that families build upon to sources within the household benefits children and manage their risks. Continuous access to and par- promotes gender equality.49 An evaluation of a cash ticipation in those markets and institutions is critical transfer program in South Africa, for instance, found for families to be successful risk managers (so much that pensions received by women improved the so in the view of the World Development Report 2014 health and nutritional status of girls but that trans- that the following four sections are devoted to as- fers received by men had no effect on either boys or sessing how they can contribute). To give just one girls.50 One important ingredient for women’s eco- example: evidence from 59 countries suggests that nomic empowerment is access to the labor market, access to programs that limit out-of-pocket health which in several contexts is limited by inadequate Risk and opportunity 23 child care infrastructure and restrictive social norms. educational campaigns to counter social norms that Some countries and regions have much room for tolerate violence or discrimination against women improvement: female labor participation rates are and children. The campaigns should target both only 20–30 percent in the Middle East, North Africa, men and women: more than 20 percent of women in and South Asia, while in most of the rest of the world all regions, except Latin America and the Caribbean, they are well above 50 percent.51 believe a husband is justified in hitting or beating his wife for reasons like going out without telling him and arguing with him.56 How can the state contribute? The state has an important role to play in providing social services and countering harmful social norms. The community Policies that empower households as a unit and poli- How can the community foster resilience and cies that empower individuals within households are prosperity? necessary. Communities are groups of people who interact Providing essential social services. Access to good, frequently and share location or identity. Neighbor- even if basic, educational and medical services hood groups, religious groups, and kinship groups can prepare people to confront major health risks, are some examples. They work through informal handle life-cycle transitions, and take advantage networks based on trust, reciprocity, and social of work opportunities. In this sense, the drive for norms—what James Coleman and Robert Putnam “equality of opportunities” can also bring about re- call “social capital.”57 In this way, communities can silience for households and individuals.52 The efforts help their members by sharing idiosyncratic risks of Thailand and Turkey to offer universal access to and confronting common risks and opportunities. quality health insurance deserve special mention. Universal access to health care is likely to require a Sharing idiosyncratic risks. Informal insurance is partnership between the public and private sectors particularly important for low-income households to ensure both fiscal sustainability and sufficient and is sometimes their only real safety net. In the vil- human resources.53 For the most vulnerable, targeted lage of Nyakatoke in Tanzania, for instance, with a safety nets can have a dramatic impact in preventing population of only 120 families, there are about 40 the coping responses that incur long-term costs— different insurance schemes (burial societies, rotat- such as reducing basic consumption, withdrawing ing savings associations, and arrangements to share children from school, selling productive assets in dis- labor and livestock).58 These practices are also rele- tress sales, or resorting to crime. Ethiopia’s Produc- vant at the country level. Indonesian households, for tive Safety Net Program is one successful example of instance, have informal insurance against 38 percent protecting the most vulnerable from food insecurity of the economic costs of serious health shocks and 71 while building community assets to better manage percent of the costs of minor illness.59 In Nigeria, in- climatic risks and raise productivity.54 formal credit and assistance make up 32 percent of all coping responses identified by households (figure 4). Increasing women’s power in the household. This can be done first through economic empowerment: en- Confronting common risks and opportunities. When couraging women’s participation in the labor force communities channel their social capital for collective and, for poor households, directly increasing their action, they can provide some publics goods (such as purchasing power. An example of the latter is con- basic transport and irrigation infrastructure) to pro- ditional cash transfer programs that make payments tect against common adverse events (such as epidem- to women directly; impact evaluations have shown ics, natural hazards, and crime and violence) and to that these programs improve family and, especially, facilitate taking advantage of common opportunities children’s outcomes, including health and cogni- (such as new markets and technologies).60 This col- tive development.55 A second route is through social lective action can be especially important when state and legal empowerment: enforcing legal measures capacity is low. The informal settlement of Orangi in against abuse and domestic violence, eliminating Karachi, Pakistan, for example, financed and orga- regulations that discriminate against women in asset nized its own sanitation, vaccination, microfinance, ownership or economic activity, and conducting family planning, and violence prevention, assisted by a local nongovernmental organization. 24 WORLD DEVELOPMENT REPORT 2014 FIGURE 4 People respond to shocks on their own and by pooling risk with others Malawi Uganda Tajikistan Uzbekistan Afghanistan Iraq Maldives Sudan Nigeria 0 20 40 60 80 100 % of all coping responses when faced with a shock Informal credit and assistance Formal credit and assistance Consumption reduction Savings and sale of assets Employment or migration Source: WDR 2014 team based on data from household surveys, various years 2004–11. What characteristics improve the How can the state contribute? community’s contribution to risk Reliance on personal interactions and informal management? means of enforcement underlies the strength of Cohesiveness. Communities with strong ties between communities, but it is also the source of their weak- their members—that is, those communities endowed ness. Communities struggle with systemic risk and with high “bonding” social capital—are better able to falter when risk management requires complex and organize collective action on behalf of the group.61 In long-term preparation. Governments can help by fact, for local problems whose solution eludes mar- providing essential public goods and promoting in- kets and governments, a cohesive community can be clusion and respect for diversity. the missing piece of the puzzle. Cohesiveness is not easy to achieve, however, when community members Providing essential public goods, such as infrastructure have different values and cultural identities, as is in- and rule of law. Communities’ autonomous coping creasingly the case in urban communities. Moreover, and insurance mechanisms do not add up to ad- community cohesiveness is seriously compromised equate risk management; they also need national and when people are excluded or discriminated against. local governments to complement their efforts. For example, neighborhoods are potentially able to main- Connectedness. Communities also need connections tain their own drains, but urban flood prevention re- to other communities and to markets; without these quires citywide drainage and land use planning that connections they remain small and insular, lack po- only city governments can provide. Similarly, neigh- litical influence, and are unable to accomplish any- borhoods can patrol against petty criminals, but they thing at scale. Communities with strong ties to one are powerless against organized crime. another—that is, those communities that have high “bridging” social capital—are more likely to collab- Promoting inclusion and respect for diversity. Commu- orate with one another on mutually beneficial risk nities are not necessarily fair or reliable and can be management projects and to coexist peacefully. Cities marked by strong inequalities in power and wealth.62 with high religious or ethnically motivated violence, They may exclude vulnerable people (chronically for example, tend to lack routine interaction among ill, widowed), new entrants (migrants, refugees), or members of different groups and to be characterized those who happen to be different (ethnic minori- by divisive local leaders, media, and criminal gangs. ties). The state can help by enacting antidiscrimina- Risk and opportunity 25 tion laws, conducting educational campaigns, and costs.64 Whereas most individuals on their own are encouraging interactions that promote cohesiveness naturally risk averse and thus reluctant to take on in the face of diversity. new ventures, in groups they become more willing Not only can governments support communities, to pursue projects that involve more risk but also but community participation can increase the qual- promise higher returns. Firms, therefore, can serve ity of the governance process and improve the per- as natural vehicles to exploit the upside of risk, with formance of government programs. People may not beneficial consequences for individuals’ resilience heed the call to evacuate when government sounds and prosperity.65 the disaster alarm, but they will run when warned by a trusted fellow community member. Mobiliz- Risk sharing. Enterprises allow risk sharing among ing communities’ voice, energy, and collective action workers through collaboration; among owners of can help overcome some of the obstacles to improv- firms through investment diversification; and be- ing risk management in countries and regions with tween workers and owners through (formal or in- weak government capacity. For example, Afghani- formal) contractual arrangements. For risk sharing stan’s National Solidarity Program is constructing within a given enterprise, achieving a certain size is rural infrastructure with community participation an advantage. The enterprise sectors of many de- and also laying a foundation for improved local veloping countries, however, are dominated by self- governance. In India and Uganda, disseminating employment (figure 5). Rates of self-employment information on health and education entitlements are around 70 percent in South Asia and exceed 80 and outcomes through community-sponsored pub- percent in Sub-Saharan Africa and are also pervasive lic meetings has improved both government services in developing countries in other regions. These high and community participation, leading to more vac- rates of self-employment suggest that the incomes of cinations, more prenatal supplements, and fewer vast numbers of workers in developing countries are excess school fees.63 vulnerable to diverse shocks—a sick child, an equip- ment failure, or a change in the weather could mean the loss of a day’s income and more. They also sug- The enterprise sector gest that the enterprise sector is not benefiting from the specialization and increased productivity that How can the enterprise sector foster resilience multiperson enterprises make possible. and prosperity? The enterprise sector comprises workers and owners, Innovation and resource reallocation. When fueled the arrangements that organize their relationships, by competition, the enterprise sector can promote and the technologies that turn production factors innovation by adopting new technologies and real- into goods and services. Enterprises, the defining locating resources. In some instances, it may require unit of the enterprise sector, range from informal exit and entry of enterprises in the economy. This to formal, from self-employment to partnerships process of “creative destruction,” as first labeled by to giant multinational corporations, and from agri- Joseph Schumpeter,66 can generate substantial ad- culture to manufacturing and services. Whereas the justment costs but may be the only way an economy owner of a single enterprise might seek to maximize remains resilient and prosperous in the face of con- its profits, the enterprise sector as a whole encom- stantly changing conditions. Improving this dynamic passes the interests of workers, owners, and consum- process can have significant effects both on reducing ers. Despite the possible important trade-offs among the risk of prolonged recessions and on increasing these interests, the enterprise sector can help people aggregate productivity. For instance, one estimate manage risk through several channels, as described finds that making resource allocation as efficient in below. China and India as it is in the United States would For workers and owners, being part of a multi- increase total factor productivity by as much as 50 person enterprise—that is, a firm—offers the possi- percent in China and 60 percent in India.67 These bility of sharing the benefits and losses from special- large gains, however, would also require developing ization, collaboration, and innovation. Indeed, this institutions and a business environment that can is one of the main motives behind the formation of support a high degree of dynamism in the enterprise firms. As Frank Knight and Ronald Coase argued sector—not an easy task. in their seminal studies, firms have an institutional advantage in providing cost-efficient ways of deal- Worker, consumer, and environmental protection. Mo- ing with uncertainty and overcoming transaction tivated by reputational considerations and properly 26 WORLD DEVELOPMENT REPORT 2014 FIGURE 5Self-employment is more prevalent in developing countries, especially in Sub-Saharan Africa and South Asia Self-employment, as percent of total employment, average 2004–06 100 Maximum 90 75th percentile Median 80 70 25th percentile 60 Percent 50 40 30 20 Minimum 10 0 OECD East Asia Europe and Latin America Middle East South Sub-Saharan and Pacific Central Asia and the and North Asia Africa Caribbean Africa Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. regulated by the state, the enterprise sector can con- enterprise sector that is flexible is more capable of tribute to people’s risk management by providing responding to shocks by allocating resources within workplace safety, consumer protection, and environ- and across enterprises, promoting risk sharing, and mental safeguards. These protections are not guaran- innovating in an ever-changing world. In the recent teed, however; and in some cases enterprises do un- global financial crisis, for instance, Denmark and dermine them and generate losses for society. These Spain were hit hard, yet their labor outcomes were harmful practices can be corrected with stewardship markedly different. In Denmark, job separations from the state, communities, and enterprises alike. were high but unemployment spells were short. In Given the right incentives, firms that make these so- contrast, in Spain the unemployment rate, which cial protections a priority can have substantial ben- stood at 25 percent at the beginning of 2013, has efits. A recent meta-analysis, for instance, found that shown few signs of abating since the start of the crisis. workplace wellness programs reduce medical and The difference is arguably explained by the rigidity absenteeism costs—gains that accrue to both work- within the enterprise sector in Spain, in contrast with ers and firms.68 Denmark’s propitious business environment. This situation has prompted a serious debate and recent What characteristics improve the enterprise reform proposals in Spain to remedy the situation. sector’s contribution to risk management? More generally, the evidence indicates that countries with less flexibility in their enterprise sectors suffer Two characteristics enhance the ability of the enter- deeper and more prolonged recessions when nega- prise sector to contribute to people’s resilience and tive shocks occur.69 prosperity: flexibility and, over time, formality. Formality. For enterprises, formality is defined as Flexibility. Flexibility is the capacity of the entire compliance with laws and regulations. Whether enterprise sector (owners, workers, technologies) formality is beneficial (for enterprises and the econ- to adjust to changing conditions. It should not be omy) or not depends on the quality of the norms confused with the simple ease of firing workers. An dictated by the state and the quality of the public Risk and opportunity 27 services it offers. When these norms and services are moving a country from the quintile with the greatest sound, the enterprise sector is characterized by less labor rigidity to the one with the least rigidity im- self-employment and larger, more stable, and more proves the speed of adjustment to shocks by one-half formal firms. These characteristics are all related. In- and increases productivity growth by as much as 1.7 formal mechanisms may be effective for small firms percentage points.73 Furthermore, strong and inclu- and simple transactions, but they are insufficient sive social insurance is necessary so that flexibility in for larger firms and complex relations with workers the enterprise sector does not come at the expense of and markets. With adequate public regulations and the well-being of workers, their households, or their services, formal firms can benefit from better legal communities (box 6). protection (such as contract enforcement) and better use of public infrastructure (such as ports for inter- Stronger and enforceable regulations for worker, con- national trade). That, in turn, can promote risk shar- sumer, and environmental safety. While in many ing and innovation among enterprises. Moreover, areas regulations can be excessive and disruptive of it can make enterprises more easily accountable for market forces, stronger and enforceable regulations their impact on worker safety and on consumer and are needed to ensure workplace safety, consumer environmental well-being.70 protection, and environmental preservation. Market There are both synergies and trade-offs between failures derived from externalities and asymmetric flexibility and formality. In countries with effective information are pervasive in these areas, requiring state institutions, formality enhances flexibility. In direct intervention by the state. The deadly gar- countries with weak state institutions and cumber- ment factory collapse in Bangladesh in 2013—which some regulatory regimes, however, the cost of for- claimed the lives of more than 1,100 workers—is a mality can be too large for the majority of enterprises sad reminder of the importance of the state’s moni- and workers. In this case, informality is a means for toring and enforcement of regulations that cannot be the economy to achieve a certain degree of flexibility overseen by people on their own. These regulations and for workers to access a practical safety net.71 Fig- are important, particularly in states whose low in- ure 6 provides a typology of countries based on the stitutional capacity requires them to prioritize their flexibility and formality of their product and labor interventions carefully. markets. The financial system How can the state contribute? How can the financial system foster resilience Public policy for the enterprise sector requires re- and prosperity? forms that balance the economy’s need for flex- ibility with society’s need for legal and regulatory Through the provision of useful financial tools and protections. responsible management of its own risks, the fi- nancial system can shield people from the impact A better business environment. Several of the ways in of negative shocks and better position them to which the state can contribute to productivity and pursue opportunities. Saving instruments (such as innovation can also enhance the resilience derived bank deposits and liquid securities) enable people from the enterprise sector. A better investment cli- to accumulate buffers for rainy days. Credit instru- mate can improve risk management in the enter- ments (such as education or mortgage loans) allevi- prise sector by encouraging adherence to sensible ate financing constraints, helping people to smooth rules and regulations and by increasing the sector’s consumption following negative shocks but also to capacity to adjust to new conditions. Most basi- exploit opportunities with greater flexibility. Finally, cally, secure property rights and regulatory certainty, market insurance (such as health and residential in- along with low costs for firm entry and exit, are es- surance) provides a means to cover the costs of dam- sential. In addition, although labor market reforms aging adverse events. in isolation are unlikely to be successful, reducing the burden of labor taxes and streamlining regulations What characteristics improve the financial is a critical component of a comprehensive set of system’s contribution to risk management? reforms—where the overall effect is larger than the sum of their parts.72 Alongside such complementary Inclusion and depth. As Merton Miller and numer- reforms, recent cross-country evidence finds that ous followers have argued persuasively, when finan- 28 WORLD DEVELOPMENT REPORT 2014 FIGURE 6 Countries vary widely in the flexibility and formality of their product and labor markets More Australia Germany Netherlands flexible Austria Hong Kong Norway Belgium SAR, China Poland Canada Ireland Singapore Armenia Albania Chile Israel Sweden Azerbaijan China Costa Rica Japan Switzerland Cambodia Czech Republic Korea, Rep. United Kingdom Kazakhstan Denmark Latvia United States Peru Estonia Lithuania Finland Malaysia Product and labor market flexibility France Mauritius Benin The Gambia Morocco Botswana Georgia Sri Lanka Brazil Kyrgyz Namibia Republic Hungary Portugal Slovenia Burkina Faso Guatemala Tanzania Bulgaria South Africa Macedonia, Italy Slovak Spain Cameroon Guinea Thailand Indonesia Ukraine Republic FYR Jordan Turkey Côte d'Ivoire Lebanon Uganda Kenya Uruguay Mongolia El Salvador Madagascar Zambia Vietnam Bangladesh Ghana Pakistan Bolivia Haiti Philippines Algeria India Russian Argentina Dominican Lesotho Federation Croatia Bosnia and Honduras Senegal Herzegovina Republic Libya Trinidad and Greece Jamaica Sierra Leone Tobago Burundi Ecuador Mexico Iran, Islamic Rep. Mali Venezuela, RB Yemen, Rep. Chad Egypt, Arab Rep. Romania Nepal Zimbabwe Colombia Nicaragua Less flexible Less formal More formal Production and labor formality Source: WDR 2014 team based on data from World Bank Pensions (database); World Bank World Development Indicators (database); World Economic Forum 2012; and Schneider, Buehn, and Montenegro 2010. Note: Economies in the top row are high (above the median value) in both product market flexibility and labor market flexibility; in the middle row they are high in one or the other of the two; and in the bottom row they are low (below the median value) in both flexibility indicators. Similarly, economies in the first column on the left are low in both formal production and formal labor; in the middle column they are high in one of the two formality indicators; and in the last column on the right they are high in both formality indicators. Only economies with data for all four indicators are considered, and median values are calculated within this sample. cial markets are competitive and function without using financial savings tools in high-income coun- distortions, they can efficiently provide more and tries), and credit is used by about 8 percent (com- better tools and services to more people.74 Indeed, pared with 14 percent in high-income countries)— financial markets can provide instruments and although great heterogeneity exists across countries services that help people face risks of varying fre- (figure 7). quency, intensity, and nature, either idiosyncratic or systemic. However, about 70 percent of people in Stability. The Achilles’ heel of the financial system low- and middle-income countries do not use es- is its propensity for crisis. As observed in the semi- sential financial tools at all, compared with about nal work of Douglas Diamond and Phillip Dybvig, 40 percent in high-income countries. Data on in- the mismatch between the duration of banks’ assets dividuals’ financial portfolios show that financial (long-term) and liabilities (short-term) makes the savings and insurance are each used by only about financial system inherently unstable.75 If the finan- 17 percent of people in low- and middle-income cial system fails to manage the risk it retains, it can countries (compared with 45 percent of people hurt people—directly by hindering their access to Risk and opportunity 29 B OX 6 Should access to social insurance be tied to work status? The provision of basic insurance against the risks associated with ill- • Allowing people to participate in health and old-age insurance ness and old age—especially for the vulnerable—is arguably a fun- regardless of work status (employed or unemployed, and formal damental goal for public policy. But how is social insurance funded or informal), requiring reasonably short vesting periods and por- and whom does it benefit? Traditionally, it has been funded through table benefits. mandatory payroll taxes levied on employers and employees, and it • Making additional contributions to health and pension schemes has benefited contributing workers. The problem with this approach voluntary and clearly linked to predictable benefits that are is its limited coverage: in most developing countries, formal workers beyond the basic provisions granted by the state. Involving the (who contribute and benefit from social insurance) make up less than private sector in the management and provision of the voluntary half the labor force (and much less in Sub-Saharan Africa and South portion of social insurance contributions and benefits. Asia). The traditional approach thus ends up excluding many work- ers—mostly those who are low-income, self-employed, or work in • Providing basic health care and old-age pensions funded by the agriculture.a state and directed to vulnerable populations but potentially open To close the coverage gap, several countries have set up noncon- to everyone (at least for health care).c tributory systems for health and old-age pension insurance. Is it a • Funding this basic provision through general government reve- good idea to combine noncontributory and mandated contributory nues and user fees (for health care), to a level consistent with fiscal systems? If the benefits from contributing to social insurance are sustainability. uncertain and the enforcement of mandated payments is weak, hav- • Clearly communicating with the public the characteristics and ing these parallel systems may undermine the incentives for employ- limitations of basic provisions and the additional costs and bene- ers to hire formally and for employees to seek formal employment. A fits of voluntary contributions. vicious circle could then ensue: informality breeds low coverage, and • Promoting financial literacy and fostering trust in the financial sys- the response to low coverage breeds further informality.b tem regarding its insurance function by macroprudential actions One possibility that merits discussion is delinking social insur- and policy certainty. ance from work status. This uncoupling would involve the following public action: Too ambitious or far-reaching? Maybe so—but worth discussing. Source: WDR 2014 team. a. Ribe, Robalino, and Walker 2012. b. Evidence from Chile, Colombia, and Mexico shows that the interplay of contributory and noncontributory systems has led to declines in formal employment, and there is widespread evidence that smaller, informal firms tend to be less productive and pay lower wages. See Levy and Schady 2013; Pagés-Serra 2010; ILO 2009; La Porta and Shleifer 2008. c. Developing countries such as Mauritius and South Africa already rely primarily on noncontributory systems for pensions, while several other countries— including China, India, Thailand, Turkey, and Vietnam—have also begun to offer universal access to health insurance. See Holzmann, Robalino, and Takayama 2009. finance, or indirectly by hampering available credit nual credit growth of 25 percent, 40 percent, and 70 for enterprises and straining public finances, thereby percent, respectively. Providing the right amount of contributing to loss of jobs, income, and wealth. The credit—not too much and not too little—is a major experience from 147 banking crises that struck 116 concern for all countries. countries from 1970 to 2011 (map 1) is telling: the av- erage cumulative loss of output during the first three How can the state contribute? years of crises was 33 percent of GDP in advanced economies and 26 percent in emerging markets.76 Providing sound financial infrastructure. Financial in- Both synergies and trade-offs may exist between frastructure consists of institutions that facilitate fi- financial inclusion, depth, and stability. By making nancial intermediation, including payment systems, greater and more diversified domestic savings avail- credit information bureaus, and collateral registries. able to banks (and thereby reducing reliance on re- Financial infrastructure also includes a regulatory versible foreign capital), greater financial inclusion framework that fosters both consumer protection and depth can enhance the stability of the financial and competition among financial institutions. Mex- system.77 But excessive financial inclusion and rapid ico and South Africa, for instance, have enacted ef- deepening can endanger stability. This applies es- ficient consumer protection frameworks, which in- pecially to credit markets. For instance, the bank- clude ombudsmen to resolve disputes in consumer ing crises in Thailand (1997), Colombia (1982), finance.78 Competition can lead to innovation in and Ukraine (2008) were preceded by excessive an- financial inclusion, as in the Philippines, which has 30 WORLD DEVELOPMENT REPORT 2014 FIGURE 7 Financial inclusion in savings, credit, and insurance across developing countries at different income levels 0.7 THA 0.6 Index of financial inclusion in 2011, MNG 0.5 CHN between 0 and 1 MYS LKA MUS 0.4 BGD ZAF LAO ALB LTU BOL SWZ KEN VNM DOM MNE LBN 0.3 URY HTI PRY CRI RWA KHM ECU AZE LBR UGA ZWE GHA PHL GTM MKD BLR BWA VEN MOZ IDN ARM PER PAN 0.2 SLE NPL COM NGA IND KSV BIH COL KAZ BRA LVA TZA ZMB MAR SRB ROM RUS CHL MWI IRQ SLV GIN TCD MRT CMR NIC HND MEX KGZ GEO BGR ARG AFG SDN UKR 0.1 BFA MLI BEN LSO MDA TUN GAB TUR BDI TGO TJK SEN COG EGY DZA ZAR NER MDG PAK UZB 0 CAF YEM TKM 5 6 7 8 9 10 GNI per capita in 2011, in natural logarithm Sub-Saharan Africa Middle East and North Africa South Asia Latin America and the Caribbean East Asia and Pacific Europe and Central Asia Source: WDR 2014 team based on data from World Bank Global Findex (database) and World Bank World Development Indicators (database). Note: The index of financial inclusion is calculated based on Global Findex data on the use of savings (percentage of adults who saved money at a financial institution in the past year); credit (percentage of adults who borrowed from a financial institution in the past year); and insurance (percentage of adults who personally paid for health insurance, and percentage of adults working in agriculture who purchased agriculture insurance). GNI = gross national income. allowed mobile network operators to take on many prudential supervision and intervene with timely and banking operations.79 Moreover, to promote finan- robust policy tools, as the Republic of Korea did in cial inclusion, the government can lead by example 2011 in the wake of the international financial crisis through innovative practices. An interesting case is by imposing a levy on bank noncore financial liabili- India’s National Rural Employment Guarantee Act, ties to manage speculative capital flows. which has improved outreach to poor people living in Ideally, macroprudential regulation would pre- rural areas through the introduction of government- vent financial crises. Some crises, however, are un- to-person payments using a bank account.80 avoidable, and a crisis resolution system is necessary. How should losses be handled? In resolving crises, Enacting macroprudential regulation for systemic risks. countries should seek to pass bank losses to exist- To better manage the potential for systemic finan- ing shareholders, managers, and in some cases un- cial crises, countries should establish strong macro- insured creditors—minimizing costs to taxpayers, prudential regulatory frameworks—frameworks that threats to fiscal stability, and future moral hazard. To consider the interconnectedness of financial institu- facilitate recovery from crises, governments and the tions and markets and that address the financial sys- international community can contribute by reduc- tem as a whole.81 Making macroprudential regulators ing regulatory uncertainty through timely decisions independent, possibly by placing them under the cen- and effective global coordination. tral bank, is the first step in this direction—as in the Czech Republic, which in 2006 gave the central bank Taking the trade-offs and synergies between inclusion, explicit responsibility for fostering financial stabil- depth, and stability explicitly into account. Evidence ity. Governments can then pursue proactive macro- suggests that in 90 percent of cases, national financial Risk and opportunity 31 MAP 1 Banking crises around the world, 1970–2011 No banking crisis One crisis Two crises Three or four crises Missing data Source: WDR 2014 team based on data from Laeven and Valencia 2012. Map number: IBRD 40098. sector strategies do not address specific trade-offs in financial markets. Sound macroeconomic man- between financial development goals and the man- agement can provide an environment where house- agement of systemic risk, although more than two- holds, communities, and enterprises are able to plan thirds of countries commit to achieving both goals for the long term and undertake their own risk man- within their strategy.82 A financial policy committee agement. Furthermore, macroeconomic policy can may provide a means for a country to better take address large systemic risks, which households and trade-offs and synergies in the financial sector into other socioeconomic systems are unequipped to account and improve policy coordination. An in- handle on their own. As Robert Barro, among others, teresting example to consider is Malaysia, where the has noted, macroeconomic crises with large welfare central bank takes the lead in engaging major stake- costs have marked the world economy for decades— holders in financial sector policy, including the min- palpably so since 2007.83 Policy makers have an es- istry of finance and private sector experts. The goal sential role to play in preventing these crises or at of this engagement is to prepare a national financial least in mitigating their effects. sector strategy for Malaysia that takes into account trade-offs between promoting financial inclusion Macroeconomic stability. Business cycles are in- and development and managing systemic risk in the trinsic to modern economies, and some degree of financial sector. volatility in aggregate prices, output, and employ- ment is normal. Evidence indicates that the harm- ful effects of volatility do not derive from moder- The macroeconomy ate fluctuations but from high inflation and abrupt How can the macroeconomy foster resilience moves in economic activity. These effects percolate and prosperity? throughout the economy—reducing employment, interrupting credit, and deferring investment—and The macroeconomy is the platform where all eco- produce losses that lead to a decline in long-term nomic activity takes place: from consumption to economic growth. Indeed, analysis across a set of savings in households, from investment to produc- developed and developing countries over four de- tion in enterprises, and from borrowing to lending cades suggests that an increase in GDP volatility 32 WORLD DEVELOPMENT REPORT 2014 from normal to crisis-related levels can decrease prudential instruments—aimed at curbing financial long-run per capita GDP growth by around 2 per- imbalances and volatile capital flows—rather than centage points a year.84 through monetary policy.88 Continuous provision of public goods and services. Part Flexible exchange rate regimes. Although debated for of the reason why crises have an impact on long-run a long time, flexible exchange rates have proven to growth is that they can result in an interruption or be effective shock absorbers. That is true whether deterioration in the provision of essential public the shock originated inside or outside the domestic goods and services. These interruptions occur es- economy. Countries with flexible exchange rates tend pecially when governments are forced to undertake to adjust better—recovering more quickly and more drastic cuts in expenditures during downturns. This strongly—to deterioration in their terms of trade,89 was the case, for instance, in several Latin American natural hazards such as earthquakes and storms,90 countries during the 1980s and 1990s, with more and other shocks that may produce internal or exter- than half the fiscal adjustment consisting of spend- nal imbalances.91 ing cuts in infrastructure investment.85 Similarly, social security spending dropped in nearly half the Countercyclical and sustainable fiscal policy. World- countries in the Middle East and North Africa fol- wide, fiscal policy has not made as much progress lowing crises in the region.86 During the latest global as monetary policy in terms of effective process and financial crisis, education budgets fell sharply in the positive results. This is not surprising: fiscal policy is majority of Eastern European countries: for instance, inherently more complex—having multiple objec- by 25 percent in Serbia and 10 percent in Hungary.87 tives and instruments and being immersed in the political process. With respect to risk management, fiscal policy in developing countries has suffered What policies can best contribute to risk from a procyclical bias that has tended to amplify up- management? swings and worsen recessions.92 In the past two de- Experts have argued that macroeconomic policies cades, however, several developing countries around should be credible, predictable, transparent, and the world have put a premium on fiscal transparency sustainable. This is sensible advice. It can also be and discipline, building buffers during good times presented more concretely in terms of risk manage- with an eye toward future downturns. These insti- ment: macroeconomic policy makers should behave tutional improvements explain the recent ability of prudently during upswings to avoid costly coping a large fraction of developing countries to conduct during downturns. countercyclical fiscal policy, mainly by turning in- vestment and consumption spending in a direction Transparent and credible monetary policy. Endowed opposite to that of the cycle in general economic ac- with independence and a drive for transparency and tivity (map 2 focuses on countercyclical consump- credibility, monetary policy authorities have success- tion spending). Independent fiscal councils can pro- fully brought down inflation worldwide in the last vide an important means to further institutionalize 25 years: while 34 countries had annual inflation such discipline (box 7). greater than 50 percent in 1990–94, only 1 country Why is countercyclical fiscal policy useful? First, (Zimbabwe) registered that rate by the end of the it allows governments to continue to provide goods 2000s. Adopting a monetary policy framework that and services and to maintain their public investment creates incentives for long-term price stability, while programs in a stable fashion, even if public revenues accounting for the business cycle, has been crucial to drop (as is normal in the downside of the business defeating inflation. cycle). Second, it provides resources to increase social The 2008–09 international financial crisis and assistance and insurance to larger numbers of peo- the ensuing recession in developed countries have ple in need who are suffering from adverse cyclical tested the improvements made in monetary policy macroeconomic conditions. These two mechanisms in developing countries. All in all, they have proven make a significant contribution not only during the to be resilient. One important issue to consider in recessionary part of the cycle but also for the long- the wake of the crisis is whether financial stability run welfare of people and the economy.93 A third pos- should be included as a direct objective of monetary sible reason is to stimulate the economy. There is little policy. The jury is still out, but it can be argued that evidence, however, that discretionary fiscal stimulus financial stability is best achieved through macro- based on fueling consumption works. To the con- Risk and opportunity 33 MAP 2 Government consumption became countercyclical in more than one-third of developing countries over the past decade Always countercyclical Becoming countercyclical Becoming procyclical Always procyclical Missing data Source: WDR 2014 team estimations based on Frankel, Végh, and Vuletin 2013 methodology. Map number: IBRD 40099. Note: The map shows the evolution of the cyclical stance of fiscal policy from 1960–99 to 2000–12. The cyclical stance is measured in a regression of the (Hodrick- Prescott) cyclical component of general government consumption expenditure on its own lagged value, and the cyclical component of real GDP. The sign of the coefficient on the cyclical component of real GDP indicates whether government consumption expenditure is procyclical (positive sign) or countercyclical (nega- tive sign). The coefficient on the cyclical component of real GDP was estimated separately for the periods 1960–99 and 2000–12. Then, countries are classified as always countercyclical (in both periods); becoming countercyclical (only countercyclical in 2000–12); becoming procyclical (only procyclical in 2000–12); and always procyclical (in both periods). The likely endogeneity of the cyclical component of real GDP was controlled for by using as instruments the (current and lagged value of the) cyclical component of real GDP of the country’s main trading partners and international oil prices, as well as the lagged value of the country’s own cyclical component of real GDP. trary, estimates of the Keynesian fiscal multiplier— Dealing with contingent liabilities requires a combi- the increase in GDP for every dollar of additional nation of measures: first, governments must provide government spending—range only between 0.4 and the right incentives for self-reliance—for example, 0.6 for most developing countries and between 0.6 by replacing pay-as-you-go systems with fully capi- and 1.2 for most developed countries.94 Once the cost talized old-age pension systems, and by letting risk- of raising the necessary additional revenue (in terms takers in financial markets suffer full losses from of taxes, debt, and red tape) is factored in, the net failed ventures. Second, market solutions should be multiplier is likely to be near zero or negative. encouraged by, for example, allowing the issuance of Finally, from a risk management perspective, catastrophe bonds in international markets to insure fiscal sustainability requires being aware of contin- against natural hazards. And, third, resources should gent liabilities. Some of them are legitimate, such be provisioned for residual liabilities that the state as reconstruction and assistance in the aftermath may have to bear. of natural disasters and the larger outlays required to cover social insurance and medical treatment for an aging population. Other contingent liabilities are The international community more controversial; financial bailouts, for example, When can the international community foster can represent a large burden for the state: around resilience and prosperity? 50 percent of GDP in Indonesia and Thailand after the 1997 East Asia crisis, and over 40 percent of GDP Unmanaged risks do not respect boundaries, and no in Iceland and Ireland during the 2008–09 crisis.95 one country or agent acting alone can deal effectively 34 WORLD DEVELOPMENT REPORT 2014 B OX 7 An independent fiscal council can help overcome procyclical fiscal bias What is the problem? Fiscal authorities around the world routinely including competitive appointment and long tenure of council deviate from sustainable plans and suffer from a “procyclical” bias: board members, budget independence, and strong accountability they tend to run budget deficits and accumulate debt in good times, mechanisms (such as being evaluated by peer councils or interna- and then lack adequate resources and flexibility (“policy space”) to tional organizations).b stabilize output in bad times. Has this solution been implemented anywhere? By 2012, 22 national A proposed solution. The creation of an independent fiscal council can governments (and counting) had created fiscal councils, with varying provide the right incentives for the government to build up resources characteristics and degrees of relevance.c The Netherlands’ Centraal to cope with cyclical downturns and long-run contingencies. The fis- Planbureau and the Swedish Fiscal Policy Council are the closest to cal council would administer a set of flexible fiscal rules mandated by full-fledged fiscal councils. In Chile, two independent advisory bod- law: deciding on the allocation of deficits over time, signaling when ies provide key inputs for the projection of the “structural” revenue, countercyclical action is justified, and monitoring public debt sus- which in turn determines government expenditure through a fiscal tainability. Full delegation of policy making to an independent fiscal rule. Acting as advisory bodies, fiscal councils in Morocco, Kenya, and council is unrealistic because of the political and redistributive Uganda provide ex ante and ex post assessment of fiscal policies for nature of fiscal policy. The government, following its political man- parliament. date, would retain control over the distribution of expenditures and If a council is not feasible, is there an alternative? Establishing an the structure of taxation. However, isolating some aspects of fiscal independent fiscal council requires the political appetite for auton- policy implementation from the political process and delegating omous institutions and strong governance underpinnings and them to an independent council can enhance fiscal credibility and thus may not be possible in all countries. Where an independent accountability.a council is not feasible, a good foundation for fiscal sustainability How can this solution be implemented? Fiscal councils should be would involve adopting transparent and comprehensive fiscal designed in a way that avoids political capture, the rise of govern- frameworks, including top-down approaches to budgeting. Since ment incentives to ignore council advice, or the possibility of being the 2000s, Armenia, for instance, has formulated a three-year roll- dismantled when conflicts within government occur. An effective ing budgetary framework with expenditure ceilings and integrated fiscal council requires independence from the political process— it into budgetary law.d Source: WDR 2014 team. a. Debrun, Hauner, and Kumar 2009. b. Calmfors and Wren-Lewis 2011. c. IMF 2013. d. World Bank 2013. with a risk that crosses a national border. Once trig- countries face severe capacity constraints and have gered, pandemics and financial or economic crises weak or dysfunctional governments.96 That is espe- can circle rapidly around an increasingly intercon- cially the case in fragile and conflict-affected coun- nected world. Armed conflicts can devastate people tries, where people face the most extreme risks and and spill over into neighboring countries. Natural di- obstacles to risk management, with limited access to sasters can ruin a country or an entire region. Climate functioning markets, communities, and public insti- change is likely to intensify all these risks. Clearly, tutions. People living in fragile and conflict-affected risks that spread across and affect multiple countries countries made up 15 percent of the world popula- or generations call for international attention. tion in 2010, but about one-third of people living in The international community is a fusion of rather extreme poverty.97 Conflicts can transcend national diverse agents, including sovereign governments, in- borders, resulting in increased refugee populations, ternational organizations, the global scientific com- spread of communicable diseases, and growing pres- munity and media, and civil society. It can offer ex- sure on public goods in neighboring countries ab- pertise and knowledge; provide protection through sorbing affected populations. Sharing a border with a global rules and regulations, capacity building, and fragile state can reduce a country’s economic growth international coordination; and pool national re- by 0.4 percent annually.98 By improving economic sources to better prepare for risk and alleviate crisis prospects and the environment for health, security, situations. and education, engagement by the international community can reduce social and economic ten- Risks that exceed national capacity. The international sions that inflame and spread conflict, while nurtur- community’s engagement may be needed when ing opportunities. Risk and opportunity 35 International support is also needed when very nizations.101 Increased air travel and trade in goods large shocks, such as natural disasters and finan- and services, for instance, can provide free passage cial crises, result in losses that dwarf a country’s to pathogens that cause infectious diseases, some resources. That can happen even in large and more of which can travel around the world in less than developed countries, as the Euro Area crisis clearly 36 hours.102 Similarly, financial crises can spread demonstrates—although low-income countries are through an increasingly complex network of links disproportionately affected by economic risks and across financial systems around the world. Rapid disasters. For example, the Aceh province in Indo- economic growth that has relied heavily on carbon- nesia bore the brunt of a powerful earthquake and based energy is also related to slowly evolving risks tsunami in 2004, leaving more than 500,000 people such as climate change and environmental degrada- homeless and an estimated economic loss of 97 per- tion, with potentially irreversible consequences for cent of Aceh’s GDP. The international community set future generations. up a special multidonor fund to support reconstruc- tion and establish early warning systems, efforts that What characteristics improve the almost 10 years after the tragedy have largely proven international community’s capacity to to be a success.99 Success does not always follow, manage risk? however, as illustrated by the disappointing results of the international community’s intervention in Haiti The effectiveness of the international community after a powerful earthquake in 2010.100 depends on how well it can fill in knowledge and ca- pacity gaps, establish rules and standards that guide Risks that cross national borders. Openness and nations in managing their risks, and facilitate and modernization have made economic, social, and coordinate collective action to manage risks that go ecological systems increasingly interconnected (fig- beyond national borders. In turn, collective action ure 8). Along with opportunities for growth and is facilitated when agents within the international poverty alleviation, this interconnectedness has community are united by shared preferences and also created a set of risks that cross national bor- objectives, or when certain actors have the ability to ders and require critical risk management from the mobilize resources and enforce agreements—even in international community, including regional orga- the absence of cohesion or unity across nations. FIGURE 8 Economic, financial, and social interconnectedness are on the rise 200 500 180 450 160 400 140 350 120 300 Index Index 100 250 80 200 60 150 40 100 20 50 0 0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 Ratio of foreign claims of international banks on all countries to GDP Tourist arrivals per capita Ratio of total trade to GDP Internet users per capita (right axis) Source: WDR 2014 team based on data from World Bank World Development Indicators (database), Bank for International Settlements Consolidated Banking Statistics (database), and World Tourism Organization Yearbook of Tourism Statistics (database). Note: All series are indexed to 100, with 2000 as the base year. 36 WORLD DEVELOPMENT REPORT 2014 Cohesiveness through shared preferences and objec- If incentives are aligned: Pursue proactive and well- tives. Mutual recognition of the need to address risks coordinated interventions. When incentives are aligned enables the international community to better pre- and a course of action is clear, scaling up risk manage- pare for risks that exceed national capacity—such as ment requires proactive and well-coordinated inter- the arrangements to provide emergency lending to ventions by the international community. In dealing countries facing acute financing shortfalls, and sup- with risks such as pandemics or financial crises in port for regional insurance pools like the Caribbean an interconnected world, the effectiveness of these Catastrophe Risk Insurance Facility.103 Similarly, actions rests critically on supporting the capacity of multilateral cooperation for risks that cross bound- individual countries to monitor and contain risks aries works best when the interests of various nations in their territory. For example, while 36 donors pro- are well aligned and are not overruled by competing vided support to more than 100 developing countries domestic policy priorities. By helping to align na- to prepare for a possible pandemic of avian flu tional interests, the almost universal agreement for (H5N1) from 2005 to 2010, local monitoring was the need to eliminate smallpox facilitated its eradi- essential to contain the virus. More resources should cation. In contrast, in cases where national interests be devoted to supporting capacity building for early diverge, such as resolving climate change risks and warning, monitoring, and communication systems, alleviating the plight of people living in fragile and and to designing risk-pooling solutions that reward conflict-affected countries, progress can be slow. preparation. Power to mobilize resources and enforce agreements. If incentives are not aligned: Use incremental ap- The international community can have a substantial proaches to global solutions. When incentives are not impact on the management of risks when there is a aligned, major sovereigns are not fully engaged, and clear goal around which to mobilize resources. For the consequences of inaction are potentially cata- example, with support from the international com- strophic—as with climate change and other environ- munity, early warning systems have helped reduce mental risks such as loss of biodiversity—the inter- deaths from many types of disasters.104 Similarly, national community should embrace incremental even if complete international consensus is lacking, approaches that can increase traction toward global the international community can make progress on solutions (box 8). To preserve full participation as the risks that cross boundaries if it can devise mecha- ultimate goal, however, special attention should be nisms for enforcing agreements. That capacity de- given to steps that can help align incentives toward pends crucially on the international community’s a common objective, even if alignment seems very ability to realign incentives around shared goals difficult to achieve. For environmental risks, this ef- and to attract participation of major players. A key fort may consist of dissemination of knowledge and element in the success of both the Nuclear Non- advocacy that can help bring diverging views closer, Proliferation Treaty and the Montreal Protocol on financial and technology incentives to countries for the protection of the ozone layer, for example, were steps such as preventing deforestation and inducing the threat of security and trade sanctions, respec- the use of cleaner technologies, and investments in tively, which helped realign national interests and research and development—for example to construct facilitate participation and action. methods for counteracting greenhouse gas con- centration in the atmosphere.105 In a similar spirit, the New Deal for Engagement in Fragile States (the How can the international community Busan Partnership) recognizes that the risk of non- improve its contribution? engagement can outweigh most risks of engagement The insights from the work by Leonid Hurwicz, in fragile countries; it outlines a framework in which Roger Myerson, and Eric Maskin on mechanism de- the international community can work to help them sign for institutions are all the more important for strengthen core institutions and policies and reduce a collectivity as fluid, diverse, and complex as the the risk of reverting to conflict.106 international community. Considering incentive constraints (and not only budget and informational An institutional reform to mainstream risk constraints) is critical to devising effective mecha- management nisms for the international community to contrib- ute to risk management despite its multiple players, The World Development Report 2014 offers dozens complicated power structures, and diverging goals. of specific policy recommendations to improve risk Risk and opportunity 37 B OX 8 For certain global risks such as climate change, the international community should embrace incremental approaches that can lead to global solutions What is the problem? Management of global risks requires proactive Are there successful examples? Some remarkable examples exist. concerted action by sovereign nations. But limited progress in some The Montreal Protocol to protect the ozone layer was originally areas has cast doubt on the possibility of fostering collective action signed by 24 countries but won universal ratification during the among countries with diverging interests, capacity constraints, and 1990s with the combined efforts of governments, international orga- incentives to free ride. Global negotiations to secure agreements nizations, nongovernmental organizations, and scientists.a Likewise, with full participation have stalled—most spectacularly for climate the Limited Test Ban Treaty, whose signatories expanded from 3 to change, where persistent inaction could have catastrophic and 119 between 1963 and 1992, paved the way for the more compre- irreversible consequences. Some potentially useful international hensive Nuclear Non-Proliferation Treaty. actions—including cooperation to develop and share technologies How can it be implemented? Country governments, international and existing financial instruments—have been postponed in the organizations, and specialized entities can form a “coalition of the expectation that they will be part of a “soon-to-be-signed” global willing” to coordinate, advocate, and take action on climate change.b agreement. The coalition can create incentives for others to join over time by pro- The proposed solution. For certain global risks such as climate moting technological change and funding that lowers participation change, the international community should embrace incremental costs (cheaper ways to reduce emissions, subsidies, or technology approaches that can increase traction toward global solutions. When transfers). It can also partner with scientists, civil society, and the incentives are misaligned, major sovereigns are not fully engaged, media to induce participants to comply and nonparticipants to join and the consequences of inaction are disastrous, progress can still be in. International institutions, including an international risk board, made outside a multilateral treaty. Incremental deals and actions by can provide platforms for policy debate and monitor, report, and an initially small group of participants can serve as building blocks aggregate actions to ensure incremental efforts are on the right to global agreements. By demonstrating benefits from action, the path. Strategically, the coalition could anchor its actions to existing expectation is that the group would include progressively more par- global frameworks to demonstrate that incremental and global ticipant countries over time. deals can be connected. Source: WDR 2014 team. a. UNEP 2007. b. Falkner, Stephan, and Vogler 2010; Goldin 2013; Hale 2011. management at different and complementary levels Establish a national risk board to manage risks of society (box 9 provides a summary of these poli- in a proactive, systematic, and integrated way cies). Its overarching advice, however, is that these recommendations should be implemented in a pro- What is the problem? All too often, risk manage- active, systematic, and integrated way to optimize ment strategies and implementation prove ineffec- their effectiveness. For this purpose, the World Devel- tive (or introduce other risks) because they are not opment Report 2014 advocates establishing a national coordinated among all relevant policy stakeholders. risk board, which can contribute to mainstreaming Managing risk in a proactive and integrated way has risk management into the development agenda. This definite advantages: it can help define priorities, en- could be a new agency or come from reform of exist- sure that all contingencies have been considered, and ing bodies: what is most important is a change in ap- avoid overspending to manage one risk in isolation proach—one that moves toward a coordinated and while neglecting others. Some countries conduct na- systematic assessment of risks at an aggregate level. tional risk assessments that involve multistakeholder Implementing this recommendation may require a teams from various ministries and often include the substantial change in the way national governments private sector and civil society. The Netherlands, develop and implement their general plans, moving the United Kingdom, and the United States have from planning under certainty to considering change completed this exercise, and other countries, such as and uncertainty as fundamental characteristics of Morocco, have begun a process toward it. However, modern economies. A national risk board can help this exercise is usually carried out by a temporary, governments overcome the political economy ob- ad hoc group that exists only while the assessment is stacles they face when managing risks at the country taking place. Other countries have created multimin- or even international levels. istry bodies in charge of information exchange and 38 WORLD DEVELOPMENT REPORT 2014 B OX 9 Selected policy recommendations from the WDR 2014 The state has an important role in supporting the contributions of For the financial system: all social and economic systems to people’s risk management. The • Sound financial infrastructure (payment systems, credit informa- following summarizes selected policy recommendations from the tion) to facilitate financial inclusion and depth WDR 2014, organized by system, as they are discussed in the Report: • Enforceable regulations that foster both consumer protection and For the household: competition among financial institutions • Macroprudential regulation, for the financial system as a whole, to • Public health insurance, run in partnership with the private sector, lessen financial crises and avoid bailouts with emphasis on preventive care and treatment of contagious • A national financial strategy that addresses trade-offs between diseases and accidents financial inclusion, depth, and stability • Public education, run in partnership with the private sector, with a focus on flexible skills, adaptable to changing labor markets For the macroeconomy: • Targeted safety nets for the poor, for instance conditional cash • Transparent and credible monetary policy, oriented to price stabil- transfers with payments directly to women ity and conducted by an autonomous central bank • Enforceable laws against domestic abuse and gender discrimina- • For the majority of countries, a flexible exchange rate regime, in a tion, accompanied by educational campaigns context of transparent and credible monetary policy For the community: • Countercyclical and sustainable fiscal policy, aided by an indepen- dent fiscal council • Public infrastructure for the mitigation of disaster risks, built in • Provision for contingent liabilities, such as natural disasters, finan- consultation with surrounding communities cial crises, and pensions of an aging population • Transportation and communication infrastructure, especially to integrate and consolidate isolated communities For the international community: • Police protection against common and organized crime, espe- • Engagement in bilateral, regional, and global agreements to share cially targeted to communities under threat risks across countries, enhance national capacity, and confront • Enforceable laws against racial or ethnic discrimination, accom- common risks, favoring proactive and coordinated interventions panied by educational campaigns • For elusive global risks such as climate change, formation of a For the enterprise sector: “coalition of the willing” with like-minded country governments, creating incentives for other countries to join in. • Secure and respected private property rights • Streamlined and predictable regulations for taxation, labor mar- The WDR 2014 advocates that these recommendations be imple- kets, and entry and exit of firms mented in a proactive, systematic, and integrated way. For this pur- • Enforceable regulations for workplace safety, consumer protec- pose, it proposes establishing a national risk board to help main- tion, and environmental preservation stream risk management into the country’s development programs • Consider the possibility of delinking social insurance (that is, and suggests the possibility of an international risk board to support health and old-age pension) from work status the “coalition of the willing.” Source: WDR 2014 team. coordination for risk management, but these bod- tions about appropriate policies to be implemented. ies usually deal with a single risk—most often with Institutionalizing the national risk board should add natural disasters, as in Peru, or national security, as value by enabling risk management to be integrated in Israel. Few countries actually have an integrated, across all sectors, by challenging inaction stemming permanent risk management agency that deals with from political interests, and by introducing clear ac- multiple risks. countability mechanisms for implementing agreed risk management measures.107 What is the solution? To facilitate proactive and inte- grated risk management at the country level, a na- How can it be implemented? The national risk board tional risk board can be set up as a standing (per- should bring together a wide range of stakeholders. manent) committee. It can analyze risks, including It could be either part of government or an autono- trade-offs across risks and across risk management mous agency. The board composition would include policies; consider and publish assessments of risk both policy makers (to reflect political priorities) management practices in the country; define pri- and independent experts (to incorporate techni- orities in risk management; and make recommenda- cal knowledge and private sector perspectives). It Risk and opportunity 39 would have the power to issue “act-or-explain” rec- ommendations to relevant authorities responsible Balancing the trade-offs in the institutional D I AG R A M 4 for implementing policy—that is, relevant authori- ties would have to act upon the board’s recommen- design of a national risk board dations or explain why they had decided to reject Independent them. Although the appropriate institutional design experts of the board will depend on the country’s political Lower Lower relevance and institutional context, the board’s composition legitimacy and powers should strive to achieve an adequate balance of expertise, credibility, relevance, and legit- imacy—that is, to fall within the “balanced” region in diagram 4. Balanced Advisory Implementation The board’s policy makers could be nominated role region role by the executive branch of government, and the in- dependent experts could come from academia, the business community, and civil society organizations. The board’s expertise would cover the areas of mili- Lower Lower credibility expertise tary, security and terrorism risk; economic risk; envi- ronmental, health, and technological risk; and social Policy risk. To avoid becoming a powerless body, the board makers should have sufficient prominence in the public eye. Source: WDR 2014 team. And it should be held accountable by regularly pub- lishing its recommendations accompanied by analy- sis and statements of policy priorities and by being subjected to annual hearings in front of a legislative the Homefront Crisis Executive Group, comprising committee. senior representatives from ministries and govern- While an autonomous national risk board may ment agencies. This multirisk framework is comple- have certain advantages, the board could also func- mented by agencies focused on specific risks, such tion as part of government. Indeed, countries as as the National Security Coordination Secretariat. different as Jamaica, Mali, Mexico, Morocco, and Singapore’s institutional arrangement for integrated Rwanda are considering establishing an integrated risk management involves a great deal of specializa- risk management function within the government tion and a complex coordination process that has structure—in part following a proposal by the World evolved over time. For developing countries, a sim- Economic Forum to establish a country risk officer, pler arrangement that involves less specificity and similar to the position of chief risk officer that has specialization in its institutional design (and requires been created in many multinational companies. 108 less demanding coordination mechanisms) may be a This institutional design could be practical in coun- good starting point. tries with a strong framework for an effective and in- Finally, two important questions should be ad- dependent civil service, with the national risk board dressed. First, what can motivate a government to members appointed as expert technocrats with guar- institute a national risk board? An initial impulse is anteed positions for periods that extend beyond the necessary for leaders to overcome opposing incen- political cycle. tives and establish a long-term institution. This im- Singapore’s Whole-of-Government Integrated pulse can come from within the country, through Risk Management framework is an example of an reform-minded political leaders and technocrats, approach that has overcome “silos” within the gov- and from outside, through incitement and support ernment.109 The institutional umbrella of the frame- from the international community. Once created, work is the Strategy Committee, composed of per- the national risk board can challenge inaction or manent secretaries from various ministries across poor practices by introducing clear accountability government and chaired by the Head of Civil Ser- mechanisms for risk management. A reformist gov- vice. In addition, the Homefront Crisis Management ernment interested in the continuation of its ben- system includes a ministerial committee chaired by eficial legacy may want future governments to be the Minister of Domestic Affairs and supported by accountable for their actions or their lack of action. 40 WORLD DEVELOPMENT REPORT 2014 The second question is whether a similar body the enterprise sector to grow, develop, and provide can be created at the global level—an international risk management resources to the entire population. risk board—to help address risks that cross national Third, an internally fragmented government that boundaries. An international risk board could in- lacks organization and coordination may end up volve the scientific and expert community around with ambivalent policies or ineffective implemen- the world to pool all available knowledge to iden- tation. This may occur, for instance, as result of a tify, assess, and manage major global risks. Its major defective decentralization process, where local and drawback would be that, in the absence of a gov- regional governments do not have the necessary re- erning body at the international level, it could lack sources and capacities to fulfill their responsibilities, implementation relevance. That could be remedied, do not share the priorities and preferences of the na- however, if the international risk board were to work tional government, or attempt to free ride on other in conjunction with the “coalition of willing” coun- local and regional governments. tries (see box 8), setting priorities on issues to be Finally, the government may be guided by ideol- tackled urgently and offering credibility and legiti- ogy, wishful thinking, or simple desperation when macy to its efforts. confronting difficult and genuine problems, instead of relying on measures based on good evidence and analysis. A common example is labor market regu- In conclusion: Five principles of public lations that purport to defend workers’ interests but action for better risk management wind up protecting only a few and contributing to Analysis throughout the World Development Report the roots of a large informal sector. Inflationary fi- 2014 suggests that, to improve the quality and de- nancing of budget deficits or variable and inconsis- livery of social protection, public goods, and public tent macroeconomic policies in the face of crisis are policy that are essential to supporting people’s risk other examples: sooner rather than later, both paths management, public action can usefully be guided lead to increased uncertainty, macroeconomic insta- by some key principles. The five principles that fol- bility, and possibly even protracted recessions. low reflect the lessons from best practice around the world and are relevant for different types of risks 2. Provide the right incentives for people and and countries. Their application should be tailored institutions to do their own planning and to specific contexts, however. Although at first glance preparation, while taking care not to impose these principles may appear uncontroversial, in ap- risks or losses on others plication they involve tensions and trade-offs that make their implementation a challenge. The challenge for public policy is to create incentives for people to do their own risk planning and prepa- ration, avoiding circumstances in which benefits 1. Do not generate uncertainty or unnecessary are privately appropriated but losses are imposed on risks others. The state’s policies and actions should strive to re- Consider financial bailouts. They are detrimen- duce risks and lessen uncertainty. At a minimum, the tal not only because they can produce a large fiscal state should not worsen them. How or why would burden but also because they provide incentives a government do that? First, through its policies, for excessive risk taking. Yet bailouts are sometimes it may perpetuate social norms that discriminate necessary to prevent a systemic collapse of financial against certain groups and make them more vulner- intermediation. Bailouts should be avoided—most- able. For example, state policies that promote gen- ly by using well-established, clear, and transparent der inequality or ethnic favoritism harm, rather than macroprudential policies—but if bailouts occur, they help, household and community resilience. should be designed to avoid providing the wrong Second, the government may favor the group that incentives for the future. Good examples of orderly supports it politically, whether a small elite or large financial bailouts are hard to find, but the Turkish constituency, against the legitimate interests of oth- experience in the wake of the 2000–01 banking crisis ers. For instance, states that expropriate financial as- (and especially the unwavering stance of the coun- sets (like savings and pension funds) or private infra- try’s bank regulatory and resolution agencies) offers structure (like residential buildings or factories) from a case to analyze and follow.110 some households may obtain short-run gains but end In a very different realm, social protection can be up hampering the ability of the financial system and criticized for not encouraging personal self-reliance Risk and opportunity 41 and being an unsustainable burden to the state. The 4. Promote flexibility within a clear and evidence, however, demonstrates that these prob- predictable institutional framework lems can be avoided by a design that takes people’s incentives directly into account. Well-designed safety Flexibility in adjusting to new circumstances is es- nets—such as conditional cash transfers or workfare sential to promoting resilience and making the most programs, as implemented in Bangladesh, Brazil, of opportunities. Prime examples include household India, and Mexico, to name a few—have promoted migration in response to shifting economic trends, better household practices in the areas of education, rural communities’ adaptation to climate change, health, and even entrepreneurship, while remaining and enterprise renewal in the face of technological fiscally sustainable.111 and demand shocks. Flexibility should not imply In all cases, to manage risks effectively, two arbitrary discretion or haphazard responses, how- changes in people’s mindset related to individual ever. A challenge for the state is to promote flexibility and social responsibility are critical: moving from while preserving a sensible, transparent, and predict- dependency to self-reliance, and from isolation to able institutional structure. cooperation. Providing the right incentives can con- For enterprises, the Danish model of “flexicurity” tribute in both regards. offers such balance, combining ease of hiring and fir- ing of workers alongside a strong social safety net and reemployment policies. The result is a dynamic econ- 3. Keep a long-run perspective for risk omy with high turnover in employment but short management by building institutional spells of unemployment. For the macroeconomy, mechanisms that transcend political cycles inflation targeting regimes with floating exchange A major challenge for public action is to establish rates offer a good model of flexible yet institutionally institutional mechanisms that induce the state to sound monetary policy. By 2012, 27 countries around keep a long-run perspective that outlasts volatile the world had adopted an inflation targeting regime. shifts in public opinion or political alliances. For in- With the onset of the European Monetary Union in stance, the state’s provision of education and health 1999, many countries that had practiced inflation services is a large investment in risk preparation for targeting in the 1990s abandoned the regime. Given families and communities that must be funded on the prolonged recession and uncertainty in the Euro a continuous and sustainable basis to succeed: that Area, monetary flexibility could have been a useful entails long-run planning. In the case of health ser- tool these countries no longer have. vices, Thailand and Turkey offer successful examples with their recent shift to universal health insurance 5. Protect the vulnerable, while encouraging programs. self-reliance and preserving fiscal Consider also the following two examples from sustainability financial and macroeconomic policy. For the fi- nancial system to support risk management, it is The harsh reality is that throughout the world, many essential to strike the right balance between inclu- people do not have the material resources and in- sion and stability. This balance can be assessed only formation necessary to confront the risks they face. through comprehensive long-run planning, like that The everyday struggle to eke out a living can make being done in Malaysia, where the strategy for the planning ahead hard for the poor. The challenge for financial sector is prepared by the central bank, in the state is to protect the vulnerable while preserving collaboration with the ministry of finance and the fiscal sustainability—and encouraging self-reliance. private sector. Countercyclical monetary and fiscal For households that remain highly vulnerable to policies also require a long-run perspective, which shocks, the state can provide safety nets to replace allows them to manage the business cycle by using the costly coping mechanisms that undermine con- resources built over a prolonged time and in dif- sumption, human capital, and productive assets. ferent scenarios. Best practice suggests targeting a Safety nets are possible even in low-income coun- long-run budget balance, as Chile, Colombia, and tries, provided the support is targeted to vulnerable Norway, among others, are doing. Institutional populations and is designed to incentivize work ef- mechanisms that transcend the political cycle— fort. Ethiopia’s Productive Safety Net System, for ex- such as a national risk board and an independent ample, demonstrates how a well-designed safety net fiscal council—can help maintain a long-run focus can protect millions of households from food inse- on risk management. curity while investing in community assets. 42 WORLD DEVELOPMENT REPORT 2014 The international community can also provide and opportunities that lie at the heart and core of support to vulnerable populations with resources modern life. So too can the fate of communities and and expertise. Although much criticized, foreign aid countries improve, if they share the continuous re- has been successful when provided in coordination sponsibility required to manage risk successfully. with accountable local institutions. Such was the case when foreign aid helped rebuild infrastructure and “I grew up in a war environment. And what I learned establish early warning systems in Indonesia after the is that you can plan your fate, at least to some degree, if 2004 tsunami. you assess your risk and do something about it.” At the end of the day, protection of the vulnerable —Klaus Jacob, disaster risk management expert entails taking the measures necessary for sustainable at Columbia University and World War II development—development that eliminates extreme survivor112 poverty and allows people to escape vulnerability through the sustained growth that risk management can offer. “There was a time I used to walk to work every day. The route I had to take was dangerous, and many peo- ple were victims [of] robbery and physical abuse. So, Some closing thoughts yes, I have overcome risk to pursue opportunity.” The fate of individuals and families can change for —Kariuki Kevin Maina, student, Kenya the better if they plan and prepare to face the risks Contribution to the WDR 2014 website Risk and opportunity 43 Notes exacerbated the Great Depression (Eichengreen and Irwin 2010). However, the international community has been less 1. Dercon, Hoddinott, and Woldehanna 2005. successful in avoiding export restrictions during the food 2. Buvinić and Morrison 2000. price crises in recent years (Martin and Anderson 2012). 3. World Bank 2012d. 37. Knight 1921. 4. WHO 2013. 38. Hallegatte and others 2012. 5. Baulch 2011 offers a useful survey. 39. Čihák and others 2012. 6. Paul 2009. 40. For a review of the literature on risk sharing and family and 7. Didier, Hevia, and Schmukler 2012. network formation, see Fafchamps 2011. 8. Dercon and Christiaensen 2011. 41. Acemoglu and Robinson 2012 compile many examples of 9. See Karlan and others 2012 for Ghana; and Cole, Giné, and such failures. Vickery 2013 for India. 42. Reinikka and Svensson 2005; Speer 2012; Devarajan, Khemani, 10. Hoddinott, Rosegrant, and Torero 2012. and Walton 2011. 11. Hallegatte 2012a. 43. Becker 1993. 12. While the costs of preparing for risk must be incurred pre- 44. Oviedo and Moroz 2013 for the WDR 2014. dominantly up front, the benefits tend to accrue over time 45. De Weerdt and Hirvonen 2013 for Tanzania; Yang and Choi and are therefore more uncertain. The probability of a risk 2007 for the Philippines; Paulson 2000 for Thailand. materializing is thus central to any assessment of a potential 46. WDR 2014 team based on OECD Programme for Interna- intervention. In formal benefit-cost analyses, this probability tional Assessment (PISA). is usually taken into account either implicitly (by basing cal- 47. Xu and others 2003. culations of averted costs on average historical data) or ex- 48. Demirgüç-Kunt and Klapper 2012. plicitly (by weighting the potential benefit of a risk manage- 49. Duflo 2003; Thomas 1990. ment intervention in the event of a shock by the probability 50. Duflo 2003; see also Thomas 1990 and Lundberg, Pollak, and of that shock occurring). See Wethli 2013 for the WDR 2014. Wales 1997. 13. See, for example, Bodie, Kane, and Marcus 2011. 51. WDR 2014 team based on World Bank World Development 14. See Kuznets 1955 and Dasgupta and others 2002. The trade- Indicators (database). offs mentioned in the text are perceived to exist by some ex- 52. World Bank 2005. perts and a large share of the public, as reflected in opinion 53. Thoresen and Fielding 2011 show how expanding health polls, but may not be present in reality. Recent analyses find, care coverage can put considerable pressure on the sustain- for instance, that economic growth and environmental pro- ability not only of fiscal resources, but also human resources. tection as well as social inclusion are often complementary. 54. Premand 2013 for the WDR 2014. See World Bank 2012b. 55. Paxson and Schady 2007 and references therein; Macours, 15. Hoddinott, Rosegrant, and Torero 2012. Schady, and Vakis 2008; see also references in Fiszbein and 16. Dethier, Hirn, and Straub 2011; Kehoe and Prescott 2007. Schady 2009. 17. Bruno and Easterly 1998. 56. WDR 2014 team based on Demographic and Health Surveys. 18. Bernoulli 1738; von Neumann and Morgenstern 1944. 57. Coleman 1988; Putnam 1993. 19. Allais 1953; Kahneman and Tversky 1979. 58. De Weerdt 2001. 20. Black 1948; Buchanan and Tullock 1962; Olson 1965. 59. Gertler and Gruber 2002. 21. Hurwicz 1960; Myerson 1979; Maskin 1999. 60. For example, Aldrich 2011 shows that social capital plays 22. The concept of a risk chain is discussed and illustrated in a key role in communities’ ability to recover from natural Alwang, Siegel, and Jørgensen 2001. See also Barrett 2002; disasters. Heltberg, Siegel, and Jørgensen 2009. 61. Alesina, Baqir, and Easterly 1999. 23. Ehrlich and Becker 1972. See also the extension in Muer- 62. Narayan, Pritchett, and Kapoor 2009; Bowles and Gintis mann and Kunreuther 2008 and the applications in Gill 2002. and Ilahi 2000; Holzmann and Jørgensen 2001; and Packard 63. Bjorkman and Svensson 2009; Pandey and others 2007. 2002. 64. Knight 1921; Coase 1937. 24. Khokhar 2013 for the WDR 2014. 65. For seminal papers on the topic, see Baily 1974 and Azariadis 25. World Bank World Development Indicators (database). 1975. 26. World Bank 2012c. 66. Schumpeter 1942. 27. For a rich discussion of the potential complementarity be- 67. Hsieh and Klenow 2009. tween insurance and protection, see Erlich and Becker 1972. 68. Baicker, Cutler, and Song 2010. 28. Economist 2013. 69. Bergoeing, Loayza, and Repetto 2004. 29. Hallegatte 2012b. 70. An interesting example of formalization leading to both en- 30. Carter and others 2007. hanced environmental protection and higher incomes has 31. FEMA 2010. recently occurred in Peru. In recent years, informal mines in 32. Gourinchas and Obstfeld 2012; Schularick and Taylor 2012. Peru have sprung up in response to rising gold prices. Ignor- 33. Hallegatte 2012b. ing existing regulations, these informal mines have caused 34. La Porta and others 1998. significant deforestation. The mercury used in the extrac- 35. See, for example, Tornell and Velasco 1992. tion process has contaminated rivers and the atmosphere 36. This kind of trade retaliation was avoided during the and threatened human health. In the La Libertad region, the 2008–09 global financial crisis, in part because of successful Poderosa Mining Company took an innovative approach to coordination by the international community—in contrast the problem after informal miners invaded one of its mining to the well-known “beggar-thy-neighbor” trade policies that concessions. The company began to formalize the invading 44 WORLD DEVELOPMENT REPORT 2014 miners, signing agreements that allowed them to continue 92. Kaminsky, Reinhart and Végh 2005. mining under its direction. The agreements, which meet in- 93. Parker 2011. ternational environmental management quality standards, 94. See Kraay 2012 and Ilzetzki and Végh 2008 for surveys of the have increased the small miners’ income and decreased literature on developing countries; and Barro and de Rugy the harm from deforestation and mercury contamination. 2013 and Ramey 2011 for surveys of the literature on devel- UNEP 2012. oped countries. 71. World Bank 2012d; Loayza and Rigolini 2011. 95. Laeven and Valencia 2012. 72. Calderón and Fuentes 2012. 96. DFID 2005; OECD 2011a, 2012; World Bank 2011. 73. Caballero and others 2013. 97. OECD 2012. 74. Miller 1986. 98. DFID 2005. 75. Diamond and Dybvig 1983. 99. See “Resilience Stories” at the Sendai Dialogue website at 76. Laeven and Valencia 2012. https://www.gfdrr.org/node/1308. 77. Han and Melecky 2013 for the WDR 2014; Cull, Demirgüç- 100. Larrimore and Sharkey 2013. Kunt, and Lyman 2012. 101. Not all risks that exceed national borders are truly global. 78. Brix and McKee 2010. Some risks, such as armed conflict between neighboring 79. Gupta 2013. countries or disputes over natural resources, may affect only 80. World Bank 2012a. a few countries. Such risks may be more appropriately or ef- 81. Borio 2003 provides a discussion of the differences between fectively managed by regional institutions. a traditional, microprudential regulatory framework and a 102. Jonas 2013 for the WDR 2014. macroprudential regulatory approach. 103. Mahul and Cummins 2009. 82. Maimbo and Melecky 2013 for the WDR 2014. 104. World Bank and United Nations 2010. 83. Barro 2009. 105. Royal Society 2009. 84. Hnatkovska and Loayza 2005. 106. OECD 2011b. 85. Easterly and Servén 2003. 107. Graham and Wiener 1995; World Economic Forum 2007. 86. Prasad and Gerecke 2010. 108. World Economic Forum 2007. 87. Education International 2009. 109. OECD 2009. 88. Svensson 2012; Bruno and Shin 2013. 110. Damar 2007; Ersel and Ozatay 2008. 89. Edwards and Levy Yeyati 2005. 111. Fiszbein and Schady 2009; Alderman and Yemtsov 2012. 90. Ramcharan 2007. 112. Quoted in Eric Klinenberg, “Adaptation: How Can Cities be 91. Edwards 2004; Lane and Milesi-Ferretti 2012; Ghosh, ‘Climate-proofed’?” The New Yorker, January 7, 2013, 33. Qureshi, and Tsangarides 2013. Risk and opportunity 45 References Black, Duncan. 1948. “On the Rationale of Group Decision- Making.” Journal of Political Economy 56 (1): 23–34. Acemoglu, Daron, and James A. Robinson. 2012. Why Nations Bodie, Zvi, Alex Kane, and Alan J. Marcus. 2011. Investments. 9th Fail: The Origins of Power, Prosperity, and Poverty. New York: ed. New York: McGraw-Hill. Random House. Borio, Claudio. 2003. “Towards a Macroprudential Framework Alderman, Harold, and Ruslan Yemtsov. 2012. “Productive Role for Financial Supervision and Regulation.” CESifo Economics of Safety Nets.” Social Protection and Labor Discussion Paper Studies 49 (2): 181–215. 1203, World Bank, Washington, DC. Bowles, Samuel, and Herbert Gintis. 2002. “The Inheritance of Aldrich, Daniel P. 2011. “The Power of People: Social Capital’s Inequality.” Journal of Economic Perspectives 16 (3): 3–30. Role in Recovery from the 1995 Kobe Earthquake.” Natural Brix, Laura, and Katharine McKee. 2010. “Consumer Protection Hazards 56: 595–611. Regulation in Low-Access Environments: Opportunities to Alesina, Alberto, Reza Baqir, and William Easterly. 1999. “Public Promote Responsible Finance.” Focus Note 60, Consultative Goods and Ethnic Divisions.” Quarterly Journal of Economics Group to Assist the Poor, Washington, DC. 114 (4): 1243–84. Bruno, Michael, and William Easterly. 1998. “Inflation Crises and Allais, Maurice. 1953. “Le Comportement de l’Homme Rationnel Long-Run Growth.” Journal of Monetary Economics 41 (1): devant le Risque: Critique des Postulats et Axiomes de l’Ecole 3–26. Americaine.” Econometrica 21 (4): 503–46. Bruno, Valentina, and Hyun Song Shin. 2013. “Assessing Mac- Alwang, Jeffrey, Paul B. Siegel, and Steen L. Jørgensen. 2001. “Vul- roprudential Policies: Case of Korea.” Working Paper 19084, nerability: A View from Different Disciplines.” Social Protec- National Bureau of Economic Research, Cambridge, MA. tion Discussion Paper 0115, World Bank, Washington, DC. Buchanan, James M., and Gordon Tullock. 1962. The Calculus Azariadis, Costas. 1975. “Implicit Contracts and Underem- of Consent: Logical Foundations of Constitutional Democracy. ployment Equilibria.” Journal of Political Economy 83 (6): Ann Arbor: University of Michigan Press. 1183–202. Buvinić, Mayra, and Andrew R. Morrison. 2000. “Living in a More Baicker, Katherine, David Cutler, and Zirui Song. 2010. “Work- Violent World.” Foreign Policy (118): 58–72. place Wellness Programs Can Generate Savings.” Health Af- Caballero, Ricardo J., Kevin N. Cowan, Eduardo M. R. A. Engel, fairs 29 (2): 304–11. and Alejandro Micco. 2013. “Effective Labor Regulation and Baily, Martin N. 1974. “Wages and Employment under Uncertain Microeconomic Flexibility.” Journal of Development Economics Demand.” Review of Economic Studies 41 (1): 37–50. 101: 92–104. Barrett, Christopher B. 2002. “Food Security and Food Assistance Calderón, César, and J. Rodrigo Fuentes. 2012. “Removing the Programs.” In Handbook of Agricultural Economics, vol. 2B, Constraints for Growth: Some Guidelines.” Journal of Policy edited by Bruce L. Gardner and Gordon C. Rausser, 2103–90. Amsterdam: Elsevier Science. Modeling 34 (6): 948–70. Barro, Robert J. 2009. “Rare Disasters, Asset Prices, and Welfare Calmfors, Lars, and Simon Wren-Lewis. 2011. “What Should Fis- Costs.” American Economic Review 99 (1): 243–64. cal Councils Do?” Economic Policy 26 (68): 649–95. Barro, Robert J., and Veronique de Rugy. 2013. “Defense Spending Carter, Michael R., Peter D. Little, Tewodaj Mogues, and Workneh and the Economy.” Mercatus Center at George Mason Univer- Negatu. 2007. “Poverty Traps and Natural Disasters in Ethio- sity, Arlington, VA. pia and Honduras.” World Development 35 (5): 835–56. Barro, Robert J., and Jong-Wha Lee. 2010. “A New Data Set of Čihák, Martin, Sònia Muñoz, Shakira Teh Sharifuddin, and Kalin Educational Attainment in the World, 1950–2010.” Working Tintchev. 2012. “Financial Stability Reports: What Are They Paper 15902, National Bureau of Economic Research, Cam- Good For?” Working Paper 12/1, International Monetary bridge, MA. Fund, Washington, DC. Barro, Robert J., and Jose F. Ursúa. 2012. “Rare Macroeconomic Coase, Ronald H. 1937. “The Nature of the Firm.” Economica 4 Disasters.” Annual Review of Economics 4 (1): 83–109. (16): 386–405. Baulch, Bob. 2011. Why Poverty Persists: Poverty Dynamics in Asia Cole, Shawn, Xavier Giné, and James Vickery. 2013. “How Does and Africa. Cheltenham, U.K.: Edward Elgar. Risk Management Influence Production Decisions? Evidence Becker, Gary S. 1993. A Treatise on the Family. Cambridge, MA: from a Field Experiment.” Working Paper 13-080, Harvard Harvard University Press. Business School, Boston, MA. Bergoeing, Raphael, Norman Loayza, and Andrea Repetto. 2004. Coleman, James S. 1988. “Social Capital in the Creation of Human “Slow Recoveries.” Journal of Development Economics 75 (2): Capital.” American Journal of Sociology 94: S95–S120. 473–506. Cull, Robert, Asli Demirgüç-Kunt, and Timothy Lyman. 2012. “Fi- Bernoulli, Daniel. 1738. “Specimen Theoriae Novae de Mensura nancial Inclusion and Stability: What Does Research Show?” Sortis.” Commentarii Academiae Scientiarum Imperialis Pet- Brief 71305, Consultative Group to Assist the Poor, Washing- ropolitanae 5: 175–92. Translated in 1954 in “Exposition of a ton, DC. New Theory of the Measurement of Risk.” Econometrica 22 Damar, H. Evren. 2007. “Does Post-Crisis Restructuring Decrease (1): 23–36. the Availability of Banking Services? The Case of Turkey.” BIS (Bank for International Settlements). Consolidated Banking Journal of Banking & Finance 31 (9): 2886–905. Statistics (database). BIS, Basel, http://www.bis.org/statistics/ Dasgupta, Susmita, Benoit Laplante, Hua Wang, and David consstats.htm. Wheeler. 2002. “Confronting the Environmental Kuznets Bjorkman, Martina, and Jakob Svensson. 2009. “Power to the Curve.” Journal of Economic Perspectives 16 (1): 147–68. People: Evidence from a Randomized Field Experiment on Debrun, Xavier, David Hauner, and Manmohan S. Kumar. 2009. Community-Based Monitoring in Uganda.” Quarterly Journal “Independent Fiscal Agencies.” Journal of Economic Surveys 23 of Economics 124 (2): 735–69. (1): 44–81. 46 WORLD DEVELOPMENT REPORT 2014 Demirgüç-Kunt, Asli, and Leora F. Klapper. 2012. “Measuring Alberto Bisin, and Matthew O. Jackson, 1255–79. San Diego: Financial Inclusion: The Global Findex Database.” Policy Re- Elsevier. search Working Paper 6025, World Bank, Washington, DC. Falkner, Robert, Hannes Stephan, and John Vogler. 2010. “Interna- Dercon, Stefan, and Luc J. Christiaensen. 2011. “Consumption tional Climate Policy after Copenhagen: Towards a ‘Building Risk, Technology Adoption and Poverty Traps: Evidence from Blocks’ Approach.” Global Policy 1 (3): 252–62. Ethiopia.” Journal of Development Economics 96 (2): 159–73. FEMA (Federal Emergency Management Agency, United States). Dercon, Stefan, John Hoddinott, and Tassew Woldehanna. 2005. 2010. “Local Official Survey Findings on Flood Risk.” FEMA, “Shocks and Consumption in 15 Ethiopian Villages, 1999– Washington, DC. 2004.” Journal of African Economies 14 (4): 559–85. Ferreira, Francisco H. G., Julian Messina, Jamele Rigolini, Luis- Dethier, Jean-Jacques, Maximilian Hirn, and Stéphane Straub. Felipe López-Calva, Maria Ana Lugo, and Renos Vakis. 2013. 2011. “Explaining Enterprise Performance in Developing Economic Mobility and the Rise of the Latin American Middle Countries with Business Climate Survey Data.” World Bank Class. Washington, DC: World Bank. Research Observer 26 (2): 258–309. Fiszbein, Ariel, and Norbert Schady. 2009. Conditional Cash Trans- Devarajan, Shantayann, Stuti Khemani, and Michael Walton. fers: Reducing Present and Future Poverty. Washington, DC: 2011. “Civil Society, Public Action and Accountability in World Bank. Africa.” Policy Research Working Paper 5733. World Bank, Foa, Roberto. 2013. “Household Risk Preparation Indices– Washington, DC. Construction and Diagnostics.” Background paper for the De Weerdt, Joachim. 2001. “Community Organizations in Rural World Development Report 2014. Tanzania: A Case Study of the Community of Nyakatoke, Frankel, Jeffrey A., Carlos A. Végh, and Guillermo Vuletin. 2013. Bukoba Rural District.” Economic Development Initiatives, “On Graduation from Fiscal Procyclicality.” Journal of Devel- Bukoba. opment Economics 100 (1): 32–47. De Weerdt, Joachim, and Kalle Hirvonen. 2013. “Risk Sharing and Gertler, Paul, and Jonathan Gruber. 2002. “Insuring Consumption Internal Migration.” Policy Research Working Paper 6429, against Illness.” American Economic Review 92 (1): 51–70. World Bank, Washington, DC. Ghosh, Atish R., Mahvash S. Qureshi, and Charalambos G. Tsan- DFID (Department for International Development, United King- garides. 2013. “Is the Exchange Rate Regime Really Irrelevant for External Adjustment?” Economics Letters 118 (1): 104–109. dom). 2005. “Why We Need to Work More Effectively in Frag- Gill, Indermit S., and Nadeem Ilahi. 2000. “Economic Insecurity, ile States.” DFID, London. Individual Behavior and Social Policy.” Working Paper 31522, Diamond, Douglas W., and Philip H. Dybvig. 1983. “Bank Runs, World Bank, Washington, DC. Deposit Insurance, and Liquidity.” Journal of Political Economy Goldin, Ian. 2013. Divided Nations: Why Global Governance Is 91 (3): 401–19. Failing and What We Can Do about It. Oxford, U.K.: Oxford Didier, Tatiana, Constantino Hevia, and Sergio L. Schmukler. University Press. 2012. “How Resilient and Countercyclical Were Emerging Gourinchas, Pierre-Olivier, and Maurice Obstfeld. 2012. “Stories Economies during the Global Financial Crisis?” Journal of of the Twentieth Century for the Twenty-First.” American Eco- International Money and Finance 31 (8): 2052–77. nomic Journal: Macroeconomics 4 (1): 226–65. Duflo, Esther. 2003. “Grandmothers and Granddaughters: Old- Graham, John D., and Jonathan B. Wiener. 1995. Risk versus Risk: Age Pensions and Intrahousehold Allocation in South Africa.” Tradeoffs in Protecting Health and the Environment. Cam- World Bank Economic Review 17 (1): 1–25. bridge, MA: Harvard University Press. Easterly, William R., and Luis Servén 2003. The Limits of Stabi- Gupta, Sunil. 2013. “The Mobile Banking and Payment Revolu- lization: Infrastructure, Public Deficits, and Growth in Latin tion.” European Financial Review (February–March): 3–6. America. Palo Alto, CA: Stanford University Press. Hale, Thomas. 2011. “A Climate Coalition of the Willing.” Wash- Economist. 2013. “Gizmos That Track Driving Habits Are Chang- ington Quarterly 34 (1): 89–101. ing the Face of Car Insurance.” February 23. Hallegatte, Stéphane. 2012a. “A Cost Effective Solution to Reduce Education International. 2009. “Education and the Global Eco- Disaster Losses in Developing Countries: Hydro-Meteorolog- nomic Crisis: Summary of Results of the Follow-up Survey.” ical Services, Early Warning, and Evacuation.” Policy Research Education International, Brussels. Working Paper 6058, World Bank, Washington, DC. Edwards, Sebastian. 2004. “Thirty Years of Current Account Im- ———. 2012b. “An Exploration of the Link between Develop- balances, Current Account Reversals, and Sudden Stops.” IMF ment, Economic Growth, and Natural Risk.” Policy Research Staff Papers 51 (Special Issue): 1–49. Working Paper 6216, World Bank, Washington, DC. Edwards, Sebastian, and Eduardo Levy Yeyati. 2005. “Flexible Ex- Hallegatte, Stéphane, Ankur Shah, Robert Lempert, Casey Brown, change Rates as Shock Absorbers.” European Economic Review and Stuart Gill. 2012. “Investment Decision Making under 49 (8): 2079–105. Deep Uncertainty: Application to Climate Change.” Policy Ehrlich, Isaac, and Gary S. Becker. 1972. “Market Insurance, Self- Research Working Paper 6193, World Bank, Washington, DC. Insurance, and Self-Protection.” Journal of Political Economy Han, Rui, and Martin Melecky. 2013. “Financial Inclusion for 80 (4): 623–48. Stability: Access to Deposits and Deposit Growth in the 2008 Eichengreen, Barry, and Douglas A. Irwin. 2010. “The Slide to Crisis.” Background paper for the World Development Report Protectionism in the Great Depression: Who Succumbed and 2014. Why?” Journal of Economic History 70 (4): 871–97. Heltberg, Rasmus, Paul B. Siegel, and Steen L. Jørgensen. 2009. Ersel, Hasan, and Fatih Ozatay. 2008. “Fiscal Dominance and “Addressing Human Vulnerability to Climate Change: Toward Inflation Targeting: Lessons from Turkey.” Emerging Markets a ‘No-Regrets’ Approach.” Global Environmental Change 19 Finance and Trade 44 (6): 38–51. (1): 89–99. Fafchamps, Marcel. 2011. “Risk Sharing between Households.” In Hnatkovska, Viktoria, and Norman Loayza. 2005. “Volatility and Handbook of Social Economics, vol. 1B, edited by Jess Behabib, Growth.” In Managing Economic Volatility and Crises: A Prac- Risk and opportunity 47 titioner’s Guide, edited by Joshua Aizenman and Brian Pinto, La Porta, Rafael, Florencio Lopez de Silanes, Andrei Shleifer, and 65–100. New York: Cambridge University Press. Robert W. Vishny. 1998. “Law and Finance.” Journal of Political Hoddinott, John, Mark Rosegrant, and Maximo Torero. 2012. Economy 106 (6): 1113–55. “Investments to Reduce Hunger and Undernutrition.” Paper La Porta, Rafael, and Andrei Shleifer. 2008. “The Unofficial Econ- prepared for the 2012 Global Copenhagen Consensus. omy and Economic Development.” Brookings Papers on Eco- Holzmann, Robert, and Steen Jørgensen. 2001. “Social Risk Man- nomic Activity 39 (2): 275–363. agement: A New Conceptual Framework for Social Protec- Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises tion, and Beyond.” International Tax and Public Finance 8 (4): Database, an Update.” Working Paper WP/12/163, Interna- 529–56. tional Monetary Fund, Washington, DC. Holzmann, Robert, David A. Robalino, and Noriyuki Takayama. Lane, Philip R., and Gian-Maria Milesi-Ferretti. 2012. “External 2009. Closing the Coverage Gap: The Role of Social Pensions and Adjustment and the Global Crisis.” Journal of International Other Retirement Income Transfers. Washington, DC: World Economics 88 (2): 252–65. Bank. Larrimore, J. T., and Brielle Sharkey. 2013. “Haiti Continues Hsieh, Chang-Tai, and Peter J. Klenow. 2009. “Misallocation and to Struggle Three Years after the Earthquake.” Council on Manufacturing TFP in China and India.” Quarterly Journal of Hemispheric Affairs, January 18, http://www.coha.org/ Economics 124 (4): 1403–48. haiti-continues-to-struggle-three-years-after-the-earthquake/. Hurwicz, Leonid. 1960. “Optimality and Informational Efficiency Levy, Santiago, and Norbert Schady. 2013. “Latin America’s Social in Resource Allocation Processes.” In Mathematical Methods in Policy Challenge: Education, Social Insurance, Redistribu- the Social Sciences, edited by Kenneth J. Arrow, Samuel Karlin tion.” Journal of Economic Perspectives 27 (2): 193–218. and Patrick Suppes, 27–46. Stanford, CA: Stanford University Loayza, Norman, and Jamele Rigolini. 2011. “Informal Employ- Press. ment: Safety Net or Growth Engine?” World Development 39 ILO (International Labour Organization). 2009. The Informal (9): 1503–15. Economy in Africa: Promoting Transition to Formality: Chal- Lundberg, Shelly J., Robert A. Pollak, and Terence J. Wales. 1997. lenges and Strategies. Geneva: ILO. “Do Husbands and Wives Pool Their Resources? Evidence Ilzetzki, Ethan, and Carlos A. Végh. 2008. “Procyclical Fiscal Policy from the United Kingdom Child Benefit.” Journal of Human in Developing Countries: Truth or Fiction?” Working Paper Resources 32 (3): 463–80. 14191, National Bureau of Economic Research, Cambridge, Macours, Karen, Norbert Schady, and Renos Vakis. 2008. “Cash MA. Transfers, Behavioral Changes, and Cognitive Development in IMF (International Monetary Fund). 2013. “Fiscal Adjustment in Early Childhood Evidence from a Randomized Experiment.” an Uncertain World.” Fiscal Monitor, October, IMF, Washing- Policy Research Working Paper 4759, World Bank, Washing- ton, DC. ton, DC. Jonas, Olga. 2013. “Pandemic Risk.” Background paper for the Mahul, Olivier, and J. David Cummins. 2009. Catastrophe Risk World Development Report 2014. Financing in Developing Countries: Principles for Public Inter- Kahneman, Daniel, and Amos Tversky. 1979. “Prospect Theory: vention. Washington, DC: World Bank. An Analysis of Decision under Risk.” Econometrica 47 (2): Maimbo, Samuel, and Martin Melecky. 2013. “Financial Policy 263–91. Formulation: Addressing the Tradeoff between Development Kaminsky, Graciela L., Carmen M. Reinhart, and Carlos A. Végh. and Stability.” Background paper for the World Development 2005. “When It Rains, It Pours: Procyclical Capital Flows and Report 2014. Macroeconomic Policies.” In NBER Macroeconomics Annual Martin, Will, and Kym Anderson. 2012. “Export Restrictions and 2004, edited by Mark Gertler and Kenneth Rogoff, 11–82. Price Insulation during Commodity Price Booms.” American Cambridge, MA: MIT Press. Journal of Agricultural Economics 94 (2): 422–27. Karlan, Dean, Robert Darko Osei, Isaac Osei-Akoto, and Christo- Maskin, Eric S. 1999. “Nash Equilibrium and Welfare Optimality.” pher Udry. 2012. “Agricultural Decisions after Relaxing Credit Review of Economic Studies 66 (1): 23–38. and Risk Constraints.” Working Paper 18463, National Bureau Miller, Merton H. 1986. “Financial Innovation: The Last Twenty of Economic Research, Cambridge, MA. Years and the Next.” Journal of Financial and Quantitative Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2010. Analysis 21 (4): 459–71. “The Worldwide Governance Indicators: Methodology and Muermann, Alexander, and Howard Kunreuther. 2008. “Self- Analytical Issues.” Policy Research Working Paper 5430, World Protection and Insurance with Interdependencies.” Journal of Bank, Washington, DC. Risk and Uncertainty 36 (2): 103–23. Kehoe, Timothy J., and Edward C. Prescott. 2007. Great Depres- Myerson, Roger B. 1979. “Incentive Compatibility and the Bar- sions of the Twentieth Century. Minneapolis: Research Depart- gaining Problem.” Econometrica 47 (1): 61–73. ment, Federal Reserve Bank of Minneapolis. Narayan, Deepa, Lant Pritchett, and Soumya Kapoor. 2009. Mov- Khokhar, Tariq. 2013. “Leveraging New Technology for Data- ing Out of Poverty: Success from the Bottom Up, vol. 2. Washing- Driven Risk Mitigation and Management: Selected Examples ton, DC: World Bank. and Summaries.” Background paper for the World Develop- OECD (Organisation for Economic Co-operation and Develop- ment Report 2014. ment). 2009. “Innovation in Country Risk Mangement.” Stud- Knight, Frank, 1921. Risk, Uncertainty, and Profit. Boston: Hough- ies in Risk Management, OECD, Paris. ton Mifflin Company. ———. 2011a. Managing Risks in Fragile and Transitional Con- Kraay, Aart. 2012. “How Large Is the Government Spending Multi- texts. The Price of Success? Paris: OECD. plier? Evidence from Lending by Official Creditors.” Quarterly ———. 2011b. “A New Deal for Engagement in Fragile States.” Journal of Economics 127 (2): 829–87. Paper prepared for the International Dialogue on Peacebuild- Kuznets, Simon. 1955. “Economic Growth and Income Inequal- ing and Statebuilding organized by OECD, November 29– ity.” American Economic Review 45 (1): 1–28. December 1. 48 WORLD DEVELOPMENT REPORT 2014 ———. 2012. Fragile States 2013. Resource Flows and Trends in a Schumpeter, Joseph A. 1942. Capitalism, Socialism and Democracy. Shifting World. Paris: OECD. New York: Harper and Brothers. Olson, Mancur, Jr. 1965. The Logic of Collective Action: Public Speer, Johanna. 2012. “Participatory Governance Reform: A Good Goods and the Theory of Groups. Cambridge, MA: Harvard Strategy for Increasing Government Responsiveness and University Press. Improving Public Services?” World Development 40 (12): Oviedo, Ana María, and Harry Moroz. 2013. “The Impacts of Risk.” 2379–98. Background paper for the World Development Report 2014. Svensson, Lars E. O. 2012. “Comment on Michael Woodford, Packard, Truman G. 2002. “Pooling, Savings, and Prevention: Mit- Inflation Targeting and Financial Stability.” Sveriges Riksbank igating the Risk of Old Age Poverty in Chile.” Policy Research Economic Review 2012 (1): 33–38. Working Paper 2849, World Bank, Washington, DC. Thomas, Duncan. 1990. “Intra-Household Resource Allocation: Pagés-Serra, Carmen, ed. 2010. The Age of Productivity: Transform- An Inferential Approach.” Journal of Human Resources 25 (4): ing Economies from the Bottom Up. Washington, DC: Inter- 635–64. American Development Bank. Thoresen, Stian H., and Angela Fielding. 2011. “Universal Health Pandey, Priyanki, Ashwini R. Sehgal, Michelle Riboud, David Care in Thailand: Concerns among the Health Care Work- Levine, and Madhav Goyal. 2007. “Informing Resource-Poor force.” Health Policy 99 (1): 17–22. Populations and the Delivery of Entitled Health and So- Tornell, Aarón, and Andrés Velasco. 1992. “The Tragedy of the cial Services in Rural India.” Journal of the American Medical Commons and Economic Growth: Why Does Capital Flow Association 298 (16): 1867–75. from Poor to Rich Countries?” Journal of Political Economy, Parker, Jonathan A. 2011. “On Measuring the Effects of Fiscal 100 (6): 1208–31. Policy in Recessions.” Journal of Economic Literature 49 (3): UNEP (United Nations Environment Programme). 2007. “A Suc- 703–18. cess in the Making: The Montreal Protocol on Substances that Paul, Bimal Kanti. 2009. “Why Relatively Fewer People Died? The Deplete the Ozone Layer.” UNEP, Nairobi. Case of Bangladesh’s Cyclone Sidr.” Natural Hazards 50 (2): ———. 2012. “Analysis of Formalization Approaches in the Arti- 289–304. sanal and Small-Scale Gold Mining Sector Based on Experi- Paulson, Anna L. 2000. “Insurance Motives for Migration: Evi- ences in Ecuador, Mongolia, Peru, Tanzania and Uganda: Peru dence from Thailand.” Unpublished manuscript, Kellogg Case Study.” UNEP, Nairobi. Graduate School of Management, Northwestern University, Université Catholique de Louvain. EM-DAT: The OFDA/CRED Evanston, IL. International Disaster Database. Université Catholique de Paxson, Christina, and Norbert Schady. 2007. “Does Money Louvain, Brussels, http://www.emdat.be. Matter? The Effects of Cash Transfers on Child Health and UNODC (United Nations Office on Drugs and Crime). Homicide Development in Rural Ecuador.” Policy Research Working Statistics (database). UNODC, Vienna, http://www.unodc Paper 4226, World Bank, Washington, DC. .org/unodc/en/data-and-analysis/homicide.html. Prasad, Naren, and Megan Gerecke. 2010. “Social Policy in Times von Neumann, John, and Oscar Morgenstern. 1944. Theory of of Crisis.” Global Social Policy 10 (2): 218–47. Games and Economic Behavior. Princeton, NJ: Princeton Uni- Premand, Patrick. 2013. “From Risk Coping to Risk Management: versity Press. Productive Safety Nets in Africa.” Background paper for the Wethli, Kyla. 2013. “Benefit-Cost Analysis for Risk Management: World Development Report 2014. Summary of Selected Examples.” Background paper for the Putnam, Robert D. 1993. “The Prosperous Community: Social World Development Report 2014. Capital and Public Life.” American Prospect 4 (13): 35–42. WHO (World Health Organization). 2013. World Health Statistics Ramcharan, Rodney. 2007. “Does the Exchange Rate Regime Mat- 2013. Geneva: WHO. ter for Real Shocks? Evidence from Windstorms and Earth- World Bank. 2005. World Development Report 2006: Equity and quakes.” Journal of International Economics 73 (1): 31–47. Development. Washington, DC: World Bank. Ramey, Valerie A. 2011. “Can Government Purchase Stimulate the ———. 2011. World Development Report 2011: Conflict, Security, Economy?” Journal of Economic Literature 49 (3): 673–85. and Development. Washington, DC: World Bank. Reinikka, Ritva, and Jakob Svensson. 2005. “Fighting Corruption ———. 2012a. “Financial Inclusion Strategies Reference Frame- to Improve Schooling: Evidence from a Newspaper Campaign work.” Prepared for the G20 Mexico Presidency. World Bank, in Uganda.” Journal of the European Economic Association 3 Washington, DC. (2–3): 259–67. ———. 2012b. Inclusive Green Growth: The Pathway to Sustainable Ribe, Helena, David A. Robalino, and Ian Walker. 2012. From Right Development. Washington, DC: World Bank. to Reality: Incentives, Labor Markets, and the Challenge of Uni- ———. 2012c. “Resilience, Equity, and Opportunity.” The World versal Social Protection in Latin America and the Caribbean. Bank 2012–2022 Social Protection and Labor Strategy, World Washington, DC: World Bank. Bank, Washington, DC. Royal Society 2009. “Geoengineering the Climate: Science, Gover- ———. 2012d. World Development Report 2013: Jobs. Washington, nance and Uncertainty.” Royal Society, London. DC: World Bank. Schneider, Friedrich, Andreas Buehn, and Claudio E. Montenegro. ———. 2013. Beyond the Annual Budget: Global Experience with 2010. “Shadow Economies All over the World: New Estimates Medium-Term Expenditure Frameworks. Washington, DC: for 162 Countries from 1999 to 2007.” Policy Research Work- World Bank. ing Paper 5356, World Bank, Washington, DC. ———. Global Findex (database). World Bank, Washington, DC, Schularick, Moritz, and Alan M. Taylor. 2012. “Credit Booms Gone http://data.worldbank.org/data-catalog/financial_inclusion. Bust: Monetary Policy, Leverage Cycles, and Financial Crises, ———. PovcalNet (database). World Bank, Washington, DC, 1870–2008.” American Economic Review 102 (2): 1029–61. http://iresearch.worldbank.org/PovcalNet/index.htm. Risk and opportunity 49 ———. Pensions (database). World Bank, Washington, DC, http:// World Tourism Organization. Yearbook of Tourism Statistics www.worldbank.org/pensions. (database). World Tourism Organization, Madrid, http:// ———. World Development Indicators (database). World Bank, statistics.unwto.org/en/content/yearbook-tourism-statistics. Washington, DC, http://data.worldbank.org/data-catalog/ Xu, Ke, David B. Evans, Kei Kawabata, Riadh Zeramdini, Jan world-development-indicators. Klavus, and Christopher J. L. Murray. 2003. “Household World Bank and United Nations. 2010. Natural Hazards, UnNatu- Catastrophic Health Expenditure: A Multicountry Analysis.” ral Disasters: The Economics of Effective Prevention. Washing- Lancet 362 (9378): 111–17. ton, DC: World Bank. Yang, Dean, and HwaJung Choi. 2007. “Are Remittances Insur- World Economic Forum. 2007. Global Risks 2007: A Global Risk ance? Evidence from Rainfall Shocks in the Philippines.” World Network Report. Geneva: World Economic Forum. Bank Economic Review 21 (2): 219–48. ———. 2012. The Global Competitiveness Report 2012–2013. Geneva: World Economic Forum. PART I Fundamentals of risk management Cyclone warning in Bangladesh. With good preparation, simple but effective interventions can save lives and avert damages. © Amir Jina/UN ISDR CHAPTER 1 Risk management can be a powerful instrument for development Risk and opportunity cent of annual gross domestic product (GDP) for countries in Latin Amer- When food prices spiked in 2008, ica, without even considering the riots broke out in more than a impact of crime on lost output be- dozen countries in Africa and cause of reduced investment and Asia. As food prices, particularly labor participation.5 Loss of em- bread prices, continued to rise ployment in countries as different in the Arab Republic of Egypt, as Argentina, Bulgaria, and Guyana Rashad Fahti, a factory worker, has not only lowered income and struggled to feed his wife and four consumption but has also reduced children on his monthly salary of people’s ability to find new work, wors- $34.1 A continent away, in Indonesia, ened social cohesion, and in some cases in- the village of Montei Baru-Baru lost more creased domestic violence.6 Whether adverse events than one in five residents—67 people—when it was come from systemic or idiosyncratic risks, they may hit by a large tsunami that followed an earthquake in destroy lives, assets, trust, and social stability. 2010.2 Globally, in the aftermath of the 2008–09 Risk is everywhere. While risks in some areas and global financial crisis, an estimated 53 million addi- some regions have diminished in recent years—in tional people will remain stuck in extreme poverty part because of improved macroeconomic and fi- by 2015 who otherwise would not have been so nancial management and better preventive health poor.3 The major economic crises and natural disas- care in developing countries—people in developing ters that have occurred in recent years underscore and developed countries alike continue to face a how vulnerable people are to systemic risks, which multitude of risks. Some types of risk—including cut across large groups of people—especially in de- those related to natural hazards, crime, environmen- veloping countries. tal challenges, and food prices—have become more Idiosyncratic risks, which are specific to individu- pronounced in recent decades (box 1.1). als or households, are no less important to people’s It is often when risks are mismanaged that the welfare. Losing a job or not finding one because of consequences become severe, turning into crises lack of skills, falling victim to disease or crime, or with dire results. Poor outcomes do not always reflect suffering a family breakup from divorce or forced bad risk management, however: extremely large and migration can all be overwhelming, particularly for unexpected shocks can overwhelm even the best vulnerable households. Households in Ethiopia preparation. Such crises have damaging effects be- whose members experienced serious illness, for ex- cause they not only affect people’s current living ample, had to reduce their consumption by almost conditions but also weaken their ability and willing- 10 percent and continued to be negatively affected ness to pursue new opportunities. Recognizing that a three to five years later.4 Health costs from high lev- negative shock can push them into destitution, bank- els of crime and violence amount to 0.3 to 5.0 per- 53 54 WORLD DEVELOPMENT REPORT 2014 B O X 1.1 A risky world: Risks vary over time and across regions In the past few decades, the patterns of risk that people have faced demics over the past three decades, although the incidence of have diverged. The incidence of natural disasters, food price shocks, shocks is generally lower than in developing countries. By contrast, and risks from climate change have increased substantially. By con- developing countries experienced fewer economic recessions in the trast, fewer risks have materialized in other areas—including mater- 2000s than in the 1980s and 1990s, but they faced an increasing inci- nal health, where the mortality rate has declined in all regions. dence of shocks in other areas, notably in epidemics in Sub-Saharan For some risks, progress has varied across regions. Developed Africa, and homicide in Latin America. countries have experienced more large recessions and health epi- a. Large recessionsa b. Incidence of natural disastersb 0.4 4.5 1981–1990 4.0 1981–1990 Proportion of decade 1991–2000 3.5 1991–2000 0.3 Annual average 2001–2010 3.0 2001–2010 in recession 2.5 0.2 2.0 1.5 0.1 1.0 0.5 0 0 OCED EAP ECA LAC MENA SAR SSA OCED EAP ECA LAC MENA SAR SSA c. Incidence of epidemicsc d. Maternal mortality Rate per 100,000 live births 0.8 800 1981–1990 1990 0.7 700 1991–2000 2000 Annual average 0.6 600 2001–2010 2010 0.5 500 0.4 400 0.3 300 0.2 200 0.1 100 0 0 OCED EAP ECA LAC MENA SAR SSA OCED EAP ECA LAC MENA SAR SSA e. Homicides f. Food prices and global sea level rise 25 70 20 Rate per 100,000 people, 1981–1990 Change in global mean sea level Standard deviation of 60 20 1991–2000 Food price volatility (right axis) 16 50 monthly index annual average Millimeters 2001–2010 40 15 12 30 10 20 8 10 5 4 0 0 –10 0 OCED EAP ECA LAC MENA SAR SSA 1992 1997 2002 2007 2012 Source: WDR 2014 team based on data from World Bank World Development Indicators (database); EM-DAT OFDA/CRED International Disaster Database; Nerem and others 2010; United Nations Office on Drugs and Crime Homicide Statistics (database); Food and Agricultural Organization Food Price Index (database). Note: Figures show the simple average across countries in each region. Organisation for Economic Co-operation and Development (OECD) countries in the figures are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. a. Large recessions are identified by following Barro and Ursúa 2012 and using a 5 percent decline in GDP per capita growth from peak to trough as a threshold. There were no large recessions in South Asia from 1991 to 2010. b. Natural disasters include droughts, earthquakes, floods, and storms. c. Epidemics refer to either an unusual increase in the number of cases of an infectious disease, which already exists in the region or population concerned, or the appearance of an infection previously absent from a region. Risk management can be a powerful instrument for development 55 ruptcy, or crisis, people may stick to technologies and Why is risk management relevant for livelihoods that appear safe but are also stagnant. development? Risk need not be harmful, however, and is not al- ways a burden. In many cases, people hoping to im- Risk management is an essential tool for develop- prove their standards of living may voluntarily take ment because people in developing countries are on risk. Indeed, risk taking is essential to the pursuit exposed to many risks, and an inability to manage of opportunity. But those opportunities may bring those risks can jeopardize development goals, in- their own risks. A country that opens its borders to cluding economic growth and poverty reduction. foster international integration and higher economic The prevalence of risk in everyday life in the devel- growth may also increase its exposure to interna- oping world is apparent in table 1.1, which presents tional shocks. An enterprise that upgrades to more data from household surveys that count the number advanced technologies to enhance its profitability of respondents who have been affected by various may also become more indebted and financially vul- shocks.7 A majority of households across a sample of nerable. Farmers who adopt new crops and use more developing countries report having been exposed to inputs in expectation of higher yields may face larger a shock in the preceding year, and a substantial pro- losses if rainfall is low. A rural household that mi- portion were exposed to more than one. The shocks grates to the city seeking better health care and edu- most frequently reported are natural hazards (such cation may expose its members to higher crime and as droughts and floods) and health risks. Rural areas less communal support. These actions are motivated tend to be more severely affected by shocks, espe- by the quest for improvement, but the results are sel- cially by droughts and floods. One exception is em- dom guaranteed. ployment shocks, which tend to be concentrated As the world changes, new opportunities and pos- in urban areas (possibly reflecting a greater share sibilities, as well as risks and complications, continu- of informal employment in rural areas). Middle- ally arise. Rejecting or ignoring change can lead to income countries (such as Peru) report a smaller stagnation and impoverishment. In contrast, em- share of people affected by shocks than do low- bracing change and proactively dealing with risks income countries. can open the way to sustained progress. Risk man- Surveys also show that people in developing coun- agement should therefore be a central concern at all tries feel susceptible to risk and are concerned by it. levels of society. By improving resilience, risk man- Figure 1.1 presents data from the latest World Values agement has the potential to bring about a sense of Survey, which asks respondents to provide a relative security and the means for people in developing judgment about risks that have materialized or that countries and beyond to achieve progress. concern them.8 Once again, regions with more low- TA B L E 1.1 Households in developing countries face many shocks Percentage of respondents reporting type of shock Afghanistana Indiab Lao PDR Malawi Peru Uganda Shocks Urban Rural Rural Urban Rural Urban Rural Urban Rural Urban Rural One or more 16.4 48.9 61.6 34.4 72.1 40.0 66.8 20.7 34.4 29.7 56.2 Two or more 8.7 39.2 23.4 11.9 36.1 12.7 40.4 1.4 1.9 5.6 15.6 Natural disasters (drought, flood) 10.6 42.2 57.3 5.6 36.0 10.4 47.2 2.6 21.5 19.9 52.1 Price shocksc 0.2 3.0 — 4.4 4.9 21.1 42.0 — — 1.7 3.2 Employment shocks 6.4 4.3 — 9.3 3.1 7.7 3.4 6.4 1.5 1.9 0.7 Health shocks (death, illness) 6.9 14.0 30.2 23.2 33.8 10.1 18.0 9.1 8.9 11.8 14.9 Personal and property crime 1.8 6.6 0.9 5.8 1.9 8.5 8.4 3.2 3.1 6.6 8.7 Family and legal disputes — — 1.9 0.0 0.9 1.7 4.3 0.7 0.3 — — Source: WDR 2014 team based on data from household surveys, various years 2005–11. Note: — = not available. a. The 2005 Afghanistan National Risk and Vulnerability Survey aims to be statistically representative at the national level. However, to the extent that it is difficult to access households most acutely affected by insecurity, the data may underestimate shocks for those households. Conversely, it shows that the risks faced by the households that were surveyed are not unlike those in other developing countries. b. Data for India are based on representative surveys from rural Karnataka, Madhya Pradesh, and Orissa. c. Price shocks refer to strong or unexpected changes in the price of agricultural outputs or inputs, or the price of staple food items. 56 WORLD DEVELOPMENT REPORT 2014 F I G U R E 1.1 Households in developing countries feel susceptible to risk and are concerned by it a. How many times in the past 12 months have you: Been without food? Felt unsafe from crime? Been without medicine Gone without income? 60 or health treatment? % responding “often” or “sometimes” 50 40 30 20 10 0 A nd th Am Asia d an d A nd th Am Asia d an d A nd th Am Asia d an d A nd th Am Asia d an d a a a a n be an n be an n be an n be an ric ric ric ric st a tin al a st a tin al a st a tin al a st a tin al a Ce Eur sia Ce Eur sia Ce Eur sia Ce Eur sia Af Af Af Af Ea th La ntr pe rib ca Ea th La ntr pe rib ca Ea uth La tr pe rib ca Ea uth La tr pe rib ca u u Ca ri Ca ri Ca ri Ca ri o o o o So So So So e e e e e e e e n n b. To what degree are you worried about: Losing your job or not Not giving your children War, a terrorist attack, or 100 finding one? a good education? civil war? % responding “very much” or “a great deal” 90 75th percentile 80 70 Median 60 50 25th percentile 0 As nd tin l As and an d As nd tin l As and an d As nd As nd an d a a a be an be an be an ric ric ric st a st a st a La tral e a Ce Eur a th Am ia ia th Am ia Ce Eur ia th Am ia i Af Af Af Ea uth ra e rib ca Ea th ra e rib ca Ea uth rib ca nt op nt op n op u Ca ri Ca ri Ca ri So So So e e e e e e Ce ur E tin La La Source: WDR 2014 team based on data from the World Values Survey, 2010–12. income countries are the most severely affected with crime. Forward-looking survey questions show (panel a). A large number of people in Sub-Saharan that a majority of people are concerned by risks that Africa report having gone without cash income, food, might emerge in the future (panel b). Indeed, in some or health treatment in the preceding year. Latin cases, the number of people who worry about future America, although a region with relatively more mid- risks exceeds the number who have been affected by dle-income countries, faces a particular problem that risk in the past: more than 50 percent of people Risk management can be a powerful instrument for development 57 in all regions express concern about losing their jobs, over 300,000 lives, one in 1991 claimed just under for instance, while far fewer report having gone with- 140,000, and one in 2007 claimed around 4,000.12 out income in the previous year. This differential un- This great reduction in casualties is a result of a na- derscores the very real psychological and emotional tionwide program to build shelters, along with im- toll that risk can have on people. proved forecasting capacity and a relatively simple but effective system for warning the population.13 Risk management saves lives Risk management averts damages and Failure to prevent and prepare for risk can have tragic prevents development setbacks consequences—often leading to widespread loss of life. Mortality is frequently higher in developing Crises can have substantial economic costs and lead countries and disproportionately affects the poor. to large-scale loss of property, infrastructure, and be- Developing countries tend to be more exposed to longings. In the past 15 years, a number of developing natural hazards, have less robust building structures, countries—including the Dominican Republic, Ec- and have low capacity to prevent disasters. One stark uador, Indonesia, Jamaica, Thailand, and Turkey— statistic sums this up: more people die from drought have faced banking crises with fiscal costs equal to in Africa than from any other natural hazard, whereas 20 percent of GDP or more.14 The value of damages virtually no one has died from drought in developed from natural disasters is often higher in developed countries in the past four decades.9 Similarly, the countries, where property and infrastructure are mortality rate from illness and injuries is far higher more costly to rebuild and repair. However, relative to in developing countries than developed countries. the size of their economies, the economic impact is The mortality rate for adults under age 60 is two and often much larger for developing countries.15 House- a half times higher for men and four times higher holds and firms may also be more acutely affected in for women in low-income countries than in high- developing countries because a smaller proportion of income countries, while the rate for children under their damages are insured.16 The costs of idiosyn- age 5 is almost twenty times higher.10 Diseases that cratic risks can also be high. Households in develop- affect the poor take the biggest toll: the mortality rate ing countries may spend up to 20 percent of their an- from preventable infectious diseases including lung nual income on the direct costs of treating a disease infections, diarrheal diseases, HIV/AIDS, and ma- such as tuberculosis, for example.17 laria is more than twenty times higher in low-income Large shocks can also cause serious long-term countries than in high-income countries.11 damage to human, social, and physical capital— The loss of life that results from crises can often be especially for the poor. When shocks are large relative avoided or reduced at moderate cost. For example, in to a country’s economy, they may have crippling January 2010, an earthquake measuring 7.0 on the long-term effects. For example, the hurricane that hit Richter scale occurred close to Port-au-Prince, Haiti; Honduras in 1998 is estimated to have caused total 230,000 people died. By contrast, a month later, a direct and indirect damages equal to 80 percent of much larger earthquake, measuring 8.8 on the Rich- GDP, leaving a legacy of substantially weaker public ter scale, struck off the coast of central Chile. While finances and current account deficits.18 At times the destruction was considerable, the total estimated effects from crises are permanent. A growing body of death toll was far lower: 525 fatalities. One significant research documents the role that shocks—above all, reason for the different outcomes is Chile’s enforce- health and weather shocks and economic crises— ment of building codes: buildings were more robust play in pushing households into poverty and keeping to ground tremors. When rebuilding following such them there.19 Following the 1999–2000 drought in events, the decision of how much to invest in better Ethiopia, households in the two lowest income quin- preparation depends in part on the probability of a tiles lost an estimated 60 to 80 percent of their assets; similar event occurring in the future. What is essen- the wealthiest quartile lost just 6 percent.20 Despite tial, however, is thinking in advance about the possi- poverty reduction, a substantial proportion of peo- bility of such an event and deciding how to prepare. ple in developing countries are vulnerable to falling Bangladesh provides a good example, where im- into poverty when they are hit by negative shocks proved preparation for natural hazards has dramati- (box 1.2). cally reduced loss of life from cyclones. In the past Proactive risk management can help prevent or four decades, three major cyclones of similar magni- lessen damages. For example, early warning systems tude have hit Bangladesh. A cyclone in 1970 claimed can curb the potential damage from natural hazards 58 WORLD DEVELOPMENT REPORT 2014 B O X 1. 2 While poverty has declined, many people around the world remain vulnerable to poverty In a significant achievement, poverty in developing countries has While all regions have reduced the shares of the population steadily declined over the past two decades. The share of people that live in or are vulnerable to poverty, progress has varied across living below $2.50 a day has dropped from 72 percent in 1990 to 50 regions. In Europe and Central Asia, where poverty was already rel- percent by 2010. Nonetheless, a substantial proportion of people atively low, substantial progress has been made in reducing vulner- remain vulnerable to poverty, with 89 percent of people in develop- ability to poverty. In East Asia and the Pacific, the rate of poverty ing countries living on less than $10 a day in 2010, compared with 94 was cut in half from 1990 to 2010, from 88 percent 40 percent, but percent in 1990.a While chronic poverty has declined significantly, 92 percent of the population continues to live in poverty or be vul- the large share of people in developing countries that live very close nerable to it. Similarly, in the South Asia and Sub-Saharan Africa to poverty highlights the potential for substantial increases in tran- regions, 98 percent of the population lived on less that $10 a day as sient poverty—which can have long-run consequences for people’s of 2010. health and livelihoods—when people are hit by negative shocks. Share of population in developing countries living in or vulnerable to poverty 1990 2010 100 80 % of population 60 Population vulnerable to Population poverty 40 in poverty (<$10 a day) (<$2.50 a day) 20 0 All East Asia Europe Latin Middle South Sub- developing and and America East and Asia Saharan countries Pacific Central and the North Africa Asia Caribbean Africa Source: WDR 2014 team based on data from World Bank PovcalNet (database). a. $1.25 a day is a widely used measure of extreme poverty. However, $2.50 a day is considered a more relevant measure of extreme poverty for some regions, such as Latin America and the Caribbean. The $10-a-day measure is an approximate threshold for measuring vulnerability to poverty across regions, which cor- relates with asset holdings. The measure is based on studies suggesting that, for some regions, income of at least $10 a day is necessary to achieve the degree of economic stability and resilience to shocks that characterizes middle-class households. By contrast, those living below $10 a day are vulnerable to poverty, in the sense that they face the possibility of remaining in poverty or easily entering into poverty. See, for example, López-Calva and Ortiz-Juarez 2011; Ferreira and others 2013. by moving people away from the areas likely to be although not greater than that in high-income coun- most affected and by preparing buildings and infra- tries, as had been the case in previous global crises. structure in advance. Forecasting capacity can also be Low-income countries, which were less exposed to helpful in minimizing the damage from other natu- financial markets, witnessed a more moderate de- ral hazards. The introduction of seasonal forecasting cline in their GDP growth, both relative to other models in the Philippines, for instance, helped farm- countries and to their own past experience. In both ers adjust their agricultural production plans ahead cases, sound macroeconomic management before of the El Niño drought in 2002–03.21 Similarly, im- the crisis—including better-controlled inflation, proved macroeconomic management can reduce the smaller fiscal and current account deficits, and in- severity of economic shocks by creating fiscal and creased international reserves—created a buffer that monetary buffers to help lessen the impact of shocks. allowed countries to use countercyclical policies in In the 2008–09 financial crisis, middle-income coun- response to their growth downturns, contributing to tries experienced a sharp decline in their GDP a much quicker recovery compared with previous growth, similar to that in high-income countries— global financial crises (see chapter 7).22 Risk management can be a powerful instrument for development 59 Risk management unleashes opportunity riskier, higher-yield types of production.29 Similarly, farmers in India who were offered rainfall insurance No less important than people’s concern about nega- shifted their production from drought-resistant to tive shocks is their desire to improve their circum- higher-yield crops and also made greater investments stances. Some 68 percent of people in developing in fertilizer, hired labor, and other inputs (figure countries said that a high level of economic growth 1.2a).30 Each year in rural Bangladesh, households should be the first priority for their country; 70 per- face the prospect of famine and increased poverty in cent said work was “very important” in their life.23 A the lean season, when there is little agricultural work quarter of people in developing countries said they and grain prices are high, yet relatively few people would like to permanently emigrate from their coun- migrate to urban areas to find employment for those try: an aspiration to improve their standards of living few months. The offer of support, in the form of a was cited as the most important motivator by a sub- modest loan or grant, led to substantially higher rates stantial majority.24 of temporary migration, with corresponding in- While the desire for improvement is strong, the creases in consumption for remaining members of changes needed to bring improvement about can en- the family. Significantly, seasonal migration also re- tail substantial risk. Sometimes fear of risk—which is mained higher in subsequent years, when an incen- often particularly acute for the poor—means that tive to migrate was not provided, suggesting an im- productive opportunities are not pursued. For ex- portant self-learning element that can help reduce ample, low-income households in developing coun- fear of risky activities (figure 1.2b). tries disproportionately choose to grow low-risk In many instances, risk management may provide crops, which are also low return, thereby perpetuat- a means both to increase economic returns and re- ing poverty.25 Farmers may choose not to use fertil- duce the propensity for crises, especially in develop- izer because they fear negative shocks such as low ing countries. For example, improvements in educa- rainfall or diminished demand, thus preferring to tion not only increase productivity and income but retain savings as a cushion rather than investing in also enhance risk management, because highly skilled intermediate inputs.26 Similarly, fear of failure may workers are less likely to be unemployed or under- inhibit rural households from taking opportunities employed.31 Investments in nutrition and preventive to migrate temporarily during the lean season, de- health allow people to be more productive, while also spite potentially large gains from doing so.27 reducing communities’ susceptibility to illness and Beyond their own forgone opportunity, individu- disease.32 A more flexible enterprise sector is both als’ choices to avoid risk because they fear loss can more productive and better able to respond to eco- also have national repercussions when aggregated. nomic shocks (chapter 5). Policies to substantially For example, evidence suggests that aversion to reduce high inflation in developing countries can risk—which is especially strong in developing coun- help reduce volatility as well as enhance long-run tries, where people often live close to subsistence economic growth.33 These win-win examples illus- levels—can explain two-thirds of the difference in trate the potential to simultaneously manage risk and the use of intermediate inputs (such as fertilizer) be- enhance development: there need not necessarily be a tween developing and developed countries, amplify- trade-off between resilience and growth. ing differences in agricultural and overall productiv- ity by as much as 50 and 80 percent, respectively, compared to a model without the risk of agricultural What does risk management entail? shocks.28 Similarly, reluctance to undertake innova- This section discusses in detail two important com- tion in the absence of risk management tools can ponents of the analytical framework introduced in have implications for national economic growth, the overview. First, it explores the process by which especially because innovation often brings positive decisions about risk are made, and the environment spillover effects, with the benefits accruing to society in which risks and opportunities arise. Then it de- and not just the individual. scribes what a strong risk management strategy looks Risk management can provide a means for people like—encompassing actions to both prepare for and to better manage the potential downside from taking cope with risk. The obstacles that often make such a on risk—thereby fostering opportunity and ulti- strategy difficult to achieve in practice are explored mately reducing poverty. For example, by mitigating in chapter 2. The discussion that follows attempts to the potential for loss, insurance enabled farmers in provide a unifying structure and set of terms to dis- Ghana to both invest more in production and pursue cuss risk and risk management across different areas 60 WORLD DEVELOPMENT REPORT 2014 F I G U R E 1. 2 Risk management tools can help people pursue opportunity a. Rainfall insurance in India b. Seasonal migration in Bangladesh Effect of rainfall insurance on the amount of: Migration rate Change in consumption 60 for family at origin Fertilizer 50 40 Seeds Percent 30 Hired labor 20 Borrowing 10 for inputs 0 50 30 10 10 30 50 2008 2009 Food Nonfood Caloric expenditure expenditure intake % of households % of households that invest less that invest more Incentivized group Non-incentivized group Source: WDR 2014 team based on data from Cole, Giné, and Vickery 2013 (panel a) and Bryan, Chowdhury, and Mobarak 2012 (panel b). Note: The bars in panel a represent the self-reported investment decisions of 749 farmers who were provided rainfall insurance in a semi-arid area of India. of expertise in development—although the termi- Whether a risk management option is ultimately nology and use of concepts may differ from applica- judged to be excessively costly depends in part on the tions in some disciplines (glossary 1.1). relative prices of different actions and the availability of alternatives—particularly on whether people are able to rely on family, the state, or other networks for Managing risk under uncertainty support, and on whether they can reasonably expect The world is constantly changing, and with change to borrow money if necessary. The perceived benefit comes uncertainty. Amid this uncertainty, peo- of risk management depends, in turn, on assess- ple must consider different options for ments of the likely size of a future shock, how to prepare for risks they may the probability of it occurring, and how face. They must decide the proper much people care about the future balance between taking on risk Risk management can (when the benefits of risk manage- and preparing for it, or acting both increase economic ment would accrue) compared only after a shock has occurred. returns and reduce the with the present (when the costs In some instances, the choice of propensity for crises: there are undertaken). Their personal how much risk to take on will be preferences (including their toler- need not be a trade-off affected by a trade-off between ance and appetite for risk) will risk and return—by reducing the between resilience and also affect their decision. These riskiness of an undertaking, people growth. preferences can be influenced by sev- may also diminish the potential re- eral factors, including income and cul- turn they can get. That is often the case tural norms. At very low levels of income, for financial investments, for example. In many for example, risk aversion might be synonymous cases, however, risk management can be growth en- with loss aversion (box 1.3). hancing, allowing people both to decrease the risk they face and to improve their return. The choice of Understanding the environment in which risks which actions to take then turns on how the up- and opportunities arise front cost of preparation compares to the likely ben- efit,34 as well as any potential obstacles to people’s Change can generate both negative shocks (such risk management—including lack of information as natural hazards or financial crises) and positive and economic constraints. shocks (such as resource booms or improvements in Risk management can be a powerful instrument for development 61 G L O S S A R Y 1.1 Terms related to risk management Risk The possibility of loss. It can be imposed from outside or taken on voluntarily in the pursuit of opportunities. Opportunity The possibility of gain. It can be regarded as the upside of risk. Systemic risk Risk that is common to most members of an entire system. Idiosyncratic risk Risk that is specific to some members of a system. Risk management The process that involves confronting risks, preparing for them (ex ante risk management), and coping with their effects (ex post risk management). Shock A change in the world that may be positive or negative and that may occur gradually or suddenly. Exposure The external environment that determines the shocks to which a system is subject. Vulnerability A high susceptibility to loss from negative shocks resulting from a system’s exposure, internal conditions, and risk management. Resilience The ability of a person or system to recover from negative shocks while retaining or improving their functioning. Crisis A situation in which the adverse outcomes from risk become so severe and generalized that the functioning of the system is threatened. Uncertainty The situation of not knowing what the outcome will be. Source: WDR 2014 team. technology). Whether risk is imposed or taken on earthquake measured 7.8 on the Richter scale). In voluntarily, the impact of shocks can be amplified addition, people’s external environment may expose or reduced depending on people’s external environ- them to earthquakes to a greater or lesser extent ment, their internal conditions, and their risk (they may live in a densely populated region prone management. to earthquakes, for example). Their internal condi- Consider the tragic case of Sichuan province in tions (age, health, education, and so on) may also China where 69,000 people, including thousands of play a role. Their own preparation matters. For ex- children, died following a large earthquake in May ample, do children in schools practice emergency 2008. What factors contribute to a death toll in an responses to earthquakes? Once a shock occurs, the event like this? Clearly, the intensity of the initial outcome also depends on people’s ability to cope— shock can have a major influence (in this case, the on how quickly emergency responders are able to B O X 1. 3 When risk aversion becomes loss aversion: A view from utility theory In economic models, agents’ risk aversion is represented by the cur- By contrast, people who are very poor and credit constrained vature of their utility function. Agents who are risk averse, for exam- may be particularly fearful of risk. For these people, risk aversion ple, have concave utility functions: they get greater utility from out- essentially becomes loss aversion: the possibility of loss weighs comes that occur with certainty than from outcomes that have the much more heavily in their minds than the possibility of gain. In same average value but are uncertain. When constructing models, these circumstances, it may be better to consider utility functions economists must make a choice about what type of utility function where (relative) risk aversion is not constant with income. In Stone- to use. Constant (relative) risk aversion utility functions are com- Geary utility functions, for example, utility depends not only on con- monly used and relatively easy to work with. One drawback of this sumption but on the difference between current consumption and class of utility functions, however, is the characteristic that agents’ a minimum level of subsistence. Loss aversion has a strongly dis- risk aversion to (proportional) variations in consumption does not couraging effect on people’s willingness to pursue new ventures change with their income. It may be reasonable to think, however, and opportunities. It can be mitigated, however, by access to finan- that agents with different levels of income but otherwise similar cial markets (in the form of insurance and credit) and the availability characteristics have different preferences for risk. For example, peo- of safety nets, especially in times of distress. ple with higher levels of income may feel they have less to lose and be willing to take on more risk. Source: WDR 2014 team. 62 WORLD DEVELOPMENT REPORT 2014 get to the scene, and what equipment they have, for shocks depend on the external environment and example. In Sichuan, questions have been raised people’s internal conditions—and, to a considerable about whether poor enforcement of building codes extent, on their preparation for risk and how they in rural schools made the disaster worse than it cope once a risk has materialized. might have been. Coping was also made more diffi- While this discussion has suggested that risks cult by damage to major highways in the region, and are propagated in a linear fashion, in reality the re- by landslides and mudflows, which made it hard lationships represented in the risk chain involve to access the affected areas after the earthquake several feedback effects (see diagram 1.1). The out- occurred. come of past shocks may affect people’s exposure By contrast, the discovery of valuable natural re- to shocks. For example, a family that moves to an sources provides an example of a positive shock. urban area following a severe drought will be ex- Paradoxically, this windfall is sometimes seen as a posed to a whole new set of potential shocks. The kind of curse, although it need not be—as Chile outcome of past shocks may also affect the propen- has shown. Chile has been a major producer of sity for future shocks to occur. Contracting HIV/ copper for the past century, and copper continues AIDS makes the risk of tuberculosis much more to account for more than half of all exports. Chile’s likely, for instance. While outcomes in a small system, ability to benefit from this resource has depended such as a household, are unlikely to have large effects in part on external conditions, including the level on their own, they may have a considerable effect and volatility of the world copper price. Internal when they are sufficiently correlated across systems. conditions have also mattered. In the postwar While one household with substantial debt may period, relatively weak technological ca- not seem too problematic, for instance, pacity and a shortage of high-skilled household indebtedness can be a source workers arguably hampered the The goal of risk of instability at an aggregate level performance of domestic copper when many people are overlever- management is to producers.35 Over time, improve- aged. People’s risk management ments in risk management have mitigate the losses and can also greatly affect the propen- also played an important role. improve the benefits sity for future shocks. The use of While imprudent spending by that people experience insecticide-treated bed nets can elites may partly explain why substantially reduce the number of when they face risk and strong copper production did not mosquitoes in an area, for example, feed through to stronger economic opportunity. decreasing the risk of malaria; man- development for much of the twenti- aging soil erosion reduces the risk of eth century,36 the government began to landslides; and effective macroprudential take an active and positive role in managing copper regulation can reduce the likelihood of future finan- revenues follow-ing a renewed increase in Chilean cial crises (see chapter 6). copper production in the 1980s and 1990s. It now uses a fiscal rule linked to the price of copper, con- The goal of risk management tributing to the government’s Economic and Social Stabilization Fund and Pension Reserve Fund (see In this context, risk is defined as the possibility of chapter 7). Today, Chile is seen as a leading example loss. Even when risk is taken on in the pursuit of op- of managing a natural resource responsibly for the portunity, the results are not guaranteed: risk thus benefit of its citizens. implies a possibility of loss. By contrast, opportunity More generally, the interactions between shocks, is defined as the possibility of gain (it can be regarded the external environment and internal conditions, as the upside of risk). People’s exposure to risk is de- approaches to risk management, and outcomes can termined by their external environment. For exam- be represented by a risk chain (diagram 1.1).37 The ple, whether a house is exposed to the risk of coastal source of any risk is the initial shock. Shocks, either flooding depends on its location. Some people may positive or negative, may occur suddenly (such as be vulnerable—that is, especially susceptible to losses natural hazards), or gradually (such as demographic from negative shocks—as a result of their exposure, transitions or technological changes). Some shocks internal conditions, and risk management.38 For ex- are systemic, while others are idiosyncratic, affecting ample, a highly leveraged financial institution that only certain individuals or households. As high- has taken high-risk positions without counterbal- lighted in the examples above, the outcomes of ancing hedges may be vulnerable to an economic or Risk management can be a powerful instrument for development 63 The risk chain: The nature and extent of outcomes depend on shocks, D I A G R A M 1.1 exposure, internal conditions, and risk management External environment Risk ks Shocks management anagement Outcomes Internal conditions Source: WDR 2014 team. Note: The feedback arrows in the risk chain diagram represent the potential for the outcomes of past shocks to affect exposure and internal conditions, as well as the propensity for future shocks. Similarly, the effectiveness of people’s risk management can significantly affect the nature of and propensity for future shocks. financial shock. Likewise, a poor household with few Risk management requires preparation assets may be especially vulnerable to a food price and coping shock. Resilience is characterized by people’s ability to To achieve that goal, risk management needs to com- recover from negative shocks while retaining or bine the capacity to prepare for risk with the ability improving their functioning. One outcome of in- to cope once a risk has materialized. Preparation (or creased resilience is likely to be reduced volatility ex ante risk management) includes a combination of household consumption and income growth (box of three actions that can be taken in advance: acquir- 1.4). A considerable body of the emerging literature ing knowledge (gathering information and making on risk in a development context emphasizes how judgments about risk); obtaining protection (to in- resilience to negative shocks can be increased through fluence the likelihood and magnitude of risk); and better risk management. However, risk management obtaining insurance (to transfer resources between also has an essential role in increasing prosperity by good and bad periods). Risk cannot—and should helping people and countries successfully manage not—be eliminated altogether, however, and excep- positive shocks. Indeed, successfully managing posi- tional shocks can always occur. Thus, once a risk (or tive shocks is a critical part of increasing people’s re- an opportunity) materializes, people need to take ac- silience to negative shocks over time. Ignoring this tion to cope with what has occurred (that is, engage aspect is particularly unsatisfactory in the context of in ex post risk management) (diagram 1.2).40 Coping chronic poverty, because it suggests that the best that actions include updating knowledge and then de- a poor household can achieve through risk manage- ploying any insurance and protection. ment is to not become any poorer over time.39 In- stead, the goal of risk management should be to in- Knowledge crease the benefits as well as decrease the losses that Because people face uncertainty when they confront people experience when they face risk. risk, increased knowledge is an essential component 64 WORLD DEVELOPMENT REPORT 2014 B O X 1. 4 Developing countries have increased their resilience over time One feature that is likely to be characteristic of resilient households is most relevant for household welfare. Developed countries were the the smoothness of their consumption and income growth over time. most resilient in this period, with lower volatility in both household A household with good preparation and diversified assets will have consumption and income growth than in developing countries. In less income volatility and will be able to smooth consumption when addition, consumption growth was relatively more stable than faced with shocks. By contrast, a household that is not resilient is income growth in developed countries. By contrast, most develop- more likely to have large drops (or increases) in consumption and ing countries have struggled with both unstable consumption and income. income.b Panel a illustrates the volatility of consumption and income The stability of consumption growth has changed over time growth per capita from 2000–11 around the world.a Countries closer (panel b). Countries that are located to the right of the 45° line had to the axis origin had more stable income and consumption growth. more stable per capita consumption growth during the 2000s than Moreover, in countries that are located to the right of the 45° line, in the 1990s. Although volatility of consumption remains high over- per capita consumption growth was more stable than income all, several developing countries across all regions have become growth—an important characteristic, given that consumption is more resilient in the past decade. a. Volatility of income and consumption growth b. Volatility of consumption growth is higher in developing countries has stabilized over time 45 45 SD of HH consumption per capita 32.0 SD of HH consumption per capita 32.0 16.0 16.0 growth, 2000–11 growth, 2000s 8.0 8.0 4.0 4.0 2.0 2.0 1.0 1.0 0.5 0.5 0.5 1.0 2.0 4.0 8.0 16.0 32.0 0.5 1.0 2.0 4.0 8.0 16.0 32.0 SD of GDP per capita growth, 2000–11 SD of HH consumption per capita growth, 1990s OECD East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: The data are presented on a logarithmic scale. Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. SD = standard deviation; HH = household; GDP = gross domestic product. a. Since accurate household consumption and income data are not widely available across countries, household final consumption expenditure per capita and GDP per capita provide imperfect proxies. Volatility is measured by the standard deviation of the respective growth rates over the period. b. This finding is consistent with the business cycle literature, which finds that output volatility is significantly higher in developing countries than in developed countries. See Agénor, McDermott, and Prasad 2000. of risk management. Increased information about education, which are often crucial to attaining better risk can help people better understand the nature standards of living. Consider a family that is contem- and likelihood of risks they may face, thus reducing plating moving to a new location where the parents uncertainty. Knowledge of risk goes beyond simply may be able to get better jobs, but which is also prone obtaining information: knowledge also involves us- to malaria. The parents may want to learn about ing that information to assess potential risks and training that could improve their chances of getting then deciding how to act. Furthermore, not only can work in the new area and, if they decide to undertake better knowledge of risk help people prepare for the opportunity, they could learn about the risk of negative shocks, it is also relevant to the management malaria and then decide what actions to take to pro- of positive shocks. For example, better knowledge tect themselves. In this way, confronting risk can im- can inform decisions about investments in skills and prove knowledge by creating a richer understanding Risk management can be a powerful instrument for development 65 D I A G R A M 1. 2 The interlinked components of risk management Insurance To transfer resources across people and over time, from good to bad states of nature Knowledge To understand shocks, internal and Coping external conditions, and potential To recover from losses and outcomes, thus reducing uncertainty make the most of benefits Protection To reduce the probability and size of losses and increase those of benefits Preparation Coping Source: WDR 2014 team. of the potential consequences and informing future regulations by providing regulatory stability and action. policy predictability. Health care is one area that has Even with increased knowledge, many decisions greatly benefited from increased provision and qual- must be made with imperfect information. In most ity of information on potential health risks, as well as cases, therefore, although people know what the pos- improved knowledge of how to manage those risks. sible outcomes are and can assess their probabilities, More broadly, new technologies can further help im- there is still uncertainty about what will actually hap- prove knowledge of potential shocks and inform re- pen. Beyond this, however, some areas (such as nu- sponses to them (box 1.5). Increased information is clear energy safety or climate change) are affected by not a sufficient condition for better risk manage- “deep uncertainty,” where either very little is known ment, however, especially when people have diffi- or even the experts cannot agree on underlying trends culty interpreting information or acting on it (see and possible outcomes, let alone the probabilities chapter 2). surrounding them. The presence of uncertainty re- quires devising strategies that can successfully man- Protection age risks in a wide variety of scenarios (see chapter 2). Protection, in turn, includes any actions that lower Acquiring knowledge (and thereby reducing un- the probability and size of negative outcomes or in- certainty) depends not only on the information that crease the probability and size of positive outcomes. people can access themselves but also on the quality Thus protection includes action to prevent negative of information that is provided by other social and shocks from occurring or to mitigate their impact economic systems. Indeed, because uncertainty can (especially for negative shocks that cannot be pre- be a substantial obstacle to people’s risk manage- vented)—or, in some cases, both. Similarly, it in- ment, public policy has an important role in improv- cludes actions to increase the propensity for positive ing access to, and presentation of, information on shocks and gains from them. Protection can be self- risk, particularly through the provision of timely and provided, purchased from the market, or provided reliable data on risk. Governments can also reduce publicly by the community or the state. Continuing the uncertainty associated with their policies and with the example of malaria, family members could 66 WORLD DEVELOPMENT REPORT 2014 B O X 1. 5 Leveraging new information and communication technologies for risk management New technologies that help to capture, assess, and communicate mobile phones. Farmers receiving information saw increases in data more quickly and with greater reach have become more their income of 10 to 30 percent. widely accessible over the past five years. Mobile phones, aerial • Respond to risk more quickly. To assist recovery following a 2008 and satellite imagery, social networks, and online platforms for col- earthquake in Rwanda, citizens in unaffected parts of the country lective and distributed work can improve risk management by used mobile phones to transfer “mobile money” to people in the enabling people to: affected area. • Evaluate the effectiveness of risk management and adjust their strate- • Be better informed of risks. To help pinpoint people in need of assis- gies accordingly. To assess whether residents heeded warnings tance following the 2010 earthquake in Haiti, the Ushahidi crowd- from the Mexican government to remain at home during the H1N1 sourcing platform enabled citizens to report incidents via text flu outbreak in 2009, researchers have tracked population move- messages to volunteers who tracked and mapped areas affected ments using data from mobile phone towers. by the disaster (map). • Better evaluate evolving risks. To identify areas hardest hit by Hurri- New technologies can make new types of information available, cane Sandy in the United States in 2012, more than 3,000 online improve its timeliness, provide more flexible ways of handling infor- volunteers helped inform official responders by conducting rapid mation, and cut costs significantly. New technologies may also bring damage assessments using fresh aerial imagery of houses in new challenges, however, including concerns about privacy, difficul- affected areas. ties in judging the validity of information on the Internet, and a risk of • Better manage risks in pursuit of opportunity. To improve their information being used for violent or oppressive ends. The challenge response to changes in agricultural prices and demand, farmers for policy makers is to leverage the benefits of new technologies in Ghana can receive specific market information through their while respecting privacy and protecting sensitive information. Disaster mapping in Haitia étion-Ville Pétion-Ville Source: Khokhar 2013 for the WDR 2014. a. WDR 2014 team based on data from Ushahidi, http://community.ushahidi.com/index.php/deployments/. The red circles in the map depict localities in need of support from emergency responders. use bed nets or wear long-sleeved clothing to avoid government authorities, such as spraying insecticide being bitten by mosquitoes (self-protection). They (state protection). could purchase protection from the market, such Different forms of protection may be relatively as paying to treat the family’s house with insect- more effective for different types of risk. Self- repelling paint (market protection). They could join protection is mostly effective for frequent risks that with local community members in draining standing have a relatively low impact, although it can also be water sources (community-based protection). Fi- relevant for some risks with potentially large losses nally, the family could benefit from activities by local (such as driving safely, or using a condom to protect Risk management can be a powerful instrument for development 67 against sexually transmitted diseases). Investments purchased by people who are the most prone to risk in human capital, especially health and education, (adverse selection). In addition, the cost of recovery are an important means of helping people improve from some (infrequent) shocks may be so large that their self-protection. However, individuals may not the likely payouts in the event of a shock would be able to do much to protect themselves against greatly exceed the amount that can reasonably be some risks with very large impacts (which also tend collected from insurance premiums. These factors to be less frequent), especially systemic risks. Such may make the price of market insurance prohibi- risks often require assistance from communities or tively expensive or eliminate market insurance for the state. State protection for systemic risks such as specific risks altogether. natural hazards or economic crises includes physical In developing countries, new technology and investments (dikes, sea walls, better roads, sanitation, distribution networks have contributed to sub- and so on), as well as investments in early warning stantial growth in formal insurance in recent years, indicators and contingency planning to improve but access still remains fairly limited overall. That emergency response. places a larger burden on self-insurance, which is often pursued through relatively costly and inef- Insurance ficient means, such as holding durable assets (like To the extent that protection cannot completely jewelry) that can be sold in the event of a shock. eliminate the risk of negative outcomes, insurance Many households in developing countries also can help to cushion the blow from adverse shocks. participate in informal risk-sharing schemes, but Insurance includes any instruments that trans- coverage is often incomplete.41 Given the lack fer resources between good and bad of market insurance in many develop- times (savings, formal insurance con- ing countries, the state can help fur- tracts, loans, credit lines, hedging ther improve access to, and use of, With good preparation, instruments), as well as means of financial risk management tools, transferring resources to those only minimal coping may be and in some cases directly inter- especially in need in bad times needed to recover quickly— vene, by, for example, providing (social safety nets, community leaving more resources credit subsidies and guarantees support, or other risk-pooling available for investment (see chapter 6). While being mechanisms). It can be self- in risk management and careful not to crowd out private provided; achieved by pool- initiatives, the state may provide reducing vulnerability to ing risk with others (formally some forms of insurance directly, through a market, or informally); future shocks. notably by using public resources to or provided by the state. Continu- provide safety nets for the most vulner- ing with the malaria example, the family able (see chapter 3). Communities and the could save to provide a financial buffer in case of state can also provide support for extreme shocks illness (self-insurance) or buy health insurance to (such as large natural hazards or financial crises) that cover potential treatment costs (market insurance). are not covered by market insurance. Public insurance (by the community or the state) As strategies to manage risk, insurance and pro- might include building social networks that could tection may create synergies or require trade-offs. A provide support to the family in case of illness, of- considerable body of economic literature is premised fering medical treatment in subsidized state hos- on the view that insurance reduces people’s incentive pitals, and providing unemployment insurance if to try to prevent bad states of nature from occurring workers in the family contract malaria. (in other words, insurance leads to moral hazard). To Self-insurance in the form of savings is an effec- the extent that moral hazard occurs, insurance and tive way to insure against frequent shocks that have a protection act as substitutes for each other. An alter- relatively small impact, but savings can quickly be native view, however, is that protection and insur- exhausted as the size of potential losses grows. Mar- ance may sometimes be complements. That happens ket insurance can thus provide a useful means to in- when the steps that people take to attain protection sure against larger shocks. However, market insur- are observable to insurers, who can then vary the ance does not offer complete coverage for all types of premiums they charge different individuals (for ex- risk for several reasons. Insurers may fear that having ample, lower prices for people who do not smoke insurance will make people reckless (a problem de- than for people who do). In such cases, protection scribed as moral hazard) or that insurance will be can make it cheaper to insure against adverse out- 68 WORLD DEVELOPMENT REPORT 2014 B O X 1. 6 Protection and insurance can provide complementary means of managing risk The relationship between the probability and severity of risk can be Self-insurance (savings) helps primarily to cover small losses. characterized as downward sloping: small losses tend to be frequent, Market insurance is better for risks that are less frequent but have while large losses are rare. Such a monotonically decreasing proba- larger losses, while community and state support may be needed bility density function is represented in the figures below. for risks that are so large that market insurance is not provided. Protection can decrease the likelihood of very severe losses, Increased protection can complement insurance by increasing the thereby increasing the probability of routine losses relative to availability of market insurance—both by reducing the cost of insur- extraordinary losses (pivot from the solid to the dashed line in panel ing for bad outcomes and increasing the supply of insurance for a). While self-protection is important for some risks with potentially some risks that were not previously covered (pivot to dotted line large losses, public protection (either by the community or state) and expansion of market insurance in panel b), in turn reducing the may be particularly important for reducing the probability of sys- burden on self-insurance and the need for community and state temic risks with severe consequences (additional pivot to dotted support. line in panel b). a. Protection increases the likelihood of normal times b. Protection can increase the availability of insurance Private protection Public protection Probability of loss Private and public protection Probability of loss Self- Market Public insurance insurance insurance Routine Severity of loss Extraordinary Routine Severity of loss Extraordinary Source: WDR 2014 team based on Ehrlich and Becker 1972; Gill and Ilahi 2000. comes by reducing the likelihood of their occurrence Brazil and Mexico—and social reforms in several and may also induce insurers to insure some large other developing countries—have improved prepa- shocks that were previously not covered because the ration for and resilience to shocks.44 Growing num- risk was too great. Thus protection and insurance to- bers of children are immunized against infectious gether may provide complementary means of man- diseases; and households in developing countries in- aging risk (box 1.6).42 While this aspect of observ- creasingly buy old-age, health, and agricultural in- ability of action is already highly relevant for informal surance. Furthermore, although international do- risk sharing in communities, technology may also nors continue to spend predominantly on disaster make it increasingly relevant for formal insurance. response, their spending on disaster preparation has New devices for cars, for instance, can allow insurers increased in recent years (see chapter 8).45 The extent to vary the insurance premiums they charge based of people’s preparation for risk tends to be correlated on the quality of people’s driving.43 with national income across countries. However, in- teresting variations within regions highlight the im- Preparation portant role of policy in determining preparation for Together, knowledge, insurance, and protection con- risk, over and above access to resources (box 1.7). stitute preparation (or ex ante risk management). Important progress to increase preparation has been Coping made in some areas, which has helped prevent some Coping (ex post risk management) encompasses all risks from developing and has averted some serious actions that are taken once a risk (or alternatively an losses. The institutions and instruments that were opportunity) has materialized. These actions include established to support conditional cash transfers in updating relevant knowledge by assessing the new Risk management can be a powerful instrument for development 69 B O X 1. 7 Preparation for risk varies within and across regions and continents Index of risk preparation across countries Least prepared quintile Most prepared quintile Missing data People’s preparation for risk at the country level includes actions by and above income and access to resources, in determining prepara- and contributions from all social and economic groups and institu- tion for risk. tions, including the state. An index of preparation across countries is The index comprises measures from four important categories: charted on the map above. The index, developed for this Report, human capital, physical and financial assets, social support, and comprises measures of assets and services that influence prepara- state support.a A household’s human capital, in the form of knowl- tion for risk, which in turn affects outcomes. edge, skills, and health, plays a role in providing flexibility to prepare The index shows that people’s preparation for risk tends to be for risk. Physical and financial assets—whether in the form of accu- correlated with national income, but only to a certain extent. On mulated savings or access to credit—provide a buffer in the face of average, people tend to be the most prepared in high-income shocks. Social support also plays a role in allowing households to countries (particularly in North America and western Europe). By respond to specific shocks. This support may include formal pro- contrast, people are the least prepared in low-income countries grams, such as pension schemes, health insurance, and unemploy- (especially in Africa). Substantial variation exists within regions, how- ment compensation, or informal institutions, such as the presence ever, even for countries with similar levels of income per capita. For of family and friends willing to provide care and support. Finally, example, Chile is reasonably well prepared for risk, while its neighbor state support has a critical bearing on risk preparation, through to the east, Argentina, has only average risk preparation despite hav- both the provision of public goods that help manage risk (such as ing a similar level of income per capita. Likewise, Ethiopia has better public health and environmental protection) and the fiscal capacity risk preparation than other countries in the region with similar or of the state to intervene to counteract systemic shocks. The specific relatively higher income per capita (Central African Republic, Sudan, indicators selected are highly correlated with and representative of Uganda). This variation underscores the importance of policies, over a wide selection of indicators in these categories.b Source: Foa 2013 for the WDR 2014. Map number: IBRD 40097. a. Each indicator is rescaled to range between zero and one. The index, which is the average of the eight indicators, thus maintains the cardinal properties of the indicators, rather than simply being an average of rankings across the components. This approach follows in part the methodology used in the construction of the Worldwide Governance Indicators (see Kaufmann, Kraay, and Mastruzzi 2010). If necessary, each indicator is transformed so that an increase in its measure represents an improvement. b. Component indicators for the risk preparation index: Human capital: average years of schooling; immunization rate (measles). Physical and financial assets: proportion of households with less than $1,000 in net assets; access to finance index. Social support: contributors to a pension scheme (as percent of workforce); proportion of respondents stating that “in general, people can be trusted.” State support: access to improved sanitation facilities (percent of population with access); gross public debt (as a percentage of revenues). 70 WORLD DEVELOPMENT REPORT 2014 situation and then implementing necessary and sources available for future risk management, wors- available responses. Continuing with the malaria ex- ening vulnerability to shocks, and weakening house- ample, if members of the household contract the dis- holds’ ability to undertake new opportunities. That ease, coping could include making use of the family’s may have a particularly harmful effect on the poor, health insurance and drawing on savings to pay for who can become trapped in poverty as they face treatment. The family could also get treatment at multiple shocks with little protection.49 Similarly, public hospitals if available, or, if necessary, borrow at the national level, declines in public infrastruc- money from friends in its social network. ture (especially in health, sanitation, and education), employment, and social cohesion following disas- ters can weaken countries already in precarious Links between preparation and coping positions. Coping can be minimal, and lead to quick recovery, Preparation and coping can also be affected by when good preparation for risk is in place, or more deep uncertainty. When knowledge is severely limited extensive, when preparation is limited or a shock is by uncertainty, it is difficult to predict how shocks unexpectedly large. For example, a home with a re- will unfold and what the consequences might be and inforced roof and protected windows may suffer thus how best to prepare for them; that, in turn, af- relatively little damage in a hurricane. Similarly, fects coping because it is difficult to anticipate what good macroeconomic management allowed a num- actions will be needed after a shock. For example, un- ber of developing countries to recover relatively certainty about how climate change is likely to affect quickly from the 2008–09 global financial crisis. In different geographical areas can make it very difficult some cases, some preparations that could help to to prepare effectively for floods or drought, which minimize a shock might not be taken because of the may lead to ad hoc and chaotic coping. To avoid crisis costs involved, but other preparations can help make and effectively manage risk in these areas, preparation coping more efficient if a shock does occur. For ex- needs to include contingency planning—and, more ample, while it may not always be cost-effective for broadly, processes, expertise, and institutions that can cities to build high flood defenses, deciding in ad- facilitate a flexible response to unexpected events (see vance how to deploy disaster relief can help avoid chapter 2). For example, regulators may not always be conflicts over resource allocation following major able to predict where and when risks in the financial floods and make coping more efficient. When prep- system will arise, but by putting in place response aration is weak, however, ex post risk management procedures and coordination mechanisms, they can must deal with unexpected, new, and uncertain situ- respond quickly to emerging risks (see chapter 6). In ations. Under those circumstances, coping can be- some cases, even when investment in risk prevention come ad hoc and often requires very costly mea- has been extensive, disaster response can be subopti- sures. For example, households unprepared for mal if it is uncoordinated and inflexible (box 1.8). income shocks may have to resort to measures such as cutting back food consumption or taking up haz- Risk management is cost-effective— ardous work.46 yet not always feasible The contrast between minimal and costly coping highlights the potential for vicious or virtuous cir- Not only can risk management save lives, avert dam- cles in risk management. When effective preparation ages, and unleash opportunities, but preparation for limits the damages from shocks, the need for coping risk often has high returns. A regimen of mineral is minimal and fewer resources are used for disaster supplements designed to reduce malnutrition and its response—leaving more resources available for in- related health risks, for example, may yield benefits vestment in risk management, reducing vulnerabil- 15 or more times greater than the cost of the pro- ity to future shocks, and so on. Evidence suggests gram.50 Similarly, improving early warning systems that preparation for risk may accelerate economic in developing countries could yield estimated bene- growth by reducing losses from disasters and de- fits 4 to 36 times greater than the cost.51 More gener- creasing economic volatility during crises.47 For ex- ally, benefit-cost analyses suggest that risk prepara- ample, analysis across a set of developed and devel- tion is often beneficial in averting costs, sometimes oping countries over four decades suggests that overwhelmingly so, as illustrated by high median “crisis volatility” can decrease per capita GDP benefit-cost estimates across a number of areas (fig- growth by as much as 2.2 percentage points a year.48 ure 1.3). Such analyses typically compare the likely Conversely, very costly coping may leave few re- cost of an intervention with the expected benefit in Risk management can be a powerful instrument for development 71 B O X 1. 8 A “man-made” disaster: The Fukushima nuclear accident in Japan In 2011 Japan was hit by joint disasters that were unprecedented in Lack of flexibility and poor coordination proved especially prob- scale and complexity. One of the largest earthquakes ever recorded, lematic in managing the response to the disaster. Local authorities measuring 9.0 on the Richter scale, struck on March 11, 2011, generat- were unprepared for a nuclear disaster alongside a natural hazard, ing an enormous tsunami that swept over part of Japan’s eastern and plant technicians were initially isolated because of transport coastline. The earthquake’s tremors, together with the flooding and communication failures. Both situations highlight the need to caused by the tsunami, resulted in a total loss of power at the Fuku- build capacity to respond flexibly to unexpected events. Respond- shima Daiichi Nuclear Power Plant. As workers struggled to cool the ers and decision makers must be in a position to respond rapidly nuclear reactors, a massive nuclear leak occurred on March 15. and flexibly—albeit within a well-established institutional frame- To provide an impartial assessment of the events that led to the work—to events that may unfold in ways never previously imag- radioactive leak, the Japanese government created the Fukushima ined. Moreover, problems of coordination among the national gov- Nuclear Accident Independent Investigation Commission. Although ernment, regulators, local authorities, and the plant’s operator the earthquake and tsunami were the proximate causes of the leak, underscore the importance of establishing coordination mecha- the commission concluded that the nuclear disaster was in fact nisms in advance. These mechanisms should include establishing “man-made.” Both the regulators and plant operator were aware of the chains of authority and means of coordination that will be the potential risk of a power outage in the event of a tsunami reach- deployed in the event of a crisis, and putting in place disaster coor- ing the plant, and of the need for structural reinforcement of the dination teams plant, but failed to act. Source: WDR 2014 team based on Fukushima Nuclear Accident Independent Investigation Commission 2012. F I G U R E 1. 3 The benefits of risk management often outweigh the costs 31 29 14 12 Break even point Benefit-cost ratios 10 8 6 75th percentile 4 Median 2 25th percentile 0 Early Improved Vaccines Nutritional Measures to reduce damage from: warning water and interventions systems sanitation earthquakes floods tropical storms Source: Wethli 2013 for the WDR 2014. Note: The figure shows the 25th percentile, median, and 75th percentile of benefit-cost ratios across a range of studies in each category (with a minimum of at least four estimates in each category). Above the dotted line, expected benefits exceed expected costs. The 75th percentiles for early warning systems and nutritional interventions are 31 and 29, respectively. terms of averted loss of life or damages should a Spending on preparation is costly, however, while shock occur. Estimates vary considerably, reflecting the benefits are not always immediately evident, and the different local contexts and assumptions of the people in developing countries often have limited re- underlying studies.52 Despite this caveat, such analy- sources with which to manage their risks. That may ses provide a useful sense of the orders of magnitude help explain why, despite the potential to save lives involved and support the assertion that preparation and avert damages, more is not spent on preparation, is often cost-effective. and why spending on disaster response is often much 72 WORLD DEVELOPMENT REPORT 2014 higher than spending on preparation. Although funds spent on disaster prevention and preparation spending on preparation is likely to be underesti- (see chapter 8).53 More generally, a number of inter- mated (and has increased more recently), interna- nal and external obstacles may prevent people from tional aid data shows that in the past three decades successfully managing the risks they face. These and more than 96 percent of the total annual spending by many other obstacles that confound risk manage- international donors on disaster management has ment are discussed in detail in chapter 2, along with been directed to emergency response and recon- the potential role and limitations of the state in help- struction relief, on average, with only 3.6 percent of ing to alleviate them. Risk management can be a powerful instrument for development 73 Notes intervention. In formal benefit-cost analyses, this probability is usually taken into account either implicitly (by basing calcu- 1. “Egypt Grants Bonuses after Deadly Food Riots,” Associated lations of averted costs on average historical data) or explicitly Press, April 4, 2008. (by weighting the potential benefit of a risk management in- 2. “Death Toll from Indonesian Disasters Nears 430,” Jakarta tervention in the event of a shock by the probability of that Globe, October 29, 2010. shock occurring). See Wethli 2013 for the WDR 2014. 3. World Bank and IMF 2010. 35. Maloney 2007. 4. Dercon, Hoddinott, and Woldehanna 2005. 36. Maloney 2007. 5. Buvinić and Morrison 2000. 37. The concept of a risk chain is discussed and illustrated in 6. World Bank 2012c. Alwang, Siegel, and Jørgensen 2001. See also Barrett 2002; 7. Self-reported data on shocks entail a subjective assessment by Heltberg, Siegel, and Jørgensen 2009. households. In addition, household surveys do not provide 38. This definition is related to, but distinct from, how the term information on the magnitude of shocks. In some cases, the vulnerability is used in various disciplines. For example, mi- shocks reported may be relatively small. croeconomic studies, which typically focus on income shocks, 8. These perceptions appear comparable with the actual inci- define vulnerability in terms of the probability that a person dence of risk (see box 1.1) for risks that are similarly defined will enter into poverty following a shock. For a full discussion and that occur frequently. For some risks, however, especially of the many ways in which vulnerability is defined in different those that evolve slowly, perceptions of risk can sometimes disciplines, see Alwang, Siegel, and Jørgensen 2001. diverge from reality (see chapter 2). 39. See, for example, Christopher Barrett and Mark Constas, 9. EM-DAT OFDA/CRED International Disaster Database. “Resilience to avoid and escape chronic poverty: Theoretical 10. WHO 2013. foundations and measurement principles,” http://www.dyson 11. WHO 2013. .cornell.edu/faculty_sites/cbb2/presentations.htm. 12. EM-DAT OFDA/CRED International Disaster Database. 40. The seminal paper in this field is Ehrlich and Becker 1972. 13. Paul 2009. See also the extension in Muermann and Kunreuther 2008, 14. Laeven and Valencia 2012. and the applications in Gill and Ilahi 2000; Holzmann and 15. World Bank and United Nations 2010. Jørgensen 2001; and Packard 2002. 16. For example, the ratio of total losses to insured losses from 41. Townsend 1994. natural disasters in the United States is typically between 42. For a rich discussion of the potential complementarity be- 2 and 4, whereas in China it has often been close to 50. See tween insurance and protection, see Erlich and Becker 1972. Kunreuther and Michel-Kerjan 2012. 43. Economist 2013. 17. Russell 2004. 44. World Bank 2012a. 18. Mechler 2004. 45. World Bank 2012b. 19. Baulch 2011 provides a useful survey. 20. Little and others 2004. 46. Heltberg, Hossain, and Reva 2012. 21. Subbiah, Bildan, and Narasimhan 2008. 47. Hallegatte 2012b; Hnatkovska and Loayza 2005. 22. Didier, Hevia, and Schmukler 2012; IMF 2010. 48. Crisis volatility (as opposed to the small but frequent cycles 23. 2010–2012 World Values Survey; median across 24 developing that are typical of business cycles) is defined in terms of down- countries in survey. ward deviations below a common threshold. Careful instru- 24. Gallup 2006–08, 2010–12. Twenty-seven percent of people mental variable analysis is needed to control for simultaneity surveyed said they would like to permanently move to another between volatility and growth. See Hnatkovska and Loayza country (median across 72 developing countries in survey). In 2005. the former Soviet Union, the main reasons for wanting to 49. For evidence on how loss of assets from natural disasters can move were to improve their standard of living (52 percent), create poverty traps, see Carter and others 2007. for their children’s future (13 percent), or to get a good job 50. Hoddinott, Rosegrant, and Torero 2012. (10 percent); in Latin America, the main reasons were to es- 51. Hallegatte 2012a. cape from poverty (38 percent), make money and help loved 52. The set of assumptions used for benefit-cost analysis varies by ones back home (29 percent), or to achieve higher personal study. Many studies provide a range of estimates, based on and professional growth (17 percent). differing assumptions about the value of lost life and discount 25. See Rosenzweig and Binswanger 1993 for evidence from In- rates. Most studies do not make specific assumptions about dia, and Dercon 1996 for Tanzania. governments’ risk aversion, or the extent to which govern- 26. Dercon and Christiaensen 2011. ments are credit constrained. Benefit-cost ratios typically 27. Brown, Mobarak, and Zelenska 2013 for the WDR 2014. increase with the time horizon considered, reflecting the in- 28. Donovan 2012. creased likelihood of shocks over a longer time period. See 29. Karlan and others 2012. Wethli 2013 for the WDR 2014 for further discussion. 30. Cole, Giné, and Vickery 2013; Mobarak and Rosenzweig 2012. 53. WDR 2014 team based on data from AidData. Disaster pre- 31. World Bank 2012c. vention and preparation includes donor funding for early 32. Hoddinott, Rosegrant, and Torero 2012. warning systems and protection of critical infrastructure, 33. Bruno and Easterly 1998. among other items. However, other spending that may im- 34. While the costs of preparing for risk must be incurred pre- prove preparation for disasters—for example, changing the dominantly up-front, the benefits tend to accrue over time location of roads—may often be classified as more general and are therefore more uncertain. The probability of a risk development spending. To that extent, these figures underesti- materializing is thus central to any assessment of a potential mate donor spending on preparation. 74 WORLD DEVELOPMENT REPORT 2014 References Ferreira, Francisco H. G., Julian Messina, Jamele Rigolini, Luis- Felipe López-Calva, Maria Ana Lugo, and Renos Vakis. 2013. Agénor, Pierre-Richard, C. John McDermott, and Eswar S. Prasad. Economic Mobility and the Rise of the Latin American Middle 2000. “Macroeconomic Fluctuations in Developing Countries: Class. Washington, DC: World Bank. Some Stylized Facts.” World Bank Economic Review 14 (2): Foa, Roberto. 2013. “Household Risk Preparation Indices— 251–85. Construction and Diagnostics.” Background paper for the Alwang, Jeffrey, Paul B. Siegel, and Steen L. Jørgensen. 2001. World Development Report 2014. “Vulnerability: A View from Different Disciplines.” Social Pro- Fukushima Nuclear Accident Independent Investigation Com- tection Discussion Paper 0115, World Bank, Washington, DC. mission. 2012. “The Official Report of the Fukushima Nuclear Barrett, Christopher B. 2002. “Food Security and Food Assistance Accident Independent Investigation Commission.” National Programs.” In Handbook of Agricultural Economics, vol. 2B, ed- Diet of Japan, Tokyo. ited by Bruce L. Gardner and Gordon C. Rausser, 2103–90. Gill, Indermit S., and Nadeem Ilahi. 2000. “Economic Insecurity, Amsterdam: Elsevier Science. Individual Behavior and Social Policy.” Working Paper 31522, Barro, Robert J., and José F. Ursúa. 2012. “Rare Macroeconomic World Bank, Washington, DC. Disasters.” Annual Review of Economics 4 (1): 83–109. Hallegatte, Stéphane. 2012a. “A Cost Effective Solution to Re- Baulch, Bob. 2011. Why Poverty Persists: Poverty Dynamics in Asia duce Disaster Losses in Developing Countries: Hydro- and Africa. Cheltenham, U.K.: Edward Elgar. Meteorological Services, Early Warning, and Evacuation.” Brown, Julia K., Ahmed Mushfiq Mobarak, and Tetyana V. Policy Research Working Paper 6058, World Bank, Washing- Zelenska. 2013. “Barriers to Adoption of Products and Tech- ton, DC. nologies that Aid Risk Management in Developing Coun- ———. 2012b. “An Exploration of the Link between Develop- tries.” Background paper for the World Development Report ment, Economic Growth, and Natural Risk.” Policy Research 2014. Working Paper 6216, World Bank, Washington, DC. Bruno, Michael, and William Easterly. 1998. “Inflation Crises and Heltberg, Rasmus, Naomi Hossain, and Anna Reva. 2012. Living Long-Run Growth.” Journal of Monetary Economics 41 (1): through Crises: How the Food, Fuel, and Financial Shocks Affect 3–26. the Poor. Washington, DC: World Bank. Bryan, Gharad, Shyamal Chowdhury, and Ahmed Mushfiq Mo- Heltberg, Rasmus, Paul B. Siegel, and Steen L. Jørgensen. 2009. barak. 2012. “Seasonal Migration and Risk Aversion.” Discus- “Addressing Human Vulnerability to Climate Change: Toward sion Paper 8739, Centre for Economic Policy Research, a ‘No-Regrets’ Approach.” Global Environmental Change 19 London. (1): 89–99. Buvinić, Mayra, and Andrew R. Morrison. 2000. “Living in a More Hnatkovska, Viktoria, and Norman Loayza. 2005. “Volatility and Violent World.” Foreign Policy (118): 58–72. Growth.” In Managing Economic Volatility and Crises: A Practi- Carter, Michael R., Peter D. Little, Tewodaj Mogues, and Workneh tioner’s Guide, edited by Joshua Aizenman and Brian Pinto, Negatu. 2007. “Poverty Traps and Natural Disasters in Ethio- 65–100. New York: Cambridge University Press. pia and Honduras.” World Development 35 (5): 835–56. Hoddinott, John, Mark Rosegrant, and Maximo Torero. 2012. Cole, Shawn, Xavier Giné, and James Vickery. 2013. “How Does “Investments to Reduce Hunger and Undernutrition.” Paper Risk Management Influence Production Decisions? Evidence prepared for the 2012 Global Copenhagen Consensus. from a Field Experiment.” Working Paper 13-080, Harvard Holzmann, Robert, and Steen Jørgensen. 2001. “Social Risk Man- Business School, Boston, MA. agement: A New Conceptual Framework for Social Protection, Dercon, Stefan. 1996. “Risk, Crop Choice, and Savings: Evidence and Beyond.” International Tax and Public Finance 8 (4): from Tanzania.” Economic Development and Cultural Change 529–56. 44 (3): 485–513. IMF (International Monetary Fund). 2010. Emerging from the Dercon, Stefan, and Luc J. Christiaensen. 2011. “Consumption Global Crisis: Macroeconomic Challenges Facing Low-Income Risk, Technology Adoption and Poverty Traps: Evidence from Countries. Washington, DC: IMF. Ethiopia.” Journal of Development Economics 96 (2): 159–73. Karlan, Dean, Robert Darko Osei, Isaac Osei-Akoto, and Christo- Dercon, Stefan, John Hoddinott, and Tassew Woldehanna. 2005. pher Udry. 2012. “Agricultural Decisions after Relaxing Credit “Shocks and Consumption in 15 Ethiopian Villages, 1999– and Risk Constraints.” Working Paper 18463, National Bureau 2004.” Journal of African Economies 14 (4): 559–85. of Economic Research, Cambridge, MA. Didier, Tatiana, Constantino Hevia, and Sergio L. Schmukler. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2010. 2012. “How Resilient and Countercyclical Were Emerging “The Worldwide Governance Indicators: Methodology and Economies during the Global Financial Crisis?” Journal of Analytical Issues.” Policy Research Working Paper 5430, World International Money and Finance 31 (8): 2052–77. Bank, Washington, DC. Donovan, Kevin. 2012. “Agricultural Risk, Intermediate Inputs, Khokhar, Tariq. 2013. “Leveraging New Technology for Data- and Cross-Country Productivity Differences.” Job Market Driven Risk Mitigation and Management: Selected Examples Paper, Arizona State University. and Summaries.” Background paper for the World Develop- Economist. 2013. “Gizmos That Track Driving Habits Are Chang- ment Report 2014. ing the Face of Car Insurance.” February 23. Kunreuther, Howard, and Erwann Michel-Kerjan. 2012. “Policy Ehrlich, Isaac, and Gary S. Becker. 1972. “Market Insurance, Self- Options for Reducing Losses from Natural Disasters: Allo- Insurance, and Self-Protection.” Journal of Political Economy cating $75 Billion.” Paper prepared for the Copenhagen 80 (4): 623–48. Consensus. FAO (Food and Agricultural Organization). FAO Food Price Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises Index (database), FAO, Rome, http://www.fao.org/worldfood Database, an Update.” Working Paper WP/12/163, Interna- situation/FoodPricesIndex/en/. tional Monetary Fund, Washington, DC. Risk management can be a powerful instrument for development 75 Little, Peter D., M. Priscilla Stone, Tewodaj Mogues, A. Peter deficiency Virus/Acquired Immunodeficiency Syndrome.” Castro, and Workneh Negatu. 2004. “ ‘Churning’ on the Mar- American Journal of Tropical Medicine and Hygiene 71 (2): gins: How the Poor Respond to Drought in South Wollo, 147–55. Ethiopia.” BASIS Brief 21, Collaborative Research Support Subbiah, A. R., Lolita Bildan, and Ramraj Narasimhan. 2008. Program, Madison, WI. “Assessment of the Economics of Early Warning Systems for López-Calva, Luis F., and Eduardo Ortiz-Juarez. 2011. “A Vulner- Disaster Risk Reduction.” Background paper submitted to ability Approach to the Definition of the Middle Class.” Policy the World Bank Group Global Facility for Disaster Reduction Research Working Paper 5902, World Bank, Washington, DC. and Recovery. Maloney, William F. 2007. “Missed Opportunities: Innovation and Townsend, Robert M. 1994. “Risk and Insurance in Village India.” Resource-Based Growth in Latin America.” In Natural Re- Econometrica 62 (3): 539–91. sources: Neither Curse nor Destiny, edited by Daniel Lederman UNODC (United Nations Office on Drugs and Crime). Homicide and William F. Maloney, 141–82. Washington, DC: World Statistics (database). UNODC, Vienna, http://www.unodc. Bank. org/unodc/en/data-and-analysis/homicide.html. Mechler, Reinhard. 2004. Natural Disaster Risk Management and Université Catholique de Louvain. EM-DAT: The OFDA/CRED Financing Disaster Losses in Developing Countries. Karlsruhe: International Disaster Database. Université Catholique de Verlag Versicherungswirtschaft GmbH. Louvain, Brussels, http://www.emdat.be. Mobarak, Ahmed Mushfiq, and Mark Rosenzweig. 2013. “Infor- Wethli, Kyla. 2013. “Benefit-Cost Analysis for Risk Management: mal Risk Sharing, Index Insurance and Risk-Taking in Devel- Summary of Selected Examples.” Background paper for the oping Countries.” American Economic Review: Papers and Pro- World Development Report 2014. ceedings 2013 103 (3): 375–80. WHO (World Health Organization). 2013. World Health Statistics Muermann, Alexander, and Howard Kunreuther. 2008. “Self- 2013. Geneva: WHO. Protection and Insurance with Interdependencies.” Journal of World Bank. 2012a. “Resilience, Equity, and Opportunity.” 2012– Risk and Uncertainty 36 (2): 103–23. 2022 Social Protection and Labor Strategy, World Bank, Wash- Nerem, R. Steven, Don P. Chambers, Chong-Deok Choe, and Gary ington, DC. T. Mitchum. 2010. “Estimating Mean Sea Level Change from ———. 2012b. “The Sendai Report: Managing Disaster Risks for the Topex and Jason Altimeter Missions.” Marine Geodesy 33 a Resilient Future.” World Bank, Washington, DC. (S1): 435–46. ———. 2012c. World Development Report 2013: Jobs. Washington, Packard, Truman G. 2002. “Pooling, Savings, and Prevention: DC: World Bank. Mitigating the Risk of Old Age Poverty in Chile.” Policy Re- ———. PovcalNet (database). World Bank, Washington, DC, search Working Paper 2849, World Bank, Washington, DC. http://iresearch.worldbank.org/PovcalNet/index.htm. Paul, Bimal Kanti. 2009. “Why Relatively Fewer People Died? The ———. World Development Indicators (database). World Bank, Case of Bangladesh’s Cyclone Sidr.” Natural Hazards 50 (2): Washington, DC, http://data.worldbank.org/data-catalog/ 289–304. world-development-indicators. Rosenzweig, Mark, and Hans Binswanger. 1993. “Informal Risk World Bank and IMF (International Monetary Fund). 2010. Sharing, Index Insurance and Risk-Taking in Developing Global Monitoring Report 2010: The MDGs after the Crisis. Countries.” Economic Journal 103 (416): 56–78. Washington, DC: World Bank. Russell, Steven. 2004. “The Economic Burden of Illness for World Bank and United Nations. 2010. Natural Hazards, UnNatu- Households in Developing Countries: A Review of Studies ral Disasters: The Economics of Effective Prevention. Washing- Focusing on Malaria, Tuberculosis, and Human Immuno- ton, DC: World Bank. S P OT L I G H T 1 Preparing for the unexpected: An integrated approach to disaster risk management in the Philippines and Colombia The frequency and severity of disasters resulting from natural hazards have been increasing. Losses from disasters amounted to $3.5 trillion between 1980 and 2011, with one-third occurring in low- and middle-income countries. The complexity of problems posed by natural hazards cannot be addressed by single-sector development planning. Thus many countries are responding with multisectoral approaches and are moving quickly toward mainstreaming the management of risks from natural hazards into all aspects of development planning and in all sectors of the economy. Recognizing that the risks from natural hazards can never be completely eliminated, a balanced approach incorporates structural measures, as well as community-based prevention, emergency preparation, insurance, and other nonstructural measures, such as education and training or land use regulation. Two of the most effective systemic approaches to disaster risk management (DRM) have been developed in Colombia and in the Philippines. An inclusive, innovative, and coordinated approach tions, and the private sector, and is complemented by Re- in the Philippines gional and Local Councils. This multistakeholder composi- Located along the western rim of the Pacific Ring of Fire tion is preserved even at the provincial and municipal and the Pacific typhoon belt, the Philippines is vulnerable levels, where Disaster Risk Reduction Councils operate in to earthquakes, tsunamis, volcanic eruptions, landslides, coordination with the national council. Local government floods, tropical cyclones, and drought. With 268 recorded units are in charge of disaster preparedness, prevention, disaster events over the past three decades and more than mitigation, and response, and since the 1970s have been 40 million people affected between 2000 and 2010, the committed to working with communities to effectively Philippines ranks eighth among countries most exposed promote resilience. to multiple hazards, according to the World Bank’s Natural Innovation and inclusiveness also guide the approaches Disaster Hotspot list. taken in risk assessment and communication. In 2006, five As early as 1941, the Philippines established the Civilian technical agencies, which traditionally had not worked to- Emergency Administration to formulate and execute poli- gether, started collaborating on multihazard mapping of cies and plans to protect the population in emergencies. the 27 provinces most vulnerable to disasters. The READY Since then, the institutional and disaster management sys- project marked the first attempt to approach disasters in a tems have focused on emergency response, with impor- multihazard fashion. It included capacity-building activi- tant measures defined and implemented for short-term ties in the provinces and established community-based forecasting, early warning and evacuation, and postdisas- early warning systems for tsunamis, floods, and landslides, ter relief. More recently, the DRM system has been en- which have been used extensively. Launched by the De- hanced through a shift in the policy framework that focuses partment of Science and Technology in 2012, the Nation- on prevention and mitigation, above and beyond emer- wide Operational Assessment of Hazards (NOAH) project gency relief and response. The Disaster Risk Reduction and aims to improve the disaster management capacity of local Management Act of 2010 adopted a comprehensive and governments by spreading out risk assessment and hazard integrated approach that promotes the involvement of all mapping that can trigger protective actions and early evac- sectors and all stakeholders at all levels, especially the local uation. By 2014 NOAH will provide high-resolution flood community. A national risk financing strategy is being un- hazard maps and install automated rain gauges and water- dertaken to establish appropriate risk transfer instruments level measuring stations for 18 major river basins of the to complement resources at the national and local levels, Philippines. It provides not only information about weather including a contingency credit line (the Catastrophe De- conditions, the amount of rainfall, and potential flooding in ferred Drawdown Option, or CAT DDO).1 a specific area but also timely warnings about severe The approach to DRM in the Philippines is distin- weather, earthquakes, and floods, reaching out to a wide guished by inclusiveness, innovation, and coordination. segment of the population. Overall policy and coordination comes through the Na- Every year, a National Disaster Risk Reduction and Man- tional Disaster Risk Reduction and Management Council, agement Fund, formerly called the Calamity Fund, is ap- which consists of 39 members from national government propriated in the national budget for disaster aid, relief, agencies, local governments, nongovernmental organiza- and rehabilitation services. A similar fund has been set up Spotlight 1 77 at the local level. Before 2010, most of the fund was used for postdi- Some cities are well advanced in their ability to carry out effective saster activities. With the enactment of the Disaster Risk Reduction disaster risk management plans and implement them well. Since the and Management Act, 70 percent of the fund can be allocated for pre- 1990s, Bogotá has conducted various studies to identify hazards and disaster preparedness activities. This shift indicates that the govern- assess risks. Detailed maps of hazards related to floods, landslides, ment is moving toward an agenda more geared to risk reduction. and forest fires, as well as a seismic microzoning, have been produced. As a consequence, unstable zones have been identified and buffer Integrating disaster management into the development zones have been established. The district planning department de- process in Colombia signed an integrated rehabilitation, reconstruction, and sustainable Colombia has established itself as a leader in Latin America in devel- development plan in 2005. A three-stage methodology was devel- oping a comprehensive vision for risk and disaster management. oped to support the resettlement process, which includes community Colombia’s advanced DRM system is anchored on investments in engagement and awareness, support with preparation for the move structural measures, risk assessments, early warning and emergency (including a special housing subsidy), and monitoring and follow-up response, institutional support, and financial and fiscal measures at after resettlement. Once families turned over their original properties, the national and municipal levels, as well as the organization of na- the process of rehabilitation and restoration of those high-risk lots tional and local entities for emergency response. As a result of these started. This successful methodology has been replicated in other cit- measures, mortality rates per natural phenomenon have dropped by ies in Colombia and elsewhere in Latin America. almost half from the 1970s to the 2000s, from 4,025 to 2,180. Housing The actions in Colombia and the Philippines represent significant damages increased almost fivefold during that period, however, steps toward a holistic and multistakeholder approach to DRM, but mainly because of unplanned urbanization, which brought almost 80 more is needed. An even greater focus on risk reduction is required, percent of the population into cities, and lack of enforcement of especially at the local level, along with a better definition of roles, re- building codes in some areas of the country. sponsibilities, and coordination among players, and additional invest- Colombia’s long history in organizing and designing risk manage- ments in specific sectors that are not fully integrated into the DRM ment measures started with instruments such as the National System system, such as housing, finance, and agriculture. for Disaster Prevention and Response (1985) and the National Plan for Disaster Prevention and Response (1998). Recently, Colombia ap- Notes 1. The CAT DDO is a World Bank financial instrument that offers eligible middle- proved a new national policy and a National System for Disaster Risk income countries immediate liquidity of up to $500 million, or 0.25 percent of Management. Law 1523 (2012) reflects a paradigm shift in which di- gross domestic product (whichever is less), in case of a natural disaster. The instru- saster risk management is explicitly recognized as a part of the devel- ment was designed by the World Bank to provide affected countries with bridge opment process, and stronger incentives for local governments to financing while other sources of funding are mobilized. invest in risk reduction and strengthen technical assistance are pro- vided. It also recognizes that natural disasters are an implicit contin- Sources Campos Garcia, Ana, Niels Holm-Nielsen, Carolina Díaz Geraldo, Diana M. Rubiano gent liability of the state (see chapter 7), and it establishes a fiscal risk Vargas, Carlos R. Costa Posada, Fernando Ramírez Cortés, and Eric Dickson, eds. management strategy, which includes sophisticated risk transfer 2012. Analysis of Disaster Risk Management in Colombia: A Contribution to the mechanisms, such as the CAT DDO. Creation of Public Policies. Bogotá: World Bank and the Global Facility for Disaster Decentralization and a growing focus on prevention are guiding Reduction and Recovery (GFDRR). Correa, Elena, ed. 2011. Preventive Resettlement of Populations at Risk of Disaster: Expe- the approach to DRM in the country. Since 1997, Colombia has re- riences from Latin America. Washington, DC: World Bank and GFDRR. quired that land use plans be developed at the municipal level; these Fernandez, Glenn, Noralene Uy, and Rajib Shaw. 2012. “Community-Based Disaster plans must consider the location of critical hazards and risk areas for Risk Management Experience of the Philippines.” In Community-Based Disaster purposes of disaster prevention. One of Colombia’s risk prevention Risk Reduction, edited by Rajib Shaw, 205–31. Bingley, U.K.: Emerald Group Pub- lishing Ltd. strategies is to resettle the at-risk population in safe areas, when risk World Bank and the National Disaster Coordinating Council of the Philippines. 2004. cannot be mitigated by other means or only by methods that are “Natural Disaster Risk Management in the Philippines: Enhancing Poverty Alle- more costly than resettlement. Enforcement of building codes is viation through Disaster Reduction.” World Bank, Washington, DC. weak, and retrofitting of existing buildings is costly and inefficient— to the point that resettlement policies have been preferred. Whether on the road or confronting economic, natural, or health risks, local conditions can pose obstacles to proactive risk management. © Kullez CHAPTER 2 Beyond the ideal: Obstacles to risk management and ways to overcome them Missed opportunities for (map 2.1). Significant progress in risk good risk management analysis has been made in recent decades, thanks to new tools such Nearly every year, Mumbai is hit as remote sensing and satellite im- by heavy rains, and for years, re- agery, better weather forecasting ports have spelled out precisely systems allowing for more relia- what to do to reduce the risk ble warning, new epidemiological of flooding. Twenty years ago, a knowledge to better target public master plan (the Brimstowad Re- health interventions, and more ex- port) provided a list of recommen- perience about how to deal with vio- dations to make the city more resil- lence or macroeconomic crises. Why ient to floods, and nearly $200 million isn’t more being done with this knowledge? was approved to implement the plan. But 12 As the Mumbai story illustrates, even the first “no years after the report was published, in 2005, only a brainer” actions such as cleaning up the drainage fraction of this sum had been spent. Then an excep- system are sometimes challenging to implement. To tional monsoon event hit the city. Almost half the cite some other distressingly common examples: average yearly rainfall fell in a single day, leaving in Hand washing is unquestionably a good investment its wake more than 400 deaths and extensive dam- in good health; nevertheless, individuals often fail to ages to buildings and infrastructure. After the 2005 do it. Early warning systems provide a cost-effective devastation, the government established a fact-find- means of mitigating the damage from natural haz- ing committee (the Chitale Committee) to investi- ards, with benefits exceeding costs by a margin of gate the causes of the disaster and propose solutions. four to one at the global level, but investment in and Perhaps not surprisingly, their recommendations implementation of early warning systems remain were very similar to those of the Brimstowad Report. limited.2 There is widespread and vigorous consen- These measures were supposed to be implemented sus on the damage caused by excessive risk taking in by 2015. But as of 2012, only about one-fourth of the the financial sector, but implementing strong regula- 58 projects in the 1993 Brimstowad Report had been tions has proven difficult. On the other hand, some- completed, while the tendering process for four ma- times too few risks are taken, as when firms are re- jor projects had not even begun.1 The city remains luctant to take on the risk of innovation for new highly vulnerable to the heavy rains that occur al- products or technologies, when farmers do not shift most every year, despite well-identified solutions to to planting more productive seeds, or when banks reduce the risk (photo 2.1). refuse to finance viable but risky economic activities. As in Mumbai, many crises—in many countries, In all these cases, desirable steps to manage risk are in many sectors, and at many scales—are repeated not taken, leading to an excess of risk taking or an that could have been prevented or at least mitigated excess of prudence. 79 80 WORLD DEVELOPMENT REPORT 2014 This lack of action suggests that the risk manage- ment framework introduced in chapter 1 is an ideal and that its implementation, in practice, is impaired by a number of obstacles. The main ones include lack of resources and information, biases in behav- ior, and constraints that can be traced to social norms, market failures, and governance shortcom- ings. Fortunately, public action can help alleviate these constraints, especially by focusing on cost- effective interventions and general capabilities, im- proving coordination across levels of government and between the public and private sectors, and aiming for robust and adaptive policies in areas of deep uncertainty (glossary 2.1). A public risk management strategy involves more than simply identifying and assessing risks. Indeed, the mere existence of a risk—even a large one—does not mean that public action is necessary. If individu- P H O T O 2 .1 Difficulties implementing known and even low-cost solutions. als and firms are taking this risk based on an informed Cleaning clogged drainage systems would mitigate the risk of flooding. However, assessment of its potential costs and benefits and are such well-identified and cost-effective solutions often fail to be implemented. The Mumbai case illustrates a common problem (as seen in this picture from Jakarta). able to cope with the consequences, there is no reason © Farhana Asnap/World Bank to prevent them from doing so. On the other hand, M A P 2 .1 Crises repeat themselves in the absence of effective prevention Flooding has resulted in widespread damage in New Orleans for more than 90 years. Source: Grossi and Muir-Wood 2006. Note: Red arrows indicate the breaches in the levees. Blue areas indicate areas flooded by hurricanes. Beyond the ideal: Obstacles to risk management and ways to overcome them 81 G L O S S A R Y 2 .1 Economics terms used throughout the Report Asymmetric A situation in which one party in a transaction has more or better information than another party. information Adverse selection A situation in which asymmetric information leads agents with privileged information to try to select products and services on advantageous terms, possibly skewing the transaction in their favor. Common-pool The problem that arises when individuals overuse common resources to which they have problem unrestricted access. Coordination failure A situation in which decision makers reach an outcome that is inferior because they are unable to jointly choose strategies that would result in a preferable outcome. Deep uncertainty A situation in which parties to a decision do not know or cannot agree on the key forces that shape the future, the probability distributions of the main variables and parameters in their models, or the value of alternative outcomes. Moral hazard A tendency for people to act less responsibly when they are protected from the harmful consequences of their behavior. Myopia A lack of long-range perspective in thinking or planning. Principal-agent The problem that arises when agents pursue their own goals, even when doing so entails poorer problem outcomes for the principals on whose behalf the agents are supposed to act. Time inconsistency A situation in which policy makers announce policies in advance to influence the expectations of in policy private decision makers, but then have an incentive to follow different policies after those expectations have been formed and acted upon. Source: WDR 2014 team. public action is needed if individuals and firms can- are best placed to estimate which risks are worth tak- not manage a risk or its consequences properly. That ing (such as moving to the city to find better-paying is the case, for instance, if they take a risk without the jobs and better public services) and which ones are proper information, if they cannot manage the too costly if things go wrong. Yet they face consequences if the risk materializes into many obstacles in their assessment of and losses, or if the people or groups taking Many crises that preparation for risk. on the risk are not the same ones occur repeatedly who would be affected by the loss. Accordingly, the development could be prevented People may lack information of a public risk management plan with existing means— and resources should be based on the identifi- but even simple “no Financial constraints. Lack of in- cation, prioritization, and correc- brainer” actions can come, assets, and resources often tion of practical obstacles to risk hinders risk management, espe- be challenging to management. This chapter therefore cially for the most vulnerable in de- presents a typology of these obstacles, implement. veloping countries. Even though some as shown in diagram 2.1. It suggests a options for risk management are cheap or methodology to prioritize the obstacles that even almost cost-free (driving safely), others require public action, starting with “soft” options can be expensive. People would prefer to live in earth- based on institutional arrangements, communica- quake-proof houses, but the construction costs may tion and information campaigns, and behavioral ap- be too high. Even when a risk mitigation investment proaches, and then looking at costlier approaches is cost-effective from an economic perspective, peo- such as providing public goods (like dikes and drain- ple or firms may find it difficult to finance because age systems). of large up-front costs and limited access to credit. Households that have limited resources and are therefore more vulnerable to risks and more severely Why aren’t people better at managing affected by losses may face poverty traps. Because of their own risk? this higher vulnerability, they cannot accumulate the Ideally, people, firms, and organizations would man- necessary resources for protecting themselves from age the risks that are within their own capacity. They further losses in the future (see chapter 3). These 82 WORLD DEVELOPMENT REPORT 2014 D I A G R A M 2 .1 Individuals, firms, and countries face many obstacles in managing risks Obstacles to risk management Individual Social obstacles beyond the Obstacles to public and obstacles control of individuals collective action Financial Missing public goods Lack of public resources constraints or services and competing needs Information Missing markets constraints and moral hazard Coordination failures Cognitive failure Distributional effects and Externalities and behavioral political economy biases Social norms Deep uncertainty Collective risks Source: WDR 2014 team. effects may even occur at the macro level, when, for drometeorological services, for example, often have instance, immediate reconstruction needs after natu- to rely on revenues from the sale of data to strengthen ral disasters crowd out longer-term development in- and maintain their observation networks. As a result, vestment at the community or regional level. the cost of one meteorological observation in Europe varies from zero (in Slovenia) to 0.40 euros (in Ro- Information constraints. Information may exist but mania).5 Costly data restrict access to valuable infor- may not be available to or known by the people who mation and reduce the social benefits that are derived have to make risk-related decisions. For instance, from them.6 only 31 percent of people living in areas prone to flooding in the United States were aware of this risk, People struggle to translate knowledge and only 33 percent knew that federally backed flood into action insurance was available, a 2010 survey by the Federal Emergency Management Agency revealed.3 Investors Individual decision making in practice can veer quite and banks may not have the knowledge to evaluate a a long way from the basic, idealized assumptions of loan application regarding innovative projects, lead- economic theory, such as the “maximization of ex- ing them to reject profitable projects and thus con- pected utility.” To the extent this gap is linked to indi- straining creative risk taking and innovation.4 vidual preferences and values, it is not a reason for Progress has been made in making information public action. But part of the gap stems from the lim- more widely available in many areas. New informa- ited time and capacity people have to process infor- tion and communication technologies help people mation on risks and to decide which risk prevention access complex information from nearly anywhere measures they will implement. As a result, they on the planet. Data collection and access remain in- sometimes make decisions that go against their own sufficient, however. Because information has aspects interests and preferences; this latter issue may justify of a public good, it is underprovided by private ac- public action. tors; thus states have a large role in producing and A variety of studies shed light on just how incon- disseminating it. Some countries have not made suf- sistent and incomplete people can be in their evalua- ficient efforts in this direction, notably where data tion of risk (box 2.1). People are biased toward the are not available for free—or not available at all. Hy- status quo and tend to choose the default option. For Beyond the ideal: Obstacles to risk management and ways to overcome them 83 B O X 2 .1 The irrational, the uncertain, and the short-sighted: Some classic experiments reveal surprises about people’s behavior A well-established economics concept known as expected utility choice, expected payoffs are the same for both lotteries, and the theory holds that people try to maximize their expected gains. Yet probability associated with the lower payoff is twice as high as for in practice, when faced with risk and uncertainty, people behave the higher payoff. In the second choice, however, the majority of differently, some classic experiments show. The first experiment participants opt for the choice with the higher payoff, rather than reveals that people are very averse to uncertainty. They value a gain the choice with the higher winning probability. These results show that is a “sure thing” more than one that is uncertain, beyond what that people treat high and low probabilities differently, which may would be predicted by the theory. In this experiment, individuals explain why people are less likely to make investments to reduce must choose between different lotteries, shown in table a. In the the risk of low-probability events at the margin (such as investing to first choice, most people (65 percent) chose the first option. In the reduce the risk of a building collapse in an earthquake) or to make expected utility framework, this means that the benefit (or “utility,” desirable trade-offs between different low-probability risks. in economic terms) of winning $4,000 is larger than four-fifths (20 A third experiment shows how people’s decisions depend on percent versus 25 percent) of the benefit of winning $3,000. But in what they consider “the default situation.” Again, people are asked the second choice, most people (80 percent) chose the “sure thing,” to choose from two lotteries. This time, the chances of winning or even though in the expected utility framework this would mean losing are equivalent, but the amounts that can be won or lost differ. that the benefit of winning $4,000 is lower than four-fifths of the First, they are given $1,000 and must choose between definitely benefit of winning $3,000 (80 percent versus 100 percent)—a direct winning another $500 or having a 50 percent chance of winning contradiction of their first choice. This aversion to uncertainty— another $1,000; 84 percent of the respondents select the sure out- beyond what risk aversion in the expected utility theory would sug- come. Second, they are given $2,000 and they must choose between gest—may be one reason behind the underinvestment in innova- definitely losing $500 and having a 50 percent chance of losing tive projects and entrepreneurship, relative to other forms of more $1,000; 68 percent of respondents select the latter lottery. Respon- secure income. dents react differently to the two choices, even though they are per- The second experiment reveals that people are not able to pro- fectly equivalent. cess differences between small probabilities and so consider all This experiment illustrates the role of the reference points. It low-probability events to be equally likely (table b). In the first also shows that individuals are often risk averse with gains and risk choice, participants can select from two lotteries, each of which pro- seekers for losses. Depending on whether the reference point is the vides the same expected payoff. An overwhelming majority (86 per- best possible outcome (and the only possibility is a loss) or the worst cent) selected Lottery 2. This shows that, in general, people’s deci- possible outcome (and the only possibility is a gain), individuals will sions account not only for the expected outcomes but also for the make different choices. Insurers have known for a long time that associated probabilities. However, the second choice shows that people are more likely to buy insurance if their reference point is the they do not do so when probabilities are very small. As in the first occurrence of a disaster. a. People are averse to uncertainty in a way that contradicts expected utility theory First choice Second choice Lottery properties Lottery 1 Lottery 2 Lottery 1 Lottery 2 Lottery option 20% chance of 25% chance of 80% chance of 100% chance of winning $4,000, winning $3,000, winning $4,000, winning $3,000 80% chance of 75% chance of 20% chance of winning nothing winning nothing winning nothing % of participants who 65 35 20 80 choose the lottery b. People treat high and low probabilities very differently First choice Second choice Lottery properties Lottery 1 Lottery 2 Lottery 1 Lottery 2 Lottery option 45% chance of 90% chance of 0.1% chance of 0.2% chance of winning $6,000, winning $3,000, winning $6,000, winning $3,000, 55% chance of 10% chance of 99.9% chance of 99.8% chance of winning nothing winning nothing winning nothing winning nothing % of participants who 14 86 73 27 choose the lottery Source: WDR 2014 team based on Kahneman and Tversky 1979. 84 WORLD DEVELOPMENT REPORT 2014 guide their decisions. That is why the “availability” of similar risks can explain why people care more about some risks and less about others. For example, be- cause of their experience with the mad cow disease crisis, Europeans may be more concerned than Americans about nontraditional food production techniques, including genetically modified crops.13 Education and communication campaigns and the provision of information in a form that individuals can easily process are thus key elements of a risk management strategy. That is why driving rules are learned not only in a theoretical setting but also through mandatory driving instruction, to the point where they become at least partially automatic.14 People are often overconfident about avoiding loss: they think they are able to drive safely under the influence of alcohol, and they think they can manage P H O T O 2 . 2 A preference for large voluntarily chosen risks compared with a flood and do not need to evacuate. They also have small externally imposed ones. People may take on large risks for their hobby (such as high-altitude mountaineering); however, they may find a much smaller risk to short memories about catastrophes, they discount be unacceptable if it is imposed on them by others (such as the construction of a the future too much and in inconsistent ways, and chemical plant in their neighborhood). © Gordon Wiltsie/National Geographic they fail to account for avoided losses that are not ob- servable. After Hurricane Katrina hit New Orleans in 2005, the number of U.S. households with flood risk example, the proportion of organ donors in coun- insurance increased more than three times more rap- tries where being a donor is the default choice (and idly than observed in previous years. However, the people must opt out if they do not want to be do- average cancellation rate remained unchanged, at ap- nors) is nearly 60 percent higher than it is in coun- proximately 33 percent a year, suggesting a short ef- tries where people must opt in to become donors.7 fect of the disaster on household behavior.15 A simu- People usually attribute a higher weight to rare lation-based study shows that the primary motivator events, but they also simply neglect the possibility of decisions to invest in protection is the size of losses that very rare ones will occur.8 Individuals disregard already experienced, not losses that were avoided.16 the possibility of very bad futures, possibly because This tendency leads to a “paradox of protection”: of the stress created in thinking about them.9 There when protection against frequent events suppresses is also a difference in how people weight the indi- losses for an extended period of time, vigilance and vidual risk they chose for themselves (when they risk awareness decrease. That leads to insufficient drive, hike, or skydive) and the collective risks that maintenance of protective measures and high invest- are imposed upon them (when a chemical plant is ments in risky areas, resulting in future losses (and built in their neighborhood): even if the benefits are losses of increasing scale) if protections collapse or similar, the level of risk is usually perceived to be are overwhelmed by an exceptional event.17 This lack higher when risks are imposed or when individuals of consistency in decision making is not unique to feel they have little control over these risks (photo risk management; it also explains why individuals 2.2).10 To account for this difference, the French gov- have so much trouble meeting their own objectives ernment recommends that the cost-benefit analyses (such as a New Year’s resolution to exercise more). It for investment in transport safety value one death is why people often try to create irreversibility in their avoided in public transport 50 percent higher than choices, by raising the cost of failing to reach their one death avoided in individual car accidents (1.5 objectives (such as paying a high annual fee at a million euros versus 1 million euros).11 gym).18 In many developing countries, this search for People, including policy makers, make many deci- irreversibility also helps explain why people save “in sions in the face of risk by using heuristics (rules of kind.” For instance, people protect their savings from thumb) or by following social norms, instead of capture not only by their extended family but also by making deliberate calculations to identify the best themselves by slowly advancing the construction of a option.12 People use risks they consider similar to house each time resources become available, even Beyond the ideal: Obstacles to risk management and ways to overcome them 85 though this practice is a very inefficient and risky way right medicine is not available. The low quality of of saving. road infrastructure is responsible for a fraction of the These biases in behavior have consequences for higher rates of traffic deaths observed in developing the design of effective risk management policies. Ex- countries. In Poland, the number of crashes at “black cessive discounting of the future, short-sightedness spots,” where accidents are frequent, decreased by 35 (myopia), and the tendency to stick to the default op- percent when danger signs were posted.19 tion can, for example, explain insufficient saving Many risk-related decisions rely at least partially where individuals are allowed to opt into a retire- on basic infrastructure. The landfall of Hurricane ment saving scheme. The biases can also explain why Katrina in New Orleans in 2005 illustrates both the flood insurance reaches large penetration only in success of a road-traffic evacuation plan—the evacu- countries where it is compulsory or during short pe- ation was quicker and smoother than previous ones riods of time following disasters. These biases can for inhabitants who owned a car—and the failure to justify specific interventions, from tax incentives to evacuate the population that relied on public trans- compulsory enrollment in insurance or pension portation.20 The lack of sanitation infrastructure is a schemes. In this context, conditional cash transfer major obstacle for individuals to manage their own programs (for instance in Mexico and Brazil) have health risks. For instance, in India open defecation proven to be highly successful in helping individuals and the absence of sanitary facilities in poor house- managing their health risks. By requiring compliance holds have been shown to be a key reason for child with certain behavior—such as adhering to a pre- stunting.21 Stunting can have a significant impact on scribed vaccination schedule for children—in order adult health, productivity, and economic prospects, to receive a monthly cash transfer, such programs and thus on development opportunities. Accord- create a direct monetary incentive for taking socially ingly, as long as sanitation infrastructure is not pro- and individually beneficial actions to reduce risk, vided, individual behavior changes and development such as medical treatments and checkups. Thus such programs such as child nutrition interventions are programs can overcome the constraints discussed likely to have limited positive impacts. This example above (from resource and information constraints to illustrates a general point: the importance of provid- behavioral biases and a tendency to postpone non- ing basic infrastructure as a basis for the success of urgent medical checkups indefinitely); they thereby further individual and collective risk management help reduce health risks for vulnerable individuals, policies (photo 2.3) their households, and their wider communities. Complicating matters, some people may be excluded from public services for risk management Obstacles beyond the control of individuals hamper their risk management Missing public goods and markets, and even social norms, may prevent people from managing their own risk taking. Above and beyond individual risks, some risks are systemic and therefore cannot be managed without collective action. Individuals must cope with market and government failures Missing public goods and services. Public goods and services that provide an essential foundation for peo- ple’s risk management are often missing. From the point of view of investors, for instance, risk can be managed only if contracts can be enforced, which re- P H O T O 2 . 3 Good infrastructure is needed for people to manage their risks. quires the rule of law and an effective judiciary. Thanks to a national project, people in Woukpokpoe, Benin, have gained access to Health insurance is of little use if poorly regulated safe, clean water, which enables them to manage health risks more effectively. health care providers are not competent or if the © Arne Hoel/World Bank 86 WORLD DEVELOPMENT REPORT 2014 because of their gender, ethnicity, political affiliation, some may make some harmful bacteria more resis- or lack of education or literacy. In Peru, for instance, tant to treatment, threatening the health of all.25 A legal proceedings are held only in Spanish, while firm that introduces a new chemical may create many farmers speak only Quechua and Aymara, health risks to others, while reaping most of the fi- making it difficult for them to rely on the judiciary nancial benefit. Disasters cause indirect losses that system to protect their rights and manage their risks. create externalities.26 In November 2012, for in- stance, Japanese automaker Honda cut the factory Missing markets and the problem of moral hazard. hours of its U.S. auto assembly workers in Ohio be- Missing markets and instruments, such as insurance cause it could not get parts from Thailand, affected and hedging markets, are key obstacles to people’s at the time by large floods. Socially optimal risk ability to manage risk. Even where instruments ex- management in one production unit (as in Thai- ist, they may be plagued by market failures. Insurers land) should take into account these supply chain offer low-deductible (and higher-premium) policies effects and the impact of interruptions in produc- to satisfy clients with high risk aversion. In Israel, tion on the ability of client factories to create value however, it has been shown that “bad” drivers who added. Such far-ranging consideration is not nor- have more accidents chose these policies more than mally the case, however, leading to insufficient risk average drivers. This is a classic case of adverse selec- management. Cases like these highlight the need to tion. The fact that people who are more vulnerable design and implement public actions (like regula- are likely to buy more insurance than individuals tion) or collective action (like supply chain manage- who are less vulnerable—and that insurers ment) to ensure that individual incentives lack information about who is and is are aligned with collective objectives. not more vulnerable, and thus cannot Identifying risks The response to the 2011 earthquake charge more to riskier customers— is not enough: the and tsunami in Japan offers exam- increases the cost of insurance for ples of such collective actions, with obstacles to risk everyone, creates affordability is- clients providing their suppliers sues, and limits the benefits from management must also free assistance to help them re- risk sharing.22 The state may need be identified, prioritized, store their production as quickly to intervene to promote the cre- and addressed through as possible.27 ation of markets and instruments private and and to regulate them in a way that public action. Social norms. Individual behavior re- supports individuals in their manage- garding risk management is embedded ment of risk. in social norms, which can present obsta- Not all risks can be covered, however. Non- cles to risk management—or facilitate it. For monetary losses, such as health and psychological im- example, use of a face mask while sick prevents pacts or personal objects and photographs lost dur- transmission of disease; in Asia, but not elsewhere, ing floods or fires, can rarely be fully compensated. wearing a mask is a commonly followed social norm. After the Bihar floods in India in 2008, for instance, By contrast, the “stigma of failure” is a social norm the elderly suffered from depression more often than that works against innovation and entrepreneurship. they did before the floods.23 Even if all impacts could To counter it, some governments and private institu- be compensated, doing so would remove all incen- tions are rewarding innovation and risk taking, even tives for individuals and firms to mitigate risks them- when it fails: for instance, through the creation of selves and would increase both adverse selection and prizes (such as India’s Tata Group award of an an- moral hazard (the fact that insurance reduces the nual prize for the best failed idea), or tax write-offs incentive for people to protect themselves against for research and development. Specific policy ap- risks) and would therefore magnify losses.24 It is thus proaches may be necessary when lawmakers chal- rarely optimal to cover losses completely, which is lenge a well-established social norm. In the United why private insurers and public schemes (or mixes States, police at first opposed enforcing rape and do- through public-private partnerships) always include mestic violence laws, until complementary measures a deductible that limits the amount of coverage. changed social norms (such as “shaming penalties” for rape and portrayals of male violence against Externalities. The actions of some actors may in- women as “cowardly” or “unmanly”).28 Changes in crease risks for others or reduce their incentives to social norms can have many origins and channels. manage their own risk. Overuse of antibiotics by Lobbies and interest groups use communication Beyond the ideal: Obstacles to risk management and ways to overcome them 87 campaigns to change perceptions about prevailing social norms. Marketing companies, seeking to max- imize sales and revenues, use advertising campaigns to shape perceptions about products and services. Sometimes these campaigns come at the expense of risk management: for example, by encouraging eat- ing and drinking habits that are detrimental to good health. In contrast, public health policies have had great successes in changing norms to improve hy- giene and prevent diseases with sanitation or hand washing. Sometimes, changes in norms have unex- pected drivers. Telenovelas (televised soap operas) in Brazil have had a large influence on fertility choices. The different life styles and ideals presented in the shows have influenced social norms, with a measur- able impact on people’s behaviors (photo 2.4).29 Some risks are collective by nature Some risks are systemic—and therefore collective— by nature. Financial crises or economic slowdowns can be managed only at the country or even interna- tional level. When industrial policies are imple- mented to support a technology or a sector, a coun- P H O T O 2 . 4 Changing social norms. Telenovelas in Latin try takes a macroeconomic and fiscal risk that is America have changed social norms by exposing people to socialized at the national level. If the technology or different lifestyles. © Globo Marcus the sector fails, the loss is shared by all taxpayers. Furthermore, in an increasingly interconnected world, many risks, such as pandemics or financial policy makers often claim that “disasters are unac- crises, are now global. In all these cases, risk must be ceptable”—especially after a catastrophe—canceling managed collectively, using public goods and ser- all risks would be prohibitively costly. Thus a certain vices such as protective infrastructure, health care amount of risk must be accepted.31 Defining an ac- systems, financial regulations, and macroeconomic ceptable level of risk is difficult because of the com- management. Many natural risks, especially in areas plexity of some issues (box 2.2) and because prefer- of geographically concentrated infrastructure and ences, values, and beliefs may differ widely. Some high density, also call for collective management. Be- individuals are more risk averse than others and may cause of various synergy effects, economic produc- prefer a more precautionary approach. Individuals tion and infrastructure tend to agglomerate geo- use their “world views” as cognitive and emotional graphically, often in at-risk zones such as coastal filters that influence how they perceive and act with areas or river flood plains.30 Moreover, protection respect to risky situations, and as a way of simplify- infrastructure is “lumpy,” meaning that it cannot be ing decision making. Working in a cultural theory increased continuously and progressively—it often setting, a study classified U.S. individuals accord- consists of a complex system (such as multiple rings ing to three fundamental world views—“fatalist/ of dikes and pumping stations) and it requires plan- hierarchical,” “individualist,” or “egalitarian”—and ning, is expensive with large up-front costs, and usu- showed that these views largely explain people’s ally covers large areas. As a consequence, individuals preferences concerning many technological and en- or firms cannot provide hard protection to their vironmental risks, such as nuclear energy, genetically houses or production facilities independently of modified crops, and climate change.32 what is put in place at the collective level. Thus man- Because factors that influence how people process aging natural risk is at least partly a collective issue. information about risks are embedded in fundamen- In cases requiring collective action, the definition tal beliefs, judgments about these risks can differ of the acceptable level of risk needs to be made at the markedly within a country and even more across social level, through a political process. Even though countries. At the international level, strong disagree- 88 WORLD DEVELOPMENT REPORT 2014 BOX 2.2 Strengthening the interaction between experts and policy makers to improve risk management Many risk assessments are based on a classic risk matrix that repre- regulated private actors. In the U.S. system, for instance, regulation sents the potential impact of an event on the horizontal axis and its tends to occur only after the damage is done. Lawsuits for damages likelihood (probability of occurrence) on the vertical axis. The risk can cost U.S. industry 1.9 percent of GDP ($180 billion) annually, in con- be considered “intolerable” if its likelihood and potential impact are trast to the United Kingdom, which has a more consensual system, too high; “acceptable” if both factors are low enough; or “tolerable,” and where liability costs industry less than 0.5 percent of GDP.a in the sense that it is not desirable to suppress it, but it nevertheless Recent work suggests the existence of considerable flexibility in needs be managed or reduced (figure). the type of risk regulations that can be implemented, regardless of Scientists and other experts alone cannot define what risks are the political and legal culture.b Regulatory instruments are dissemi- acceptable; they lack legitimacy to do so. Nor can policy makers by nated and hybridized to make them appropriate for different coun- themselves define what risks are acceptable; they usually lack tech- try contexts, helping improve risk regulation globally.c In practice, nical expertise. Thus closer and better interactions between science most countries have tried to adapt risk regulation instruments to and policy are needed to codefine what is acceptable, tolerable, and their cultural and institutional settings, in an effort to balance the intolerable. cost and the transparency of their risk management institutions. Different countries have introduced different institutional sys- tems to reach these definitions, in line with their political culture. A risk matrix can be used to map hazards according Some countries (the United States) use what is referred to as an to their probability and impact “adversarial” system, in which there is an open, procedure-based, 1 and transparent confrontation of viewpoints, and the outcome is Risk so much greater Probability of consequences Prohibition or substitution than benefit that it determined through a legal process. Other areas (parts of southern cannot be taken on Europe) rely on a “patronage” system, in which a public entity is in charge of assessing risks, relying on in-house experts and pro- cesses, with little public scrutiny and participation. Finally, a “con- Benefit worth the risk, Reduction but risk reduction sensual” or “corporatist” approach is more common in northern measures necessary Europe, based on closed-door negotiations between regulators and stakeholders, with little public control and the aim of creating Acceptance No formal intervention compromise. necessary The adversarial system is better able to manage uncertainty and ensure accountability; however, this system is also very costly, may 0 Extent of consequences ϱ exclude some stakeholders, and is not able to benefit from collabo- Acceptable risk Tolerable risk ration and information exchanges between the regulator and the Intolerable risk Not defined Source: WDR 2014 team adapted from Renn and Graham 2005. Source: WDR 2014 team based on Renn and Graham 2005. a. Loewenberg 2006. b. De Francesco 2012. c. Wiener 2013. ments over acceptable levels of risk have led to trade important for health issues; some individuals are far disputes involving high-uncertainty risks such as more sensitive to pollutants than others, for instance. beef hormones, mad cow disease, and genetically Investigating health effects of one type of air pollu- modified crops. An in-depth comparison of risk reg- tion (particulate matter), a study concluded that the ulations in the United States and Europe highlights most susceptible people (those in the 99.9th percen- the role of cultures and world views on risk manage- tile) would feel negative effects at exposures only 0.2 ment practice, as well as the complexity of the matter, to 0.7 percent of the level that would induce the same with European governments showing more risk effects in people of median susceptibility.34 In the aversion than the United States for some risks (beef presence of such heterogeneity, designing homoge- hormones), but less for other risks (mad cow dis- nous regulations is challenging and highly depen- ease).33 The difficulty in defining an acceptable level dent on considerations of equity (especially when of risk in a given context means that prescriptive rec- sensitivity is correlated with other social factors). The ommendations concerning risk management are very selected regulation is also unlikely to satisfy all indi- sensitive issues and need to be issued with caution. viduals. The same problem exists at the international Defining a social level of acceptable risk is also level. For instance, the international community difficult because of differences in individual or col- committed to avoid “dangerous climate change” as lective sensitivity. Such differences are particularly long ago as 1992, through the United Nations Frame- Beyond the ideal: Obstacles to risk management and ways to overcome them 89 work Convention on Climate Change. Countries of the population at the expense of the rest and for its have very different exposures and vulnerabilities to own benefit (for example, a government provides climate change, however, and thus have different protection to an industry in exchange for political views of what a “dangerous” change is, making it support, or a regulation is captured by the regulator). more difficult to reach an agreement on greenhouse A last category is linked to uncertainty and the impos- gas emission reductions (see chapter 8).35 sibility of identifying clear-cut solutions to a problem— regardless of the capacity, resources, and goodwill Why aren’t states better at filling of the government—because of lack of knowledge concerning the appropriate course of action; this sit- in the gaps? uation is referred to as deep uncertainty and is dis- Countries—or the international community—could cussed further later in this chapter. do much to help people overcome obstacles to their management of risk, and these actions can be very Lack of resources and technical capacity cost-effective and supportive of development and hampers public policy poverty alleviation, as discussed in chapter 1. Cor- recting market failures and other social obstacles to Resource and capacity constraints to pursue risk realign the incentives of individuals with the goals of management can be a serious barrier to public ac- society is one of the main roles of governments and tion. The expensive disaster risk preventions that local authorities. Yet this help often does not happen, have been implemented in developed countries or it happens in a very imperfect way that can make (such as flood protections in the Netherlands) are risk management counterproductive or excessively out of reach for many similarly threatened develop- costly. States face large obstacles themselves and have ing countries (such as the Arab Republic of Egypt, competing priorities that do not focus on managing Bangladesh, or Vietnam). This gap is largely the re- risks and fostering resilience. First, decision makers sult of financial constraints, including the lack of and policy makers, as individuals, are susceptible to financial instruments appropriate for long-term the biases that have been described earlier. For in- investments with large social returns but limited stance, they tend to have short memories and have a cost recovery.38 But it is also connected to the exis- short-term bias, they tend to misestimate low prob- tence of many competing needs in developing coun- abilities, and they use rules of thumb and social tries, from health and education to infrastructure norms instead of rigorous risk assessments to make development. In addition, technical capacity is often many decisions. But obstacles to public action go be- lacking in developing countries, particularly so in yond the obstacles to individual decision making, countries where the public sector has well-identified and include many other government failures. difficulties in retaining its talented and skilled work- Defining a government failure is difficult, how- ers. The competencies needed to analyze risks and ever.36 A government failure can be defined as a situ- identify relevant management actions may thus be ation where the government does not operate in the inaccessible even for governments willing to act. public interest—that is, the interests of the taxpayers Lack of resources may weaken institutions and and the users of public services—but in the interests impede enforcement of rules, which in turn may of “narrow interest groups who are able to exploit a lead to poorly designed or implemented risk man- privileged position for their own benefit.”37 This defi- agement policies. For example, the inability of gov- nition links government failure with capture by inter- ernment to enforce property rights and land titles est groups. Here, four broad categories of govern- has a negative impact on risk management: house- ment failures are discussed. The first is linked to holds with precarious tenure risk eviction and are insufficient resources and capacity and to the involun- unlikely to invest in risk-mitigating investments such tary implementation of government actions that are as flood-proofing and earthquake-proofing their less desirable than what could be achieved. The sec- houses. They are also unable to use their home as col- ond is related to coordination failure (within govern- lateral to obtain credit to finance such investments. ment or between public and private actors), includ- Corruption often thrives amid weak institutions and ing policy capture, such as a government that is is an obstacle to public risk management: for in- manipulated by an interest group to introduce regu- stance, when contractors do not respect building lation that is detrimental at the social level. A third norms for public buildings in earthquake-prone category is linked to political economy issues and vol- cities. As illustrated in chapter 4, community-based untary decisions by the government to favor a subset solutions can help in instances when governments 90 WORLD DEVELOPMENT REPORT 2014 and local authorities are unable or unwilling to man- measures such as land use plans and building norms age risks. Where school construction does not re- at the local level (a task often led by local authorities). spect building norms because local authorities are In the absence of cross-scale regulations, a “public unable to enforce them, the involvement of the com- moral hazard” may emerge, if local authorities rely munity—and the parents who will send their chil- on national support in case of disasters, reducing the dren to the school—may be a solution. In situations incentive to implement preventive actions.40 where enforcement and compliance are weak, more Coordination is also required between public and effective enforcement of existing building norms can private actors. The impact of a flood is highly depen- potentially have significant benefits. For instance, in- dent on the ability of private actors to reallocate surance experts estimate that insured losses in the resources and the ability of utilities and transport United States from Hurricane Andrew in 1992 could companies to restore basic services. The impact of have been reduced by 25 percent if building norms epidemics is dependent on the ability of companies had been fully enforced.39 and organizations to maintain operations with a re- Lack of resources not only influences what can be duced workforce. In sectors where states often rely achieved but also the type of solutions and measures on private sector expertise (cyberattacks, finance), that are desirable. For risk management, as for other public-private cooperation is a critical ingredient in productive investments, the “best” technology de- the design of a strategy. Such cooperation is often pends on the relative scarcity of production factors. difficult to establish because of differences in culture In developing countries where capital and skilled la- and work habits, issues related to privacy and com- bor are scarce and unskilled labor is underused, risk mercial secrecy, risks of capture and rent-seeking be- management solutions will necessarily be different haviors from private actors, and lack of incentives on from those selected by more developed countries both sides. The recent financial crisis illustrates the where capital is cheap and labor is expensive. In par- difficulty regulators face in determining the best ticular, risk management strategies are likely to be course of action when their main advisers—profes- based more on hard infrastructure (large dike sys- sionals from the financial sector—have a large stake tems) in higher-income countries. The appropriate in the decision (see chapter 6). risk management actions also depend on institu- To promote and improve coordination, multiple tional and enforcement capacity. Where enforcement stakeholders need to be involved in decision-making capacity is limited, strengthening building norms processes for risk assessment and implementation may worsen the situation. Increasing compliance (box 2.3). Stakeholder involvement is useful not only costs can increase the number of noncompliant, to disseminate information and increase the accept- high-vulnerability buildings, with an impact on ag- ability of risk management policies; it is also a means gregate risk larger than the risk reduction from of enhancing the technical quality of the analysis and more-resistant, compliant buildings. ensuring that risk management strategies are reason- able and well developed.41 A collective approach al- lows the transfer of risks to the actors that are best Coordination failures impair risk able to manage them—for instance, because of their management access to knowledge and resources. In the manage- Coordination failures between different state agen- ment of natural disaster risks, for example, a set of cies may also hinder risk management. Horizontal promising initiatives has been implemented to im- coordination is needed to ensure that actions from prove coordination, based on the creation of multi- different ministries are consistent and synergetic. For ministry bodies in charge of information exchange instance, the ministry of finance may create and regu- and coordination. The responsibility for risk man- late a health care insurance system, but the usefulness agement is located in the highest office (the prime of the insurance will depend on the availability of minister’s or president’s office) in about 25 percent competent health care providers, a responsibility of of the countries and in a central planning or coordi- the health ministry. Vertical coordination is also cru- nation unit in 10 percent of them.42 In Peru, the re- cial because risk management must be shared across sponsibility for disaster risk management resides in a different levels (from the neighborhood to the coun- new agency within the president’s office and is there- try and the global communities). A public-private fore able to coordinate across ministries. The time insurance scheme (regulated by the ministry of fi- and resources consumed by coordination actions nance at the national level) cannot be designed inde- should not be underestimated, however, and the cost pendently of the implementation of risk reduction of doing so may be important in countries where Beyond the ideal: Obstacles to risk management and ways to overcome them 91 BOX 2.3 Institutions to improve risk management: National risk assessments National risk assessments (NRAs) to improve policy related to pre- pates trade-offs, helping to avoid conflicts between stakeholders venting and planning for crises and emergencies have been con- during a crisis. In the Netherlands, that is the case during floods, ducted in the United Kingdom and in the Netherlands since 2005 when several regions compete for access to limited resources and 2007, respectively.a In both countries, NRAs are used to assess needed for emergency management. Third, it helps involve new the main risks faced by the country, regardless of the type and origin actors. The private sector has a key role in risk and crisis manage- of risk (natural, technological, terrorist, other). They are based on ment, and NRAs have been used to involve them in the develop- similar approaches: identifying risks, generating scenarios, assessing ment of risk management strategies. Finally, an NRA influences the the probability or plausibility and impacts of the risks, and construct- distribution of resources dedicated to managing different risks. ing a national risk matrix. The matrix in the figure summarizes the Following a similar approach, the government of Morocco is main risks and organizes them according to their likelihood (x-axis) considering a multirisk approach. With the support of the World and severity of impact (y-axis). Bank and the Global Facility for Disaster Reduction and Recovery, Several major benefits emerge from conducting such an assess- Morocco has conducted risk assessments in three key areas: natural ment. First, it helps with coordination and cooperation across minis- disaster risk, volatility in commodity prices, and risks in the agricul- tries and organizations, thus avoiding “silo” effects. Ministries in tural sector. The country is seeking to adopt an explicitly integrated charge of one risk (for instance, the ministry of health, for epidemics) approach to assessing its key risks and is now beginning to develop have found that the NRA helped them mobilize other ministries to options on how best to mitigate the identified risks, including provide information and design their own response plan (for exam- through developing a national risk management strategy and sup- ple, the ministry of education needs to set up a response plan to porting institutions. cover the event of many teachers becoming sick). Second, it antici- The United Kingdom has adopted a comprehensive risk management framework 5 Pandemic influenza Coastal flooding 4 Effusive volcanic Overall relative impact score eruption Severe space Other infectious weather Major industrial Major transport diseases 3 Low temperatures accidents accidents and heavy snow Inland flooding Heatwaves Zoonotic animal Explosive diseases volcanic eruption 2 Storms and gales Drought Public disorder 1 Non-zoonotic Disruptive animal diseases industrial action Between Between Between Between Greater than 1 in 20,000 and 1 in 2,000 and 1 in 200 and 1 in 20 and 1 in 2 1 in 2,000 1 in 200 1 in 20 1 in 2 Relative likelihood of occurring in the next five years Source: Cabinet Office 2012. Source: WDR team based on Vastveit 2011. a. Ministry of the Interior and Kingdom Relations 2009; Cabinet Office 2012. 92 WORLD DEVELOPMENT REPORT 2014 public resources are scarce. That is particularly the government should ensure free access to data; free case where the public sector struggles to attract dissemination of results through media, the Internet, skilled and motivated workers. and social networks; and some legal protection for whistleblowers. Many countries provide protection for individuals who report alleged dishonest or ille- Political economy problems hinder risk gal activities that have been occurring in a govern- management ment department or private company or organiza- Political economy obstacles. Even when resources are tion (in the United States, the first whistleblower law available, politicians may be reluctant to devote them was passed in 1863). Recent progress in this direction to risk management because the costs of risk man- has been achieved by many countries, including agement are immediate, concentrated, and observ- Jamaica and India. able, while the benefits are longer term, distributed more broadly, and often less visible. When regulating Lack of well-accepted indicators for risk. The lack of the use of new chemicals or the development of a well-accepted indicators for risk makes it difficult to new area, for instance, public decision makers have a measure the performance of decision makers and to strong and immediate influence on the revenues of hold them accountable for their risk management one or a few firms (when regulating chemicals) or on choices. Controlling and coordinating the delivery the value of people’s assets (landowners).43 Affected of public services is difficult when the potential firms and people will naturally tend to oppose any for competition in quasi-markets—that is, markets constraint and be very vocal about it. On the other created and organized by the government to create hand, the people protected by the regulation—peo- competition among public service providers—is ple negatively affected by pollution or future buyers limited and when the quality of the service is not eas- of apartments in the newly developed areas—are of- ily observable.47 Health care is a common illustration ten not aware that the regulation may eventually pro- of this problem: service quality is not directly observ- tect them and therefore rarely support it. Even more able, and results can evolve over the long term and complicated are cases such as climate change, where are always very uncertain. Risk management is no beneficiaries are not born yet, because the benefit different: while forgone profits and lost jobs from a from risk management extends over the very long chemical firm can be measured and published, the term (see chapter 8). reduction in risk from banning some potentially car- The existence of dispersed interests is a classic is- cinogenic product cannot be easily measured. In sue in institution building and is responsible for general, the fact that disaster relief is immediate and many government failures, especially when public pertinent while prevention is less visible and more goods are concerned.44 Policy trade-offs are often difficult to measure makes it impossible to enforce determined by the ability of various interests to the accountability of decision makers, leading to organize themselves: for instance, through lobbying biased spending decisions toward less cost-effective organizations. Evidence shows that increasing trans- ex post action. parency and providing a voice to dispersed interests The long horizon of risk management actions help avoid capture by interest groups and improve and the lack of indicators mean that it is also difficult policy decisions.45 These political issues have beset to use competition to control public risk manage- many well-known efforts to control risks, such as ment. In theory, competition across localities should those relating to asbestos, lead paint, and tobacco. be an incentive to risk management: localities often These cases show that powerful lobbies can block rely on local taxes on economic activity, and the risk health regulation even in the presence of well- level can be a determining factor for a private actor established scientific evidence of negative health im- who wants to invest in one locality or another. But pacts. They also demonstrate that nongovernmental the risk level is often not directly observable, and de- organizations, scientific organizations, and citizen cades can pass before a good risk management action associations play a key role in bringing these issues to translates into a lower risk level. Competition can the public and creating broad support for regulating thus hardly be used to discriminate between good these risks.46 Risk management is thus more likely to and bad risk management. In such a context, regula- be efficient where strong civil society organizations tory approaches have high potential.48 One promis- are able to conduct independent risk audits and as- ing option is the creation of national risk boards in sessments and to communicate their results to the charge of conducting risk assessments and assessing wider public. To make these possible, however, the the quality of risk management of various agencies Beyond the ideal: Obstacles to risk management and ways to overcome them 93 and organizations (including local authorities and fects, and the risk management policy mix may need their land use plans) through risk audits and bench- to include redistributive measures to be accepted by a marking (see box 2.3; see also the “Focus on policy majority and not harm the poorest.52 reform” at the end of this Report). Risk assessments In this context, conditional cash transfers have could be used to create indicators that would help proven to be able to deliver both distributional ben- populations reward risk-sensitive policy making; efits to the poor, as well as positive impacts in terms they would also trigger risk-based competition of risk management. The Bolsa Família Program in across agencies and localities to encourage good risk Brazil, for instance, has not only been a main driver management. National risk boards could also help for significantly reducing inequality and extreme with vertical coordination issues, by mitigating the poverty in recent years but has also improved in- public moral hazard created by the national support dividual risk management. The underlying idea of to affected subnational entities. the program is to provide poor households with monthly cash transfers that are conditioned on Preference for one policy in advance but another when compliance with certain risk management behav- the time comes to implement it (time inconsistency). iors, such as completing health check-ups, monitor- Sometimes after a shock or crisis, a government will ing growth of children, or pursuing adequate care have an interest in acting in a way that contradicts its for pregnant women. The program has been able to commitments before the event. To cite a recent and achieve compliance rates for these health condition- major example: to avoid excessive risk taking in the alities of close to 100 percent, thereby significantly financial sector, the government may promise improving the management of health risks of not to bail out bankrupt financial institu- 11 million poor households. tions; but if a large financial institu- tion does go bankrupt, the govern- Under deep uncertainty, policies Uncertainty is sometimes ment will have an interest in severe bailing it out, regardless of its need to be robust in a previous commitments. These large range of possible Sometimes, information about incentives reduce the credibility how to manage risks does not of the entire strategy and create scenarios, and able exist, and decisions involve a a strong moral hazard issue. to be revised as condition known as deep uncer- These problems are amplified by future circumstances tainty. These cases occur when the lag between the short period of warrant. experts cannot agree on which many political mandates and the lon- models to use (disagreement on how ger period needed for risk management to transfer the results of analyses of the results to be observable. Disaster relief can health impact on animals to human beings, even be used opportunistically, by being distributed for example); on the probability distributions of key close to an election or targeted to areas that vote for uncertain parameters (the probability of a long pe- the ruling party.49 riod without economic growth); or on the values of alternative outcomes (the acceptability of a total loss Distributional problems. All risk management policies of the Amazon forest in the event of significant cli- redistribute wealth and power; at their worst, they mate change).53 In such cases, it may be impossible can harm the poorest and the most vulnerable, rais- to define a probability for alternative outcomes, or ing important equity concerns. For instance, increas- even to identify the set of possible futures (including ing building construction costs to improve earth- highly improbable events—like the famous “black quake and flood resilience may make it even more swan”).54 Or it can be impossible to reconcile differ- difficult for inhabitants of informal settlements to ent views through a common estimation of proba- obtain decent housing. In Jakarta, flood- and earth- bilities of different outcomes. A situation of deep quake-proofing a typical home costs $3,100, on aver- uncertainty is different from a situation of “large age, only slightly less than the annual per capita GDP uncertainty,” in which different actors can agree on in Indonesia.50 Even excise taxes on cigarettes, an ef- the probabilities and values of different outcomes, ficient tool to prevent young people from starting even if the range of possible outcomes is very broad smoking and protect them from addiction, have been because knowledge is limited. In situations of deep criticized for their cost to the poor.51 Complementary uncertainty, different stakeholders or experts can policies may be needed to mitigate these negative ef- have divergent opinions and may not even agree on 94 WORLD DEVELOPMENT REPORT 2014 the existence of large uncertainty. These situations plementations, but these climate models are based lead to gridlock and lack of consensus, strong politi- on the same widely accepted laws of physics. And cal opposition to any action, and therefore to paraly- while these models agree on the large patterns of cli- sis. They are particularly difficult to manage when mate change, they can point in opposite directions at large and irreversible damages are possible, when the local scale and for some parameters. For example, decisions cannot be postponed until more informa- depending on the model, rainfall in West Africa tion is available, and when policy or technical op- could increase or decrease by 25 percent by the end tions are “brittle”: that is, very sensitive to small of this century (map 2.2). Such uncertainty is clearly errors in design. an obstacle to the design of water infrastructure able To further complicate matters, uncertainty sur- to deal with floods and droughts in the region. The rounds not only the risk itself, but also the risk man- experts’ ability to forecast future energy demand has agement measures that are implemented and their also been disappointing,58 and few anticipated the efficiency and side effects. In practice, anticipating all rapid technological developments that have occurred consequences of risk management policies is impos- in medicine, energy, or information and communi- sible, and some policies may have unacceptable side cations technology. effects or create other risks. The Koka reservoir in There are many other examples of cases where ex- Ethiopia illustrates this problem: it was built to store perts cannot agree on the assumptions that lie at the water for agriculture and improve food security, but heart of their analysis, and thus reach results that can- its impact on the mosquito population and thus on not be reconciled (a situation labeled “dueling certi- health was not anticipated: as a result, malaria case tudes” by economist Charles Manski). These exam- rates within three kilometers of the reservoirs are 2.3 ples include the evaluation of the fiscal consequences times as great as for those living six to nine kilometers of health care reforms—with large uncertainties con- from the reservoir.55 Around micro-dams, malaria cerning how medical practice will evolve—and the prevalence is as much as 7 times greater than in the impact of various policies to reduce cocaine con- rest of Ethiopia.56 Such side effects cannot always be sumption.59 In these cases, different stakeholders can avoided and need to be monitored and managed.57 rarely come to agreement on the “most likely” future Uncertainty is especially deep in “emerging risks” on which planning should be based, or on the “most or in areas where scientific uncertainty is the greatest likely” consequences of a given action. Moreover, do- (genetically modified crops, hydropower dams, nu- ing so would be dangerous if future risks and events clear energy, climate change). A common example is deviate from this most likely case. the uncertainty about future changes in local cli- A special case of deep uncertainty that can para- mates. Different scientific teams develop simulations lyze action is regulatory and policy uncertainty (see of climate systems that differ in their technical im- chapter 5). Firms working on renewable energy tech- M A P 2 . 2 Deeply uncertain futures: Different climate models project very different changes for precipitation in Africa CC SM3 G F D L- C M 2 . 0 Source: IPCC (Intergovernmental Panel on Climate Change) 2007. Note: The maps show simulations from two different climate models. The brown areas indicate decreases in annual precipitation. The green areas indicate increases in annual precipitation. Beyond the ideal: Obstacles to risk management and ways to overcome them 95 nologies, for instance, are dependent on environ- mental regulations and policy-determined carbon An iterative process of decision making D IAG R AM 2.2 prices. They are highly vulnerable to policy reversals, and this uncertainty is a major obstacle to taking risk to prompt robust action in the face of uncertainty and innovating.60 In situations of deep or large uncertainty, many Multistakeholder analysis traditional methods for decision making under of uncertainties and possible scenarios uncertainty are difficult to apply. And many well- identified biases toward overconfidence and against What are the possible scenarios? the provision of estimates of uncertainty become particularly dangerous.61 Alternative tools are useful Identification of the for communicating the presence and degree of un- Implementation vulnerabilities of existing plans certainty and for finding compromises, more con- and learning In what scenarios does sensual solutions, and options that are less brittle my plan fail? and more acceptable for stakeholders with different beliefs and values. In these cases, it is preferable to Adjustment of plans and introduction of implement adaptive and robust policies that lead to monitoring systems acceptable outcomes in a large range of scenarios and that can be revised when new information is avail- Can I change my plan to avoid failure in these scenarios? able or when the context changes.62 When will I know that my plan is at risk of failure? Many methodologies have been proposed for de- What will I be able to do at that time to correct course? signing such policies, and the best approach depends on the context.63 “Robust decision making” is one of Source: WDR 2014 team. these methodologies. The iterative process, by which multiple stakeholders can identify vulnerabilities and options to reduce them, and then implement them and projects in these cases. Often low- or zero-cost with appropriate monitoring and revision using new options that reduce vulnerability in these extreme knowledge, is shown in diagram 2.2. Its advantage is scenarios can be identified and implemented. For ex- that all actors do not have to agree on what is the ample, an early warning system is useful even when most likely future and on the value of different pos- all floods are supposed to be prevented by hard sible outcomes before a decision is discussed. More- protections. over, it helps identify the uncertainties that do not A balance needs to be struck between the plausi- matter for a given decision, thus focusing the process bility of scenarios and the need to explore the range on what is most important (box 2.4). This approach of possible, if uncertain, events. This balance is dif- also explicitly recognizes that different actors have ficult to define (as illustrated by applications of the different values and beliefs, and it makes the influ- precautionary principle, discussed in box 2.5). Deci- ence of these values on the decision much more ap- sion makers have a tendency to be too conservative parent than with other methods (such as the cost- in their assessment of what is possible and plausible. benefit analysis, in which values and preferences are The U.K. national risk assessment deals with this is- captured in complex valuation techniques). As a re- sue through the use of “reasonable worst-case sce- sult, it helps create a dialogue among stakeholders narios,” but there is no objective way of designing and facilitates reaching an agreement on solutions such scenarios, and subjective judgment will always that are more widely acceptable. be necessary. This subjectivity implies that the de- Above all, decision makers should avoid plans velopment and selection of scenarios must be done that are designed for the most likely outcomes but by policy makers working closely with experts (see that increase the vulnerability to less likely events. box 2.2). Huge dikes built to guard against tsunamis and ty- phoon storm surges may encourage investment in Putting it all together: A policy sequence coastal areas and increase vulnerability to exceptional events that exceed the design level of the dikes.64 Diagram 2.3 presents a set of screens that aid in ana- Taking into account extreme cases requires defining a lyzing a risk in a given context, starting from the as- set of scenarios—including low-probability, high- sessment of risk itself. But designing effective public impact ones—and evaluating the robustness of plans policy must go beyond risk assessment, to analyze 96 WORLD DEVELOPMENT REPORT 2014 B O X 2 . 4 Applying a robust decision-making methodology to deal with the risk of flooding in Ho Chi Minh City Ho Chi Minh City, a low-lying and fast-growing metropolis of 7.4 mil- trends; the uncertainty surrounding these aspects is therefore not a lion people, faces significant and growing flood risk. The city’s plan- problem for designing a flood protection system. However, the cur- ners are seeking to implement an integrated flood risk management rent plan may not be sufficiently robust to plausible impacts from strategy. However, many factors that would affect their choice of climate change. Relying exclusively on the currently planned infra- strategy are deeply uncertain, such as future population, economic structure (“baseline”) would keep risk below recent levels only if growth, and the effects of climate change on rainfall and the sea rainfall intensities increase by no more than approximately 5 percent level. In partnership with Ho Chi Minh City’s Steering Center for and if the Saigon River rises less than 45 centimeters (figure). Various Flood Control, researchers undertook a demonstration project to scientific estimates suggest that both these thresholds may be analyze how robust decision making (RDM) could improve flood exceeded by mid-century. The RDM analysis considered additional management, using the Nhieu Loc-Thi Nghe canal catchment area as measures to ensure risk reduction for increases in rainfall intensity of a case study. The analysis explored 12 different risk management up to 35 percent and increases in the level of the Saigon River of up portfolios, each consisting of combinations of options such as raising to 100 centimeters. The cost of these measures could be reduced if homes and retreating from low-lying areas. Each portfolio was simu- the city implements an adaptive plan, which adds some measures lated in 1,000 scenarios, where each scenario consisted of up to six now and more in the future if needed. The results of the RDM analy- different assumptions about socioeconomic development and cli- sis allow policy makers to evaluate robustness gains of certain strat- mate change in the future. egies, and consider the associated trade-offs against their risk pref- The RDM analysis found that the current infrastructure plan erences and available budgets. Overall, the findings suggest that reduces risk in best-estimate future conditions. Moreover, this plan is these additional actions would significantly improve the robustness robust to a wide range of possible future population and economic of Ho Chi Minh City’s risk management plans. Various risk management strategies, and their robustness to increases in river levels and rainfall intensity 110 Relocate to Safe Areas All Options (17%, 100 cm) (32%, 100 cm) Capture Rainwater 100 (10%, 100 cm) Elevate and Relocate with Adaptive 90 Groundwater and Groundwater and Groundwater and Rainwater Rainwater Rainwater with Adaptive (26%, 100 cm) (14%, 85 cm) Elevate and Relocate Elevate and Relocate 80 Increase in Saigon River levels (cm) (17%, 100 cm) (21%, 100 cm) Lempert, Sriver, and Keller 70 SLR estimate (75 cm) Manage Groundwater 60 (7%, 55 cm) 50 Elevate Buildings (23%, 55 cm) 40 Baseline (6%, 45 cm) 30 MONRE SLR Estimate (30 cm) 20 10 IPCC SREX IPCC SREX Mid Value (20%) High Value (35%) 0 0 5 10 15 20 25 30 35 40 45 50 55 60 % increase in rainfall intensity Source: WDR 2014 team based on Lempert and others 2013. Note: Different estimates of potential rainfall and river levels in the figure are indicated by the blue dashed lines. MONRE = Ministry of Natural Resources and the Environment (Vietnam). SLR = sea level rise. SREX = IPCC special report (IPPC 2012). Beyond the ideal: Obstacles to risk management and ways to overcome them 97 BOX 2.5 Precaution in the face of risk: Striking a balance between costs and benefits with the precautionary principle Precaution in the face of risk has been formalized in many countries The European Union (EU) seems to be following the first version in the form of a “precautionary principle.” Three versions of the prin- of the principle in some areas. The Maastricht treaty (1992) states ciple, in increasing level of stringency, have been identified by that EU policy on the environment “shall be based on the precau- Wiener and Rogers: tionary principle” (Art 130R). A communication from the European Commission published in 2000 states the following: “Where there • “Uncertainty does not justify inaction” (it is possible to regulate are threats of serious or irreversible damage, lack of full scientific without full scientific certainty). certainty shall not be used as a reason for postponing cost- • “Uncertainty does justify action” (regulations are required when effective measures to prevent environmental degradation.” The there is a possibility of danger). communication provides implementation guidance, designed to ensure that the precautionary principle does not lead to paralysis • “Uncertainty requires shifting the burden and standard of proof” or self-contradiction: the measures it recommends are supposed to (potentially risky activities are prohibited until the proponent of be temporary (and to include a plan to collect the information the activity can prove it poses no risks or acceptable risks). needed to conduct a more classical risk analysis); proportional (the Implementation of the precautionary principle may be either threat must be “serious” and the measures “cost-effective”); based explicit or implicit. German laws have mentioned the Vorsorge- on significant evidence of the existence of a risk (“full certainty” is prinzip since the 1970s, and France introduced the precautionary not necessary, but the threat needs to be credible); and based on principle in its national constitution in 2005. In the United States, the an assessment of benefits and costs. concept of “precautionary actions” is regularly mentioned in court Putting the precautionary principle into practice has been diffi- decisions and in laws and regulations. For instance, the Environmen- cult in some cases. Precaution can prevent some risks but can also tal Protection Agency banned leaded gasoline in the 1970s, when pose new risks as well as economic costs. Debates over the precau- the benefits of doing so were still unclear; this measure is now tionary principle have been heated at times (such as when critics known to be cost-effective and responsible for decreasing the num- perceive high costs, low risks, or disguised trade protectionism). But ber of children with lead-related learning deficits in the United the reality of actual application in policy making has been quite States. But the United States has not adopted an overarching pre- diverse, within the United States and the European Union and else- cautionary principle for all regulation; since the late 1970s, every U.S. where, suggesting that concerns about costs and risks are shaping president has instead required impact assessments of the benefits real policy choices. These issues also suggest that more learning, and costs of regulation. The precautionary principle is recognized at and an iterative process, are needed about how to apply and the international level: for instance, by the 1992 declaration of the put into operation a precautionary approach, especially in lower- United Nations Conference on Environment and Development. income environments. Source: WDR 2014 team based on Charnley and Rogers 2011; European Commission 2000; Wiener and Rogers 2002; Wiener and others 2011. the obstacles to risk management that are described 4. Stakeholders may be willing to improve their risk in diagram 2.1 and discussed earlier. A sequence that management, but may be unable to do so because policy makers might consider in conducting a risk they lack resources. In this case, the provision management analysis follows: of additional resources—directly or indirectly— would help. 1. Individuals and firms may fail to manage risks be- cause decision makers (private or public) face in- This sequence is only an example: in practice, the centives that are distorted toward too much or too second, third, or fourth of these considerations could little risk taking. These situations need to be iden- happen first or at the same time as the others. How- tified and corrected as a priority, since other mea- ever, applying these screens in that order can help sures are unlikely to be effective in the presence of identify critical gaps that need to be addressed first bad incentives. and reveal early in the process relatively low-cost in- terventions that are highly effective (such as correct- 2. If incentives are correct, then providing good in- ing incentives through market instruments or better formation is critical to improving the quality of coordination). decisions. In theory, these measures to correct incentives 3. Sometimes, correct incentives and good informa- should pay for themselves over time, since they cor- tion are unable to trigger a change in behavior. rect externalities and market and government fail- Specific measures are required to correct cogni- ures. But in the short run, costs may be high, and tive failures and behavioral biases. over time, there may be tough trade-offs, large dis- 98 WORLD DEVELOPMENT REPORT 2014 D IAG R AM 2.3 A set of screens for assessing obstacles to risk management, and formulating policy responses Assessment Assessment Assessment Assessment Assessment of incentives of access to Policy design of risks of behavior of resources information How much Are bad incentives leading to too Are decision Are Are What policies should risk are much or too little risk taking? makers ill behavior resources be implemented? we facing? informed? biases and access to Because of Because of impairing risk resources too market government management? limited? failures? failure? Introduce Build institutions Improve data Launch Provide Adopt norms and collection and education and public goods multistakeholder Build capacity regulations distribution communication and services iterative decision (e.g., land use Improve campaign (e.g., dikes making plans) Launch and drainage vertical and education Introduce systems) Choose robust Create market horizontal and norms and and flexible solutions instruments coordination communication regulations Build markets Consider worst-case (e.g., risk- Correct bad campaign (e.g., building Provide public scenarios based incentives norms) Introduce norms support for insurance Invest in monitoring Introduce and regulations low-income premium) systems redistribution (e.g., land use and vulnerable instruments plans) households Regularly revise (e.g., buy out Provide policies i programs) international aid focused on prevention Source: WDR 2014 team. tributive impacts, or high political costs (when some In making those choices, policy makers should actors oppose the measure). A low or negative aggre- choose steps that can be taken now but allow for im- gate cost does not mean that a policy is easy to imple- provements later. This practical approach leads to ment. Complementary measures may be needed to specific priorities, building on the five policy princi- cope with political obstacles or with negative side ef- ples discussed at the end of chapter 1. fects over the short term.65 Other measures will be more costly (such as direct investment in building Be realistic dikes), and they can be considered at a later stage. Keep it simple. Simple risk management instruments A way past the obstacles: Choosing policy should be preferred when capacity is low, even though they may be less efficient in theory. Simple priorities regulations, for instance, may be easier to enforce How can policy makers build a strong foundation for than sophisticated (and theoretically more efficient) improved risk management over time? Given the ob- approaches based on market instruments. An exam- stacles they face, they must make hard choices. They ple is car insurance in South Africa, where a fuel tax must be practical in these choices. Their choices helps pay for third-party insurance for motor vehi- sometimes must be second-best: less than ideal, and cles. Such an approach provides a simple and effi- more limited than desired. cient way of compensating accident victims in a Beyond the ideal: Obstacles to risk management and ways to overcome them 99 weak institutional environment where enforcing disaster—and that cannot be easily reversed when the compulsory car insurance is difficult.66 Selecting so- memory of the event has disappeared. Doing so helps lutions that are appropriate for local capacity and prevent some of the negative consequences of peo- resources is particularly critical to ensure that risk ple’s short memory about risk and disaster and com- management does not cause unnecessary harm and pensates for the implementation issues related to excessive costs or create new risks through unin- political economy challenges. While policy makers tended negative policy consequences or political should not wait to initiate such institutional changes economy backlash. until a contingency occurs, the increased public awareness in such a situation will increase the mo- Tailor technology to local circumstances. Adoption of mentum and support for institutionalizing best- the most recent vintage of technologies can help de- practice risk management. For instance, the Nether- veloping countries manage their risks. Technology lands reacted to the 1953 floods by implementing transfers can help: for instance, by making available local maximum acceptable flood risk levels; these globally the most recent drugs and vaccines. What is limitations are fixed by law, making it harder for peo- required, however, is not a simple technology trans- ple to disregard flood management as time goes by.69 fer but the adaptation of technology to fit local The creation of a national risk board is an option to needs and social norms and constraints and to thus create an irreversible institutional change that incen- maximize adoption (see spotlight 1). Successful in- tivizes long-term risk management and helps coordi- novations in risk management in developing coun- nate risk reduction actions. tries have often relied on modern technologies, but always through a complex process of adaptation. Start with soft measures that change incentives or The Bangladesh hurricane early warning system, for make them more effective. Hard measures (such as instance, combines modern hurricane track forecast dikes) or complex risk-sharing mechanisms are very technologies with low-tech, locally designed com- unlikely to be efficient and sustainable if incentives munication tools, including handheld bullhorns, are distorted toward too much or too little risk tak- bicycle-mounted loudspeakers, and house-to-house ing. Obstacles to risk management related to incen- contacts.67 tives need to be identified and corrected as priori- ties, through institutional reforms and economic Concentrate on low-hanging fruit and win-win instruments (from regulation to market instru- solutions. Relatively low-cost interventions that are ments), to communication and information cam- highly effective should be favored. Cleaning the storm paigns, and behavioral approaches. Starting with drains in Mumbai helps flood control. But beyond soft measures can correct for a bias in risk manage- that, it improves health and hygiene and even the ment toward hard and capital-intensive solutions— quality of life in neighborhoods, leading to a virtuous even when cheaper and more flexible institutional cycle. Removing bad incentives can be extremely ef- solutions are available. ficient and relatively inexpensive, even though it may not be easy to do. For instance, reforms can target fos- Choose flexible solutions and build in learning. To sil fuel subsidies that promote energy-inefficient cope with uncertainty and differences in beliefs, val- transport and heating and thus increase health risks ues, and sensitivity, policy makers should aim for ro- from local air pollution. Strengthening the capabili- bust policies that may not be optimal in the most ties that are generally useful to manage risks of differ- likely future but that lead to acceptable outcomes in ent natures is particularly cost-effective.68 For exam- a large range of scenarios and that are adaptive and ple, the ability to manage large-scale evacuation is the flexible: that is, policies that are easy to revise as new same whether the reason is a flood or a technological information becomes available. More learning, and accident, as is the capacity to scale up cash transfers an iterative process of monitoring and learning, rapidly after a shock, whether the shock is a natural is needed about how to apply risk management hazard or an economic crisis. approaches, especially in lower-income environ- ments.70 One way to maximize learning is to learn from other domains where experience is system- Build a strong foundation for improved risk atized and internationalized. An example is the Inter- management over time national Civil Aviation Organization, a specialized Create institutional arrangements when the need for agency of the United Nations that defines the proto- them is obvious in everybody’s mind—such as after a cols for investigations of aviation accidents and 100 WORLD DEVELOPMENT REPORT 2014 shares the results, ensuring that everybody can ben- knowledge on disaster risk management and post- efit from the mistakes of the others. “Learning from disaster reconstruction with other countries vul- Megadisasters,” a knowledge-sharing project spon- nerable to disasters. Improving the ability of the sored by the government of Japan and the World international community to share information on Bank, collects and analyzes information, data, and risks—from health and road-accident risks to large- evaluations on the Great East Japanese Earthquake scale disasters—would be a useful input into the de- of March 11, 2011, with the aim of sharing Japan’s sign of more robust strategies (see chapter 8). Beyond the ideal: Obstacles to risk management and ways to overcome them 101 Notes 36. Krueger 1990. 37. James 2000, 330. 1. “Brimstowad May Miss Its 2015 Deadline,” Hindustan Times, 38. World Bank 2012. July 26, 2012. 39. Kunreuther 2006. 2. World Bank and GFDRR 2013. 40. Michel-Kerjan 2008. 3. FEMA 2010. 41. Stern and Fineberg 1996. 4. Guiso 1998. 42. UNISDR 2007. 5. See the Ecomet’s website at www.ecomet.eu. 43. Viguie and Hallegatte 2012. 6. World Bank and GFDRR 2013. 44. Olson 1965. 7. Johnson and Goldstein 2003. 45. World Bank 2000, chapter 3. 8. Kahneman and Tversky 1979. 46. Blanke 2011. 9. Banerjee and Duflo 2011. 47. Ferlie 1992; James 2000. 10. Sjöberg 2000. 48. James 2000. 11. Commissariat Général du Plan 2001. 49. Healy and Malhotra 2008. 12. Weber and Johnson 2012. 50. Kunreuther and Michel-Kerjan 2012. 13. Sunstein 2011. 51. Jha and Chaloupka 2000. 14. Engel and Weber 2007. 52. Viguie and Hallegatte 2012. 15. Michel-Kerjan, de Forges, and Kunreuther 2012. 53. Hallegatte and others 2012. 16. Meyer 2010. 54. Taleb 2010. 17. Hallegatte 2012. 55. Lautze and others 2007. 18. Banerjee and Duflo 2011. 56. Ghebreyesus and others 1999. 19. Hyder and Aggarwal 2009. 57. Graham and Wiener 1995. 20. Kiefer and Montjoy 2006. 58. Craig, Gadgil, and Koomey 2002. 21. Spears 2013. 59. Manski 2011. 22. Cohen 2005. 60. See the Climate Policy Initiative’s website at http://climate 23. Telles, Singh, and Joshi 2009. policyinitiative.org/publication/risk-gaps. 24. Laffont 1995. 61. Manski 2011. 25. Austin, Kristinsson, and Anderson 1999. 62. Pate-Cornell 2012. 26. Lall and Deichmann 2012. 63. Hallegatte and others 2012. 27. Todo, Nakajima, and Matous 2013. 64. Hallegatte 2012. 28. Kahan 2000. 65. World Bank 2012. 29. La Ferrara, Chong, and Duryea 2008. 66. Smith 2006. 30. UNISDR 2011. 67. Paul 2010. 31. Hallegatte 2012. 68. Ministry of the Interior and Kingdom Relations 2007. 32. Peters and Slovic 1996. 69. Slomp 2012. 33. Wiener and others 2011. 70. Dewey 1927. 34. Hattis and others 2001. 35. World Bank 2009. 102 WORLD DEVELOPMENT REPORT 2014 References Deep Uncertainty: Application to Climate Change.” Policy Research Working Paper 6193, World Bank, Washington, DC. Austin, Daren J., Karl G. Kristinsson, and Roy M. Anderson. 1999. Hattis, Dale, Abel Russ, Robert Goble, Prerna Banati, and Marga- “The Relationship between the Volume of Antimicrobial ret Chu. 2001. “Human Interindividual Variability in Suscepti- Consumption in Human Communities and the Frequency of bility to Airborne Particles.” Risk Analysis 21 (4): 585–99. Resistance.” Proceedings of the National Academy of Sciences of Healy, Andrew J., and Neil Malhotra. 2008. “Mass and Elite Prefer- the United States of America 96 (3): 1152–56. ences for Disaster Relief and Prevention Spending: Retrospec- Banerjee, Abhijit V., and Esther Duflo. 2011. Poor Economics: tive Voting and Failures in Electoral Accountability.” Econom- A Radical Rethinking of the Way to Fight Global Poverty. New ics Department Working Paper, Loyola Marymount University, York: PublicAffairs. Los Angeles. Blanke, D. Douglas. 2011. “Tobacco.” In Wiener and others, The Hyder, Adnan A., and Anju Aggarwal. 2009. “The Increasing Bur- Reality of Precaution, 91–120. den of Injuries in Eastern Europe and Eurasia: Making the Cabinet Office (United Kingdom). 2012. “National Risk Register Case for Safety Investments.” Health Policy 89 (1): 1–13. of Civil Emergencies.” Cabinet Office, London. IPCC (Intergovernmental Panel on Climate Change). 2007. Cli- Charnley, Gail, and Michael D. Rogers. 2011. “Frameworks for mate Change 2007: The Physical Science Basis. Contribution of Risk Assessment, Uncertainty, and Precaution.” In Wiener and Working Group I to the Fourth Assessment Report of the others, The Reality of Precaution, 361–75. IPCC. Cambridge, UK: Cambridge University Press. Cohen, Alma. 2005. “Asymmetric Information and Learning: ———. 2012. Managing the Risks of Extreme Events and Disasters Evidence from the Automobile Insurance Market.” Review of to Advance Climate Change Adaptation. A Special Report of Economics and Statistics 87 (2): 197–207. the Intergovernmental Panel on Climate Change. New York, Commissariat Général du Plan. 2001. “Transports: Choix Des In- NY: Cambridge University Press. vestissements Et Coût Des Nuisances.” Commissariat Général James, Oliver. 2000. “Regulation inside Government: Public Inter- du Plan, Paris. est Justifications and Regulatory Failures.” Public Administra- Craig, Paul P., Ashok Gadgil, and Jonathan G. Koomey. 2002. tion 78 (2): 327–34. “What Can History Teach Us? A Retrospective Examination of Jha, Prabhat, and Frank J. Chaloupka. 2000. Tobacco Control Poli- Long-Term Energy Forecasts for the United States.” Annual cies in Developing Countries. New York: Oxford University Review of Energy and the Environment 27 (1): 83–118. Press. De Francesco, Fabrizio. 2012. “Diffusion of Regulatory Impact Johnson, Eric J., and Daniel Goldstein. 2003. “Do Defaults Save Analysis among OECD and EU Member States.” Comparative Lives?” Science 302 (5649): 1338–39. Political Studies 45 (10): 1277–305. Kahan, Dan M. 2000. “Gentle Nudges vs. Hard Shoves: Solving the Dewey, John. 1927. The Public and Its Problems. New York: Henry Sticky Norms Problem.” University of Chicago Law Review 67 Holt and Company. Engel, Christoph, and Elke U. Weber. 2007. “The Impact of Institu- (3): 607–45. tions on the Decision How to Decide.” Journal of Institutional Kahneman, Daniel, and Amos Tversky. 1979. “Prospect Theory: Economics 3 (3): 323–49. An Analysis of Decision under Risk.” Econometrica 47 (2): European Commission. 2000. “Communication from the Com- 263–91. mission on the Precautionary Principle.” Communication Kiefer, John J., and Robert S. Montjoy. 2006. “Incrementalism be- COM(2000) 1 final, European Commission, Brussels. fore the Storm: Network Performance for the Evacuation of FEMA (Federal Emergency Management Agency). 2010. “Local New Orleans.” Public Administration Review 66 (s1): 122–30. Official Survey Findings on Flood Risk.” FEMA, Washington, Krueger, Anne O. 1990. “Government Failures in Development.” DC. Journal of Economic Perspectives 4 (3): 9–23. Ferlie, Ewan. 1992. “The Creation and Evolution of Quasi Markets Kunreuther, Howard. 2006. “Disaster Mitigation and Insurance: in the Public Sector: A Problem for Strategic Management.” Learning from Katrina.” ANNALS of the American Academy of Strategic Management Journal 13 (S2): 79–97. Political and Social Science 604 (1): 208–27. Ghebreyesus, Tedros A., Mitiku Haile, Karen H. Witten, Asefaw Kunreuther, Howard, and Erwann Michel-Kerjan. 2012. “Policy Getachew, Ambachew M. Yohannes, Mekonnen Yohannes, Options for Reducing Losses from Natural Disasters: Allocat- Hailay D. Teklehaimanot, Steven W. Lindsay, and Peter Byass. ing $75 Billion.” Copenhagen Consensus 2012 Challenge 1999. “Incidence of Malaria among Children Living near Paper, Copenhagen Consensus Center, Washington, DC. Dams in Northern Ethiopia: Community Based Incidence La Ferrara, Eliana, Alberto Chong, and Suzanne Duryea. 2008. Survey.” British Medical Journal 319 (7211): 663–66. “Soap Operas and Fertility: Evidence from Brazil.” Working Graham, John D., and Jonathan B. Wiener. 1995. Risk Versus Risk: Paper 172, Bureau for Research and Economic Analysis of De- Tradeoffs in Protecting Health and the Environment. Cam- velopment, Durham, NC. bridge, MA: Harvard University Press. Laffont, Jean-Jacques. 1995. “Regulation, Moral Hazard and Insur- Grossi, Patricia, and Robert Muir-Wood. 2006. “Flood Risk in New ance of Environmental Risks.” Journal of Public Economics 58 Orleans: Implications for Future Management and Insurabil- (3): 319–36. ity.” Risk Management Solutions, Newark, CA. Lall, Somik V., and Uwe Deichmann. 2012. “Density and Disasters: Guiso, Luigi. 1998. “High-Tech Firms and Credit Rationing.” Jour- Economics of Urban Hazard Risk.” World Bank Research Ob- nal of Economic Behavior & Organization 35 (1): 39–59. server 27 (1): 74–105. Hallegatte, Stéphane. 2012. “An Exploration of the Link between Lautze, Jonathan, Matthew McCartney, Paul Kirshen, Dereje Development, Economic Growth, and Natural Risk.” Policy Olana, Gayathree Jayasinghe, and Andrew Spielman. 2007. Research Working Paper 6216, World Bank, Washington, DC. “Effect of a Large Dam on Malaria Risk: The Koka Reservoir in Hallegatte, Stéphane, Ankur Shah, Robert Lempert, Casey Brown, Ethiopia.” Tropical Medicine & International Health 12 (8): and Stuart Gill. 2012. “Investment Decision Making under 982–89. Beyond the ideal: Obstacles to risk management and ways to overcome them 103 Lempert, Robert, Nidhi Kalra, Suzanne Peyraud, Zhimin Mao, Stern, Paul C., and Harvey V. Fineberg. 1996. Understanding Risk: Sinh Bach Tan, Dean Cira, and Alexander Lotsch. 2013. “En- Informing Decisions in a Democratic Society. Washington, DC: suring Robust Flood Risk Management in Ho Chi Minh City.” National Academy Press. Policy Research Working Paper 6465, World Bank, Washing- Sunstein, Cass R. 2011. “Precautions against What? Perceptions, ton, DC. Heuristics, and Culture.” In The Reality of Precaution, Loewenberg, Samuel. 2006. “US Chemical Companies Leave Their 492–517. Mark on EU Law.” Lancet 367 (9510): 556–57. Taleb, Nassim N. 2010. The Black Swan: The Impact of the Highly Manski, Charles F. 2011. “Policy Analysis with Incredible Certi- Improbable. New York: Random House. tude.” Economic Journal 121 (554): F261–F89. Telles, Shirley, Nilkamal Singh, and Meesha Joshi. 2009. “Risk of Meyer, Robert. 2010. “Why We Still Fail to Learn from Disasters.” Posttraumatic Stress Disorder and Depression in Survivors of In The Irrational Economist: Making Decisions in a Dangerous the Floods in Bihar, India.” Indian Journal of Medical Sciences World, edited by Erwann Michel-Kerjan and Paul Slovic, 63 (8): 330–34. 124–31. New York: Public Affairs. Todo, Yasuyuki, Kentaro Nakajima, and Petr Matous. 2013. “How Michel-Kerjan, Erwann. 2008. “Disasters and Public Policy: Can Do Supply Chain Networks Affect the Resilience of Firms to Market Lessons Help Address Government Failures?” Pro- Natural Disasters? Evidence from the Great East Japan Earth- ceedings of the 99th National Tax Association Annual Confer- quake.” Discussion Paper Series 13-E-028, Research Institute ence, Boston, MA. of Economy, Trade and Industry, Tokyo. Michel-Kerjan, Erwann, Sabine Lemoyne de Forges, and Howard UNISDR (United Nations International Strategy for Disaster Re- Kunreuther. 2012. “Policy Tenure under the U.S. National duction). 2007. “Hyogo Framework for Action 2005–2015: Flood Insurance Program (NFIP).” Risk Analysis 32 (4): Building the Resilience of Nations and Communities to Disas- 644–58. ters.” United Nations, Geneva. Ministry of the Interior and Kingdom Relations (the Nether- ———. 2011. “Global Assessment Report on Disaster Risk Reduc- lands). 2007. National Security: Strategy and Work Programme tion: Revealing Risk, Redefining Development.” United Na- 2007–2008. The Hague: Ministry of the Interior and Kingdom tions, Geneva. Relations. Vastveit, Kirsti R. 2011. “The Use of National Risk Assessments in ———. 2009. Working with Scenarios, Risk Assessment and the Netherlands and the UK.” Master’s dissertation, University Capabilities in the National Safety and Security Strategy of the of Stavanger, Stavanger, Norway. Netherlands. The Hague: Ministry of Interior and Kingdom Viguie, Vincent, and Stéphane Hallegatte. 2012. “Trade-Offs and Relations. Synergies in Urban Climate Policies.” Nature Climate Change 2 Olson, Mancur. 1965. The Logic of Collective Action: Public Goods (5): 334–37 and the Theory of Groups. Cambridge, MA: Harvard University Weber, Elke U., and Eric J. Johnson. 2012. “Psychology and Behav- Press. ioral Economics Lessons from the Design of a Green Growth Pate-Cornell, Elisabeth. 2012. “On ‘Black Swans’ and ‘Perfect Strategy.” Policy Research Working Paper 6240, World Bank, Storms’: Risk Analysis and Management When Statistics Are Washington, DC. Not Enough.” Risk Analysis 32 (11): 1823–33. Wiener, Jonathan B. 2013. “The Diffusion of Regulatory Over- Paul, Bimal K. 2010. “Why Relatively Fewer People Died? The sight.” In The Globalization of Cost-Benefit Analysis in Environ- Case of Bangladesh’s Cyclone Sidr.” Natural Hazards 50 (2): mental Policy, edited by Michael A. Livermore and Richard L. 289–304. Revesz, 123–41. New York: Oxford University Press. Peters, Ellen, and Paul Slovic. 1996. “The Role of Affect and World- Wiener, Jonathan B., and Michael D. Rogers. 2002. “Comparing views as Orienting Dispositions in the Perception and Accep- Precaution in the United States and Europe.” Journal of Risk tance of Nuclear Power.” Journal of Applied Social Psychology Research 5 (4): 317–49. 26 (16): 1427–53. Wiener, Jonathan B., Michael D. Rogers, James K. Hammitt, and Renn, Ortwin, and Peter Graham. 2005. “Risk Governance: To- Peter H. Sand. 2011. The Reality of Precaution: Comparing Risk wards an Integrative Approach.” White Paper 1, International Regulation in the United States and Europe. Washington, DC: Risk Governance Council, Geneva. RFF Press. Sjöberg, Lennart. 2000. “Factors in Risk Perception.” Risk Analysis World Bank. 2000. World Development Report 2000/2001: Attack- 20 (1): 1–12. ing Poverty. Washington, DC: World Bank. Slomp, Robert. 2012. “Flood Risk and Water Management in the ———. 2009. World Development Report 2010: Development and Netherlands: A 2012 Update.” Rijkswaterstaat, Ministry of In- Climate Change. Washington, DC: World Bank. frastructure and the Environment, The Hague. ———. 2012. Inclusive Green Growth: The Pathway to Sustainable Smith, Stephen. 2006. “Taxes on Road Transport.” In Excise Tax Development. Washington, DC: World Bank. Policy and Administration in Southern African Countries, World Bank and GFDRR (Global Facility for Disaster Reduction edited by Sijbren Cnossen, 117–50. Pretoria: University of and Recovery). 2013. “Weather, Climate and Water Hazards South Africa Press. and Climate Resilience: Effective Preparedness through Na- Spears, Dean. 2013. “How Much International Variation in Child tional Meteorological and Hydrological Services.” World Height Can Sanitation Explain?” Policy Research Working Bank, Washington, DC. Paper 6351, World Bank, Washington, DC. S P OT L I G H T 2 Protecting the food consumption of the poor: The role of safety nets in Ethiopia and El Salvador Around the world, 870 million people are chronically undernourished. Environmental degradation, climate change, and food price spikes put additional pressures on the ability of poor people to meet their basic food consumption needs. The experiences of Ethiopia and El Salvador show that investments in safety nets can be an effective mechanism to reduce chronic food insecurity and respond to spikes in food prices. Building resilience to droughts in Ethiopia provide funding during a crisis should the contingency For more than 30 years, emergency food aid was the pri- budget be exhausted; and an emergency response system mary response to food insecurity in Ethiopia. The aid saved to cover districts that are not participating in the PSNP. lives, but it did not prevent asset depletion, nor did it help Evaluations have shown that the program has reduced households rebuild their livelihoods after droughts. To household food insecurity. Receiving PSNP transfers for move from crisis response to prevention, the government five years is associated with an increase of food security for of Ethiopia launched the Productive Safety Net Program 1.05 months a year compared with having received no (PSNP) in 2005. Its main objective is to protect food con- transfers, and 1.53 months when households receive both sumption and prevent asset depletion of rural households PSNP and HABP services. The PSNP also reduced distress that lack food security. The program includes a public sales of assets, from 51 percent of beneficiary households works component and provides direct grants for those at the start of the program to 34 percent by 2010. In addi- who cannot work. By 2009 the PSNP had become the larg- tion, the public works projects helped create valuable com- est program of its kind in Sub-Saharan Africa, supporting munity assets such as roads, schools, and clinics, and im- 7.6 million chronically food insecure people. In 2011, when proved agricultural productivity through water and soil the Horn of Africa suffered its worst drought in 60 years, conservation. the PSNP was expanded to cover 9.6 million people. It re- The program has outperformed traditional humanitar- lies on donor contributions but is managed by the govern- ian responses in timeliness of disbursements during major ment; its budget is about 1.2 percent of gross domestic droughts. During the 2011 drought, the government pro- product. vided aid within two months after a drought warning, us- The program provides predictable cash and food trans- ing the risk financing facility—the typical humanitarian re- fers for six months of the year corresponding to the so- sponse is eight months. called “lean season,” when households tend to run out of The PSNP has demonstrated that receipt of timely and food. The transfers cover about 40 percent of a recipient’s predictable assistance enables households to manage risks annual food needs. A complimentary initiative, the House- more effectively and reduce the use of costly coping mech- hold Asset Building Program (HABP), provides microcredit, anisms. The program also has positive community-wide agricultural extension, and business advice to help PSNP impacts through creation of public assets and environmen- recipients improve their asset base, so that eventually they tal restoration, which should facilitate long-term improve- can manage without aid. About three of every four PSNP ments in livelihoods. It also demonstrates the benefits of beneficiaries also had access to some HABP services from shifting the donor community’s approach from meeting 2006 to 2010. short-term food needs through emergency relief to ad- Program participants are chosen through a combina- dressing the underlying causes of household food insecu- tion of geographic and community targeting. At the start rity and investing in permanent systems with the ability to of the program, the government identified the most food manage risk proactively. insecure districts based on a long history of emergency food aid. Within these districts, household eligibility is de- Ensuring food security in the face of natural disasters termined through a community-based selection process and economic shocks in El Salvador according to predefined criteria. The wage rate for the El Salvador is exposed to multiple shocks that threaten public works program was set low, to discourage better-off food security, including natural disasters—floods, earth- households from applying. quakes, and hurricanes—as well as shocks to the economy, Steps were also taken to respond to the needs of the which is small, dollarized, and dependent on remittances. transitory poor. These included annual retargeting to re- To mitigate the impact of these shocks, the government spond to changes in the relative positions of households; a has introduced several safety nets: conditional cash trans- contingency budget to cover households that might be- fers (CCTs) and pensions in rural areas, school feeding in all come chronically food insecure; a risk-financing facility to public schools, and an income support program in cities. Spotlight 2 105 While the programs were designed with donor support, they are fully PATI had a positive impact on improving job readiness and willing- owned and managed by the government. These safety nets were ef- ness to start a business. Once considered a short-term crisis response fective in protecting the consumption of recipients during food price program, PATI is now a core element of El Salvador’s new Comuni- spikes in 2007–08, job losses and reduction of remittances due to the dades Solidarias Urbanas Strategy, modeled after the CSR but adapted global economic crisis in 2009, and severe storms in 2010. to an urban context. The urban CCT program is currently being pi- The Comunidades Solidarias Rurales, or CSR (formerly Red Soli- loted in 25 municipalities, with plans for expansion. daria), launched in 2005, provides short-term assistance to the ex- El Salvador is an example of a small, relatively poor country that treme poor in rural areas and increases incentives for investments in has been able to design and implement several complex safety net human capital. Pregnant women and mothers of children up to age 5 programs. The government learned from international practice but receive monthly cash transfers of $15. Families with children aged adapted the programs to the local context and added some innova- 6 to 18 also receive the transfers if they comply with regular health tions. These include combining CCTs with activities to improve infra- checkups and preschool and school attendance requirements. In addi- structure and generate income, and conditioning short-term income tion, the program supports improvement of basic services—water, support on training to facilitate longer-term improvements in liveli- sanitation, electricity, health and nutrition, and income-generating hoods of beneficiaries. Political commitment to reducing food insecu- activities—in target localities. In 2012, CSR had about 85,000 benefi- rity and poverty was an important element of El Salvador’s success; ciaries in 100 of the poorest rural communities. the programs were developed continuously under several different CSR was effective in smoothing food consumption and prevent- administrations. A sequenced approach was another characteristic. ing increases in stunting among children in beneficiary households Beginning with a small rural safety net in 17 poor rural communities, when food price inflation doubled in 2008. Beneficiaries of the CCTs the government gradually added more rural areas as administrative reported improvements in nutrition and health as a result of program capacity improved, then a pension component, and is now expand- participation. In 2009, as part of the crisis response program, the gov- ing to urban areas. The government is currently working with donor ernment also introduced monthly pensions of $50 for people older agencies on building a more integrated social protection system that than 70 in CSR-supported communities. An evaluation by the Interna- could provide protection against risks across the entire life cycle. tional Food Policy Research Institute has shown that CSR is well tar- geted, ranking third in its targeting effectiveness in Latin America af- Sources ter programs in Brazil and Chile. Ethiopia The government has also taken steps to address rising food Hobson, Matt, and Laura Campbell. 2012. “How Ethiopia’s Productive Safety Net Pro- insecurity and poverty in cities following job losses and the increase gramme Is Responding to the Humanitarian Crisis in the Horn.” Humanitarian Exchange Magazine 53 (March). of food prices associated with the global economic crisis. A Tempo- Subbarao, Kalanidhi, Carlo del Ninno, Colin Andrews, and Claudia Rodriguez-Alas. rary Income Support (PATI) Program, launched in 2009, provides six 2012. Public Works as a Safety Net: Design, Evidence and Implementation. Wash- months of income support to the urban poor who participate in com- ington, DC: World Bank. munity projects ranging from rehabilitating infrastructure to provid- World Bank. 2010. “Designing and Implementing a Rural Safety Net in a Low-Income ing social services. An innovative feature of the program is that it con- Setting: Lessons Learned from Ethiopia’s Productive Safety Net Program 2005–2009.” World Bank, Washington, DC. ditions income support on attendance of training activities, and thus ———. 2012. “Before Crisis Hits: Can Public Works Programs Increase Food Security?” aims to address income vulnerability and longer-term employability From Evidence to Policy. World Bank, Washington, DC (September). with one integrated instrument. Experienced social workers teach El Salvador participants not only technical skills such as sewing, cooking, or car De Brauw, Alan, and Edmundo Murrugarra. 2011. “How CCTs Help Sustain Human repair, but also entrepreneurship skills and soft skills such as how to Capital during Crises. Evidence from Red Solidaria in El Salvador during the Food prepare for job interviews or write resumes. Price Crisis.” Unpublished manuscript, World Bank, Washington, DC. IFPRI (International Food Policy Research Institute), and FUSADES (Fundación Salva- PATI targets municipalities with the highest urban poverty rates, doreña para el Desarrollo Económico y Social). 2008. “Evaluación de Impacto giving preference to two groups with above-average unemployment Externa de la Red Solidaria: Informe de la Eficacia de la Focalización.” IFPRI and rates: young people and female household heads. The program FUSADES, Washington, DC. started as a pilot in two municipalities in 2009, was rapidly expanded Soares, Fábio Veras. 2012. “What Is Happening with El Salvador’s CCT Programmes?” to 11 municipalities affected by severe storms in 2010, and reached One Pager 168, United Nations Development Programme, Brasilia. World Bank. 2010. “Accessing Good Quality Jobs: Priorities for Education, Social 40,000 beneficiaries in 36 municipalities by 2012. It has proven helpful Protection, Science and Technology.” World Bank, Washington, DC. in rehabilitating urban infrastructure and smoothing beneficiaries’ ———. 2012. “The Evolution of Poverty and Equity in El Salvador, 2000–2010.” food consumption. Preliminary World Bank evaluations show that Unpublished manuscript, World Bank, Washington, DC. PART II The role of key social systems Three generations living together in Indonesia. Household members can increase their resilience by pooling individual energy and resources to manage risks, but internal and external obstacles limit their ability to do so efficiently. © Curt Carnemark/World Bank CHAPTER 3 Households are the first line of support to confront risk and pursue opportunity Sharing good times and enough for her daughter’s marriage. bad times She also feels vulnerable to illness and other negative shocks because of Shamsun Nahar and her family her limited assets.1 have helped one another in good This story illustrates some of times and bad as they struggle to the risks that vulnerable families eh leave poverty behind. Shamsun Hous old face and how they attempt to build is 44 years old and lives in a vil- resilience together. Large shocks lage in central Bangladesh with can force poor people to use costly her 16-year-old daughter. Two sons coping measures that set them back live with their wives nearby but keep and undermine their ability to escape separate households; her eldest son and poverty. In contrast, families that are able to his family live in Dhaka. invest in human capital, accumulate financial and Fifteen years ago, when Shamsun’s husband, physical assets, and share risks among their members Mobarak Molla, was 35, he felt the first symptoms of can become resilient to shocks and are better posi- tuberculosis. Unable to afford proper treatment, he tioned to pursue opportunity. died at age 40. The couple had owned about an acre Jean-Jacques Rousseau wrote in The Social Con- of land and an ox. When Mobarak died, Shamsun tract that “the oldest of all societies, and the only had to sell their ox to cover funeral expenses and buy natural one, is that of the family.” Indeed, for most food. Her three sons continued to work in the field, people, the members of their household constitute but because they had no ox, they had to share the the main source of material and emotional support. plot with another farmer, so they harvested a smaller A household is defined here as a group of individuals crop. Slowly, Shamsun managed to improve her sit- related to one another by family ties (kinship).2 They uation, thanks to a few loans she obtained from her might live under the same roof or not, and they village savings group and a local nongovernmental might be a small nuclear family of parents and chil- organization (NGO). This money enabled her to dren, or a large extended family including grandpar- send her eldest son, Masud, to work for part of the ents and other relatives. In any event, households year as a rickshaw driver in Dhaka—where he even- form a very strong and tightly knit community, tually stayed permanently—as well as to set up a very where members often pool their resources to con- small shop, where she sells necessities such as soap sume, invest, and care for the most vulnerable, and biscuits. Having two sons nearby and one in the among them children and elderly adults. city means Shamsun can get help in times of crisis. Many households, however, particularly poor The steady income from her shop enables her to ones, struggle to help individuals cope with shocks provide the basics for herself and her daughter. and are unable to support their search for opportu- However, Shamsun worries about being able to save nities. As units, they face the challenges both of pro- 109 110 WORLD DEVELOPMENT REPORT 2014 tecting and insuring their members against common individuals; and the actions households can take to shocks, such as illness and income losses, and of ac- improve their preparation in the face of risk and op- cumulating sufficient assets and human capital to portunity. It also focuses on public policies to help grow their income.3 To meet these challenges, house- households prepare for and cope with risk. It starts holds need to have sufficient resources and to be by describing the main shocks that affect households closely connected to their community, to markets, in developing countries, the different strategies that and to good-quality public goods and services. In ad- households use to manage risk, and the obstacles dition, family dynamics and social norms sometimes they encounter. It then discusses how a systemic ap- limit the extent to which members can collaborate proach to policy for risk management should con- effectively, increasing the vulnerability of certain in- sider the multiple risks that households face, and all dividuals within the household—typically women, the different instruments that they need to manage children, and elderly adults—in the face of shocks. these risks effectively. This systemic approach has Government policies can strengthen households’ the benefit of highlighting complementarities and ability to manage risk by facilitating their access to synergies across policies, in particular for protecting information, financial tools, and labor markets. Pub- households against risk while enhancing their access lic policies should also ensure access to education to opportunity. and provide basic protection against health and in- come risk, especially for the poor. And they can ad- What risks do households face and how do dress inequalities within the household through a they cope? mix of regulation and interventions to empower and protect the most vulnerable members. Every day, millions of people all over the world fall This chapter focuses on the internal and external sick, lose their jobs, fall victim to crime, or are hit by obstacles that prevent households from building ad- natural hazards. The wide variation in both the inci- equate preparation and from sharing risk within the dence and the nature of the shocks that affect house- household; the implications of those obstacles for hold members is shown in figure 3.1, which reports the vulnerability and resilience of households and survey data from six countries documenting shocks F I G U R E 3 .1 Shocks to households vary considerably across countries Percentage of households reporting shocks in each category 8.5 3.5 32.8 5.3 4.9 2.2 25.2 11.5 13.9 13.9 1.5 11.4 1.9 18.5 34.2 11.1 5 0.6 38.8 3.3 4.4 32.4 8.8 23.5 25.2 15.8 6.7 3.1 0.6 2.3 1.4 4.2 Malawi Lao PDR Afghanistan Uganda China Peru Asset or crop loss Employment shock Disasters Illness Price shocks Crime or safety Source: Heltberg, Oviedo, and Talukdar 2013 for the WDR 2014. Note: Multiple answers per household were possible, so totals may add up to more than 100 percent. The recall period was 12 months. Households are the first line of support to confront risk and pursue opportunity 111 that household members suffered over the previous found that households protect their consumption— 12 months. Disasters, illness, and asset losses are the at least in part—after illness shocks by using several most common shocks across countries, followed by strategies, including increasing labor supply within price shocks. Rural and urban households are ex- the household.5 A recent study for Indonesia reveals posed to different risks: weather shocks typically af- that while households face significant income risk fect rural households disproportionately, whereas from several kinds of shocks, they manage to achieve urban households are more exposed to price shocks, a level of insurance representing at least 60 percent crime, and unemployment (see chapter 1).4 of this risk. On average, however, expected con- sumption represents 65 percent of the expected income, suggesting that households are willing to Household members can help one another incur a large cost to insure their consumption.6 manage risk and pursue opportunity Looking more closely at how the composition of the When household members share risks, they can in- household affects shock responses, a study of house- crease their own resilience and that of the household. holds in Senegal shows that in urban areas, larger Economic theory suggests that individuals should be households that include extended family members able to smooth consumption over their life cycle. As experience smaller declines in basic consumption chapter 1 argues, they are better able to do so when after being hit by a negative shock, compared with protection and insurance mechanisms—either for- smaller households. At the same time, heads of mal or informal—are available to help them absorb household supporting extended families experience income shocks and maintain stable consumption. In the largest relative declines in (food) consumption, particular, when perfect credit and insurance mar- to preserve the consumption of the other household kets are not available, household members can in- members.7 crease their resilience by pooling individual energy Concern and attention to the most vulnerable and resources to invest in protection and insurance, members distinguish the household from other and to cope with shocks—particularly with idiosyn- groups. Several factors, including altruism, reciproc- cratic shocks. ity, and social norms, explain why household mem- By and large, empirical research shows that bers care for one another (box 3.1). Whatever the households manage to protect their consumption motivation, abundant evidence shows that in all so- from shocks, albeit not fully. Research in Bangla- cioeconomic and cultural contexts, the family is seen desh, Ethiopia, India, Mali, and rural Mexico has as a key pillar of support to the individual. To cite an B O X 3 .1 Altruism, exchange, or social norms: What motivates family members to care for one another? Why do parents invest in their children’s education? Why do adult people provide assistance now in the expectation that they might children take care of their elderly parents? Why do spouses pool their require assistance later.d resources, siblings lend money to one another, and extended family Empirical evidence suggests that each of these theories has members check on their relatives? Economists and sociologists have merit.e For example, reciprocal arrangements such as time-for- long pondered these questions and have come up with three broad money exchanges between parents and children are more common theories for why family members care for one another. in countries where government support is weak and access to mar- According to the theory of altruism, an individual’s welfare kets is limited, suggesting the presence of exchange motives. Kin- depends on the welfare of others. Maximizing utility then involves ship norms may explain why grandparents frequently care for their transferring a portion of one’s resources to others.a More recent grandchildren in certain countries in Sub-Saharan Africa and in research suggests that evolutionary forces such as genes or socio- China. Altruism explains why parents have been found to transfer cultural influences may be behind this altruism.b The second theory money to children on the basis of need, and children to devote suggests that social norms define how family members should greater amounts of time to parents with the worst health. Other evi- help one another.c The third theory, one emphasizing exchange, dence, however, reveals that motives are not always altruistic: the hypothesizes that familial support is rooted in reciprocal arrange- extent of parental investment in children has been shown to influ- ments, which can (but need not) be reinforced by social norms: ence children’s support for their parents. Source: WDR 2014 team. a. Becker 1974. b. Alger and Weibull 2010. c. Alesina and Giuliano 2013. d. Bernheim, Shleifer, and Summers 1985. e. Silverstein and Giarrusso 2010. 112 WORLD DEVELOPMENT REPORT 2014 FIGURE 3.2 Elderly people frequently live with other family members Share of households with at least one member aged 60+ years, and share of population aged 60+ years 50 40 Percent 30 20 10 0 a u es n an n R ico q ria h ia an da il i m aw az in PD r ta da s Ira an na Pe v de ist ist ge an ex Ch kis di Br al Su nz et o jik an la Ug Ni M al M be La Vi ng Ta M Ta gh Uz Ba Af % of households Share of population aged 60+ years Source: WDR 2014 team based on data from household surveys (years range between 2002 and 2011) and United Nations 2009. example, evidence on living arrangements in 17 de- with shocks once they hit—a strategy that increases veloping countries shows that on average 28 percent their vulnerability to future shocks even more. of households have members over the age of 60, a Moreover, complex dynamics within the house- significantly higher share than this age group repre- hold may increase the vulnerability of some mem- sents in the total population (figure 3.2). In many bers. Empirical evidence shows that household circumstances, children and grandchildren take care members—who often have different preferences— of older relatives. For example, in China, 90 percent allocate resources (labor, capital, and output) follow- of elderly with disabilities living with their ing a bargaining process that in many instances families have access to support, compared appears to be inefficient.10 For example, in with 73 percent of elderly living Burkina Faso, husbands and wives cul- alone.8 In other cases, the elderly tivate separate plots, following the Pooling risk take care of younger relatives. In traditional division of labor among Sub-Saharan Africa and China, within the rural couples. Plots run by wives for instance, grandparents are of- family has been a are significantly less productive, ten the main caregivers of their basic form of which implies that household in- grandchildren.9 come could be increased with a insurance from time Obstacles internal and external different allocation of labor across to the household limit the ability of immemorial. plots. In Côte d’Ivoire, husbands its members to manage risks effi- and wives cultivate “gender-specific” ciently, however. As chapter 2 describes crops in separate plots, and they strictly in detail, financial constraints, information allocate the income of each crop to specific constraints, and an inability to translate information consumption categories, such as personal consump- into knowledge and knowledge into action all limit tion, food, and education. Thus, as the income from the ability of individuals to manage risks effectively. one crop fluctuates, so do the consumption expendi- Social obstacles such as missing markets and public tures tied to that income. Consumption decisions goods, moral hazard, externalities, and social norms based on pooled incomes would increase the stability limit the range of instruments that households have of consumption for both spouses as well as for their at their disposal to manage risk. In this context, poor children.11 Experimental evidence from games com- households in particular try to reduce their exposure paring the behavior of husbands and wives also to shocks by opting for low-risk, low-return activi- shows that in many cases their choices fail to maxi- ties and are forced to take costly measures to cope mize joint income.12 Households are the first line of support to confront risk and pursue opportunity 113 To respond to shocks, poor households use poverty. However, households with higher wealth costly strategies, with unequal effects on and better access to infrastructure and services more household members frequently report using their savings or borrowing money, either formally or informally. When shocks damage substantial proportions of The impacts of the more costly coping strategies households’ assets and slash their income, house- can be long term and even permanent, particularly holds with limited savings or insurance mechanisms for children. Empirical evidence from studies of the must take difficult steps to maintain a minimum impact of large shocks (usually natural hazards) on level of consumption, mainly of food. The shock sur- economic and human development outcomes typi- vey data mentioned earlier reveal that more than half cally finds that while all household members experi- of households affected by health and income shocks ence hardship, adult outcomes tend to revert to their in Afghanistan, China, and Tajikistan, and almost long-term trends eventually, whereas children can half of households in Lao People’s Democratic Re- suffer permanent effects, especially during the criti- public and Uganda, reported having cut consump- cal development period of the first two years of life.14 tion.13 Regression analysis of the data shows that in These effects, which tend to be more prevalent in several countries, poor households tend to use costly poor households, translate into lower earnings and coping mechanisms, including selling a productive worse health in adult years. In some cases, larger neg- asset such as livestock, cutting food consumption, ative impacts have been documented for girls than or consuming lower-quality food (table 3.1). Selling for boys. productive assets is particularly harmful because it Shocks can cause long-term damages to human curtails the household’s ability to generate adequate capital when the nutrition of very young children is income for a long time after the shock. Strategies compromised. Children who are heavier and longer such as working longer hours, taking on more work, at birth tend to develop better cognitive skills that or migrating seem to be less clearly associated with enable them to attain more education and get better TA B L E 3 .1 Poorer households are more likely to report using costly mechanisms to cope with shocks Type of household most likely to report coping strategy Assistance Sell Reduce Use savings/ Work more/ (government/family/ productive consumption Country credit/assets migrate community/NGOs) assets quantity/quality Afghanistan Richer Richer Richer Poorer Poorer China — Richer — Richer Richer Iraq Richer Richer — — Poorer Malawi Richer — — — — Mexico Poorer — — — Poorer (credit/asset sales) Nigeria Richer — — — — Peru Richer — — — Sudan Richer Poorer Poorer Poorer Poorer Tajikistan — — — Richer Poorer Uganda Richer (savings/sell — Richer — Poorer assets), poorer (credit) Uzbekistan Poorer (credit) Poorer — Poorer — Source: Heltberg, Oviedo, and Talukdar 2013 for the WDR 2014. Note: The table presents the results of regression estimations where socioeconomic indicators (either consumption quintile or asset-based measure of wealth) significantly affected the probability of reporting the corresponding coping strategy (significance at least 5 percent). Regressions include region and urban fixed effects, household size, gender, education and occupation of head (where available), dependency ratio, consumption quintile, and the principal factor of access to piped water, quality of roof/floor, having a cell phone, and distance to public services and main roads. — = Socioeconomic indicators not significant. No data were available for the sale of productive assets in Peru. 114 WORLD DEVELOPMENT REPORT 2014 jobs. Birth weight is closely related to the quality of of child labor has permanent consequences for their the mother’s nutrition during pregnancy. Therefore, human capital. In northern Tanzania, children who reducing food intake during pregnancy can cause had to work an additional 5.7 hours a week after a large and irreversible damage to the development of rainfall shock attained one year less of school, com- children in utero. Extreme shocks, such as the 1919 pared with those who did not work more.17 influenza pandemic and the great Chinese famine of Stress, fueled by shocks, can increase domestic 1959–61, caused losses in height, cognitive develop- abuse. Adult stress levels increase significantly with ment (measured in years of schooling), and overall shocks. The Asian financial crisis of 1997 increased health outcomes for the generation born during depression and anxiety in Indonesia and Thailand, those years. Shocks can also compromise the quality particularly among the less educated, urban, and of nutrition available to children in their first two landless populations. In Kenya, farmers experience years, which is also essential to physical and cognitive increased levels of cortisol (a hormone produced in development. The 1994–95 drought in rural Zimba- times of stress) when rainfall is too low.18 Focus bwe cut growth by 1.5 to 2.0 centimeters among chil- group participants in Cambodia, Jamaica, and Mon- dren aged 12 to 24 months. Children who were 3 golia reported that the hardship generated by the years old or younger during the 1998–2000 economic 2008 crisis had increased men’s violence toward crisis in Ecuador had a significantly lower height- wives and children.19 Shocks can also lead to in- for-age score and a lower vocabulary test score than creased abuse of elderly people. A study in rural Tan- children of same age in noncrisis times.15 Moreover, zania found that during years of low rainfall, the in a few countries, shocks have been found to hurt number of murders of elderly women—accused of early nutritional and developmental outcomes for witchcraft and killed by their family members— girls disproportionately.16 nearly doubles.20 In the face of disaster, some households also sacri- Physical and psychological abuse experienced fice investments in education in exchange for having during childhood can have long-lasting effects on an additional member—usually a school-age child— identity and behavior, not only undermining the enter the workforce. School attendance dropped by self-esteem that is crucial for decisions regarding risk almost 7 percent among those households more taking and pursuing opportunities but also increas- heavily hit by two strong earthquakes in El Salvador ing the likelihood of violent behavior in adulthood.21 in 2001, while children in these households were two A survey of men in six countries found that those and a half times more likely to be working after the who had been victims of abuse during childhood earthquake than before (the share rose from 6.5 per- were twice as likely to have been violent toward their cent to 16.5 percent). Sometimes, the temporary use partners (figure 3.3). F I G U R E 3 . 3 Men who experienced violence in the household when young are more likely to act violently as adults Percentage of men who have perpetrated violence on an intimate partner as adults 50 40 % within group 30 20 10 0 India Rwanda Chile Brazil Croatia Mexico Did not suffer household violence, psychological and/or physical Suffered household violence, psychological and/or physical Source: WDR 2014 team based on data from Contreras and others 2012. Households are the first line of support to confront risk and pursue opportunity 115 How do households prepare to manage Third, mobile phones can reduce information asym- risks, and what obstacles do they face? metries and price uncertainties, enabling farmers to increase their surplus. Finally, mobile banking offers To confront risk and pursue opportunity, household opportunities for household members to transfer members acquire knowledge and invest in protection money to one another and undertake other financial and insurance. The quality of their risk management transactions in a safe and cost-efficient way.23 depends on their access to information, markets, public services, and infrastructure, as well as on the Investing in human capital to increase level of risk sharing within the household. protection and access to opportunity Better nutrition, sanitation, and access to preventive Acquiring and sharing knowledge about risk health care increase productivity and reduce the risks and opportunity of morbidity and mortality. Historically, one of the Obtaining information is crucial for managing risk. leading causes of premature death in the developing News about such matters as prices, weather risks, world has been the high risk of maternal mortality better agricultural technologies, and job openings and the exposure of young children to malnutrition can immediately improve the ability of a household and disease. In recent years, however, more invest- to prepare for and respond to risk. Mobile phones ment in prevention, better health services, and higher have increasingly become indispensable tools to ob- income have led to a significant decline in infant, tain and exchange information for many households child, and maternal mortality. Immunization rates in the developing world, which account for nearly for measles, for example, which were as low as 60 two-thirds of the world’s 4.77 billion users. First, percent in South Asia and 64 percent in Sub-Saharan by drastically reducing communication costs, mo- Africa in 1990, are now above 75 percent in every bile phones improve cohesiveness of disperse social region of the world (figure 3.4a). Since 2002, infant groups and networks, enabling people to respond mortality rates have declined significantly as well more quickly to the income shocks of other family (figure 3.4b). On the other hand, dietary risks (lead- members.22 Second, readily available information ing to obesity) and smoking continue to increase, on risks, such as weather updates and early warn- which has made noncommunicable disease a leading ings, can assist households in preparing for disasters. cause of death (figure 3.5). FIGURE 3.4 Immunization rates have increased and infant mortality has declined everywhere a. Measles immunization rate b. Changes in infant mortality 100 0 % of children age 12–23 months 90 –5 –10 Change, % 80 –15 70 –20 –25 60 –30 50 –35 1990 2000 2010 cifi ia l A nd be nd ric d CD ia ric n Af ara Af an Pa As As c sia an a a ra a rib a a OE nt pe rth ast h d st h Ca ric Low income Sa Lower middle income ut an Ea Ce ro No e E e e So b- Eu th Am dl Su Upper middle income High income id tin M La 5-year change 10-year change Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: The 5-year change indicates the change in the infant mortality rate between 2007 and 2011. The 10-year change indicates the change between 2002 and 2011. Organisation for Economic Co-operation and Development (OECD) countries in the panel b are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. 116 WORLD DEVELOPMENT REPORT 2014 FIGURE 3.5 The risk of death from noncommunicable disease is growing in all developing regions 100 80 60 Percent 40 20 0 2000 2011 2000 2011 2000 2011 2000 2011 2000 2011 2000 2011 2000 2011 cifi ia l A nd be nd ric d CD ia ric n Af ara Af an Pa As As c sia an a a ra a rib a a OE nt pe rth ast h d st h Ca ric Sa ut an Ea Ce uro No le E e e So b- th Am Su E d id tin M La Injuries Communicable, maternal, perinatal, Noncommunicable diseases and nutritional conditions Source: World Health Organization Global Health Estimates Summary Tables: Deaths by Cause, Age and Sex, available at http://www.who.int/healthinfo/global_ burden_disease/en/ (regional tabulation prepared specifically for the World Development Report 2014). Note: Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. Despite the clear benefits of disease prevention Education helps people achieve better health out- in reducing risk and improving welfare, demand for comes. For example, young people with more tools to reduce the risk of disease can be surpris- education are less likely to engage in substance abuse, ingly low among poor households.24 For example, violence, and unprotected sex.25 In Taiwan, China, the experimental trials that have provided several such 1968 expansion of compulsory education from six to tools—including insecticide-treated bed nets, wa- nine years reduced the likelihood that girls of primary ter disinfectants, soap, multivitamins, and improved school age at the time of the reform would give birth cookstoves—to poor households show that demand to underweight babies as adults, compared with girls falls sharply in response even to small increases in who were not affected by the reform.26 Education also price (figure 3.6). Resource constraints may not be increases productivity and income.27 The education the only explanation for this high price elasticity, system provides children with critical generic skills however; as chapter 2 discusses, many behavioral (literacy and math), as well as “soft” or socioemo- and cognitive biases reduce investment in protec- tional skills, such as effective communication, the tion. Nonetheless, experiments in Guatemala, Kenya, ability to work in teams, and the ability to learn new India, and Uganda that tried to disentangle the fac- concepts and methods. These skills are crucial in the tors behind this behavior found that cash constraints transition from school to work, and they enable peo- explain most of it, whereas education and peer effects ple to adapt to a rapidly changing work environment. have negligible effects. Differences in preferences and Generally speaking, an additional year of education is bargaining power can also hinder the ability of some associated with a 12 percent increase in earnings, al- household members to invest more in protection. In though returns vary greatly across income levels.28 the case of improved cookstoves offered to house- Educational attainment is rising, but low-income holds in Bangladesh, women—who benefit the most countries still lag behind. While much progress has from this technology—were unable to purchase them been made toward the Millennium Development because they have little say over financial decisions. Goal of ensuring that all children complete primary Households also invest in education to manage school, less progress has been made for higher grades risks better and to take advantage of opportunity. of attainment (figure 3.7a). The low attainment, Households are the first line of support to confront risk and pursue opportunity 117 FIGURE 3.6 Demand for preventive health care products falls steeply as the price increases 1 0.9 0.8 0.7 Take-up rate 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Price in 2009 US$ Deworming, Kenya Bednets in clinics, Kenya Water disinfectant, Zambia Soap, India Water disinfectant, Kenya Bednet vouchers, Kenya Source: Abdul Latif Jameel Poverty Action Lab (J-PAL) 2011. F I G U R E 3 . 7 Low-income countries still lag in educational attainment, and some middle-income countries suffer from gaps in quality a. Educational attainment of 15–24 year-olds b. Attainment in mathematics by income 10 500 Average PISA math score 8 400 Years of education 6 300 4 200 2 100 0 0 1960 1970 1980 1990 2000 2010 East Asia Europe and Latin America OECD and Pacific Central Asia and the Low income Lower middle income Caribbean Upper middle income High income Source: WDR 2014 team based on data from Barro and Lee 2010 (panel a) and OECD Programme for International Student Assessment (database) (panel b). Note: In panel b, the red line indicates overall average score. Organisation for Economic Co-operation and Development (OECD) coun- tries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. PISA = Programme for International Student Assessment. 118 WORLD DEVELOPMENT REPORT 2014 FIGURE 3.8 Educational attainment is still uneven for boys and girls from poor households, especially in lower-income countries Gender gap in primary education by level of income Belarus Serbia Kazakhstan Montenegro 8,000 Macedonia, FYR Dominican Republic Colombia Bosnia and Herzegovina Thailand Ukraine GNI per capita, 2005 PPP 6,000 Namibia Egypt, Arab Rep. Swaziland Azerbaijan Armenia 4,000 Indonesia Honduras Mongolia Moldova Vietnam Pakistan Guyana Congo, Rep. India 2,000 Kyrgyz Republic Senegal Cambodia Bangladesh Côte d’Ivoire Tajikistan Gambia, The Tanzania Haiti Mali Nepal Benin Rwanda Zambia Sierra Leone Guinea Ethiopia Uganda Congo, Dem. Rep. Liberia 0 −20 −10 0 10 20 Percentage-point difference between males and females ages 15–19 in the poorest quintile attaining grade 6 Favors females Favors males Source: WDR 2014 team based on data from World Bank EdAttain (database). Note: Positive differences denote a lower percentage of females attaining grade 6 relative to males. Household poverty is based on an index of assets and housing characteristics. Years are between 2005 and 2008. GNI = gross national income. PPP = purchasing power parity. especially among the poorest, can be seen as a supply Assessment (PISA) math test (figure 3.7b). These and demand problem. On the supply side, insuffi- quality gaps reduce the potential effect of education cient and inadequate infrastructure, lack of teacher on skill acquisition and job opportunities. training, and weak monitoring and enforcement of Moreover, gender gaps in education continue to basic standards (such as teacher attendance) dimin- exist within poor households in certain countries, ish the quality of education. On the demand side, limiting the opportunities for women to participate resource constraints, lack of employment opportu- in the labor market years later. Gaps in education en- nities, or limited information about returns to edu- rollment and attainment have narrowed impressively cation add to the low quality, reduce the perceived in the developing world, and they have even reversed value of education, and increase dropout rates.29 Re- in some groups, with women surpassing men in at- search shows that conditional cash transfer (CCT) tainment in several countries.31 However, among programs have been successful in increasing the de- poor households, some parents are still reluctant to mand for education. But better employment oppor- invest as much in education for girls as they do for tunities also act as a powerful incentive to invest in boys. In many low-income countries, for example, education, as research from Bangladesh and Mexico fewer girls than boys in the poorest quintile achieve shows.30 six years of education (figure 3.8). Low educational While middle-income countries are catching up attainment also affects women’s participation in the to high-income countries on attainment, there are labor force; in Nicaragua, for example, women with still large differences in quality between countries, as complete tertiary education were almost twice as shown by the average scores of 15-year-olds in the likely to participate in the labor force as women with standardized Programme for International Student only primary education.32 Households are the first line of support to confront risk and pursue opportunity 119 Accumulating financial and physical assets to hold, leaving her and her children more vulnerable build insurance and investment opportunities to shocks and less able to pursue opportunities.37 For instance, in southern Ethiopia, where divorces are The ability to maintain liquid savings and tap credit rare and divorced wives get no share of joint assets, in safe and flexible ways is important for managing women from poor households—but not their hus- risks and investments. Even among the poor, saving bands—reduce their food consumption when they rates and the number of financial instruments used get sick and are unable to work.38 (mostly informal) are high.33 But many obstacles im- Few households in developing countries rely on pede poor households from keeping significant market insurance products. As noted in chapter 1, amounts of liquid savings to manage risks more ef- the risks that concern people the most relate to insur- fectively or to undertake investment opportunities. able events: illness, loss of income, and loss of assets. First, for very poor households, satisfying immediate Although these events can be difficult to insure needs takes most of their income, making the oppor- against when they affect large numbers of people, as tunity costs of saving very high. Second, informal in the case of an epidemic or a natural disaster, insur- risk-sharing mechanisms and the associated pressure ance products are widely available in high-income to share income might affect decisions about how countries. Health, property, and unemployment in- much to save and what instruments to use.34 Despite surance are common, and in some cases even man- these obstacles, having access to saving and credit op- datory. Yet in developing countries, only about 1 per- tions is highly valued, which is evident from the sub- cent of total asset losses from natural hazards were stantial fees that some people are willing to pay to be formally insured between 1980 and 2004, compared able to save safely; these include rotating savings and with 30 percent in high-income countries.39 The low credit associations, which do not pay interest and penetration of market insurance products in devel- bear significant risk for loss, and deposit collectors, oping countries results from high transaction costs who charge fees, rather than pay interest, to keep (for assessing claims), which translate into high pre- the customers’ savings. The growing number of cli- miums. Instead, most people rely on informal risk- ents—137.5 million in 2010—of microfinance pro- sharing arrangements with their extended family viders is also a sign of the substantial demand and and community members (see chapter 4). the potential benefits of expanding financial prod- A low supply of formal insurance only partly ex- ucts among the poor.35 As chapter 6 discusses, how- plains low coverage. Demand for insurance tends to ever, wider access to financial products for house- be low among many households even when people holds, if not well-managed, can increase demand for repeatedly suffer from or are exposed to shocks such credit beyond amounts these households can reason- as illness or disasters. Several explanations have been ably handle and lead to overindebtedness. That, in advanced for this low observed demand. First, re- turn, can affect aggregate financial stability. source constraints often restrict people’s ability to Physical assets—while less efficient than liquid purchase insurance. Second, people with limited ed- savings—are another important resource for manag- ucation and low numeracy skills might find the con- ing risk. Most poor households save in part by accu- cept of insurance complex and therefore prefer in- mulating physical and productive assets such as live- formal reciprocity arrangements. Third, subscribers stock, jewelry, or appliances, all of which have low must trust that the provider will deliver the payment liquidity and uncertain returns. Physical and pro- if the shock occurs. Building trust becomes more ductive assets may be exposed to risk from disasters, challenging if the institutional environment offers crime, or expropriation, and some forms of assets, few avenues to enforce the contract. Finally, the high such as land, may not be transferable. Asset price noncovered risk (or “basis” risk) of many insurance fluctuations can also hurt the ability of households schemes reduces the expected payment and under- to use assets effectively as insurance mechanisms.36 mines the value of the insurance policy and the trust Moreover, the ability to own and accumulate as- associated with it.40 sets is unequal within many households. Laws in most of the world allow women to own assets, but several countries—particularly in South Asia and Building insurance informally through family Sub-Saharan Africa—still have gender-specific own- formation, fertility, and marriage ership rights that limit women’s ability to acquire, In places where social protection and access to finan- sell, transfer, or inherit property. Such laws weaken cial markets are limited, the process of family forma- the bargaining position of the woman in the house- tion can be highly related to risk.41 Agreements for 120 WORLD DEVELOPMENT REPORT 2014 mutual support among family members can be one Guatemala; and 94 percent in Udaipur, India, report of the limited options available in contexts where ac- earning income from more than one type of activ- cess to other forms of support are missing, either ity.43 Having a diversified income portfolio including from the market or the state (see box 3.1). In Andhra farm and nonfarm activities might reduce income Pradesh and Maharashtra in southern India, for ex- fluctuations, but it does not always do so—in part be- ample, many parents marry their daughters into cause incomes from different activities tend to be households in distant villages to diversify income more correlated during crises.44 In addition, because risks among households exposed to different cli- household effort is divided into many different ac- matic shocks. In these cases, a daughter’s marriage tivities, diversification often leads to lower average becomes an informal insurance mechanism to pro- incomes. Greater access to markets and safety nets tect consumption. In countries where parents must reduces the need for people to diversify their income provide a dowry to their daughters, the financial activities to lower their exposure, as a recent study in pressure can lead them to marry their daughters at Bangladesh found, and it also opens possibilities for younger ages or to other members of the extended them to enter higher-income activities.45 family (family ties act as a form of credit because par- Other households—typically those with no access ents can commit to later payments). These practices to credit markets, formal or informal—opt for activi- limit the potential for investing in daughters’ human ties that have low risk exposure but also have low re- capital and expose them to risk of abuse by their turns, such as drought-resistant crops, which tend to spouses, as well as increasing health risks for future have low yields. Households with very few assets can- children when girls marry biological relatives. not self-insure against shocks, either by selling these Parents may also have to rely on their children to assets or by using them as collateral for credit. At the confront the risk of income loss.42 For example, rural same time, diversifying income sources often requires households in Bangladesh with higher risk exposure, a minimum amount of starting capital (say, for pur- fewer credit sources, and weaker ties to their com- chasing an animal). As a result, many asset-poor munity have higher rates of fertility than similar households have no other option but to opt for ac- households in India. One of the driving factors be- tivities where income risk is minimal. Studies in In- hind this difference is that women in rural Bangla- dia, Tanzania, and other countries have found that desh, unlike women in rural India, lack job opportu- poor rural households grow disproportionately more nities outside the home. Women who cannot low-risk, low-return crops, such as sweet potatoes.46 participate in the labor market face more difficulties Increasing their labor supply can help households in responding effectively to large shocks—such as cope with shocks, provided that household members widowhood—and hence must rely more heavily on can work and that sources of employment are avail- support from their children. High fertility has nega- able. In such cases, households with excess labor sup- tive consequences for human capital accumulation ply, which can be readily tapped as needed, can pro- in developing countries, where children from larger tect consumption more effectively. Regression analysis families receive fewer vaccinations and have lower shows that, for example, members of larger house- school attainment. This quantity-quality trade-off holds in China, Iraq, Peru, and Uzbekistan respond to suggests that these families have less leeway to make shocks by increasing the hours they work, taking on adjustments in their resource allocation as the num- more jobs, or working in a different location.47 ber of children increases. The ability of a household to increase its labor supply either temporarily or permanently depends critically on the ability of women to participate in the Diversifying income sources and increasing labor force. Female participation has increased signifi- labor supply cantly in recent decades, but a large share of the female In developing countries, where exposure to income population still remains outside the labor force. Only shocks is large and formal insurance is unavailable, one in five women in the Middle East and North Af- households often diversify their sources of income. rica and less than one in three in South Asia were ei- Throughout the developing world, household in- ther working or looking for work in 2011 (figure 3.9). come often comes from more than one sector (for Economic factors including higher labor demand, bet- example, farming and services), location (urban and ter infrastructure, and higher educational attainment rural, domestic or foreign), or product. For exam- play a pivotal role in increasing female labor market ple, 10 to 20 percent of households in Mexico, Nica- participation. However, because women tend to be the ragua, Panama, and Timor Leste; 50 percent in In- main providers of child care, lack of good-quality donesia; 72 percent in Côte d’Ivoire; 84 percent in child care alternatives can create a trade-off between Households are the first line of support to confront risk and pursue opportunity 121 F I G U R E 3 . 9 Female labor force participation is still limited in some regions Percentage of females in the labor force 80 70 60 50 Percent 40 30 20 10 0 East Asia OECD Sub-Saharan Europe and Latin America South Middle East and and Pacific Africa Central Asia and the Asia North Africa Caribbean Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: Data are as of 2011. Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. female participation in the labor force and child devel- Experimental evidence shows that poor house- opment. In addition, social norms that limit the par- holds fail to take advantage of migration opportuni- ticipation of women in economic activity may slow ties that would improve their response to shocks. A down gains in female participation in the regions with study in a famine-prone rural area of Bangladesh historically low participation.48 found that people are not willing to migrate tempo- Migration is also an effective way to diversify risk rarily to the city during the lean season, despite the and can be a coping response to a shock. Although harsh conditions. For them, the cost of migrating migration can weaken social ties within the family and the possibility of not finding a job opportunity and with the rest of the community, its net economic in the city make it too risky an enterprise, even if ex- effect is still likely to be positive for the entire house- pected returns are positive. As part of the study, a hold—a fact reflected in the growing movement of small cash incentive was offered to farmers to mi- people within and across countries and in the vol- grate. Those who took the incentive and migrated ume of remittances worldwide. Temporary migra- increased the consumption of their family members tion for work from rural to urban areas is the domi- by 30 percent, increased the calorie intake by 550– nant form of migration; there are 740 million internal 700 calories a person a day, and were also more likely migrants worldwide, nearly four times the number to migrate again in following years, when the incen- of international migrants.49 Most studies on migra- tive was no longer offered.51 tion, which focus on international migration, find positive and substantial contributions to migrant in- How can government strengthen come, although the effects on family members who protection and foster better opportunities remain behind are less clear. Still, in the Philippines for households? for example, an unexpected increase in a migrant’s income leads to higher educational expenditures, en- Households face multiple risks every day, and they hanced entrepreneurship, and lower poverty rates at use as many tools as they have available to manage the origin. A growing body of literature finds that them. Poor households, however, tend to have only remittances are an informal insurance against con- limited access to protection and insurance mecha- sumption shocks at the origin. For instance, in the nisms provided by markets and public services. Philippines, Tanzania, and Thailand, migrant work- These limitations increase vulnerability, particularly ers are more likely to send remittances to members of for risks that are not equally shared within the house- their family who are hit by a shock.50 hold. Government policies can substantially improve 122 WORLD DEVELOPMENT REPORT 2014 TA B L E 3 . 2 A systemic approach for policies to address multiple risks using multiple instruments Types of policy instruments Access to Facilitate Social Cash Information/ Wage Access to credit/ Goal savings insurance transfers training subsidies services grants Preparing and coping with: Illness X X X X X Disability X X X X X X X Old age X X X X X X Death X X X Unemployment X X X X X X Weather shocks/ X X X X X X disasters Investment in X X X human capital Poverty reduction X X X X X Source: WDR 2014 team based on Robalino, Rawlings, and Walker 2012. household risk management and increase house- ing features that enable a quick response to changes holds’ access to better opportunities by taking a sys- in households’ needs. The following rec- temic approach to risk management. Such an ap- ommendations describe different policy instruments proach implies a sound understanding of the risks that can help households address barriers to building that households face, the interactions between them, better protection, insurance, and coping strategies, as the range of obstacles to better risk management, well as instruments that can improve risk sharing and the right combination of instruments that can within the household, keeping in mind that single- strengthen risk management while increasing access instrument solutions are rare and that the most suc- to opportunities. Table 3.2 shows how various kinds cessful policies involve a combination of them. of policy instruments address multiple sources of risk and suggests how they might be used from a sys- Designing policies with a risk temic perspective; for example, providing health ser- management lens vices in combination with other instruments, such as insurance and information on preventive practices, Policies often have indirect effects on people’s behav- can deliver better outcomes in protecting people ior, and understanding these effects can be useful for against the risk of disease and the financial burden it improving incentives to invest in protection at the causes. Some tools, such as cash transfers, also con- household level. Property rights are a telling case. Se- tribute to investment in human capital and poverty cure property rights increase the value of the asset to alleviation. the owner, because the asset can be safely transferred Taking a systemic approach to policy making for and also used as collateral, increasing access to credit. risk management also implies that instruments need Beyond that, secure land tenure rights increase the to be put in place before shocks take place and that value of investments in land conservation and infra- coordination is essential. In practice, different pro- structure, which reduce risk exposure and increase grams and policies are the responsibility of different productivity. Secure land tenure rights have also government agencies; institutional mechanisms are been found to increase labor market participation thus needed to coordinate these programs and poli- and reduce child labor.52 Another example is cash cies to align goals, set monitoring and performance transfers. They are a very direct way to help house- standards, and establish common infrastructure that holds overcome financial constraints to invest in hu- enables risk management tools to function seam- man capital, but incentives to invest are even greater lessly. Moreover, just as households need to prepare when the transfers are conditioned on making the in anticipation of shocks, governments need to have investments. A recent experimental comparison of these systems in place before shocks hit, incorporat- conditional and unconditional cash transfers for Households are the first line of support to confront risk and pursue opportunity 123 girls in Malawi shows that conditioning plays an im- Addressing disparities within households portant role in improving school attendance and Public policies can also redress the balance of power even learning outcomes.53 and reduce inequities within the household. In many cases, a combination of regulatory reforms, targeting Leveraging technology and partnering with of public programs, and social norms that empower the private sector women to take greater control over decisions regard- Systemic policies addressing multiple risks require ing family planning, work, and financial manage- not only close coordination among different govern- ment can increase their bargaining power in the ment agencies but also partnering with other actors household, while reducing the vulnerability and im- in the society. The ability to share data and to track proving opportunities for children. beneficiaries across programs is one example—sim- Many women in poor households—particularly ple in theory—that many countries struggle with in in poor and rural areas—have little control over fer- practice. The availability of affordable technologies tility decisions, in part because good-quality family to collect biometric data is opening opportunities for planning information and services are not available. governments to identify beneficiaries and deliver ser- Women in these areas should have reliable access to vices to them, in particular for the poor, who are of- these services through health care providers, together ten “invisible” populations for lack of proper proof with a range of contraceptive options. This task re- of identity. India’s pioneering identification project quires putting in place an efficient supply chain. In aims to issue a unique identification number, or addition, providers need to communicate effectively Aadhaar, to every resident of the country, linked to and transparently with their patients about the ben- basic demographic information and biometrics, as a efits and potential side effects of different contracep- formal proof of identity. The scheme has assigned tion methods, while respecting women’s preferences more than 300 million Aadhaars so far, at a cost of and privacy.55 To be truly effective, however, family less than $3 per capita. The Aadhaar is a gateway to planning services need to be accompanied by other both public and private services: the government interventions to increase women’s bargaining power uses it to deliver public benefits directly to individ- in the household, notably those that increase their uals with fewer leakages, and the private sector— economic clout and legal standing. particularly financial services providers—can use the Access to labor markets for women is particularly Aadhaar to expand access to financial services.54 important because it allows households to diversify More generally, public service delivery can benefit their sources of income and improve the risk man- from public-private partnerships in many areas, agement of entire households.56 But women’s access from health to social assistance (box 3.2). to the labor market is extremely limited in some BOX 3.2 Improving service delivery by partnering with the private sector Public-private partnerships are an increasingly important compo- demic and vocational schools. The program was cost-effective and nent of efficient and effective service delivery. In India, government- increased student achievement. sponsored health insurance schemes—which aim to expand access The private sector can also play an important role in government to health insurance to half the country’s population by 2015—have transfer programs. In Brazil, banks now pay the government for the engaged private sector firms as both administrators and health care valuable right to distribute social security benefits; this is a reversal providers. In São Paulo, Brazil, private nonprofit operators were per- from the past, when the Brazilian government would pay banks. In mitted to run new hospitals based on a performance-based contract making the transition, the government has saved money, beneficia- model. While implementation was not without challenges, impact ries have gained more places to obtain their benefits, and banks evaluations have found that the nonprofit hospitals have been more have acquired additional customers for their credit products. Condi- efficient than for-profit ones, without sacrificing quality. Improve- tional cash transfer (CCT) programs in Ecuador and South Africa also ments in human resources and management practices seem to be partner with private financial institutions to deliver payments. CCT responsible. In education, public-private partnerships can cut costs programs in Bangladesh, Chile, and Colombia allow beneficiaries to and improve student achievement. Colombia’s Programa de Ampli- use private providers to fulfill their education and health commit- ación de Cobertura de la Educación Secundaria (PACES) provided ments. The private sector may be critical for meeting the increased 125,000 children with vouchers to attend private secondary aca- service demands stimulated by CCTs in many countries. Source: WDR 2014 team based on La Forgia and Nagpal 2012; Lewis and Patrinos 2012; Fiszbein and Schady 2009; La Forgia and Harding 2009; and Ortiz d’Avila Assumpção 2012. 124 WORLD DEVELOPMENT REPORT 2014 places. Public policies can help, starting by ensuring Saharan Africa, more than 20 percent of women that girls, in particular in poor households, complete believe that a husband is justified in hitting or their education. Providing child care alternatives and beating his wife for commonplace reasons such as promoting family-friendly workplace poli- going out without telling him and arguing cies can encourage women to stay in the with him (figure 3.10). As argued ear- labor force when they have children. lier, domestic violence can be both Public action (through the media, Abuse and the outcome of an environment for example) can also help coun- dominated by risk as well as a discrimination ter social norms that keep women source of risky behavior. Better at home (see chapter 2). within the family risk management tools can re- Giving women more voice in makes it a source of, duce the incidence of violence by household decision making has rather than a reducing the stress factors associ- positive consequences for risk ated with risk. Beyond that, legal solution to, risk. management and investment in sanctions against violent behavior human capital.57 Women’s empower- provide a strong signal that, regard- ment can be achieved not only by en- less of social norms, domestic violence hancing access to the labor market but also has serious consequences. by making women the beneficiaries of cash transfer and other social programs. Moreover, regulatory re- Enhancing access to labor markets forms that increase women’s land tenure and inheri- tance rights, such as the reforms to the Hindu Suc- To increase investment in skills, interventions should cession Act, have been found to increase women’s tackle demand and supply constraints. The first step bargaining power as well as human development is providing information about job opportunities outcomes for girls.58 and returns to education. Research from the Domin- Legal action against domestic violence is neces- ican Republic shows that receiving information on sary to counter social norms that tolerate violence returns to education significantly reduces the likeli- against women and children. In East Asia and the hood that students will drop out of school before Pacific, the Middle East and North Africa, and Sub- completion. In India, providing recruitment services F I G U R E 3 .10 Social norms in many regions tolerate domestic violence Percentage of women who say that a husband is justified in hitting or beating his wife 45 40 35 30 Percent 25 20 15 10 5 0 Sub-Saharan Middle East and East Asia South Europe and Latin America and Africa North Africa and Pacific Asia Central Asia the Caribbean Argues with him Goes out without telling him Neglects the children Source: WDR 2014 team based on data from demographic and health surveys, for the most recent year available 2000–11. Households are the first line of support to confront risk and pursue opportunity 125 for jobs to women delayed their decision to marry in specific sectors, such as agriculture, construction, and have children and increased their reported desire or health care. Receiving and sending countries es- to obtain training and work more steadily.59 tablish agreements on quotas, wages, and duration of Second, education and training systems should stay. If workers fulfill the requirements, they can mi- help students develop the skills that employers want, grate legally for a preestablished period, with a fixed which include soft skills, such as communications contract and wage. Workers can reapply to the pro- ability, in addition to reading and numeracy skills, gram upon their return, which lowers their incentive according to employer surveys throughout the to stay illegally once their contract expires.61 world. Evaluations of training programs for out-of- Lower transaction costs for remittances can im- school youth that combine soft skills with technical prove households’ ability to mitigate income losses, training have shown positive results in increasing while opening the door to better income opportuni- beneficiaries’ chances of job placement and the ties. Worldwide in 2011, migrants sent approximately quality of their jobs. Evidence also suggests that the $372 billion in remittances to their families in devel- delivery model that works best involves partnering oping countries (see chapter 6). Migrants often pay with (accredited) private providers. Finally, many hefty transaction costs to send small amounts of people need specific assistance to navigate the labor money through large carriers. Lowering transaction market and find the best opportunities available. costs by encouraging competition and promoting Countries with sufficient implementation capacity transparency can increase the benefits of remittances can establish employment services to facilitate the significantly for receiving countries. Simply provid- matching process of employers and employees; ing information about fees for different carriers can these services have proven highly effective in high- help. The project Envía CentroAmérica, financed in income countries, particularly for unskilled work- part by the World Bank and the Inter-American ers.60 These skill-enhancement strategies can deliver Development Bank, provides detailed information better job opportunities, however, only if there is an about the fee variation across carriers for sending accompanying demand for these skills. For that, a money to seven Central American countries. For ex- dynamic enterprise sector is crucial (see chapter 5). ample, fees for sending $200 from Washington, D.C. to Guatemala ranged from 1.29 percent to 17.42 per- cent, depending on the institution and the modality Facilitating migration and remittances of payment. Mobile phones have also enabled many Lowering barriers to domestic and international mi- migrants, especially in countries like Kenya, to send gration also helps households diversify incomes and remittances in a safe and affordable way.62 respond to shocks. As discussed, domestic migration provides millions of families with alternative income Increasing access to financial products opportunities outside farming. But for poor house- holds, even temporary migration is a risky venture Financial products are a crucial component of a that many are not willing to pursue. In many cases, household’s risk management strategy, and govern- indirect policies that, for instance, lower transporta- ments can facilitate access to formal financial prod- tion costs, can encourage temporary migration. In- ucts in several ways. As discussed earlier, poor ternational migration also presents people with bet- households face significant barriers to access formal ter income opportunities, but increasing movements financial services. In addition to promoting the fi- of undocumented migrants across countries have nancial inclusion of the poor through an appropri- increased social tensions and violence and even cost ate regulatory framework (such as consumer protec- lives. To mitigate these risks, sending countries can tion laws) and by expanding financial literacy (see play a more proactive role in facilitating legal migra- chapter 6), governments can leverage social pro- tion, protecting the rights of migrants abroad, and grams to connect beneficiaries with the financial respecting migration policies of receiving countries. sector through payment systems. This link has been A few countries have implemented programs of reg- made successfully in Brazil, Ecuador, and South Af- ulated migration. While these programs might not rica, where beneficiaries of pensions and cash trans- always be able to ensure proper treatment of mi- fer programs receive their payments electronically grants, they offer an avenue to migrate legally. Pro- through the financial system (see box 3.2). In addi- grams in Morocco, the Philippines, St. Lucia, and tion, carefully designed public subsidies can help Tonga, for example, offer workers the opportunity to expand the supply of certain commercial financial migrate temporarily to countries that demand labor products, such as index insurance, that significantly 126 WORLD DEVELOPMENT REPORT 2014 BOX 3.3 Index-based insurance: The potential and the challenges Index-based insurance can be a viable instrument to manage agri- small farmers, has managed to increase its client base by insuring cultural risk. Index-based, or parametric, insurance provides pay- inputs instead of harvests, using “aggregators” such as cooperatives ments based on physical triggers (such as variation in rainfall) rather to insure groups rather than single farmers, creating premium- than loss claims. This type of insurance is less subject to moral hazard sharing arrangements between farmers and agribusinesses, selling and has significantly lower transaction costs. Although some farm- through local businesses that are frequented by the farmers, and ers bear significant “basis risk” because their risk is imperfectly cor- paying claims immediately using mobile phones. related with the risk insured by the index contract, several studies Governments also have a role to play. The Mexican Catastrophe show that index-based insurance increases investment and improves Climate Contingency Insurance Program provides state govern- yields. In Tamil Nadu in India, for example, offering farmers index- ments with funding for the purchase of insurance, most of which is based insurance made them more likely to plant higher-yield (but index-based and targeted at subsistence producers below the riskier) rice varieties and less likely to plant lower-yield but drought- threshold for commercial agricultural insurance. India’s Weather tolerant ones. And when basis risk is large, having an informal net- Based Crop Insurance Scheme has significantly expanded the use of work can help by providing insurance against basis risk. Thus the index-based insurance by subsidizing premiums. However, govern- presence of informal risk sharing actually increases demand for ment premium subsidies create tensions in the market that are dif- index-based insurance in the presence of basis risk. ficult to resolve. For example, selling subsidized index-based insur- Still, the coverage of index-based insurance remains low. In par- ance to landless agricultural laborers (who technically do not ticular, providers need to find better ways to market it by taking into possess an insurable interest) opens the market to others (such as account the context in which farmers operate, the variety of risks urban residents) to gamble with the product because the subsi- that they face, and their lack of experience with formal financial dized premiums make the insurance product, in effect, look like an products. For instance, the studies for India show that selling insur- attractive lottery ticket. Government resources might be better tar- ance to landless laborers, not just land owners, provided significant geted to covering the up-front cost of installing weather stations protection to their income and their ability to invest, because they needed to monitor rainfall at high density (to reduce basis risk), and bear a disproportionate share of agricultural risk. In Kenya and to scaling up these investments, rather than subsidizing the price of Rwanda, Kilimo Salama, an index-based insurance program for the premiums. Source: WDR 2014 team based on Brown, Mobarak, and Zelenska 2013 for the WDR 2014; Alejandro de la Fuente for the WDR 2014; “Fact sheet: Kilimo Salama (“Safe Agriculture”),” available at http://kilimosalama.files.wordpress.com/2010/02/kilimo-salama-fact-sheet-final11.pdf. improve risk management for the poor but that face regressive.63 As a result, poor people receive substan- scaling-up challenges (box 3.3). dard treatment while bearing large out-of-pocket costs. Many countries have undertaken efforts to re- form their health care systems to deliver better ser- Building health and social protection systems vices to the poor—while maintaining sustainabil- that protect the most vulnerable ity—by moving toward a model of universal health Because health and income shocks can be particu- insurance (see the cases of Turkey and the Kyrgyz larly destructive for the poor, protecting them against Republic in spotlight 3). A growing consensus is fo- these risks is a priority. This section discusses how cusing on three basic goals for reform: enhancing countries can expand the coverage of health and so- risk pooling to ensure that health expenditures do cial protection, starting with the most vulnerable not overwhelm the household’s saving capacity; en- populations, while striving to improve service deliv- suring financial sustainability and equitable access ery and results and still maintain fiscal sustainability. by defining specific benefit packages and providing Health insurance reforms are increasingly im- insurance at a low cost (or free) for the poor; and proving protection for the most vulnerable. In many improving efficiency by delinking financing from developing countries, public health systems are frag- service provision. mented, inefficient, and inequitable. Typically, health Table 3.3 summarizes some recent efforts in differ- insurance systems are available to a minority of peo- ent countries to expand access to health insurance. ple, usually workers in the formal sector, while ev- While many features respond to the unique context of eryone else has access to lower-quality national each country, a few patterns are apparent. First, these health care systems financed by general revenues. efforts provide subsidies so that poor and vulnerable Such duplication puts financial pressure on health populations can obtain access to insurance. For ex- systems, creates tiered-quality services, and may be ample, China’s rural health insurance program subsi- Households are the first line of support to confront risk and pursue opportunity 127 TA B L E 3 . 3 Common features of programs to expand coverage of health insurance What Who How Use of general Colombia The Régimen Subsidiado health insurance program offers free and heavily subsidized health care to the revenue financing poorest. to include the poor India The Rashtriya Swasthya Bima Yojna (RSBY) insurance program requires payment of a small nominal fee (5 percent of the combined registration fee and premium). Indonesia The Jamkesmas program covers the poor and near-poor population at no cost. Mexico The Seguro Popular program offers a full subsidy to informal sector households. Thailand Any person outside the formal/civil service sector is covered for a nominal fee. Turkey Premiums of the poor are covered by the state. Vietnam Those below the poverty line and other selected groups are fully subsidized; the near-poor receive partial subsidization. Higher quality India Authorizations and case management in RSBY and Rajiv Aarogyasri are fully electronic. and efficiency Defined service Brazil The Sistema Único de Saúde (SUS) offers a comprehensive package that includes essential drugs and dental package care. Mexico Package includes coverage of catastrophic illness. South Africa The Antiretroviral Program provides testing, monitoring, and treatment for HIV/AIDS, based on a financial sustainability study. Turkey Comprehensive basic benefits package includes diagnostic services, inpatient treatment, and hospitalization for emergency care. Incentives on Brazil SUS transfers to municipal governments depend on meeting performance and coverage targets. the supply side Colombia Capitation (per capita fees) and fee-for-service are commonly used in the Régimen Subsidiado. India Private insurers are selected through competitive bidding and paid on the basis of enrollment. Indonesia Providers are paid capitation at the primary level, and negotiated fees at the secondary level. Services tailored Brazil The Family Health Strategy uses outreach activities to expand use of primary care and to identify and treat to vulnerable common diseases. populations Ethiopia The Health Extension Workers program trains households to adopt best practices and to become role models in their community. Data-driven India Biometric data collected at enrollment are used for monitoring use and outcomes. Kyrgyz Payment and utilization are analyzed to identify outliers and barriers to access and to forecast needs and costs. Republic South Africa Expenditures on personnel, drugs and supplements, lab services, and information systems are actively monitored. Source: WDR 2014 team based on Cotlear, forthcoming, for Brazil, Colombia, Ethiopia, India, the Kyrgyz Republic, Mexico, South Africa, Turkey, and Vietnam. For India, Indonesia, and Thailand, see Adam Wagstaff 2011, “Health reform: A Consensus Emerging in Asia?” Let’s Talk Development (blog), April 12. http://blogs.worldbank.org/ developmenttalk/. dizes at least 80 percent of the premium cost, and it fee to everyone who is not in the civil service or the aims to cover the entire rural population. In India, the formal sector. Second, most insurance programs Rashtriya Swasthya Bima Yojna scheme requires ben- cover only a defined range of services, which can vary eficiaries to pay only a nominal registration fee to widely from country to country. Some countries have join; the premium is paid by taxpayers. Indonesia’s focused on covering primary care, while others cover Jamkesmas scheme also covers the poor and near- only catastrophic illness, and still others offer a mix. poor at taxpayers’ expense. In Thailand, the Universal Finally, several countries include the private sector in Coverage scheme provides insurance for a nominal service delivery and make innovative uses of technol- 128 WORLD DEVELOPMENT REPORT 2014 Noncontributory pension programs have expanded coverage in F I G U R E 3 .11 developing countries, especially for the poorest HUN SVK 15,000 HRV LTU POL COL TUR GNI per capita, 2005 PPP ARG 10,000 MUS VEN BGR DOM BRA SRB BLR MNE CRI THA BIH ECU PER ALB SLV UKR 5,000 MDV GTM JOR ARM LKA PRYMAR GEO BOL BTN HND PHL MNG VNM INDPAK MDA LAO TLS NIC RWA KHM NGA TJK KGZ NPL GHA 0 MWIAFG 0 20 40 60 80 100 Coverage of poorest 40% of households with elderly members, % Source: WDR 2014 team based on data from Evans, forthcoming. Note: Years vary between 2003 and 2010. Countries marked in blue have noncontributory pensions. GNI = gross national income; PPP = purchasing power parity. ogy to improve efficiency. It is still too early to pass employed and agricultural workers who work in the definitive judgment on the impact of these trends on informal sector. In other countries, where contribu- health outcomes and household health expenditures, tory pension systems are supporting a rapidly grow- as well as their fiscal sustainability. But learning from ing elderly population, declining ratios of workers these cases will be crucial as more countries move to- to retirees and a higher life expectancy make the ward universal health insurance. current model unsustainable. As a result, more and more countries are expand- ing coverage of basic pensions by introducing non- Providing income support for old age contributory pensions financed by general revenues. Life-cycle transitions such as old age reduce the abil- For example, 13 countries in Latin America and the ity of individuals to earn sufficient income to re- Caribbean now have noncontributory pensions for main out of poverty. In addition, time-inconsistent those not covered by the contributory system.64 Oth- behaviors discussed in chapter 2 justify policies to ers, like Mauritius, South Africa, and several high- encourage people in their earning years to save for income countries, have historically relied on non- the future. In many countries, however, social insur- contributory pensions. As figure 3.11 shows, aside ance systems (mostly for pensions and health) cover from the former socialist countries in Eastern Europe a minority of the population, usually the nonpoor. and Central Asia, all developing countries that cover This exclusion occurs because so-called contribu- more than half of the poorest 40 percent of house- tory insurance systems are financed by payroll taxes holds with elderly members have noncontributory and contributions levied on employers and employ- systems.65 Several studies show that noncontributory ees, typically only in the formal sector—which ef- pensions have increased coverage and reduced pov- fectively denies access to the large proportion of self- erty among the elderly.66 Households are the first line of support to confront risk and pursue opportunity 129 When contributory and noncontributory systems Offering safety nets for bad times coexist, however, incentives to participate in man- datory contributory systems may diminish. That Safety nets are also crucial to help the most vulnerable occurs because contributory systems may be ill- households manage risk and protect consumption in designed or because contributory and noncontribu- the face of shocks. Studies show that safety net pro- tory systems may not be well integrated. Workers grams such as public works and cash transfers help who move between formal and informal jobs or in people build assets, take more risk in their productive and out of the labor force might not contribute suf- activities, and accumulate human capital. Beneficia- ficiently to be eligible to receive benefits or might ries of programs in Bangladesh, Brazil, India, Mexico, achieve very low income-replacement rates.67 Work- Nicaragua, and other countries have invested more ers in rapidly aging countries might contribute to- resources in high-risk, high-return ventures (such as ward increasingly uncertain benefits, creating a per- fertilizer), have diversified their income away from ception of contributions as a pure tax on labor and agriculture, and have gained access to credit.69 encouraging them to underreport wages or opt for For example, public works programs provide la- informal employment. bor income while contributing to local economic de- To ensure that social insurance systems are equi- velopment. These programs are particularly benefi- table, fiscally sustainable, and minimize distortions cial in countries with large rural populations, like in the labor market, countries should reconcile the those in Sub-Saharan Africa. Indeed, the region has need to expand coverage with the need to encourage more than 150 public works programs and more private savings. Specifically, noncontributory sys- than 120 cash transfer programs.70 Public works pro- tems that provide a basic level of benefits should be grams protect households’ consumption in the face financed through general revenues just like any other of income losses, while enabling them to build crucial basic government function. But not all countries are growth-enhancing assets; examples include building in a position to provide adequate benefits universally water infrastructure and improving land manage- in a fiscally sustainable manner. In practice, many ment at the local level. Moreover, public works pro- developing countries may be able to provide only a grams have positive local economic spillovers, similar minimum level of benefits and possibly to only a tar- to other safety nets. For instance, the Extended Public geted population. That is true in particular for coun- Works program in South Africa boosted local econo- tries where the old-age dependency ratio is growing mies, because 67 percent of the beneficiaries pur- fast. Thus countries need to consider their long-term chased food from local shops.71 Most of these pro- fiscal capacity in relation to their future commit- grams are targeted geographically, and beneficiary ments to decide what the appropriate levels of cover- selection is done through self-targeting (by offering age and benefits are. Importantly, they need to con- low wages, the programs attract those who most need sider the different options to collect the necessary tax the income, while encouraging them to continue to revenues (see the “Focus on policy reform” at the end seek other work). These programs also have the flex- of this Report). ibility to respond to specific adverse conditions in Contributory systems can help to increase the certain households; for instance, the Ethiopian Pro- adequacy of insurance benefits, but they need to be ductive Safety Net Program uses family targeting, designed in a way that does not create distortions which allows the program to adjust the quota of days in the labor market. In some contexts, distortions that a family can benefit from the program to the size will be avoided only if contributions are made vol- of the family. untary and open to all, regardless of work status or A systemic approach to social protection is help- if the mandatory contribution rates are reduced. In ing governments exploit synergies across instruments all cases, contributory systems should provide ben- and deliver better service to all those who need it.72 efits that are clearly linked to contributions. Further, For instance, several programs have recently imple- incentives to save—automatic enrollment, matching mented strategies that combine protection with ac- contributions, simplifying processes, and lowering cess to opportunity.73 These programs seek to “grad- information barriers through financial literacy— uate” beneficiaries sustainably by building pathways can have a significant impact. New Zealand’s to better income through self-employment or wage KiwiSaver scheme is an interesting example of an employment opportunities. In most cases, these automatic enrollment program (with an “opt-out” strategies require combining a range of instruments option) that increased retirement savings for about beyond simply providing cash transfers, such as pro- half the population.68 viding links to financial services or training. Safety 130 WORLD DEVELOPMENT REPORT 2014 net beneficiaries in Côte d’Ivoire, Rwanda, and Tan- into two complementary groupings: policies to em- zania are encouraged to save part of their earnings by power households, and policies to empower individu- obtaining access to bank accounts and community als within households to better manage their risks. saving groups. Public works programs in El The first set of policies addresses obstacles that Salvador, Sierra Leone, and South Africa households face as units, such as lack of also provide some basic and technical information, lack of resources, and skills training.74 In Cameroon, ben- Both policies that limited access to labor and financial eficiaries of the new cash transfer empower households markets. The second group ad- program attend financial literacy dresses challenges to risk sharing and business training activities. as a unit and policies within the household and imped- Evaluation from pilot programs that empower iments that increase the vulnera- that provide cash and training to individuals within bility of certain members, includ- beneficiaries in Ethiopia, Ghana, households are ing underinvestment in human and Nicaragua show that the com- capital, using children to manage bination of instruments leads to en- necessary. risk, excluding women from financial try into higher-income activities in the decision making, and exposure to do- short to medium term. mestic violence. Table 3.4 presents a sum- mary of the policy recommendations discussed in Putting it all together: Guidelines for this chapter, highlighting how combining different instruments may contribute to strengthening risk policy implementation management, and the possible complementarities The policy recommendations to improve house- between them. This systemic approach requires holds’ risk management can be roughly categorized strong coordination across government institutions TA B L E 3 . 4 Policy priorities to improve risk management at the household level POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Basic literacy and training Secondary education and training Higher education and training Media and community campaigns Teaching preventive health in schools Facilitating informed fertility decisions Access to mobile technology Protection Sanitation infrastructure and preventive health care (including women’s health) Migration assistance/access to labor and other markets (especially for women) Regulation to guarantee equal property Policies to promote gender parity in rights for women leadership positions Promulgation and enforcement of domestic violence and abuse laws Insurance Index insurance Financial inclusion of the poor Lower remittance costs Health insurance Pensions Unemployment (old age, disability, death) insurance Coping Self- and community-targeted y-targ community geted income support ted income support Means-test Means-tested o wom Transfers targeted to men women Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing poli- cies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably and that can be improved over time. Households are the first line of support to confront risk and pursue opportunity 131 as well as between the government and the other eco- across agencies, such as unified registries and data- nomic and social actors, and depends on the coun- sharing protocols, and by building strong technical try’s institutional capacity. capacity among civil servants. In addition, key pro- A country’s initial conditions also affect the use grams that help people manage risk, such as social and effectiveness of each policy instrument. Hence, assistance programs, need to be properly funded policies in table 3.4 are grouped according to a coun- when they are needed the most: that is, during down- try’s initial conditions and follow the guidelines pre- turns. To do that, governments should exploit good sented in chapter 2: first, to be realistic, with policies times to set up safety nets that can be scaled up to adapted to the country’s capacity; and second, to cover more people and offer more benefits in bad build a strong foundation, with policies that address times, when more households face illness, unem- the most important obstacles first and upon which ployment, and other losses. more advanced policies can be designed and imple- mented over time. Thus countries with limited re- Promote flexibility sources and low institutional capacity can begin by focusing on the most foundational policies: ensuring Within the institutional framework proposed above, access to basic services, while also improving the effi- government policies and programs should also be cacy of informal mechanisms, for example, by facili- sufficiently flexible to adapt to changing circum- tating migration and remittances. Countries that stances. One example is the labor market. Demo- have laid the foundations for risk management can go graphic and economic changes can cause deep beyond the basics and focus on expanding access to changes in the labor market; thus labor market poli- services and raising productivity to foster the ability cies, including education and training policies, of households to take advantage of opportunity—by should be flexible enough to adapt to such changes. improving access to formal risk management prod- Similarly, safety net programs require both effective ucts, and expanding coverage of social insurance. instruments to identify the most vulnerable house- In most countries, and in particular in those with holds and individuals when crises hit and the neces- limited capacity, implementing coordinated policies sary infrastructure to deliver services in a timely using multiple instruments can be very difficult. As manner. They should also be able to scale back their discussed in chapter 2, obstacles inherent in public coverage when the crisis passes. policy undermine the effectiveness of many govern- ment actions in helping people manage risk. These Provide the right incentives obstacles include limited capacity and resources, co- ordination failures within the government and with Increasing incentives for members of the household other actors, political economy constraints, and deep to take personal responsibility is an important part uncertainty. A few basic principles—following the of empowering households to manage risk. Many so- guidelines to be realistic and build a strong founda- cial assistance programs are now taking incentives tion—can help policy makers overcome these obsta- into account: for instance, by establishing benefits cles as they design and implement policies. and setting time limits on receipt to avoid discourag- ing beneficiaries from working. Public policy should also aim to change incentives within the household Keep a long-run perspective so that members decide to pool their resources for To ensure their long-run sustainability, governments the benefit of all. This goal might require a mix of should make sure that policies are fiscally sustainable regulation reforms and specific design features in and that institutional arrangements transcend the public programs (such as targeting of beneficiaries, political cycle. Often, governments come into office combined with legal reforms). For instance, targeting eager to establish ambitious “flagship” programs, women in cash transfer programs can empower which turn out to be unaffordable, especially during them economically, but can also have negative reper- economic downturns. Moreover, many of these pro- cussions (such as an increased risk of domestic vio- grams are operated in isolation, rather than in coor- lence) if their legal protection is not guaranteed.75 dination with similar and complementary programs run by other agencies. Instead of taking this short- Protect the vulnerable term approach, governments should focus on build- ing a legacy through stable institutional arrange- The first priority for policies to improve risk man- ments that improve coordination and efficiency agement should be those households that face the 132 WORLD DEVELOPMENT REPORT 2014 largest barriers to preparation. Too often, however, ter seek to overcome obstacles to risk management the definition of vulnerability is determined by inter- for households. In some cases, however, design flaws est groups. In the United States, the government may create new barriers. For example, private saving spends 2.2 times as much on the elderly as on chil- incentives can be undermined if governments use dren, yet 22 percent of children under 18 live in pov- the funds in public saving programs to finance cur- erty, compared with 9 percent of adults aged 65 and rent expenditures. Price caps imposed on food sta- older.76 Similarly, Brazil has practically eradicated ples to keep them affordable to the poor often result poverty among the elderly, but not among children. in massive shortages and speculation, making people Transparent policies with clearly defined priorities worse off. In more extreme cases, governments enact and goals, but with the flexibility to reallocate public regulations that legitimize social norms that may funds when these goals are not met or when they weaken household risk management, such as those change, can help. that limit the economic and social participation of women. Anticipating the additional risks and other unintended consequences for risk management that Do not generate uncertainty or policies might generate should help governments unnecessary risks avoid implementing policies where “the medicine is Policies should not create new obstacles to risk man- worse than the disease.” agement. Most of the policies discussed in this chap- Households are the first line of support to confront risk and pursue opportunity 133 Notes 37. World Bank 2011. 38. Dercon and Krishnan 2000. 1. This story is adapted from Davis 2011. 39. Linnerooth-Bayer, Mechler, and Hochrainer-Stigler 2011. 2. See Hoff and Sen 2005. 40. See references in Brown, Mobarak, and Zelenska 2013 for the 3. Dercon 2002. WDR 2014. 4. Heltberg, Oviedo, and Talukdar 2013 for the WDR 2014. 41. See references in Cáceres-Delpiano 2013 for the WDR 2014. 5. See references in Oviedo and Moroz 2013 for the WDR 2014. 42. See references in Cáceres-Delpiano 2013 for the WDR 2014. 6. Genicot and Ligon 2013 for the WDR 2014. 43. See references in Brown, Mobarak, and Zelenska 2013 for the 7. De Vreyer and Lambert 2013 for the WDR 2014. WDR 2014. 8. Giles, Guo and Zhao 2013. 44. Dercon 2002. 9. Silverstein and Giarrusso 2010. 45. Bandyopadhyay and Skoufias, forthcoming. 10. See literature discussion in Alderman and others 1995. 46. Dercon 2002. 11. Udry 1996 for Burkina Faso; Duflo and Udry 2004 for Côte 47. Heltberg, Oviedo, and Talukdar 2013 for the WDR 2014. d’Ivoire. 48. See in-depth discussion in World Bank 2011. 12. See, for example, Iversen and others 2011. 49. See references in Brown, Mobarak, and Zelenska 2013 for the 13. Heltberg, Oviedo, and Talukdar 2013 for the WDR 2014. WDR 2014. 14. Oviedo and Moroz 2013 for the WDR 2014. 50. Paulson 2000; De Weerdt and Hirvonen 2013. 15. See references in Friedman and Sturdy 2011; Oviedo and 51. See references in Brown, Mobarak, and Zelenska 2013 for the Moroz 2013 for the WDR 2014. WDR 2014. 16. Dercon 2002; Friedman and Sturdy 2011. 52. Field 2007; World Bank 2003. 17. See references in Oviedo and Moroz 2013 for the WDR 2014. 53. Baird, McIntosh, and Özler 2011. 18. Haushofer, de Laat, and Chemin 2012; Friedman and Sturdy 54. Nilekani 2013. 2011. 55. World Bank 2011. 19. Heltberg, Hossain, and Reva 2012. 56. See, for example, Basu 2006. 20. Miguel 2005. 57. World Bank 2011. 21. World Bank 2011. 58. Deininger, Goyal, and Nagarajan 2013. 22. See Kusimba and others 2013, for example. 59. Jensen 2010; Jensen 2012. 23. See discussion and references in Brown, Mobarak, and Zelen- 60. Brown, Mobarak, and Zelenska 2013 for the WDR 2014; ska 2013 for the WDR 2014. Almeida, Behrman, and Robalino 2012; Almeida and others 24. See references in Brown, Mobarak, and Zelenska 2013 for the 2012. WDR 2014. 61. Ruiz 2008. 25. Heckman, Stixrud, and Urzua 2006; Grossman 2006. 62. Kusimba and others 2013. 26. Chou and others 2010. 63. Ribe, Robalino, and Walker 2012. 27. For instance, see Card 1999. 64. See Holzmann, Robalino, and Takayama 2009. 28. Barro and Lee 2010. Returns to an additional year range from 65. Evans, forthcoming; Levy and Schady 2013. 6 percent in Sub-Saharan Africa to close to 14 percent in high- 66. See, for instance, Holzmann, Robalino, and Takayama 2009; income economies. Levy and Schady 2013; Frölich and others, forthcoming. 29. See the discussion on this topic in Banerjee and Duflo 2011. 67. Ribe, Robalino, and Walker 2012. 30. See references in Brown, Mobarak, and Zelenska 2013 for the 68. Hinz and others 2013. WDR 2014. 69. Alderman and Yemtsov 2012. 31. World Bank 2011. 70. Premand 2013 for the WDR 2014. 32. World Bank 2012. 71. Alderman and Yemtsov 2012. 33. Collins and others 2009. 72. Robalino, Rawlings, and Walker 2012. 34. See Di Falco and Bulte 2011 and Baland, Guirkinger, and Mali 73. See references in Premand 2013 for the WDR 2014. 2011 for examples. 74. See spotlight 2 on promoting food security in El Salvador. 35. Ledgerwood 2013. 75. World Bank 2011. 36. Dercon 2002. 76. Isaacs and others 2012. 134 WORLD DEVELOPMENT REPORT 2014 References Natural Experiment in Taiwan [China].” American Economic Journal: Applied Economics 2 (1): 33–61. Abdul Latif Jameel Poverty Action Lab (J-PAL). 2011. “The Price Is Collins, Daryl, Jonathan Morduch, Stuart Rutherford, and Wrong: Charging Small Fees Dramatically Reduces Access to Orlanda Ruthven. 2009. Portfolios of the Poor. Princeton, NJ: Important Products for the Poor.” J-PAL Bulletin April 2011. Princeton University Press. Alderman, Harold, Pierre-André Chiappori, Lawrence Haddad, Contreras, Manuel, Brian Heilman, Gary Barker, Ajay Singh, Ravi John Hoddinott, and Ravi Kanbur. 1995. “Unitary Versus Verma, and Joanna Bloomfield. 2012. “Bridges to Adulthood: Collective Models of the Household: Is It Time to Shift the Understanding the Lifelong Influence of Men’s Childhood Burden of Proof?” World Bank Research Observer 10 (1): 1–19. Experiences of Violence.” International Center for Research on Alderman, Harold, and Ruslan Yemtsov. 2012. “Productive Role of Women, Washington, DC. Safety Nets.” Social Protection and Labor Discussion Paper Cotlear, Daniel. Forthcoming. “The Challenges of Implementing 1203, World Bank, Washington, DC. Universal Health Coverage in Developing Countries: Lessons Alesina, Alberto, and Paola Giuliano. 2013. “Family Ties.” Working from 25 Countries from around the World.” Unpublished Paper 18966, National Bureau of Economic Research, Cam- manuscript. World Bank, Washington, DC. bridge, MA. Davis, Peter. 2011. “Patterns of Socio-Economic Mobility in Rural Alger, Ingela, and Jörgen W. Weibull. 2010. “Kinship, Incentives, Bangladesh: Lessons from Life-History Interviews.” Working and Evolution.” American Economic Review 100 (4): 1725–58. Paper 197, Chronic Poverty Research Center, Bath, U.K. Almeida, Rita, Juliana Arbelaez, Maddalena Honorati, Arvo Deininger, Klaus, Aparajita Goyal, and Hari Nagarajan. 2013. Kuddo, Tanja Lohmann, Mirey Ovadiya, Lucian Pop, Maria “Women’s Inheritance Rights and Intergenerational Trans- Laura Sanchez Puerta, and Michael Weber. 2012. “Improving mission of Resources in India.” Journal of Human Resources 48 Access to Jobs and Earnings Opportunities: The Role of (1): 114–41. Activation and Graduation Policies in Developing Countries.” Dercon, Stefan. 2002. “Income Risk, Coping Strategies, and Safety Social Protection and Labor Discussion Paper 1204, World Nets.” World Bank Research Observer 17 (2): 141–66. Bank, Washington, DC. Dercon, Stefan, and Pramila Krishnan. 2000. “In Sickness and in Almeida, Rita, Jere R. Behrman, and David A. Robalino. 2012. The Health: Risk Sharing within Households in Rural Ethiopia.” Right Skills for the Job? Rethinking Training Policies for Workers. Journal of Political Economy 108 (4): 688–727. Human Development Perspectives. Washington, DC: World De Vreyer, Philippe, and Sylvie Lambert. 2013. “Household Risk Bank. Management in Senegal.” Background paper for the World Baird, Sarah, Craig McIntosh, and Berk Özler. 2011. “Cash or Con- Development Report 2014. dition? Evidence from a Cash Transfer Experiment.” Quarterly De Weerdt, Joachim, and Kalle Hirvonen. 2013. “Risk Sharing and Journal of Economics 126 (4): 1709–53. Internal Migration.” Policy Research Working Paper 6429, Baland, Jean-Marie, Catherine Guirkinger, and Charlotte Mali. World Bank, Washington, DC. 2011. “Pretending to Be Poor: Borrowing to Escape Solidarity Di Falco, Salvatore, and Erwin Bulte. 2011. “A Dark Side of Social in Cameroon.” Economic Development and Cultural Change 60 Capital? Kinship, Consumption, and Savings.” Journal of (1): 1–16. Development Studies 47 (8): 1128–51. Bandyopadhyay, Sushenjit, and Emmanuel Skoufias. Forthcom- Duflo, Esther, and Christopher Udry. 2004. “Intrahousehold ing. “Rainfall Variability, Occupation Choice, and Welfare in Resource Allocation in Côte d’Ivoire: Social Norms, Separate Rural Bangladesh.” Review of Economics of the Household. Accounts and Consumption Choices.” Working Paper 10498, Banerjee, Abhijit V., and Esther Duflo. 2011. Poor Economics: A National Bureau of Economic Research, Cambridge, MA. Radical Rethinking of the Way to Fight Global Poverty. New Evans, Brooks. Forthcoming. “An Examination of Pension Cover- York: PublicAffairs. age of Elderly in the Developing World.” Social Protection and Barro, Robert J., and Jong-Wha Lee. 2010. “A New Data Set of Edu- Labor Policy Note, World Bank, Washington, DC. cational Attainment in the World, 1950–2010.” Working Paper Field, Erica. 2007. “Entitled to Work: Urban Property Rights and 15902, National Bureau of Economic Research, Cambridge, Labor Supply in Peru.” Quarterly Journal of Economics 122 (4): MA. 1561–602. Basu, Kaushik. 2006. “Gender and Say: A Model of Household Be- Fiszbein, Ariel, and Norbert Schady. 2009. Conditional Cash Trans- haviour with Endogenously Determined Balance of Power.” fers: Reducing Present and Future Poverty. Washington, DC: Economic Journal 116 (511): 558–80. World Bank. Becker, Gary S. 1974. “A Theory of Social Interactions.” Journal of Friedman, Jed, and Jennifer Sturdy. 2011. “The Influence of Eco- Political Economy 82 (6): 1063–93. nomic Crisis on Early Childhood Development: A Review of Bernheim, B. Douglas, Andrei Shleifer, and Lawrence H. Sum- Pathways and Measured Impact.” In No Small Matter: The Im- mers. 1985. “The Strategic Bequest Motive.” Journal of Political pact of Poverty, Shocks, and Human Capital Investments in Economy 93 (6): 1045–76. Early Childhood Development, edited by Harold Alderman, Brown, Julia K., Mushfiq A. Mobarak, and Tetyana V. Zelenska. 51–83. Washington, DC: World Bank. 2013. “Barriers to Adoption of Products and Technologies Frölich, Markus, David Kaplan, Carmen Pagés-Serra, Jamele Rigo- That Aid Risk Management in Developing Countries.” Back- lini, and David Robalino. Forthcoming. Social Insurance and ground paper for the World Development Report 2014. Labor Markets: How to Protect Workers While Creating Good Cáceres-Delpiano, Julio. 2013. “Literature Review: Family Forma- Jobs. Washington, DC: World Bank, Inter-American Develop- tion and Fertility as Risk Coping Mechanisms.” Background ment Bank, and Institute for the Study of Labor. paper for the World Development Report 2014. Genicot, Garance, and Ethan Ligon. 2013. “Risk and the Extent Card, David. 1999. “The Causal Effect of Education on Earnings.” of Insurance.” Background Paper for the World Development In Handbook of Labor Economics, vol. 3A, 1801–63. Amster- Report 2014. dam: Elsevier. Giles, John, Junjie Guo, and Yaohui Zhao. 2013. “Community, Chou, Shin-Yi, Jin-Tan Liu, Michael Grossman, and Ted Joyce. Family and Household Support for the Elderly in the Wake of 2010. “Parental Education and Child Health: Evidence from a Rapid Urbanization: Evidence from Rural China.” Research in Households are the first line of support to confront risk and pursue opportunity 135 progress supported by the Knowledge for Change Program, Lewis, Laura, and Harry Anthony Patrinos. 2012. “Impact Evalua- World Bank, Washington, DC. tion of Private Sector Participation in Education.” Research Grossman, Michael. 2006. “Education and Nonmarket Out- Report, CfBT Education Trust, Reading, U.K. comes.” In Handbook of the Economics of Education, vol. 1, Linnerooth-Bayer, Joanne, Rainhard Mechler, and Stefan edited by Eric Hanushek and Finis Welch, 578–633. Amster- Hochrainer-Stigler. 2011. “Insurance against Losses from Nat- dam: Elsevier. ural Disasters in Developing Countries: Evidence, Gaps and Haushofer, Johannes, Joost de Laat, and Matthieu Chemin. 2012. the Way Forward.” Journal of Integrated Disaster Risk Manage- “Poverty Raises Levels of the Stress Hormone Cortisol: Evi- ment 1 (1): 1–23. dence from Weather Shocks in Kenya.” Unpublished manu- Miguel, Edward. 2005. “Poverty and Witch Killing.” Review of script. J-PAL, Cambridge, MA. Economic Studies 72 (4): 1153–72. Heckman, James J., Jora Stixrud, and Sergio Urzua. 2006. “The Ef- Nilekani, Nandan. 2013. “The Aadhaar Project.” Presentation at fects of Cognitive and Noncognitive Abilities on Labor Market Development Economics Lecture Series, World Bank, Wash- Outcomes and Social Behavior.” Journal of Labor Economics 24 ington, DC, April 24. (3): 411–82. OECD (Organisation for Economic Co-operation and Develop- Heltberg, Rasmus, Naomi Hossain, and Anna Reva, eds. 2012. Liv- ment). Programme for International Student Assessment ing through Crises: How the Food, Fuel, and Financial Shocks (PISA) (database). OECD, Paris, http://www.oecd.org/pisa. Affect the Poor. Washington, DC: World Bank. Ortiz d’Avila Assumpção, Rodrigo. 2012. “How Brazil Stopped Heltberg, Rasmus, Ana Maria Oviedo, and Faiyaz Talukdar. 2013. Paying the Banks: A Brazilian Perspective on Social Security “What Are the Sources of Risk and How Do People Cope? In- Payments.” Presentation at Social Protection and Labor Sector sights from Household Surveys in 15 Countries.” Background Seminar Series, World Bank, Washington, DC, October 26. paper for the World Development Report 2014. Oviedo, Ana María, and Harry Moroz. 2013. “The Impacts of Hinz, Richard, Robert Holzmann, David Tuesta, and Noriyuki Risk.” Background paper for the World Development Report Takayama. 2013. Matching Contributions for Pensions: A Re- 2014. view of International Experience. Washington, DC: World Paulson, Anna L. 2000. “Insurance Motives for Migration: Evi- Bank. dence from Thailand.” Kellogg Graduate School of Manage- Hoff, Karla, and Arijit Sen. 2005. “The Kin System as a Poverty ment, Northwestern University, Evanston, IL. Trap?” Policy Research Working Paper 3575, World Bank, Premand, Patrick. 2013. “From Risk Coping to Risk Management: Washington, DC. Productive Safety Nets in Africa.” Background paper for the Holzmann, Robert, David A. Robalino, and Noriyuki Takayama. World Development Report 2014. 2009. Closing the Coverage Gap: The Role of Social Pensions and Ribe, Helena, David A. Robalino, and Ian Walker. 2012. From Right Other Retirement Income Transfers. Washington, DC: World to Reality: Incentives, Labor Markets, and the Challenge of Uni- Bank. versal Social Protection in Latin America and the Caribbean. Isaacs, Julia, Katherine Toran, Heather Hahn, Karina Fortuny, and Washington, DC: World Bank. C. Eugene Steurle. 2012. “Kids’ Share 2012: Report on Federal Robalino, David A., Laura Rawlings, and Ian Walker. 2012. “Build- Expenditures on Children through 2011.” Urban Institute, ing Social Protection and Labor Systems: Concepts and Washington, DC. Operational Implications.” Social Protection and Labor Dis- Iversen, Vegard, Cecile Jackson, Bereket Kebede, Alistair Munro, cussion Paper 1202, World Bank, Washington, DC. and Arjan Verschoor. 2011. “Do Spouses Realise Cooperative Ruiz, Neil G. 2008. “Managing Migration: Lessons from the Phil- Gains? Experimental Evidence from Rural Uganda.” World ippines.” Migration and Development Brief 6, World Bank, Development 39 (4): 569–78. Washington, DC. Jensen, Robert. 2010. “The (Perceived) Returns to Education and Silverstein, Merril, and Roseann Giarrusso. 2010. “Aging and Fam- the Demand for Schooling.” Quarterly Journal of Economics ily Life: A Decade Review.” Journal of Marriage and Family 72 125 (2): 515–48. (5): 1039–58. ———. 2012. “Do Labor Market Opportunities Affect Young Udry, Christopher. 1996. “Gender, Agricultural Production, and Women’s Work and Family Decisions? Experimental Evidence the Theory of the Household.” Journal of Political Economy 104 from India.” Quarterly Journal of Economics 127 (2): 753–92. (5): 1010–46. Kusimba, Sibel, Harpieth Chaggar, Elizabeth Gross, and Gabriel United Nations. 2009. “World Population Prospects: The 2008 Re- Kunyu. 2013. “Social Networks of Mobile Money in Kenya.” vision.” Department of Economic and Social Affairs Popula- Working Paper 2013-1, Institute for Money, Technology, and tion Division, United Nations, New York. Financial Inclusion, Irvine, CA. World Bank. 2003. Land Policies for Growth and Poverty Reduction. La Forgia, Gerard M., and April Harding. 2009. “Public-Private Washington, DC: World Bank. Partnerships and Public Hospital Performance in São Paulo, ———. 2011. World Development Report 2012: Gender Equality Brazil.” Health Affairs 28 (4): 1114–26. and Development. Washington, DC: World Bank. La Forgia, Gerard, and Somil Nagpal. 2012. Government- ———. 2012. “Better Jobs in Nicaragua: The Role of Human Sponsored Health Insurance in India: Are You Covered? Wash- Capital.” Policy Note, World Bank, Washington, DC. ington, DC: World Bank. ———. Educational Attainment and Enrollment around the Ledgerwood, Joanna. 2013. The New Microfinance Handbook: A World (EdAttain) (database). World Bank, Washington, DC, Financial Market System Perspective. Washington, DC: World http://iresearch.worldbank.org/edattain/. Bank. ———. World Development Indicators (database). World Bank, Levy, Santiago, and Norbert Schady. 2013. “Latin America’s Social Washington, DC, http://data.worldbank.org/data-catalog/ Policy Challenge: Education, Social Insurance, Redistribu- world-development-indicators. tion.” Journal of Economic Perspectives 27 (2): 193–218. S P OT L I G H T 3 Moving toward universal health insurance coverage in Turkey and the Kyrgyz Republic The economic consequences of an illness are often devastating in developing countries. About 100 million people fall into poverty annually struggling to cover health care costs. The experiences of Turkey and the Kyrgyz Republic show that countries at all levels of development can improve access to and affordability of medical services by increasing the efficiency of government health spending and protecting the poor through publicly financed health insurance. Increasing equity in access to health care in Turkey government introduced several incentives, including rais- Turkey has achieved impressive results in access, afford- ing salaries of family doctors, introducing performance ability, and quality of health care. Health insurance covers guidelines, and regularly monitoring the quality of service 95 percent of the population, and 76 percent of Turkish delivery through facility visits and patient surveys. Provid- citizens are satisfied with health care services. Before 2003, ers risk paying up to 20 percent of their base salary in pen- however, use of health services was very uneven among alties for failure to meet certain performance targets, such regions, and health care in rural areas was both hard to ob- as immunizations and antenatal care. The government also tain and more costly than in cities. Health financing was introduced monthly bonus payments of up to 40 percent fragmented among four different insurance schemes. A of base salary for doctors who relocate to underserved separate Green Card Program for the poor covered only locations, a step that has reduced the gap in access to inpatient services and therefore was not widely used. Most health care between rural and urban locations. public health resources were allocated to costly hospital- These reforms have significantly improved access to based services, rather than primary care. services and financial protection against medical costs To address these problems, the government launched throughout the country. Use of health services has more a comprehensive Health Transformation Program in 2003. than doubled since 2003, satisfaction with the quality of All health insurance schemes were merged into a universal health care has also risen, and key health indicators—life health insurance program managed by the newly created expectancy, and child and maternal mortality—have im- Social Security Institution. Every insured person, including proved. A World Bank evaluation of the Green Card Pro- the poor, has the same benefits package, which covers in- gram showed that it provided an effective safety net for patient and outpatient services, dental care, diagnostic the poor during the economic crisis of 2008, with benefi- tests, emergency care, and pharmaceuticals. The poor are ciaries less likely than those with no insurance to reduce exempt from co-payments if they use public facilities. The their use of curative and preventive care. Improvements in expanded benefits led to greater demand for the Green access to health care were achieved without excessive Card; participation more than tripled from 2003 to 2011, public health spending: at 5.1 percent of gross domestic from 2.5 million to 9.1 million. Targeting of the program product (GDP), Turkey’s public health spending is com- has also improved substantially: Green Card benefits to parable to that of other countries at similar levels of devel- those in the lowest income quintile increased from 55 per- opment. Going forward, it will be important to strengthen cent in 2003 to 71 percent in 2012. mechanisms to contain costs and further increase effi- Premiums are based on household income and in- ciency of health spending. crease with wealth. The government pays the premiums for the poor—defined as households with per capita in- Improving affordability of health care in the come less than one-third the minimum wage, or about Kyrgyz Republic $163 a month. The poor are identified through the national At independence in 1991, the Kyrgyz Republic had a stan- Integrated Social Aid Services System, which is also used to dard Soviet health care system, characterized by a large determine eligibility for other social assistance programs. network of providers, a focus on curative hospital care The integrated system helps avoid duplication of informa- rather than preventive services, and a centrally planned, tion and improves benefits administration. The near-poor input-based financing system. Although inefficiencies (those with per capita income between one-third and the plagued this system, every Kyrgyz citizen enjoyed access to full minimum wage) are also well protected, with premi- free medical services. During the early 1990s, the young ums set at about $20 a month. The rest of the population state experienced a deep economic crisis, GDP declined by pays higher premiums, depending on income. more than half, and the government was unable to main- The government sought to strengthen primary care by tain the oversized health care system. Informal out-of- promoting family medicine. This decision was in keeping pocket payments to health care providers became com- with global evidence that systems oriented to primary care mon to compensate for low salaries. Hospitalized patients produce better health for the population at lower cost. The often had to help pay for medicine, bed linens, and even Spotlight 3 137 light bulbs. For many of the poor, health care was unaffordable and thus unused. FIGURE S3.1 Use of health care and out-of-pocket Starting in 2001, the government introduced a series of reforms to improve the efficiency of health sector spending and decrease out-of- health expenditure by income status, the Kyrgyz pocket costs. The sequencing of reforms was important to the success Republic, 2009 of the approach. First, the Mandatory Health Insurance Fund (MHIF) 12 was introduced, funded by a 2 percent payroll tax paid by employers. The government makes contributions for the retired and the unem- 10 ployed, and the self-employed can purchase health insurance for 8 Percent about $10 a year. Significant efficiency gains were obtained by con- solidating separate pools of public health care funding at the district 6 and regional levels into a single pool managed by the MHIF. This 4 arrangement has reduced overhead costs and resulted in more equi- 2 table allocation of resources across administrative units. Second, pur- chase of health services was centralized under the MHIF, which 0 Utilization of Utilization of Out-of-pocket contracts with providers across the country under output-based pay- primary care hospital care health expenditure ment mechanisms. This approach has enhanced efficiency, giving fa- among all among all of total spending households households cility managers some flexibility in how to use the funds. Third, primary care was made a priority. The oversized hospital sector was reduced Poorest 20% Richest 20% by about 40 percent, and savings were allocated to medical supplies Source: WDR 2014 team based on data from Giuffrida, Jakab, and Dale 2013. and salaries of health providers. A major outcome of the reforms was the explicit definition of benefits and regulation of entitlements. The State Guaranteed Benefit Package establishes free primary and emergency care for all citizens Kyrgyz Republic also has a rather developed health information sys- and subsidized secondary care with exemptions from co-payments tem for a low-income country, which allows the government to fore- for vulnerable groups: children under age 5, retirees older than 70, the cast income from co-payments, plan annual expenditures, and moni- disabled, pregnant women, and those with medical conditions with tor the impact of new policies. high expected use of health care (diabetes, cancer, tuberculosis, Future reforms should focus on improving sustainability of health and asthma). These groups also benefit from access to subsidized care financing. Further rationalization of health care financing will en- medications. tail introducing targeting of co-payment exemptions by poverty sta- The impact of the reforms has been very positive. Use of health tus, as is done in Turkey and many other countries. Additional effi- care is now roughly the same at all income levels (figure S3.1). House- ciency gains could be realized by reducing hospitalization rates and holds are less likely to fall into poverty as a result of illness. Out-of- overuse of medication, cutting utility costs of health facilities, and pocket health expenditures have declined among all income groups streamlining funding on drug procurement. since the start of reforms and constituted only 4.4 percent of total household spending among the poorest quintile in 2009 (see figure Sources S3.1). The incidence of catastrophic health spending (more than 20 Turkey percent of total household expenditures) declined from 8 percent in Aran, Meltem A., and Jesko S. Hentschel. 2012. “Protection in Good and Bad Times? 2000 to 5 percent in 2009. Several health indicators, such as infant and The Turkish Green Card Health Program.” Policy Research Working Paper 6178, under-five mortality rates, have improved, and the country has much World Bank, Washington, DC. better health outcomes than the average low-income nation. Public Menon, Rekha, Salih Mollahaliloglu, and Iryna Postolovska. 2013. “Toward Universal Coverage: Turkey’s Green Card Program for the Poor.” UNICO Studies Series 18. health spending constitutes about 3.5 percent of GDP, which is some- World Bank, Washington, DC. what higher than the average for low-income countries and reflects OECD (Organisation for Economic Co-operation and Development) and World Bank. the government’s prioritization of health spending. 2008. OECD Reviews of Health Systems: Turkey. Paris: OECD. The experience in the Kyrgyz Republic shows that more efficient Ministry of Health of Turkey. 2012. Health Statistics Yearbook 2011. Ankara: Ministry of Health of Turkey. use of public resources can reduce the patient financial burden in a low-income country. This experience can be particularly valuable for Kyrgyz Republic Giuffrida, Antonio, Melitta Jakab, and Elina M. Dale. 2013. “Toward Universal Cover- other transition economies with limited fiscal space and overcapacity age in Health: The Case of the State Guaranteed Benefit Package of the Kyrgyz in the health sector. The positive outcomes were achieved thanks to a Republic.” UNICO Studies Series 17. World Bank, Washington, DC. comprehensive approach rather than reliance on a single instrument. Ibraimova, Ainura, Baktygul Akkazieva, Aibek Ibraimov, Elina Manzhieva, and Bernd Introduction of strategic purchasing through the MHIF, giving provid- Rechel. 2011. “Kyrgyzstan: Health System Review.” Health Systems in Transition ers greater autonomy and allowing them to manage some of the sav- 13 (3): 1–152. WHO (World Health Organization). 2010. Health Systems Financing: The Path to ings, downsizing the hospital sector, and increasing investments in Universal Coverage. Geneva: WHO. primary care have resulted in significant efficiency gains that were Yazbeck, Abdo S. 2009. Attacking Inequality in the Health Sector: A Synthesis of directed toward greater financial protection of the population. The Evidence and Tools. Washington, DC: World Bank. Strength in numbers: members of a women’s self-help group in rural Madhya Pradesh, India, attend a skills training program. © McKay Savage CHAPTER 4 Cohesive and connected communities create resilience Communities confront such as crime, pollution, and epidem- many risks ics. Citizens are not defenseless: they may take steps to stop crime More than 750 people died when and disorder by mobilizing against a heat wave struck Chicago in gangs, mentoring at-risk youth, munity 1995, many of them elderly Com or stopping provocations from poor who avoided opening their flaring into ethnic riots; they may windows at night for fear of petition authorities for sanitation, crime. These deaths were not uni- health services, and law and order; formly distributed across the city, and they may provide mutual insur- even controlling for income. In some ance and assistance. Chicago neighborhoods, the elderly poor Communities are groups of people who perished, while in others they weathered the interact frequently and share identity or location. heat wave. Cohesive neighborhoods with strong Neighborhood groups, religious groups, and kinship social networks had lower mortality.1 groups are some examples. Community relation- In the rural community of Nyakatoke in Tanza- ships are not as close and long term as family rela- nia, a population of 120 households has formed at tionships, in part because community groups are least 40 different community groups, such as burial larger than families. People usually share gains and insurance societies, rotating savings and credit asso- losses to a larger degree within their family than ciations, and labor and livestock sharing groups. within their communities. At the same time, com- These groups offer helpful (if partial) insurance to munities involve more personal interaction than villagers, and most residents belong to several market relationships; as a result, members of the groups. Groups have verbal or written rules govern- community have better knowledge about one an- ing contributions and payouts for specific events, as other’s successes and failings, luck and misfortune. well as sanctions against noncontributors. Some This knowledge helps them enforce norms about groups complement their rules with ceremonies that contributing to the common good and sharing with emphasize unity and the importance of mutual help. those in need. Most groups offer some form of insurance, and Relying on norms, shame, guilt, and personal in- there is also a group that patrols the village at night.2 teraction allows well-functioning communities to While local communities are at the epicenter of solve many problems that markets and governments many crises, conflicts, and disasters, they are also struggle with: preventing crime, managing natural part of the solution. Unmanaged risk causes loss of resources, affecting behavior change, getting credit to life, health, and property and deters investment.3 the poor, and insuring risks that are hard for outsid- Unlocking the potential of cities as centers of pro- ers to monitor. Communities can mobilize for social ductivity and opportunity requires managing risks change and responsive governance, as in the Arab 139 140 WORLD DEVELOPMENT REPORT 2014 Spring. Community-driven projects can deliver many countries have asked people which shocks they public goods and services in fragile settings where have recently experienced: natural disasters, health government capacity is lacking, as in Afghanistan’s shocks, price shocks, and asset loss tend to be the National Solidarity Project. But communities can most frequent and severe (see also chapters 1 and 3). also exclude people and foment violent conflict with Evidence from many locations indicates that idio- neighboring communities. And they struggle to cre- syncratic shocks such as death, illness, and accidents ate insurance pools and public goods of sufficiently are common and create high costs.5 Systemic risk is large scale to address systemic risk and exploit more important as well. In Tanzania, for example, six of complex opportunities.4 the seven most frequently self-reported shocks— This chapter focuses on how communities help including spikes in food prices, floods, and water their members confront risk and pursue opportuni- shortages—are systemic, in that they affect many ties and how development actors—including gov- members of the local community (figure 4.1a). In ernments, donors, and nongovernmental organiza- rural China, individual health and more widespread tions (NGOs)—can support them in this role. The agricultural shocks are both major concerns. Agri- chapter describes how many communities have cultural shocks are more frequent, while health evolved mechanisms of insurance, protection, and shocks tend to be more severe (figure 4.1b). In Nige- coping that help them address local risks; how some ria, health shocks are the most common (35 percent communities prosper while others expose their of all severe shocks reported by urban households, members to terrible risks; how free rider problems, and 27 percent by rural ones), followed by price and unresponsive authorities, resource constraints, employment shocks in urban areas and disasters, and other obstacles obstruct better com- price shocks, and asset loss in rural areas. munity risk management; and how so- Some systemic shocks are rather local- cial cohesion and connections with For problems ized (drought, crop loss), while oth- other communities, markets, and whose solution ers are nationwide or global (swings government shape communities’ in food and input prices). chances at succeeding. The chap- eludes markets and What “community” means in ter also argues that communities governments, a cohesive the context of risk management can become much better risk community could be can be defined in two comple- managers with outside support. the missing piece mentary (and sometimes overlap- Addressing risk is increasingly ping) ways, one centered on loca- recognized as complementary to of the puzzle. tion, the other on cultural identity.6 combating poverty in its many dimen- Communities of location are exactly what sions. When agencies listen to communities, they sound like: people sharing a location, they often discover that communities are keenly such as the long-term residents of a village or urban aware of risk and request support preparing for it. neighborhood. Communities of location can provide But what does risk management entail in a com- credit, insurance, and assistance against idiosyncratic munity context? This chapter explores various ap- risk and organize protection against local risks such as proaches governments, donors, and civil society crime, disaster, and infection—for example, through organizations can use to foster communities’ resil- policing, drainage, sanitation, or solid waste collec- ience, building on what is already there. The chapter tion—provided they can overcome people’s tendency advocates the types of policies that empower com- to free ride and shirk on their responsibility to con- munities as risk managers by creating favorable legal tribute or pay for their fair share of local public goods. regimes, fostering their own capacity to manage Communities of location can either organize their local risks, promoting their “voice” and ability to own protection or petition government to provide hold government accountable and providing com- the necessary services and infrastructure. Shared plementary public goods and services. It also cau- spaces such as parks and sports facilities strengthen tions against policies that ignore communities and people’s ties to a local community. create risk and instability. The informal urban settlement of Indio Guays in Guayaquil, Ecuador, exemplifies a location-based community. The settlement sprang up in 1975, when Communities as risk managers poor people began acquiring land and building sim- People and their communities face risk from many ple bamboo-walled houses. At that time, the new sources, both idiosyncratic and systemic. Surveys in settlement sat in a swamp connected by dangerous Cohesive and connected communities create resilience 141 People and their communities are affected by both idiosyncratic and F I G U R E 4 .1 systemic shocks a. Systemic shocks (price, disaster, argicultural) are frequent in Tanzania, as are idiosyncratic health shocks Large rise in food prices 58 Death of an extended family member 39 Insufficient water 35 Drought or floods 30 Crop disease or pests 27 Large fall in crop sale prices 27 Large rise in agricultural input prices 25 Livestock death or theft 22 Death of a household member 12 Crime 11 Illness or accident of a working member 9 Household breakup 8 Household business failure 6 Loss of own land 5 Loss of salaried employment 3 0 10 20 30 40 50 60 % of households that have experienced each type of shock in the past 5 years Shock was idiosyncratic Shock was mostly systemic (affected all or in most cases most other households in the community) b. Health and agricultural shocks are common concerns in rural China 100 % of households reporting subjective 80 Job loss poverty because of shock Major illness of working member Death of a member Major illness of dependent 60 Crop failure Theft 40 Marriage Livestock death 20 Childbirth 0 0 5 10 15 20 25 % of households reporting shock Source: WDR 2014 team based on data from the Tanzania National Panel Survey 2010/11 (panel a), and the China Rural Social Protection Survey 2004 (panel b) covering rural areas in three provinces (Fujian, Gansu, and Zhejiang) and one autonomous region (Guangxi) of China. 142 WORLD DEVELOPMENT REPORT 2014 B O X 4 .1 When sharing wealth is a religious mandate: The use of zakat in Islamic communities Many religious traditions encourage or even mandate charitable giv- directly to the chosen beneficiary or delivered to the mosque or ing to the poor and needy. Zakat— one of the Five Pillars of Islam—is neighborhood leader for distribution. Survey data from the capital, a system ingrained in Islamic society in which every person has a Sana’a, show that around one-third of the poor receive zakat and duty to give a certain share of his or her income and assets to help that it is fairly well targeted, in that about 60 percent of it reaches specific categories of people defined by the Qur’an, including wid- households in the bottom 40 percent. However, poor households ows, orphans, people with disabilities, and others living in dire pov- lacking social connections are sometimes excluded. As a lump-sum erty. Interpretation varies from country to country. In some coun- donation given once or twice a year, zakat is not an effective instru- tries, zakat is collected and distributed by the government, while ment for consumption smoothing in the face of shocks. Similar in others it remains a private matter distributed by donors directly shortcomings are observed in Pakistan, where the zakat system is to the chosen beneficiary or through a community affair in which overseen by the government and implemented locally by commu- donors deliver funds to the mosque for distribution. It is normally nity committees. Implementation is fraught with problems, such as given once or twice a year, during Ramadan and Eid. mistargeting, infrequent and unpredictable delivery of benefits, In the Republic of Yemen, zakat is not an obligatory tax but a and widespread perception of patronage and petty corruption. transfer; the amount is calculated by the donor and distributed Source: WDR 2014 team based on Levin, Morgandi, and Silva 2012 and World Bank 2007. walkways with no physical or social infrastructure of stitutions to provide both spiritual comfort and ma- any kind. The new neighbors formed a cohesive self- terial relief in times of adversity (box 4.1). In Indone- help committee and elected dynamic women to run sia, for example, the financial crisis in 1997–98 it. Over the next decades, the committee successfully caused the price of rice, the main staple food, to petitioned authorities and political parties for ser- nearly triple, resulting in widespread economic dis- vices, infrastructure, and land titles. By the early tress, political transition, and social upheaval. Many 2000s, it had become a stable urban neighborhood Indonesians responded by more active participation with cement houses, paved roads, running water, in organized religion such as Qur’an study groups, lighting, sanitation, schools, clinics, and people tak- which seemed to offer both spiritual relief and access ing advantage of overseas migration opportunities.7 to informal insurance. Religious participation was Collective accomplishments, often involving risk associated with a reduced need for alms or credit and management achieved through cohesion and links to was most pronounced in locales where formal credit authorities, combined with private accumulation of was unavailable.9 assets, housing, and education to transform the The examples from Chicago, Ecuador, Indonesia, neighborhood and its people. and Tanzania illustrate how communities often are The second, complementary definition considers vitally important for helping people prepare for and community as a cultural, identity-based group such cope with risk, in urban and rural areas alike. Some- as a kinship or ethnic group. Friends and fellow times communities are the only source of assistance, members of the same kin or ethnic group are a fre- the lifeline that helps people survive disaster and quent source of help to people facing shocks. Shared food scarcity. But these examples also suggest that identity—not location—grants access to the net- community mechanisms rarely add up to adequate work. Such groups can offer assistance and insurance risk management. Communities struggle to provide against loss of income, and their composition can be effective insurance against systemic risk and public fluid: when large shocks strike, or more complex op- goods at sufficient scale and across divisions (prob- portunities (involving migration, for example) arise, lems that enterprises and governments sometimes the boundaries of groups stretch further as people are better equipped to solve). In the Tanzanian vil- seek assistance through more distant social ties.8 Cul- lage of Nyakatoke, for example, none of 40 commu- ture—language, religious rituals, shared symbols and nity groups had links to government or to NGOs, celebrations—strengthens people’s sense of shared although several had attempted to forge them, real- identity. izing that such links often are necessary to solve Religion and faith-based institutions play an im- larger problems and help people escape from pov- portant role in tying together cultural communities. erty. Communities also tend to be better at coping Around the world, people often rely on religious in- with than preparing for risk and exploring opportu- Cohesive and connected communities create resilience 143 nity. No communities are utopian ideals of equality Communities as providers of insurance and peace: hierarchies, inequalities, violence, and exclusion are always factors to reckon with, to differ- Families are usually the first port of call when some- ent degrees. one is faced with a shock such as illness, disability, or Many of the obstacles facing communities as they job loss, but when family support proves inadequate, confront risk can be traced to the informal organiza- people turn to their communities for assistance. tional mechanisms communities tend to use. The Loans and assistance from family, friends, and neigh- community sphere is where norms of morality, fair- bors are among the most frequent responses to ness, reciprocity, a sense of duty to the common shocks (of any kind) in many developing countries good, and occasionally altruism play out; markets are (figure 4.2). In Nigeria, for example, informal credit ruled by contracts and monetary reward. In the mar- and assistance was the most prevalent coping re- kets, agreements are ultimately enforced through the sponse, accounting for 32 percent of all responses. credible threat of legal sanctions; communities are Informal credit and assistance is particularly impor- just as likely to rely on shame, guilt, ostracism, or tant for low-income households, and sometimes is violence to settle disputes and to solve problems of their only real safety net. Cutting back on food is also adverse selection, moral hazard, and free riders (see common; it makes up 20 to 22 percent of all coping glossary 2.1, chapter 2). These informal mechanisms responses reported by households in Afghanistan, work better in small cohesive groups than in larger Iraq, and Uganda. Across Eastern Europe and Cen- and less dense groups. Accordingly, the need for in- tral Asia, people with strong social capital in the form group bonding and cohesion limits the size of the of trust and networks are significantly less likely to insurance pools and public goods that communities rely on costly coping such as reducing consumption are able to create; it can also lead to exclusion of cer- of stable foods and forgoing medical care, probably tain people. Communities’ risk management mecha- because of better informal support (figure 4.3). Peo- nisms, geared for small groups, therefore perform ple confronted with shocks also rely on savings and best to ease idiosyncratic and small systemic risks; asset sales and seek more work. In contrast, credit adapting them to larger and more complex problems from formal lenders and assistance from government requires research and careful organization. and NGOs usually play lesser roles. F I G U R E 4 . 2 People respond to shocks on their own and by pooling risk with others, relying on informal credit and assistance to cope with shocks Malawi Uganda Tajikistan Uzbekistan Afghanistan Iraq Maldives Sudan Nigeria 0 20 40 60 80 100 % of all coping responses when faced with a shock Informal credit and assistance Formal credit and assistance Consumption reduction Savings and sale of assets Employment or migration Source: WDR 2014 team based on data from household surveys, various years 2004–11. 144 WORLD DEVELOPMENT REPORT 2014 FIGURE 4.3 Social capital helps reduce costly coping in Eastern Europe and Central Asia Among respondents 7 with some trust in other people 24 Among respondents 9 with some distrust in other people 29 Among respondents 13 with complete distrust in other people 42 0 5 10 15 20 25 30 35 40 45 % of population in Eastern Europe and Central Asia that have done the following in response to a decline in income or other economic difficulty Postponed or skipped visits to the doctor Reduced consumption of staple foods Source: WDR 2014 team based on data from Life in Transition Survey II, 2010. The organization of informal insurance and cop- is retained.11 Indonesian households insured them- ing support varies widely around the world. Com- selves against 38 percent of the economic costs of munity insurance—sometimes also called informal more serious health shocks and 71 percent of the safety nets, risk pooling, and mutual insurance— costs of more minor illness.12 In the Philippines, re- tends to be based more on kin and reciprocal rela- mittances compensate for 65 percent of the cost of tionships than on formal contracts. Within those rainfall shocks.13 The poorest are the least insured. parameters, it can assume many different organiza- For example, in rural China, for the poorest tenth of tional forms. Membership-based groups operate the population, a loss of income of 100 yuan led rule-based insurance (as is the case with the insur- people to cut food and other expenditures by 40 ance groups in the Tanzanian village). Kinship net- yuan, while for the richest third of households, the works, often including migrants, may exchange ad same shock resulted in a consumption cut of only 10 hoc, need-based transfers. In fact, there is evidence yuan.14 Because of these limits on the effectiveness that households in disaster-prone areas are more of risk management, large shares of households likely to send migrants and to use their remittances around the world rely on costly coping responses to prepare for shocks. Remittances shoot up after di- such as skipping meals, forgoing necessary medical sasters, helping recipients cope.10 Further, many reli- care, or selling productive assets (see figure 4.2). As gions support and organize charitable giving (see discussed in chapter 3, such practices are hardest on box 4.1). Transfers can be in the form of cash, food, the youngest and can lead to infant mortality and labor, clothing, or farm inputs. People often do not chronic malnutrition. In sum, although informal in- distinguish closely between gifts and loans, altruism surance helps many people cope with smaller idio- and insurance; all form part of the reciprocal rela- syncratic shocks, it is often inadequate for other tionship between people currently in need and peo- risks and for vulnerable population groups.15 ple currently in a position to help. The size and effectiveness of informal insurance Community-based insurance provides people groups is limited by the need for social ties among with partial compensation for the impact of shocks, members. In theory, insurance requires pooling but many shocks nevertheless cause serious hard- across large groups of diverse people—but that is not ship. Studies of households that face income shocks what usually occurs in informal insurance groups, show that their consumption falls less than income: which tend to be small and homogeneous and there- in other words, some risk is insured away and some fore susceptible to aggregate risk.16 Communities use Cohesive and connected communities create resilience 145 direct observation of one another and frequent per- additional employment (12 countries), usually in sonal communication to control the scope for ad- informal occupations such as retail and services, verse selection and moral hazard and keep insurance where competition intensified and demand fell. pools financially viable. Kinship and intermarriage Social cohesion sometimes declined, with upticks help foster strong ties. Some kinship groups have in petty theft reported in 10 countries and in sex evolved strong norms about sharing with kin mem- work in 2 countries.19 bers in need, helping to extend the size of the insur- • The most vulnerable are excluded or are included on ance pool beyond the ancestral village. In-group so- very adverse terms. Exclusion keeps people with cial ties reduce problems of information asymmetry high needs and limited ability to contribute out of and enforcement of shared norms. But there are the insurance pool, partly in defense against ad- also drawbacks. Insurance arrangements that rely verse selection. One study among pastoralists in on social ties are easily challenged by in-migration Ethiopia, for example, found that poorer people (new settlers may lack the shared norms) and often lacking cattle are excluded from informal credit, in exclude minorities.17 Community arrangements are part because of repayment risk and in part be- more difficult to sustain in larger groups where peo- cause they are socially invisible.20 Chronically ill ple lack direct interpersonal connections and com- people may experience the same exclusion. A So- munication. Such arrangements therefore work best mali proverb puts it succinctly: “Prolonged sick- for small or one-off idiosyncratic shocks that are easy ness and persistent poverty cause people to hate to verify, such as funerals, and are least effective when you.”21 Bonded labor, child labor, and early mar- communitywide shocks affect many members simul- riage are often the adverse consequences of infor- taneously.18 Linking communities to market-based mal arrangements. Bonded labor results from in- credit and insurance can help them overcome limita- formal credit arrangements in which labor is used tions of small size and in turn can benefit these pro- to guarantee repayment. grams. For example, community insurance groups in Ethiopia are helping to market crop insurance to their members and, by also sharing risk within the Communities as providers of protection group, to overcome the problem of basis risk. Exposure to risk is closely tied to location. Popu- Communities’ own support mechanisms are vi- lation growth and land scarcity are pushing more tal—in fact, they are often people’s only source of and more people into neighborhoods that lack basic help in hard times—but they are not enough and infrastructure, are scarred by crime and violence, are need to be augmented with outside assistance. marred by pollution, and face disaster risk. Latin Viewed as a form of risk management, informal in- America’s 20 largest cities, for example, are all located surance suffers certain deficiencies: in areas prone to flooding or earthquakes or have • Informal insurance is insufficient in the face of sys- steep slopes; they are also plagued by crime.22 temic shocks. Small informal groups cannot effec- tively smooth large shocks. Qualitative research Confronting disasters and climate change during the global food, fuel, and financial crises that started in 2008 found that many people relied Safety, like so much else, has its price. In slums of on costly coping responses as waves of systemic Santo Domingo in the Dominican Republic, rents shocks strained informal mechanisms; reductions are almost twice as high in the safer areas than for in the quality and quantity of food and in nonfood residences near rivers and gullies.23 Almost one- consumption were reported in study sites in all 13 third of Nicaragua’s population lives in areas that countries studied; crime increased in 10 of them. residents say are exposed to disaster, flooding, or The most important sources of assistance were contamination. Among these, half a million people relatives (sites in 13 countries), friends and neigh- live on or at the bottom of hills where they are ex- bors (11 countries), and mutual solidarity groups posed to landslides. This risk is 10 times higher for (7 countries), but such informal support became the poorest 20 percent than for the richest quintile.24 less available and more rationed along ethnic or In Djibouti City, a poor neighborhood is known as religious lines as the ability of better-off com- Bach à l’eau (underneath the water) because of its munity members to contribute declined. Credit frequent flooding. dried up as banks and moneylenders lent smaller Evidence suggests that countries and communi- amounts at higher interest rates. People also sought ties are ill-prepared for the impacts of climate change, 146 WORLD DEVELOPMENT REPORT 2014 such as the higher frequency of extreme weather exist—drought-tolerant varieties, forestry, tree crops, events. Disaster losses have grown significantly in re- migration, small businesses—but lack of credit and cent decades as a consequence of climate change, resources, a preference for traditional staple foods, population pressures, environmental degradation, and similar obstacles often block change. Poorer and other factors. Countries with historically heavy households usually respond to drought and impacts exposure to tropical cyclones are better adjusted to of climate change by selling assets, eating fewer and this risk than countries with weaker exposure, but cheaper meals, borrowing at high cost, and migrat- more recent increases in cyclone risk have not been ing on a seasonal basis. Some of these practices can matched by increases in preparation. One estimate be counterproductive, resulting in a loss of human, finds that countries are unprotected against virtually physical, and environmental assets and undermining all of the additional damage caused by intensified long-term prosperity. cyclone risk.25 Proactive and cost-effective measures to manage Containing crime, violence, and conflict disaster risk are often available but not adopted because of inertia, short-sightedness, or reasons of More than one in ten people in many African and political economy. For example, early warning sys- Latin American countries are victims of crime every tems, preparedness drills, sustainable land use plan- year (figure 4.4). Some 1.5 billion people live in fragile ning, and ecosystem restoration are often good risk or conflict-affected states or in countries with high management but bad politics. In many countries, levels of violent crime. Young men are the chief per- environmental changes and unsustainable agricul- petrators. Men are more likely to be the victims of tural practices have reduced the productivity of farm robbery and assault, while women are more likely to lands. Farmers respond by cutting down forests to face sexual assaults and personal theft. More than half plant on the rich soils beneath—a practice that may of all women in places such as rural Bangladesh, Peru, yield income in the short term but reduces soil fertil- and Tanzania experience domestic violence on a reg- ity and increases sensitivity to climate change in the ular basis (see chapter 3), although it is hugely under- medium term. Changing such behaviors is hard be- reported in official statistics. Rates of crime and vio- cause the benefits are immediate while the costs are lence are usually higher in cities than in rural areas, diffuse and long term.26 Smallholder communities in and within cities, crime and violence are frequently dry areas often rely far too much on a few drought- clustered in poorer communities. In Cape Town, sensitive crops, such as corn (maize). Better options South Africa, 44 percent of all homicides occur in FIGURE 4.4 Insecurity affects more than the victims a. Have you been the b. How frequently do police or c. Have you preferred not to victim of a crime during the military interfere with people’s go out at night for reasons past year? private life? of security? Maximum 80 80 80 75th 70 70 70 percentile % responding “yes” % responding “yes” % responding “yes” 60 60 60 50 50 50 Median 40 40 40 25th 30 30 30 percentile 20 20 20 Minimum 10 10 10 0 0 0 h fic ia La l As nd be th a ric n h fic ia La l As nd be th a ric n h fic ia La l As nd be th a ric n rib nd ric rib nd ric rib nd ric Af ara Af ara Af ara ut ci As ut ci As ut ci As As + e As + e As + e ra e a ra e a ra e a tin ia a tin ia a tin ia a Ca a me Ca a me Ca a me h h h So d Pa ast So d Pa ast So d Pa ast ia Su an ia Su an ia Su an nt op nt op nt op Sa Sa Sa A A A an E an E an E b- b- b- Ce Eur Ce ur Ce ur E E Source: WDR 2014 team based on data from the World Values Survey, 2010–12. Cohesive and connected communities create resilience 147 D I A G R A M 4 .1 Three major types of violence and their spillovers Major political violence Localized collective violence Individual violence • Interstate wars • Organized crime • Assault, theft, robbery • Civil wars • Communal violence (riots, • Domestic violence, child abuse • Irregular armed conflicts programs, blood feuds) • Family or private violence (guerrilla, paramilitary groups) • Gangs and scattered attacks Source: Petesch 2013 for the WDR 2014. three neighborhoods that are among the city’s poor- contests for power among elite actors with ties to est. Rural areas often endure less violence but not nec- armed groups and criminal networks. Weak state essarily less conflict, and disputes over land, livestock, institutions permit such contests to play out vio- or infidelity can erupt into blood feuds and revenge lently. States with stronger institutions and rule killings. The weak presence of the state in many rural of law constrain these conflicts and settle them in areas leaves it up to local informal institutions to ad- a more peaceful manner. Sometimes, there are dress such feuds.27 Not that the security forces are all localized pockets of fragility with dysfunctional that helpful: in many regions, more than one in ten state institutions within otherwise well-function- people say that police or the military interfere “fre- ing states. quently or quite frequently” in people’s private life. • Grievances. Patterns of inequality and marginaliza- Collective forms of violence have far more severe tion across groups holding grievances—some- societal consequences than individual violence and times historical ones—play a major role in fueling petty crime because they also destroy social institu- many conflicts. Tensions often have historic causes tions, displace people, deter investment, trigger fear, or stem from disputes over land and natural re- and cause short-term thinking (diagram 4.1). Farm- sources, which can be worsened by environmental ers grow food crops instead of perennials. According changes. High inequality, perceptions of injustices to one estimate, half the losses stemming from the committed by the state, and politicians playing up conflict in northern Uganda result from cautious re- ethnic divisions for their own purposes can spur sponses to risk (forgone opportunities) rather than ethnic violence and conflict.30 from direct exposure to attacks—with significant risk-related losses also occurring in households that • Conflict legacies. Major political violence is often did not suffer a direct attack. Risk of violence may followed by long periods of heightened crime and have reduced per capita expenditure in the affected violence facilitated by ex-combatants, widespread region by some 70 percent.28 availability of arms, and breakdown of social Most conflicts today have strong ethnic or reli- norms. Trauma caused by past violence, coloniza- gious components, but that does not mean that tion, or a history of state oppression (perhaps un- ethnic differences necessarily cause antagonism. der the dominance of a particular ethnic group) Ethnicity is often exploited for political or economic makes it difficult to trust the state.31 Although time gain, leading to a sharpening of ethnic divisions and heals, its healing effects can be measured in de- tensions.29 The literature is divided on the causes of cades.32 Collective violence such as organized conflict and the role of ethnicity, but it appears that crime and political violence spill over to individual conflict is more likely when social groups lack insti- violence. tutions to bridge their divides, states are weak, and • Climate change. Extreme climatic events (high tem- there are strong external stressors. Four broad types peratures, unusually low rainfall) are associated of stressors stand out and result in spillovers be- with violence, instability, and state collapse. A recent tween types of violence: meta-analysis finds a strong association between • Weak states and links between politicians and crimi- climatic deviations and a wide range of measures of nals. The supply of local violence is often linked to individual violence, intergroup violence, and state 148 WORLD DEVELOPMENT REPORT 2014 collapse, both in recent time and historically. The dients of effective community risk management. mechanisms through which climate changes make Where communities operate in relative isolation, so- conflict more likely are not yet clear. But the asso- cial cohesion (or bonding social capital) helps them ciation between the two means that, in the coming cope by facilitating informal credit, insurance, and decades, large projected changes in precipitation assistance. But this type of social cohesion is rarely and temperature could help drive a major uptick in sufficient to help communities thrive and prosper. violence and conflict.33 Horizontal connections to other communities (or bridging social capital) help communities solve larger Cross-cutting social ties help communities remain problems and prevent conflict. Vertical connections peaceful, even when conflict stressors are present. to markets and governments help them access re- Most ethnically diverse localities remain peaceful, sources and opportunities. whereas others with the same diversity experience The tight-knit rural village of Santa Ana, frequent outbursts of violence. The most Paraguay is an example of a community violent cities tend to lack routine inter- that has mastered collective action. action among members of different Everybody contributes toward the Communities need groups and to be characterized common good by helping, for ex- by divisive local leaders, media, connections to other communities and to markets; ample, to maintain roads, bridges, and criminal gangs. In contrast, water supply, and electricity lines cities with strong civic networks without them communities whenever repairs are needed. to bridge ethnic groups tend to remain insular, lack political They confront cattle thieves and remain peaceful, because those influence, and are unable provide housing to a bus driver networks constrain political lead- to ensure that his route covers to accomplish anything ers from using violent means and their village. A sense of civic duty quell rumors and tensions before at scale. rather than written rules ensures they erupt into riots. These networks that all contribute. Homogeneity and are not designed with violence prevention trusted local leaders contribute to their ex- in mind: they can be film clubs, sports clubs, ceptionally high cohesion. The community did not community associations, and so on. Crucially, they do it all on its own. Roads, electricity, and water sup- span ethnic groups.34 Cross-cutting social ties also ply were constructed by government agencies (often aid in recovery following conflicts and disasters.35 with donor support); the community handles only However, community mechanisms are almost always the maintenance. Although comparative statistics are completely overwhelmed when violence is linked to lacking, the majority of the worlds’ location-based organized criminal networks. communities are far less cohesive than Santa Ana and achieve at best a fraction of its accomplishments. Cohesive and connected communities are What makes people mobilize? Common interests more effective and trusted leaders motivate people to mobilize in community groups and social movements. Risk— Risk need not imply vulnerability if communities and the perceived injustice of mismanaged risk—can can put risk protection in place, either by organizing be a forceful motivating factor. Historically, many ef- it themselves or by mobilizing to demand risk pro- forts to control risk, reduce pollution, and improve tection from local authorities. In higher-income social policies have resulted from successful social countries, local governments routinely clear drains mobilization; a campaign started by antislavery ac- and provide policing, sanitation, and clean water. In tivists and parliamentarians in 1787 led Britain to poor countries—and particularly in poor neighbor- ban its slave trade in 1807, perhaps the first-ever vic- hoods—local governments often do not provide tory by a social movement. The neighborhood com- these basic public goods of common protection. The mittee in the informal settlement of Indio Guays in alternative facing communities is therefore to mobi- Guayaquil, Ecuador, discussed earlier, was formed to lize collective action to create, demand, or maintain help settle land and other disputes and to petition mutual protection, or else suffer exposure to risk. authorities to provide basic infrastructure, health, Communities vary enormously in how effectively and other services. India’s Self Employed Women’s they manage collective risks. Association (SEWA) mobilizes 1.3 million members Social cohesion and connections to markets, around work security, income security, food security, governments, and other communities are the ingre- and social security.36 Cohesive and connected communities create resilience 149 F I G U R E 4 . 5 Membership in voluntary organizations varies widely around the world Percentage of survey respondents who belong to a voluntary organization 100 90 80 70 Maximum 60 Percent 50 40 75th percentile 30 Median 25th percentile 20 10 Minimum 0 East Asia Europe and Latin America Middle East South Asia Sub-Saharan and Pacific Central Asia and the and Africa Caribbean North Africa Source: WDR 2014 team based on data from African Barometer Round 5, 2010–11; East Asia Barometer, 2005–08; Latinobarómetro, 2005; South Asia Barometer, 2005–08; Arab Barometer, 2005–08; and Life in Transition Survey, II 2010. Figures are broadly comparable and, outside Africa, include religious organizations (but not attendance at worship). Groups vary widely in their size and degree of for- tiplies the content of the pot by a factor larger than mality. Burial societies and church groups are typi- one but smaller than the number of players and dis- cally very small and informal. Community-based or- tributes the sum equally among all players, regardless ganizations are slightly larger and sometimes employ of whether they contributed to the pot or kept the a few staff. NGOs are larger still and constituted le- money. In this game, the common good is maxi- gally. Jointly, these groups are often referred to as civil mized when all players contribute their entire society. They pursue many causes such as providing amount, although the individually rational choice is credit, insurance, assistance, and services; managing to contribute nothing. The game has been played in common resources; and providing citizen “voice” different cultures. Inevitably a significant number of demanding services, protection, and accountability players contribute—as long as they believe that oth- from authorities and employers.37 Although diffi- ers will reciprocate. The game has also been modified cult to estimate, the size and importance of civil soci- to allow for repeated interactions and punishment. ety varies significantly. Organizational membership Punishing of free riders (those who do not put funds tends to be highest in South Asia, Africa, and Latin in the common pot) leads to greater group coopera- America (figure 4.5). In Bangladesh and Sri Lanka, 80 tion, and some people are willing to punish even at a and 77 percent of people, respectively, report they are personal cost. Researchers believe that such punish- members of at least one such organization. ment of free riders is crucial for sustaining collabora- Unless they receive outside support, communities tion in the real world.38 need to overcome free rider problems to be effective. Without sustained cooperation, there would be Groups of people find it hard to take collective action no public goods, no collective management of natu- for joint benefits because some members of the ral resources, no social movements, and no society in group free ride on others’ efforts: that is, they con- any meaningful sense. Groups overcome the free sume more or pay less than their fair share of a com- rider problem by devising ways to enforce rules of mon resource. The problem is illustrated by the contribution that reward cooperation and punish “public goods game.” In this experimental game, those who fail to contribute their fair share. Many each player is given a small amount of money that he social norms and institutions also serve to reinforce or she can choose either to keep or to put in a pot mutual collaboration based on reciprocity and pun- that is shared by a group of players. A facilitator mul- ishment. The Tanzanian village of Nyakatoke, for 150 WORLD DEVELOPMENT REPORT 2014 example, operates a villagewide mutual insurance and caused forest degradation in many places. In re- scheme for funerals and hospitalization that includes cent years, many countries have devolved aspects of almost all the women in the village. To guard against forest management to communities, with promising free riding, nonmembers are categorically refused results for sustainability of forest resources.41 How- help, and there are punishments for members who ever, when communities are rife with divisions, there are caught helping nonmembers.39 is a crying need for neutral institutions with capacity Researchers, including the Nobel Prize winner to foster transparency, inclusion, and accountability. Elinor Ostrom, have documented considerable regu- In-group bonding and social cohesion sustains larity in the factors that promote or hinder local gov- communities’ risk management but often excludes ernance and public goods provision. From the per- outsiders and may lead to antagonism toward other spective of community risk management, some of ethnic and religious groups. Social exclusion stems the most important findings show that:40 from attitudes, policies, and laws that discriminate against certain groups on the basis of their ethnic, • Communities whose rights to organize and govern gender, or religious identity, for example. Attitude local affairs are legally recognized are more likely to surveys suggest that tolerance of members of “out- succeed. They must own the fruits of their labor. groups” such as immigrants and ethnic minorities is • Rules need to be enforced. Access to local and not on the rise (figure 4.6). Excluded groups face dis- timely dispute resolution mechanisms are critical crimination that increases their risks and weakens because conflicts among community members or their access to risk management. They often live in with local government are inevitable and need to areas marked by high risk and have little access to be resolved. services, markets, and institutions that could other- wise help them manage risks. They may face harass- • In larger systems, governance activities are often ment, particularly when seeking to explore new op- best organized in multiple, nested layers. For in- portunities or when demanding accountability and stance, when many communities share a common equal treatment. People displaced by violence often water supply, each may manage its own water dis- face debilitating exclusion in their new location (box tribution branch, while a higher body manages the 4.2). Exclusion can lead to risky behaviors such as ex- entire system. cessive drug and alcohol consumption, crime, riots, and violence. • Communities with low inequality, few ethnic divi- The Indian state of Kerala has often been hailed sions, equal treatment, and relatively low discrimi- for its achievements in health, education, and social nation are more likely to create public goods, in development, but it was once a bastion of exclusion. part because of dense interpersonal interactions Kerala has high ethnic and religious heterogeneity. and shared norms about in-group reciprocity. Until the middle of the nineteenth century, it was These findings imply that governments should deeply divided by rigid caste barriers that denied op- create a favorable legal regime that enables commu- portunity to most castes. A movement of assorted nities to undertake their own risk management and ethnic groups started creating a common identity, dispute resolution—provided these function in a based on shared language, culture, and values. Over reasonably fair manner. The high costs and slow pace time, demands for participation in governance grew, of the formal justice system make it unsuitable for as did petitions for collective welfare in the form of many of the small, frequent disputes of community schools and hospitals. By the 1940s, Kerala was a life. Responsibility for managing local risks and dis- forerunner among Indian states in many areas of so- putes can often be delegated to local actors. Decen- cial development.42 tralized management of disaster risk has been found Collaboration within groups is increasingly insuf- to be cost-effective, for example, in part because local ficient and in need of being complemented by cross- knowledge of risk can help keep people’s lives, assets, group collaboration. Cross-cutting social ties, or and businesses out of harm’s way. Moreover, many bridging social capital, mitigates social polarization, countries have enacted laws that undermine com- promotes trust in the system, inoculates society munities’ customary arrangements for settling dis- against identity politics and ethnic riots, and facili- putes and managing common property such as for- tates collective action. These ties need to include a ests without providing adequate formal justice and sufficiently large number of groups.43 How can they resource management systems. The resulting gover- be encouraged and supported? Tanzania under Pres- nance vacuum has weakened forest management ident Julius Nyerere from 1964 to 1985 famously Cohesive and connected communities create resilience 151 F I G U R E 4 . 6 Social exclusion based on ethnicity is not declining in many parts of the world Percentage of the population that would not welcome a neighbor of a different race or ethnic group 35 30 25 20 Percent 15 10 5 0 1980 1985 1990 1995 2000 2005 2010 East Asia and Pacific Middle East and North Africa Europe and Central Asia OECD countries South Asia Sub-Saharan Africa Latin America and the Caribbean Source: Foa 2012 based on data from the World Values Surveys and European Values Study. Note: Organisation for Economic Co-operation and Development (OECD) countries in the figure are all high-income countries that are members of the OECD. All other countries are grouped into geographic regions. downplayed the role of ethnic identity and built a Other approaches used in various countries and cit- shared national Tanzanian identity, fostered through ies include integrating schools, neighborhoods, and a national language, equitable public spending, and a economic life; more-or-less mandatory participation common discourse of the nation’s history. Thanks to in community self-help groups; “truth and reconcili- this foresight, ethnic rivalry never rose to promi- ation committees”; quotas for historically disadvan- nence in Tanzania and did not spill over into conflict taged groups; and local peace committees.44 The les- as happened in so many neighboring countries. sons are that ethnic differences need not result in BOX 4.2 Refugees and internally displaced people: Moving from managing displacement to facilitating opportunities A staggering 43 million people are forcibly displaced. Some 15.2 tional humanitarian assistance to the displaced totals around $8.4 million people are currently refugees because of violent conflict, billion annually, or 6 percent of all official development assistance.a and at least 27.5 million people are internally displaced in more than There is little accountability for long-term results. 40 countries. The displaced are often surprisingly resourceful. Leav- A better approach would recognize the long-term nature of dis- ing behind homes, assets, and familiar environments, they employ placement and help the displaced pursue economic opportunities. dynamic coping strategies to find new livelihoods and even prosper National legislation needs to recognize the displaced, remove dis- in their areas of displacement. But policies often constrain them. crimination, and lift mobility restrictions on refugees. Support could Efforts to address displacement have not been successful. The emphasize housing, education, infrastructure, livelihoods, and average length of time that a refugee or internally displaced person accessing any assets left behind by the displaced. To reduce ten- lives in displacement is close to 20 years, yet displacement is most sions, support should also benefit host communities, as is being often managed as a short-term humanitarian crisis. Many problems done in a World Bank–supported project for displaced people in arise as a result: prolonged residence in camps, dependency on Azerbaijan. Greater convergence of humanitarian and development assistance, restrictions on mobility, and failure to integrate in new funding could help promote long-term approaches. locations. Tensions with host communities are common. Interna- Source: WDR 2014 team based on Lakhani 2013 for the WDR 2014. a. OECD Development Assistance Committee estimates. 152 WORLD DEVELOPMENT REPORT 2014 divided and antagonistic societies if there is trust and instances, actions by agents of the state may result in ties across groups and that both state and nonstate social instability. How can this be countered? actors can promote such ties. Progress countering abuse and discrimination has often resulted from building greater respect for social, civil, and political rights. These include the Public policies to improve local risk right to association; the right to security of people management and property from crime and violence; the right to As discussed, communities’ risk management is nondiscriminatory practices by the state or nonstate grounded in their core strengths—cohesion, vitality, actors; the right to freedom of speech, press, and in- survival skills—but communities usually do not de- formation; and the right for affected groups to voice liver insurance and protection at a large enough their grievances and seek redress. These and other scale, leaving many risks uncovered; weaknesses such rights foster an environment in which collective ac- as exclusion and conflict also often limit community tion and social mobilization can take place and flour- risk management. Communities can become much ish so that groups of people can pursue actions that better risk managers with the right form of support. reduce their collective risks. NGOs, donors, and local and national governments can all help strengthen local risk management using Empower communities to manage risks four broad approaches. The first is to create a favor- able legal regime for communities. The second is to Mobilizing communities can be a powerful force of mobilize communities as risk managers by fostering local development and risk management. The infor- their own capacity to manage local risks. The third is mal settlement of Orangi in Karachi, Pakistan, for to promote their “voice” and ability to influence gov- example, financed its own low-cost sanitation as part ernment-provided risk management. The fourth is of the “Orangi Pilot Project,” started in 1980 by the to provide public goods and services that comple- social activist Akhtar Hameed Khan, assisted by a lo- ment communities’ capacities, involving users as ap- cal NGO. The project also helped the community propriate. These broad approaches all empower organize housing, vaccinations, microfinance, family communities in different ways, as discussed next. planning, and steps to prevent violence, and elicited local government responses to problems too large for the community to handle. Its success rested on the Create a favorable legal environment insight that when the state does not supply essential Ideally, the state provides a neutral system of law and services and public goods, communities can be mo- order where enforcement is impartial, property bilized to do so.45 The project, which has been repli- rights are enforced, crime and violence is kept in cated in other cities of Pakistan and other countries, check, members of different social groups are equally helped identify many of the essential ingredients of respected, and communities’ ability to organize and community-driven development: let communities exercise voice for risk management is protected. themselves define priorities; maintain a long-term However, that is not how many people see the state. presence; and adjust project details over time as new Instead, many people experience state institutions problems, solutions, and opportunities emerge. as unpredictable and unaccountable, one more Such spontaneous social action has inspired do- source of risk one has to navigate. In parts of Africa nors and governments to promote community- and South Asia, chaotic land governance has led to driven development projects that induce communi- corruption and land conflicts. Moreover, case studies ties to mobilize and build their capacity to plan their and media reports abound of instances where street own development. Projects put self-help groups in vendors have their goods confiscated; slum dwellers charge of resources and decision making, supported are violently evicted without notice; workers are pre- by community mobilizers. NGOs, governments, and vented from forming unions that can voice their de- donors, including the World Bank, have promoted mands for workplace safety; undignified treatment such projects in many sectors in both urban and by service providers make excluded groups reluctant rural areas in numerous countries. Indonesia, for to use basic services, leaving them less protected; example, has made the National Program for Com- governments provide arms to unaccountable local munity Empowerment (also known by its acronym, militias; and police and security forces demand PNPM), a pillar of its antipoverty strategy. The pro- bribes and commit abuses (see figure 4.4). In these gram, which started in 1998 at a time of economic Cohesive and connected communities create resilience 153 BOX 4.3 New communication technologies help communities manage violence and local conflict On the ground, patterns of violence are always varied and dynamic. tion riots. It aggregates reports sent by citizens using mobile phones Confusion, uncertainty, and misinformation can make it hard for or the web and, according to Ushahidi, had 45,000 users in Kenya at people to know what is going on and how to stay out of trouble. To the time. It has since been expanded and adapted to various other track, report, and warn of violence, people around the world are uses in crisis situations, including in Haiti after the 2010 earthquake. developing a host of information and communication technologies An initiative based in Cairo, called Harassmap, is a mobile phone– and platforms. Many of these information systems rely on a mix of based reporting system used by women who encounter harass- “crowdsourcing” and maps to report the different forms of violence ment as they move about city streets. It is raising awareness of the and crime and locate emerging hotspots, in real time and in ways problem and warning women of trouble spots. Across Latin Amer- that are both authoritative and widely accessible. The spread of cell ica, people are setting up blogs, websites, and Internet applications, phones helps community members both to report incidences of and using Twitter to report and share information on all forms of crime and violence and to access aggregated results. crime, violence, and police abuse. Helpful as these are, there are also A software platform called Ushahidi (Swahili for testimonial ) was downsides, such as deliberate misinformation. initially developed to map violence during Kenya’s 2008 postelec- View of Cairo showing incidents of sexual harassment reported by the publica 8 13 25 48 97 13 100 3 239 7 2 2 9 5 7 16 32 3 2 3 11 8 3 5 Source: Petesch 2013 for the WDR 2014. a. WDR 2014 team based on data from Harassmap at http://harassmap.org. The red circles in the figure depict the number of reported incidents in that locality. crisis, political transition, and a sharp spike in pov- Regardless of their stated objectives, projects have erty, uses a community-driven development ap- repeatedly discovered that risk management is high proach to build local infrastructure, improve health on communities’ list of priorities: for example, man- and education in the poorest regions, respond to cli- aging droughts and disasters in rural areas or vio- mate change and environmental degradation, and lence and sanitation in urban ones. Modest but per- target particularly marginalized groups. Evaluations sistent technical and organizational assistance can of community-driven development interventions reduce risk greatly. Microfinance can unlock com- have shown mixed but overall good results on pov- munities’ entrepreneurial potential. Civil society has erty; infrastructure costs and quality; and access to helped Bangladesh advance its disaster risk manage- and utilization of health, education, and drinking ment with low-cost, community-based disaster early water. Evaluations have also criticized projects for re- warning systems. Maps and electronic applications stricting themselves to short time frames and failing that show where most crime and violence occurs help to create feedback loops to learn from experience people avoid trouble spots and shame local police and adjust accordingly.46 forces to take action (box 4.3). A project in the Lao 154 WORLD DEVELOPMENT REPORT 2014 People’s Democratic Republic helped communities about how to make authorities pay attention to the integrate disaster risk management into village devel- problem.50 opment plans. In Afghanistan, the National Solidar- The disconnect between communities and gov- ity Program has constructed rural infrastructure ernment came through loud and clear in a multi- with community participation and laid the founda- country study, Voices of the Poor, which asked thou- tions for inclusive local governance in all parts of the sands of poor people around the world which country. Win-win projects can provide local employ- institutions they found most and least effective.51 ment in building protective physical assets or in man- Across locales, respondents considered kin, family, aging or restoring ecosystems as part of disaster pre- and community-based and religious organizations paredness. Community-driven development projects among the most effective support systems. Munici- have responded to disasters with speed and agility in palities, police, and ministries were considered the many countries, including Indonesia, Pakistan, and least effective and were often regarded as unaccount- the Philippines.47 able and unresponsive to the needs of the poor. Ac- Whereas some community-driven approaches counts of neglect, petty corruption, and extortion at have been on a modest scale, Andhra Pradesh and the hand of officials abounded; the police were per- several other Indian states have assisted hundreds of ceived the most negatively and were often described thousands of self-help groups with microfinance and as a source of insecurity. Overall, state institutions in diversification of local livelihoods. Groups are also constituted 33 percent of effective and 83 percent of helped to link up across villages and form associa- ineffective institutions. By contrast, civil society or- tions covering larger areas, creating a movement ganizations constituted 60 percent of the institutions made up of millions of poor people, largely women. considered effective and only 15 percent of those The resulting strength in numbers permits these as- considered ineffective. These findings are echoed— sociations to buy food, obtain credit, access social although in a slightly muted manner—in the World programs, and organize insurance on favorable Values Survey, where confidence in religious leaders terms. Crucially, it also gives them collective voice is far higher than confidence in police and govern- and influence, which they use, for example, to gain ment (figure 4.7). access to social programs or to seek election to the Local democracy, government capacity, and po- village council.48 This approach is now central policy litical economy influence how accountable decen- under India’s National Rural Livelihoods Mission. tralized government is to risks faced by low-income The NGO Bangladesh Rural Advancement Commit- groups. Local governments in many countries lack tee (BRAC) is another example of community-based the capacity, resources, and decision-making author- development that has gone to scale. BRAC, which ity needed to be effective. Electing (rather than ap- started as a relief organization in Bangladesh in 1972, pointing) local governments can help by adding an is now active in 11 countries and assists an estimated element of accountability. Research, mostly from In- 125 million people. dia, finds that decentralized electoral systems tend to deliver benefits to citizens—or at least to the “median voter”—but that they also can induce rent seeking by Empower communities with voice and opportunistic local politicians. 52 Government capac- accountability ity to deliver is often a critical bottleneck: In Peru, for Government institutions are essential for managing example, many local governments underspend their the many risks that exceed the capacity of communi- allocated budget to a considerable degree, with aver- ties, yet these institutions often disappoint. Public age spending in 2009 ranging from 63 to 97 percent health workers are absent more than one-quarter of of the allocation depending on the type of funds. the time in Bangladesh, India, Indonesia, Peru, and Many smaller municipalities lack a trained engineer Uganda.49 The Orangi Pilot Project mentioned ear- to oversee construction projects.53 lier could cost-effectively supply improved toilets Stronger accountability mechanisms where citi- and feeder sewers but had to rely on the municipal- zens, service users, and communities demand better ity’s outdated main sewer lines, making its overall services directly from police, authorities, and utilities sanitation effort less effective. Flooding, a growing can help ensure that public spending is responsive to problem in many African cities, is caused by deficient local risks and delivers results.54 For example, experi- drainage and unregulated urban development— ments in India and Uganda involved public meetings problems that only governments can solve. Commu- where information on health and education entitle- nities bear the cost of flooding but have little idea ments and achievements were disseminated. After Cohesive and connected communities create resilience 155 FIGU R E 4.7 Who do you trust? Confidence in police, government, and religious leaders 82 Sub-Saharan 55 Africa 57 85 Middle East and 63 North Africa 70 67 Latin America 41 and the Caribbean 35 75 Europe and 30 Central Asia 39 66 East Asia and 64 Pacific + South Asia 63 0 10 20 30 40 50 60 70 80 90 % of survey respondents who have reported “a great deal” or “quite a lot” of confidence Religious institutions Government Police Source: WDR 2014 team based on data from the World Values Survey, 2005. one year and four to six meetings, one intervention and social mobilization go hand-in-hand. This law, in India resulted in 27 percent more tetanus vacci- enacted after pressure from the Right to Food Cam- nations, 24 percent more prenatal supplements, 25 paign and others, creates an entitlement of 100 days of percent more infant vaccinations, and fewer excess unskilled employment per year, at minimum wage, to school fees compared with control villages where all workers in rural areas who demand it. The law also no meetings were held.55 These results indicate that provides for social audits and redress of grievances. community participation can improve outcomes for Demand for work is massive, mostly from poor and health and education investments, reduce absentee- disadvantaged groups, and at times of the year where ism, increase enrollment and health care use, and no other work is available. Not only does the program reduce maternal and infant mortality. offer a useful safety net, but it also helps spread aware- Citizen participation works best when higher-tier ness of rights and promotes dignity. Implementation government is responsive.56 Localized accountability varies across states: no state is able to offer all workers mechanisms are useful but may not be transformative all the employment they demand and are entitled to. on their own. Local accountability is more powerful A state such as Rajasthan, which promotes transpar- when the state creates complementary mechanisms to ency and accountability and has a long history of sanction corruption, inadequate services, and abuse popular mobilization, performs relatively better: in of authority. These mechanisms can entail an inde- Rajasthan, 84 percent of job seekers report being suc- pendent judiciary or auditing institutions. Moreover, cessful (against 56 percent nationwide), receiving 71 rights and social guarantees can galvanize social days of employment (against 37 nationwide), on aver- movements and foster collective action by offering age. The fact that the law is organized as a right moti- critical entry points for citizens and NGOs demanding vates job seekers’ collective action to hold authorities better risk management. Modern information and accountable for supplying employment instead of communication technologies make it simpler and siphoning off the allocated funds.57 faster for individuals and communities to mobilize. The right to assembly and political participation en- Provide infrastructure and services that ables citizens to demand economic and personal pro- complement communities’ capacities tection and better services. India’s Mahatma Gandhi National Rural Employ- Basic infrastructure enables people to manage risk ment Guarantee Act illustrates how good governance and seize opportunities. Mobile phone penetration 156 WORLD DEVELOPMENT REPORT 2014 BOX 4.4 Clean, green, and blue: Managing water and flooding in Singapore In the 1960s, Singapore imported most of its drinking water, was fre- The government closed or relocated polluting industries such as quently flooded, and suffered from highly polluted rivers, air, land, pig and duck farms, informal manufacturers, and street vendors; it and streets. Singapore’s transformation, achieved in four decades, built new food centers and markets to house food vendors; and it shows that the rapidly growing urban population in the developing engaged the population to conserve water and stop pollution and world can aspire to a safe and clean living environment with suffi- littering. Today, Singapore has put in place a diversified water sup- cient water. Singapore moved forward with a long-term vision that ply strategy known as the Four National Taps comprising water the country could achieve more control over its water supply and from local catchments, imported water, high-grade reclaimed become a livable city without sacrificing economic growth and com- water, and desalinated water. A large dam, the Marina Barrage, was petitiveness. It relied on research and development; effective imple- built to create a freshwater reservoir that has a catchment area of mentation; and a combination of engineering, political commitment, 10,000 hectares, while regulating flooding of low-lying areas in the and community-based measures. city center. In water resource management, Singapore’s efforts focused on Singapore recognizes that climate change effects could result in building up its local catchment water resources by capturing, stor- more intense rainfall and rising sea levels. To cope with more intense ing, and using more of its heavy rainfall, as well as creating sources rainfall, it is strengthening its drainage infrastructure and introduc- of water less dependent on the weather. The city built reservoirs to ing measures to better control storm water. One measure includes a store rainwater collected from rivers, streams, canals, and drains new requirement for developers to implement on-site measures to that converted large parts of the city-state into a rainwater catch- slow down surface runoff and reduce the peak flow of storm water ment area. Developing separate collection systems for used water into the public drainage system. In anticipation of rising sea levels, and rainwater at the outset allowed Singapore to integrate its drain- Singapore has added to its minimum reclamation. Since 2011, newly age system and channel rainwater to its reservoirs without affecting reclaimed land must be raised by a minimum of 1 meter in addition water quality. To maintain the cleanliness of the water supply, to the previous level of 1.25 meters above the highest recorded tide streams and culverts had to be kept free of garbage and pollution. level observed before 1991. Source: WDR 2014 team based on Soon, Jean, and Tan 2009 and contributions from PUB, Singapore’s national water agency. has skyrocketed, even in the most remote, low- necessarily reduce crime (spotlight 4). Communities income regions; meanwhile, 88 percent of the global can help create awareness, knowledge, and changes population now has access to an improved water in norms and behavior—something India is exploit- source. Singapore is an example of a city-state that ing with a campaign that discourages young women took a proactive approach to bring health and envi- from marrying men who do not have a toilet. To re- ronmental risks under control and in the process duce the spread of HIV, changes in sexual behavior created one of Asia’s most livable cities (box 4.4). are needed, in particular among high-risk groups But such an approach to managing risks to such as sex workers and men who have sex the public remains elusive in many low- with men. Community-based organiza- and lower-middle-income countries, tions can reach such groups far better where 2.7 billion people lack access Understanding the than can governments. They can tai- to sanitation (figure 4.8) and 1 bil- strengths and limitations lor interventions to key obstacles lion defecate in the open. Open blocking the adoption of risk- of communities’ risk sewers and garbage still mar reducing behaviors. In western many slums. Such environmental management can help Kenya, for example, community- conditions cause much disease. In governments design based organizations provided Sub-Saharan Africa, one of every complementary policy awareness and knowledge that eight children, and in South Asia, increased the use of condoms; in actions. one of every fifteen, die before their other localities, where knowledge was fifth birthday. already high, community-based HIV in- Many health, sanitation, security, and terventions instead targeted the stigma that other programs have concluded that involving and stops people from seeking testing and provided em- partnering with user communities improves out- powerment and motivation for behavior change.58 comes (photo 4.1); such partnerships may even be Working closely with affected communities helps necessary for success. Simply providing a toilet does programs like these discover the obstacles for low not necessarily improve sanitation (box 4.5) in the uptake of risk management and redesign approaches same way that provision of more police does not accordingly. Cohesive and connected communities create resilience 157 F I G U R E 4 . 8 More people have cell phones than have toilets—illustrating a failure to provide the most basic protection Organisation for Economic Co-operation 100 and Development Eastern Europe and Central Asia Middle East and North Africa 80 Latin America and the Caribbean East Asia and Pacific 60 South Asia 40 Sub-Saharan Africa 20 0 Mobile phone Access to subscriptions sanitation facilities, per 100 people % of population Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. Data are as of 2010. Income-support programs can often benefit from community participation, particularly in low- and lower-middle-income countries. Central au- thorities are better at identifying poor localities than at identifying the poor within them. Community participation in targeting can help by exploiting lo- cal knowledge and keeping elite capture of benefits in check—although the way the process is designed matters crucially.59 An experiment in Indonesia compared community-based targeting conducted in villagewide meetings with a proxy means test. The community method resulted in slightly worse tar- geting at the $2 poverty line, equal targeting at $1 a day, and greater satisfaction and legitimacy overall (fewer complaints, fewer difficulties in distributing the funds, and greater acceptance of the beneficiary list).60 Participation is equally important in provid- ing relief after disasters: humanitarian assistance to FPO people affected by disasters is often more effective when it takes local conditions into account and uses communities’ own capacities. Likewise, an emphasis on rapid reconstruction, without adequate commu- P H O T O 4 .1 Clean water and community-led sanitation improve hygiene and nity involvement, can lead to recovering in ways that reduce diarrheal and other diseases. Children draw water from a water pump in recreate vulnerabilities.61 North Sudan. © Fred Noy/UN Photo 158 WORLD DEVELOPMENT REPORT 2014 BOX 4.5 Using social marketing to increase access to sanitation in rural Tanzania Around 1 billion people worldwide defecate in the open, creating and complicated. Although people were very dissatisfied with their vast risks for infants and children that often play amid excreta and situation, affordable options were lacking, sanitation was not seen waste water. Poor sanitation is responsible for an estimated 1.7 mil- as a high priority, and a sense of powerlessness prevailed. lion deaths each year, is a cause of high medical bills, and is linked to The program developed an inspirational promotion platform, disease and stunting among children. It can also impair dignity, gen- called Choo Bora Chawezekana! (A Good Toilet Is Possible!) and deliv- der equality, and quality of life. Agencies have learned that simply ered the message to an estimated 160,000 people in rural districts giving households toilets does not work. Instead, sanitation needs to through radio spots, radio soap opera, and local events. District and be understood within the context of the obstacles households face. ward authorities, masons, and village committees were employed Social marketing is being used in sanitation projects in rural as front-line promoters of the Choo Bora brand. Local masons pro- Tanzania, where most toilets are basic pit latrines and diarrhea is duced an affordable cement slab for the toilet. The work of learning common among children. The World Bank’s Water and Sanitation how Tanzanian households use sanitation continues. Findings from Program conducted outreach activities with rural Tanzanians to focus groups were incorporated into the National Sanitation Cam- understand their experience using toilets. People considered paign, which aims to convince 6 million rural Tanzanians to invest in improving toilets and sanitary practices to be a low priority, costly, and use improved sanitation facilities by 2015. Community-event in rural Tanzania promoting Two women after purchasing a cement slab the Choo Bora message for a new toilet Source: Jacqueline Devine and Jason Cardosi for the WDR 2014. Photos: © Water and Sanitation Program staff. Win-win investments should be favored. Not only munities by creating and maintaining infrastructure can public works offer employment to unemployed and providing social services.63 (For more win-win workers; they can also boost preparation for crises examples, see spotlight 2.) and shocks by building local public goods or enhanc- Earlier concerns that income-support programs ing employment skills. Public works projects often might crowd out community mechanisms seem exag- build dams, shelters, drainage, and rural roads, and gerated, at least in developing countries. Some studies undertake soil, water, and forest conservation that find that public safety nets displace informal transfers protects communities against disasters. Djibouti, for to a moderate or even significant extent.64 As with example, started a workfare program when the eco- credit, some substitution of formal sources for infor- nomic crisis hit in 2008. The program offers short- mal ones is to be expected. But displacement of private term employment in community-based, labor- transfers does not translate to a net social loss when re- intensive works. It also supports improved nutrition cipients share their benefits with other poor people or practices, focusing on preschool children and preg- are able to increase their labor supply. This process has nant or lactating women.62 Many programs aim to been studied extensively for South Africa’s relatively impart skills that people can use to diversify their generous old-age social pensions. Unemployment income sources. Latvia has used workfare to create is high in South Africa and supporting unemployed temporary labor-intensive employment for people family members drags many people into poverty, who had lost their jobs but were ineligible for unem- particularly families without any pensioner. When ployment benefits. The program has benefited com- an elderly person starts receiving a pension, working- Cohesive and connected communities create resilience 159 age family members are more likely to migrate for ture, and land acquisition are taken without consult- work—not less so—as the pension finances migra- ing the people they affect, opportunities to identify tion.65 This way, access to risk management enables potential risks are missed. migration, employment, and escape from poverty. Promote inclusion and accountability. State institu- tions need to strive for neutrality and reduce dis- Putting it all together: Policy principles crimination and practices that are unpredictable, and research priorities to foster resilient abusive, or illegal. Ethnic favoritism, and perceptions communities thereof, should be avoided when targeting income Communities can become much better risk manag- support, services, and public sector employment, ers when supported by a favorable legal regime; when particularly in fragile environments. Local militias their organizational capacity is strengthened; when should not be armed, as they often become unac- there is voice, transparency, and accountability; and countable or even criminal over time. Civil society when complementary public goods and services are and the press should not be constrained in their abil- provided, including those that link them to markets ity to serve as watchdogs. The press, on the other and other communities. Many different policies can hand, needs to behave responsibly and refrain from be used to this end. These policies are summarized in playing up ethnic tensions and stereotypes. table 4.1 by type of risk management instrument and by applying the lens of the five simple policy princi- Think long run ples in the discussion that follows. Prepare for disasters and emergencies in advance. Much disaster and humanitarian response is im- Do not generate uncertainty or peded by an absence of preparation: agencies raise unnecessary risk funding for response on an ad hoc basis once emer- Consult communities on important local decisions that gencies have been declared; competition over funds involve them. When decisions on services, infrastruc- limits coordination; a disaster’s ability to grab head- TA B L E 4 .1 Policy priorities to improve risk management at the community level POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Transparency and freedom of the press Simpler conflict filters for fragile Violence and disaster hotspot maps settings Protection Law and order for all; anti-discrimination measures Local institutions that bridge divided Community-based crime prevention groups (particularly for fragile settings) Community infrastructure and capacity for disaster management Insurance Credit and savings (group-based or Credit, savings, and insurance (bank-based) microfinance) Coping Workfare for employment and basic Workfare for skill building infrastructure Making humanitarian relief accountable Support delivered via permanent for long-term results government systems Community-targeted income support Means-tested income support Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing poli- cies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably and that can be improved over time. 160 WORLD DEVELOPMENT REPORT 2014 lines determines funding; there are few incentives chapter 3.) Income support is a useful complement (and little funding) to invest ex ante in risk protec- to informal community coping and insurance— tion and preparedness; and permanent systems for which are likely to be overwhelmed by systemic delivering income and other support are missing. shocks—and can avoid costly and irreversible cop- Moreover, since most humanitarian funding is spent ing responses. Income-transfer programs need to be on complex, protracted emergencies and not on in- scalable and flexible in order to increase coverage in tensive disasters, support that is inherently meant to communities facing shocks and be scaled back once be temporary and short-term in many cases ends up crisis abates. The beneficiary selection process carrying on for the long haul (see box 4.2). Putting in should be able to identify those most affected by place a coordinated funding mechanism for disaster shocks, not just the chronic poor. Community- and humanitarian preparation and response would based targeting can help in this regard, often in con- help ensure timely responses, investment in prepara- junction with geographic targeting to affected re- tion, and accountability for long-term outcomes (see gions, or self-targeting. also chapter 8).66 A final note: Refocus research priorities from Build transparent, accountable, inclusive, and scal- diagnostics to solutions able institutions, and realize that fostering communi- ties’ cohesion and capacity to organize their own risk Refocus research priorities to investigate the underly- management takes time and patience. In fragile and ing causes of vulnerability, how to address them, and conflict-affected countries, in particular, restoring how to promote opportunity. Among all the topics confidence and institutional capacity may take a covered in this chapter, by far the most researched generation. Therefore, it is best to stay engaged for has been informal insurance in rural areas. Urban the long haul, and to use that time to experiment areas have received less attention. Preparation for with ways to scale up. risk, which arguably can have the farthest reaching effects, seems underresearched. In general, research tends to be stronger on the diagnostics—the nature Promote flexibility of shocks, their impacts, and the ways people cope— Listen to communities, allow local actors discretion in than on the particular policies to address vulnerabil- devising local solutions, adopt learning-by-doing, and ity and ways they might promote opportunity. Fi- build feedback loops. All of these activities can help nancial and social protection instruments have with adaptation of interventions over time and attracted more research interest, particularly in eco- across space as local risks evolve as well as offer op- nomics, than broader policies pertaining to partici- portunities for managing them. pation, fostering collective action, scaling up local initiatives, upgrading slums, improving governance, and fighting exclusion and discrimination; yet such Provide the right incentives “softer” approaches may well have more transforma- Promote social accountability and transparency at lo- tional impacts. cal and national levels. Local accountability measures can contribute to better and more consistent delivery Make it easier to consider communities in develop- of local services and risk management by making ment planning. Preventing local interventions from service providers perform better, ensuring that having unintended harmful consequences often re- spending meets local needs, and reducing corrup- quires detailed understanding of specific contexts, tion. To work well, local accountability measures of- knowledge that can be time-consuming for planners ten require complementary national action such as to acquire. It would therefore be useful if social sci- open flow of information, freedom of the press, and entists could develop simpler and quicker tools mechanisms for holding authorities accountable. (sometimes known as “conflict filters”) to ensure that development planning takes local situations into account. More broadly, research on organiza- Include the vulnerable in protection tional models such as associations of the poor that Focus public action on ways to protect people’s basic can scale up community-based solutions and also consumption and access to health and education dur- deal better with large, complex problems would help ing systemic shocks. (This recommendation builds on fill a major void. Cohesive and connected communities create resilience 161 Notes 33. Hsiang, Burke, and Miguel 2013. 34. Varshney 2002. 35. Narayan, Nikitin, and Petesch 2010. 1. Klinenberg 2002. 36. World Bank 2012. 2. De Weerdt 2001. 37. Mansuri and Rao 2012. 3. Dercon 2008; Jalan and Ravallion 2001. 38. Bowles and Gintis 2002; Ostrom 2000 offers an insightful sur- 4. Bowles and Gintis 2002. vey of the literature. 5. Christiaensen and Subbarao 2005 for Kenya; Dercon and 39. This refusal has serious consequences, since funerals are elab- Krishnan 2000 for Ethiopia; Heltberg and Lund 2009 for Paki- orate affairs and there are bound to be logistical disasters stan; Townsend 1994 for rural India; and Dercon and De without help. See De Weerdt 2001. Weerdt 2006 for community networks to help with illness. See 40. Ostrom 2000; Bowles and Gintis 2002; see also Habyarimana also chapters 1 and 3. and others 2007 and the sources cited therein. 6. The term community has its origins in the Latin term com- 41. Ostrom 2000; Mansuri and Rao 2012. munis—shared by all or many. Communities are defined by 42. Singh 2011. relationships and connections, not affection (Bowles and Gin- 43. Marc and others 2013; Singh 2011. tis 2002). Community denotes the group. Social capital refers 44. World Bank 2010. to aspects of relationships within the group, such as norms 45. Khan 1996. and trust. 46. Mansuri and Rao 2012. 7. Moser 2009. 47. World Bank 2006. 8. Fafchamps and De Weerdt 2011. 48. Narayan, Prennushi, and Kapoor 2009. 9. Chen 2010. 49. Chaudhury and others 2006. 10. Mohapatra, Joseph, and Ratha 2012. 50. Douglas and others 2008. 11. Fafchamps and De Weerdt 2011; Morduch 2002; Ravallion 51. Narayan and others 2000, 198–202. and Chaudhuri 1997; Townsend 1994. 52. Besley, Pande, and Rao 2012; Mansuri and Rao 2012. 12. Gertler and Gruber 2002. 53. Loayza, Rigolini, and Calvo-Gonzalez 2011. 13. Mohapatra, Joseph, and Ratha 2012. 54. Mansuri and Rao 2012 and Speer 2012 provide recent surveys; 14. Jalan and Ravallion 1999. see also World Bank 2003. 15. Dercon 2002. 55. Pandey and others 2007. Bjorkman and Svensson 2009 found 16. Fafchamps and Gubert 2007. broadly similar results in Uganda, while Banerjee and others 17. Attanasio and others 2012; Coate and Ravallion 1993; De 2010 did not find significant impacts on services of giving in- Janvry and others 2002. formation, perhaps because their experiment was of shorter 18. De Janvry and others 2002; Genicot and Ray 2003. duration. 19. Heltberg, Hossain, and Reva 2012. 56. Devarajan, Khemani, and Walton 2011; Mansuri and Rao 20. Santos and Barrett 2011. 2012. 21. World Bank 2000, 16. 57. Dreze and Khera 2011; Liu and Barrett 2013. 22. Fay 2005. 58. Rodriguez-Garcia and others 2013. 23. Fay 2005. 59. Coady, Grosh, and Hoddinott 2004; Mansuri and Rao 2012. 24. WDR 2014 team analysis based on Nicaragua Encuesta Na- 60. Alatas and others 2012. cional de Hogares sobre Medición de Nivel de Vida 2005. 61. IPCC 2012. 25. Hsiang 2012. 62. Levin, Morgandi, and Silva 2012. 26. Ashwill and Heltberg 2013 for the WDR 2014. 63. Azam, Ferre, and Ajwad 2012. 27. Petesch 2013 for the WDR 2014; World Bank 2010. 64. Cox, Hansen, and Jimenez 2004; Dercon and Krishnan 2003. 28. Rockmore 2011. 65. Ardington, Case, and Hosegood 2009. 29. Esteban, Mayoral, and Ray 2012. 66. van Aalst and others 2013 for the WDR 2014. 30. Varshney 2002; see also Lederman, Loayza, and Menendez 2002. 31. Marc and others 2013. 32. Collier 2009, 89; Petesch 2013 for the WDR 2014; World Bank 2010, ch. 2. 162 WORLD DEVELOPMENT REPORT 2014 References Dercon, Stefan. 2002. “Income Risk, Coping Strategies, and Safety Nets.” World Bank Research Observer 17 (2): 141–66. Alatas, Vivi, Abhijit Banerjee, Rema Hanna, Benjamin A. Olken, ———. 2008. “Fate and Fear: Risk and Its Consequences in and Julia Tobias. 2012. “Targeting the Poor: Evidence from a Africa.” Journal of African Economies 17 (2): 97–127. Field Experiment in Indonesia.” American Economic Review Dercon, Stefan, and Joachim De Weerdt. 2006. “Risk Sharing Net- 102 (4): 1206–40. works and Insurance against Illness.” Journal of Development Ardington, Cally, Anne Case, and Victoria Hosegood. 2009. “Labor Economics 81 (2): 337–56. Supply Responses to Large Social Transfers: Longitudinal Evi- Dercon, Stefan, and Pramila Krishnan. 2000. “In Sickness and in dence from South Africa.” American Economic Journal: Applied Health: Risk Sharing within Households in Rural Ethiopia.” Economics 1 (1): 22–48. Journal of Political Economy 108 (4): 688–727. Ashwill, Maximillian, and Rasmus Heltberg. 2013. “Is There a ———. 2003. “Risk Sharing and Public Transfers.” Economic Jour- Community-Level Adaptation Deficit?” Background paper for nal 113 (486): 86–94. the World Development Report 2014. Devarajan, Shantayanan, Stuti Khemani, and Michael Walton. Attanasio, Orazio, Abigail Barr, Juan C. Cardenas, Garance Geni- 2011. “Civil Society, Public Action and Accountability in cot, and Costas Meghir. 2012. “Risk Pooling, Risk Preferences, Africa.” Policy Research Working Paper 5733, World Bank, and Social Networks.” American Economic Journal: Applied Washington, DC. Economics 4 (2): 134–67. De Weerdt, Joachim. 2001. “Community Organizations in Rural Azam, Mehtabul, Celine Ferre, and Mohamed Ihsan Ajwad. 2012. Tanzania: A Case Study of the Community of Nyakatoke, “Did Latvia’s Public Works Program Mitigate the Impact of Bukoba Rural District.” The Nyakatoke Series 3. the 2008–2010 Crisis?” Policy Research Working Paper 6144, Douglas, Ian, Khurshid Alam, Maryanne Maghenda, Yasmin World Bank, Washington, DC. McDonnell, Louise McLean, and Jack Campbell. 2008. “Unjust Banerjee, Abhijit V., Rukmini Banerji, Esther Duflo, Rachel Waters: Climate Change, Flooding and the Urban Poor in Glennerster, and Stuti Khemani. 2010. “Pitfalls of Participa- Africa.” Environment and Urbanization 20 (1): 187–205. tory Programs: Evidence from a Randomized Evaluation in Dreze, Jean, and Reetika Khera. 2011. The Battle for Employment Education in India.” American Economic Journal: Economic Guarantee. New Delhi: Oxford University Press. Policy 2 (1): 1–30. Esteban, Joan, Laura Mayoral, and Debraj Ray. 2012. “Ethnicity Besley, Timothy, Rohini Pande, and Vijayendra Rao. 2012. “Just and Conflict: An Empirical Study.” American Economic Review Rewards? Local Politics and Public Resource Allocation in 102 (4): 1310–42. South India.” World Bank Economic Review 26 (2): 191–216. Fafchamps, Marcel, and Joachim De Weerdt. 2011. “Social Identity Bjorkman, Martina, and Jakob Svensson. 2009. “Power to the and the Formation of Health Insurance Networks.” Journal of People: Evidence from a Randomized Field Experiment on Development Studies 47 (8): 1152–77. Community-Based Monitoring in Uganda.” Quarterly Journal of Economics 124 (2): 735–69. Fafchamps, Marcel, and Flore Gubert. 2007. “The Formation of Bowles, Samuel, and Herbert Gintis. 2002. “Social Capital and Risk Sharing Networks.” Journal of Development Economics 83 Community Governance.” Economic Journal 112 (483): 419–36. (2): 326–50. Chaudhury, Nazmul, Jeffrey S. Hammer, Michael Kremer, Karthik Fay, Marianne. 2005. The Urban Poor in Latin America. Directions Muralidharan, and F. Halsey Rogers. 2006. “Missing in Action: in Development. Washington, DC: World Bank. Teacher and Health Worker Absence in Developing Coun- Foa, Roberto. 2012. “Trends in Tolerance of Social Minorities tries.” Journal of Economic Perspectives 20 (1): 91–116. across the World.” Unpublished manuscript, World Bank, Chen, Daniel L. 2010. “Club Goods and Group Identity: Evidence Washington, DC. from Islamic Resurgence during the Indonesian Financial Genicot, Garance, and Debraj Ray. 2003. “Group Formation in Crisis.” Journal of Political Economy 118 (2): 300–54. Risk-Sharing Arrangements.” Review of Economic Studies 70 Christiaensen, Luc J., and Kalinidhi Subbarao. 2005. “Towards an (1): 87–113. Understanding of Household Vulnerability in Rural Kenya.” Gertler, Paul, and Jonathan Gruber. 2002. “Insuring Consumption Journal of African Economies 14 (4): 520–58. against Illness.” American Economic Review 92 (1): 51–70. Coady, David, Margaret E. Grosh, and John Hoddinott. 2004. Tar- Habyarimana, James, Macartan Humphreys, Daniel N. Posner, geting of Transfers in Developing Countries: Review of Lessons and Jeremy M. Weinstein. 2007. “Why Does Ethnic Diversity and Experience. World Bank Regional and Sectoral Studies. Undermine Public Goods Provision?” American Political Washington, DC: World Bank. Science Review 101 (4): 709–25. Coate, Stephen, and Martin Ravallion. 1993. “Reciprocity without Heltberg, Rasmus, Naomi Hossain, and Anna Reva. 2012. Living Commitment: Characterization and Performance of Informal through Crises: How the Food, Fuel, and Financial Shocks Affect Insurance Arrangements.” Journal of Development Economics the Poor. Washington, DC: World Bank. 40 (1): 1–24. Heltberg, Rasmus, and Niels Lund. 2009. “Shocks, Coping, and Collier, Paul. 2009. Wars, Guns, and Votes: Democracy in Dangerous Outcomes for Pakistan’s Poor: Health Risks Predominate.” Places. New York: Harper Collins. Journal of Development Studies 45 (6): 831–946. Cox, Donald, Bruce E. Hansen, and Emmanuel Jimenez. 2004. Hsiang, Solomon. 2012. “Are We Well Adapted to Our Climate? “How Responsive Are Private Transfers to Income? Evidence And What Might That Mean for Future Climate Changes?” from a Laissez-Faire Economy.” Journal of Public Economics 88 Earth Magazine 57 (8): 8. (9–10): 2193–219. Hsiang, Solomon, Marshall B. Burke, and Ted Miguel. 2013. De Janvry, Alain, Rinku Murgai, Elisabeth Sadoulet, and Paul Win- “Quantifying the Climatic Influence on Human Conflict, ters. 2002. “Localized and Incomplete Mutual Insurance.” Violence and Political Instability.” Science DOI: 10.1126/ Journal of Development Economics 67 (2): 245–74. science.1235367. Cohesive and connected communities create resilience 163 IPCC (Intergovernmental Panel on Climate Change). 2012. “Man- yan, 234–85. Washington, DC: World Bank and Palgrave aging the Risks of Extreme Events and Disasters to Advance Macmillan. Climate Change Adaptation.” Special Report, IPCC, New York. Ostrom, Elinor. 2000. “Collective Action and the Evolution of So- Jalan, Jyotsna, and Martin Ravallion. 1999. “Income Gains to the cial Norms.” Journal of Economic Perspectives 14 (3): 137–58. Poor from Workfare: Estimates for Argentina’s Trabajar Pro- Pandey, Priyanka, Ashwini R. Sehgal, Michelle Riboud, David gram.” Policy Research Working Paper 2149, World Bank, Levine, and Madhav Goyal. 2007. “Informing Resource-Poor Washington, DC. Populations and the Delivery of Entitled Health and Social ———. 2001. “Behavioral Responses to Risk in Rural China.” Services in Rural India.” The Journal of the American Medical Journal of Development Economics 66 (1): 23–49. Association 298 (16): 1867–75. Khan, Akhter Hameed. 1996. Orangi Pilot Project: Reminiscences Petesch, Patti. 2013. “How Communities Manage Risks of Crime and Reflections. Karachi: Oxford University Press. and Violence.” Background paper for the World Development Klinenberg, Eric. 2002. Heat Wave: A Social Autopsy of Disaster in Report 2014. Chicago. Chicago: University of Chicago Press. Ravallion, Martin, and Shubham Chaudhuri. 1997. “Risk and In- Lakhani, Sadaf S. 2013. “Forced Displacement: Moving from Man- surance in Village India: Comment.” Econometrica: Journal of aging Risk to Facilitating Opportunity.” Background paper for the Econometric Society 65 (1): 171–84. the World Development Report 2014. Rockmore, Marc. 2011. “The Cost of Fear: The Welfare Effects of Lederman, Daniel, Norman Loayza, and Ana Maria Menendez. the Risk of Violence in Northern Uganda.” Working Papers 2002. “Violent Crime: Does Social Capital Matter?” Economic 109, Households in Conflict Network, Brighton, UK. Development and Cultural Change 50 (3): 509–39. Rodriguez-Garcia, Rosalia, Rene Bonnel, David Wilson, and Levin, Victoria, Matteo Morgandi, and Joana Silva. 2012. Inclusion N’Della N’Jie. 2013. Investing in Communities Achieves Results: and Resilience: The Way Forward for Social Safety Nets in the Findings from an Evaluation of Community Responses to HIV Middle East and North Africa. Middle East and North Africa and AIDS. Directions in Development. Washington, DC: Development Report. Washington, DC: World Bank. World Bank. Liu, Yanyan, and Christopher B. Barrett. 2013. “Heterogeneous Santos, Paulo, and Christopher B. Barrett. 2011. “Persistent Pov- Pro-Poor Targeting in India’s Mahatma Gandhi National Ru- erty and Informal Credit.” Journal of Development Economics ral Employment Guarantee Scheme.” Economic and Political 96 (2): 337–47. Weekly 48 (10): 46–53. Singh, Prerna. 2011. “We-Ness and Welfare: A Longitudinal Analy- Loayza, Norman, Jamele Rigolini, and Oscar Calvo-Gonzalez. sis of Social Development in Kerala, India.” World Develop- 2011. “More Than You Can Handle: Decentralization and ment 39 (2): 282–93. Municipal Spending Ability in Developing Countries.” Policy Soon, Tan Yong, Lee Tung Jean, and Karean Tan. 2009. Clean, Research Working Paper 5763, World Bank, Washington, DC. Green and Blue: Singapore’s Journey towards Environmental Mansuri, Ghazala, and Vijayendra Rao. 2012. Localizing Develop- and Water Sustainability. Singapore: Institute of Southeast ment: Does Participation Work? Policy Research Report. Wash- Asian Studies. ington, DC: World Bank. Speer, Johanna. 2012. “Participatory Governance Reform: A Marc, Alexandre, Alys Willman, Ghazia Aslam, Michelle Rebosio, Good Strategy for Increasing Government Responsiveness and Kanishka Balasuriya. 2013. Societal Dynamics and Fragil- and Improving Public Services?” World Development 40 (12): ity: Engaging Societies in Responding to Fragile Situations. 2379–98. Washington, DC: World Bank. Townsend, Robert M. 1994. “Risk and Insurance in Village India.” Mohapatra, Sanket, George Joseph, and Dilip Ratha. 2012. “Re- Econometrica: Journal of the Econometric Society 62 (3): mittances and Natural Disasters: Ex-Post Response and Con- 539–91. tribution to Ex-Ante Preparedness.” Environment, Develop- van Aalst, Maarten, Jan Kellett, Florence Pichon, and Tom Mitch- ment and Sustainability 14 (3): 365–87. ell. 2013. “Incentives in Disaster Risk Management and Hu- Morduch, Jonathan. 2002. “Consumption Smoothing across manitarian Response.” Background paper for the World Devel- Space: Testing Theories of Risk-Sharing in the Icrisat Study opment Report 2014. Region of South India.” United Nations University Wider Dis- Varshney, Ashutosh. 2002. Ethnic Conflict and Civic Life: Hindus cussion Paper 55, United Nations University–World Institute and Muslims in India. New Haven, CT: Yale University Press. for Development Economics Research, Tokyo. World Bank. 2000. World Development Report 2000/2001: Attack- Moser, Caroline O. N. 2009. Ordinary Families, Extraordinary ing Poverty. New York: Oxford University Press. Lives: Assets and Poverty Reduction in Guayaquil, 1978–2004. ———. 2003. World Development Report 2004: Making Services Washington, DC: Brookings Institution Press. Work for Poor People. New York: Oxford University Press. Narayan, Deepa, Robert Chambers, Meera K. Shah, and Patti ———. 2006. Hazards of Nature, Risks to Development: An IEG Petesch. 2000. Crying Out for Change. Vol. 2 of Voices of the Poor. Evaluation of World Bank Assistance for Natural Disasters. Washington, DC: World Bank and Oxford University Press. Washington, DC: World Bank. Narayan, Deepa, Denis Nikitin, and Patti Petesch. 2010. “Building ———. 2007. “Social Protection in Pakistan: Managing House- States from the Bottom Up in Conflict-Affected Countries.” In hold Risks and Vulnerability.” World Bank, Washington, DC. Rising from the Ashes of Conflict. Vol. 4 of Moving out of Pov- ———. 2010. World Development Report 2011: Conflict, Security, erty, edited by Deepa Narayan and Patti Petesch, 1–190. Wash- and Development. Washington, DC: World Bank. ington, DC: World Bank and Palgrave Macmillan. ———. 2012. World Development Report 2013: Jobs. Washington, Narayan, Deepa, Giovanna Prennushi, and Soumya Kapoor. 2009. DC: World Bank. “People’s Organizations and Poverty Escapes in Rural Andhra ———. World Development Indicators (database). World Bank, Pradesh.” In The Promise of Empowerment and Democracy in Washington, DC., http://data.worldbank.org/data-catalog/ India. Vol. 3 of Moving out of Poverty, edited by Deepa Nara- world-development-indicators. S P OT L I G H T 4 Where criminal justice is not enough: Integrated urban crime and violence prevention in Brazil and South Africa Levels of urban crime and violence differ within countries, within regions, and even within states and provinces, as do the drivers of crime, which are influenced by historical, political, cultural, and socioeconomic factors. Despite these differences, successful strategies to reduce and prevent crime and violence have common elements, often referred to as an integrated approach to urban crime and violence prevention. Three cities in Brazil and one in South Africa are emblematic of this integrated approach to urban crime and violence prevention, which includes communities and moves beyond mere policing. Understanding risk factors, combining prevention and Diadema, Belo Horizonte, and Rio de Janeiro in Brazil and in criminal justice reform, and pursuing inclusive approaches Johannesburg in South Africa highlight the benefits of that engage actors at all levels of government, in civil soci- complementing traditional criminal justice approaches ety, and in the private sector are crucial in fighting and pre- with prevention strategies. venting urban crime and violence. A transition away from a criminal justice–first approach is frequently the first step. Making cities safer in Brazil Preventive strategies—particularly those targeting young Brazil is a heterogeneous country, which is reflected in the people through education, job training, psychological sup- variation in crime rates across its different regions. While port, and early childhood development—usually follow. the homicide rate has increased significantly in areas such Because the benefits of prevention materialize in the lon- as the northeastern portion of the country, over the past 25 ger run, strategies that aim to reduce crime and violence years some regions, particularly those in the south, have more quickly are also important. Among these strategies experienced large declines in their homicide rates. In São are inclusive citizen security, public health programs such Paulo, for instance, the homicide rate fell 67 percent be- as drug and alcohol rehabilitation, and the construction of tween 2000 and 2010. youth-friendly spaces. Crime reduction approaches that have shown good Although crime rates in some cities remain very high results in the southeast include educational programs, (figure S4.1), the improvements in crime and violence in programs for youth, and gun and alcohol control, all of which involved inclusive citizen security components at the municipal level. The strategies of three different Brazil- FIGURE S4.1 Homicide rate in selected cities ian municipalities are particularly noteworthy. Diadema, a city in the state of São Paulo where violence Johannesburg 43.0 increased 49 percent between 1995 and 1998, enacted a São Paolo 10.8 Mexico City new public security policy in 2000. Important components 8.4 New York City 5.6 of the policy included a limit on alcohol sales after 11 p.m. Moscow 4.6 and monitoring of the number and legality of alcohol li- Nairobi 4.0 censes in the city. Diadema also improved public lighting Bangkok 4.0 Seoul and installed security cameras in areas with high crime 2.4 Amman 1.8 rates. The public security policy contributed to a decline in Sofia 1.8 the homicide rate from 389 per 100,000 inhabitants in 1999 Mumbai 1.3 to 167 in 2003. The alcohol policy seemed particularly ef- Colombo 1.2 fective: the homicide rate fell 44 percent and the rate of Rome 1.1 Algiers 0.5 assaults against women fell 56 percent compared with levels expected in the absence of the alcohol policy. 0 10 20 30 40 50 Homicide rates also increased significantly in the late Per 100,000 population 1990s in Belo Horizonte, Brazil’s third-largest city. Crime in the city often occurred in slums and was frequently perpe- Source: WDR 2014 team based on data from the United Nations trated by young men. Following a public outcry, city offi- Office on Drugs and Crime Homicide Statistics database. cials in 2002 piloted the Fica Vivo (Stay Alive) program in Note: All rates are for 2009, except for Algiers (2008), Amman (2006), Johannesburg (2007–08), Nairobi (2008), and Rome (2008). the city’s most violent slums. The program involved the city council; the municipal, federal, and military police; the Spotlight 4 165 public prosecutor’s office; private businesses; nongovernmental or- Forums to put this proposal into action. Consisting of schools, busi- ganizations; and local communities. Preventive actions, including nesses, and civil and religious institutions, the forums facilitate part- support for education, job training, and youth sports and arts pro- nerships and joint problem identification and problem solving be- grams, aimed to reduce violence through the combination of crime tween the police and the community. The partnerships are designed control and social development programs. Information sessions to lead to the development of multistakeholder community safety about violence, drugs, and sexually transmitted diseases were also of- plans. fered. Thirty months after the implementation of Fica Vivo, homicides Johannesburg adopted a strategy in line with the rethinking of had fallen by 47 percent and attempted homicides by 65 percent. crime prevention taking place in the rest of the country. Indeed, the The homicide rate jumped in the state of Rio de Janeiro, as well, homicide rate in Johannesburg was even higher than in South Africa more than doubling between 1980 and 1997 from 26 to 59 homicides as a whole in 2007–08, reaching almost 43 homicides per 100,000 in- per 100,000 inhabitants. Crime began to drop in the 2000s as a result habitants. The Joburg City Safety Strategy, part of Johannesburg’s of 2003 national gun control legislation and a small-arms buyback development plan, is designed to cut the city’s high crime rate. The campaign. By 2008, homicide rates in the state and in the City of Rio strategy prioritizes actions that are critical to economic development had fallen to less than 34 per 100,000 habitants. and foresees a gradual broadening of the boundaries of crime reduc- However, drug trafficking persisted in Rio’s favelas, the informal tion and prevention interventions. Key programs include targeted settlements of Brazilian cities. In 2008, an initiative combining crimi- surveillance, patrols, closed circuit television, and other tools to deter nal justice and prevention was launched in response. The initiative criminals and increase people’s perception of safety. relied on elite police units to clear favelas of drug traffickers. Once Reforms continue along these lines. In 2012, South Africa under- they were gone, UPPs (Police Pacification Units) entered the favelas took a review of the criminal justice system in an effort to increase and focused on building trust between the community and the conviction rates, speed trials, improve the rehabilitation capacity of police, while also providing a continued security presence. The UPPs the prison system, and promote released prisoners’ reentry into soci- helped accelerate the decline in the homicide and violent crime rates. ety. The country is also increasing the budget for and expanding loca- A second phase, UPP Social (UPPS), links residents of those favelas tions of the police, courts, and prisons. Civil society organizations, that have been cleared of drug traffickers with social assistance. such as the Open Society Foundation and the Center for Justice and Dialogue is encouraged among residents, service providers, govern- Crime Prevention, have helped launch safety audits in poor commu- ment agencies, and the private sector to help ensure that community nities and have assisted with the design of crime prevention pro- needs are met. To ensure the sustainability of the approach, this grams tailored to the local context. The private sector is also heavily phase also involves efforts to formalize services such as electricity, involved. For example, Business against Crime, a coalition of South gas, cable, and Internet; rehabilitate youth formerly involved in crimi- African companies, collaborates with the government to develop nal activity; and revitalize the urban area. By 2012, more than 25 UPPs public-private partnerships to reduce crime. had been established. The government plans to serve 165 communi- ties through the establishment of 45 UPPs and UPPSs by 2014. Sources Greenwood, Peter W., Karyn Model, and C. Peter Rydell. 1998. “Diverting Children from a Life of Crime: Measuring Costs and Benefits.” RAND Corporation, Santa Combining prevention and criminal justice in South Africa Monica, CA. In the past, South Africa took a more reactive approach to crime, rely- Moser, Caroline. 2006. “Reducing Urban Violence in Developing Countries.” Policy ing heavily on the criminal justice system. This reactive approach led Brief 2006-01, Brookings Institution, Washington, DC. Petesch, Patti. 2013. “How Communities Manage Risks of Crime and Violence.” to one of the highest rates of incarceration in the world but did little Background paper for the World Development Report 2014. to stem crime. For instance, the country’s homicide rate was 39 per UNODC (United Nations Office on Drugs and Crime). Homicide Statistics. UNODC, 100,000 in 2007–08, higher than Brazil’s rate of 26. A shift to a multi- Vienna, http://www.unodc.org/unodc/en/data-and-analysis/homicide.html. sector preventive approach to crime reduction appears to be making World Bank. 2011. Violence in the City: Understanding and Supporting Community some progress, however. In 1996, shortly after the end of apartheid, Responses to Urban Violence. Washington, DC: World Bank. ———. 2012a. “Bringing the State Back into the Favelas of Rio de Janeiro.” World the South African government launched a National Crime Prevention Bank, Washington, DC. Strategy, which advocated a greater focus on prevention. In recent ———. 2012b. “Making Brazilians Safer: Analyzing the Dynamics of Violent Crime.” years, the South Africa Police Services have used Community Police World Bank, Washington, DC. Assembling computers in China. Jobs are drivers of development and pillars of resilience for people. © Sinopix CHAPTER 5 Fostering resilience and prosperity through a vibrant enterprise sector Creating jobs and supporting the potential mutual benefits of work- innovation ers and firm owners) and, gradually, formality (abiding by sensible laws The fear of losing or not find- prise s and regulations)—can greatly help er ec ing a job is a primary concern Ent tor enterprises support people’s risk for most people. About three- management. quarters of respondents in de- Two examples, continents veloping countries worry a “great apart, illustrate how flexibility and deal” about being unemployed, formality can improve the enter- according to the latest World Val- prise sector’s capacity to foster peo- ues Survey. Having a job is indeed ple’s resilience and promote prosperity. valuable: it produces income to support The first example illustrates the importance of consumption and to help meet important goals, flexibility. In Europe, Denmark, Germany, and Spain such as providing education, health care, and assets are among the many countries hit hard by the recent for family members. A job also contributes to self- global economic and financial crises: Denmark from esteem, a sense of personal security, and even social an overheated labor market, Germany from a shock cohesion.1 As discussed in previous chapters, these to global demand, and Spain from a popped real es- benefits are pillars of resilience and prosperity for tate bubble. By the beginning of 2013, more than 25 households and communities alike. percent of Spain’s labor force was unemployed. Den- Not all jobs are the same for purposes of risk mark’s unemployment rate also increased, but to a management, however. From the perspective of much lower 7.4 percent. In Germany, the rate was workers, jobs that provide secure and increasing in- just 5.3 percent.2 Many factors explain the different come and a safe working environment are preferable experiences of these three countries. For one thing, to jobs that do not carry those benefits. Moreover, the shock that hit Spain was larger than those affect- jobs that produce goods and services that consum- ing Denmark and Germany. But explanations of the ers want and can rely on and jobs that respect and widely different unemployment outcomes should conserve the environment are better from society’s also consider the flexibility of the labor markets in perspective. Where can these good jobs be found? those countries. Labor market reforms in Germany This chapter argues that a vibrant enterprise sector is helped moderate wage increases before the interna- best situated to provide such jobs through its poten- tional crisis and enabled employees and firms to ad- tial to mitigate the risk of unemployment, reallocate just hours worked through work sharing without the resources to create opportunities, and contribute to need for layoffs. In Denmark, job separations were worker, consumer, and environmental protection. high, but unemployment spells were short and eased The chapter further argues that two characteris- by a robust safety net and retraining programs for tics—flexibility (the capacity to adapt to changes to the unemployed. In Spain, in contrast, where a high 167 168 WORLD DEVELOPMENT REPORT 2014 percentage of the workforce was employed in the sector out of necessity rather than choice. In Ghana, hard-hit construction sector, stringent employment for example, more than 60 percent of informal sala- regulations with significant severance costs, along ried and self-employed workers without employ- with other structural factors such as the relatively ees would rather have formal wage jobs.8 Informal high share of unskilled young labor, have added to mechanisms may be effective for small firms and the high and persistent unemployment.3 simple transactions, but are insufficient for larger The second example shows the benefits for enter- firms and complex relations with workers and mar- prises of becoming formal. In Peru in recent years, kets. That may be why wage employment as a share informal mines have sprung up in response to ris- of total employment increases as a country devel- ing gold prices. Ignoring existing regulations, these ops. Notwithstanding significant variation across informal mines have caused significant deforesta- and within regions, self-employment—a large part tion. The mercury used in the extraction process has of which is likely of a subsistence nature—is wide- contaminated rivers and the atmosphere and threat- spread in the developing world (figure 5.1). ened human health.4 In the La Libertad region, the If the enterprise sector is to fulfill its role in sup- Poderosa Mining Company took an innovative ap- porting people’s risk management, public policy for proach to the problem after informal miners invaded the sector requires reforms that balance the econo- one of its mining concessions. The company began my’s need for flexibility with society’s need for legal to formalize the invading miners, signing agree- and regulatory protections. The chapter argues for a ments that allowed them to continue mining under combination of reforms to help the enterprise sec- its direction. The agreements, which meet interna- tor become flexible as it gradually becomes formal. tional environmental management quality stan- These reforms include improving the basic foun- dards, have increased the small miners’ income and dations of the enterprise sector by strengthening decreased the harm from deforestation and mercury property rights and reducing uncertainty about gov- contamination.5 ernment policy; implementing and enforcing sound These two stories convey the chapter’s main mes- regulations; and providing inclusive social protec- sage: flexibility and formality enhance the enterprise tion. In the longer run, when sound regulations for sector’s ability to contribute to people’s resilience flexibility are in place, the government can pursue and prosperity. An enterprise sector that is flexible reforms that encourage both flexibility and formal- is more capable of responding to shocks by reallo- ity by spurring innovation; increasing the skill level cating resources within and across enterprises and of of the labor force; and enhancing worker, consumer, innovating in an ever-changing world. An enterprise and environmental protections. These reforms are sector that is formal is better situated to take advan- necessary not only to increase growth but to enhance tage of legal protection and contract enforcement people’s resilience and to promote prosperity. Al- and to make better use of public infrastructure. In though the risks that enterprises themselves face and addition, formal enterprises can be more easily held their risk-taking decisions are very important aspects accountable for their impact on worker safety and on of a vibrant enterprise sector, this chapter maintains consumer and environmental well-being. the Report’s focus on risks faced by people. It thus There are both synergies and trade-offs between focuses on the enterprise sector as a sector and on its flexibility and formality. In countries with responsi- function of supporting people and society in manag- ble and strong state institutions and streamlined reg- ing risk in different ways. ulations, formality enhances flexibility. In countries with weak state institutions and cumbersome regu- The ways that the enterprise sector can latory regimes, however, the cost of formality can be help people confront risks too large for the majority of enterprises and workers. In this case, “informal is normal,” and informality is a Let’s step back for a moment and consider what the means for the economy to achieve a certain degree of enterprise sector is. The enterprise sector comprises flexibility and for workers to access a practical safety workers, owners, the arrangements that organize net.6 A diverse array of workers and enterprises then their relationships within an individual enterprise, remains informal because they are excluded from or and the technology that turns labor and capital into choose not to join a formal sector that offers limited goods and services. Enterprises, the defining unit of benefits.7 the enterprise sector, range from informal to for- Informality is often a second-best response, how- mal, from self-employment to partnerships to giant ever. The majority of the poor work in the informal multinational corporations, and from agriculture Fostering resilience and prosperity through a vibrant enterprise sector 169 F I G U R E 5 .1 Wage employment rises with the level of economic development a. Self-employment by region b. Transition to wage employment Percentage of total employment, average 2004–06 Percentage of total employment, early 1990s to late 2000s, selected countries 100 100 Maximum 90 75th percentile 90 Median 80 80 70 70 25th percentile 60 60 Percent Percent 50 50 40 40 30 30 20 Minimum 20 10 10 0 0 c sia M ibb rica Af and a CD ia nd sia ico il p. ey ka nd n ia o cifi ric az as ta As an Re lA dl an So ica an rk ay la la OE r e Af ex Pa kis Br aF rth st Ca m nz Tu Po ai h ra e iL r ab al M a n ut Pa A Th d in nt No e E Ta M ra Sr Ar an th atin rk Ce ha Bu t, ia Sa id yp d L As e an b- Eg st Su pe d Ea an ro Average 1991–95 Average 2006–10 Eu Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: Total employment consists of wage employment and self-employment. Organisation for Economic Co-operation and Development (OECD) countries in panel a are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. to manufacturing and services. Whereas a single on their own are naturally risk averse and thus re- enterprise might seek to maximize its profits, the luctant to take on new ventures, in groups and sub- enterprise sector as a whole is not confined to this ject to contractual arrangements, they become more objective. The sector encompasses the interests of willing to pursue projects involving more risk but workers, owners, and consumers, and—despite the also promising higher returns. Firms can thus serve possible important trade-offs between these inter- as natural vehicles to exploit opportunities, with ests—has the potential to help them manage risk beneficial consequences for individuals’ resilience through several channels discussed below. Flexibility and prosperity.10 and formality are important ways of ensuring that More specifically, the enterprise sector has the po- the enterprise sector can fulfill this function. tential to support people’s risk management through Enterprises include firms—enterprises in which three channels: sharing risk; allocating resources and two or more people work together. For workers and promoting innovation; and protecting workers, con- owners, being part of a firm widens the possibility of sumers, and the environment. This potential is not sharing the benefits and losses from specialization, always realized, however. In reality, each of the chan- collaboration, and innovation. Indeed, this is one nels can entail significant costs that are often borne by of the main motives behind the formation of firms the most vulnerable. Every day, newspapers are filled and, as such, has featured prominently in economic with stories about enterprises—especially those with thinking at least since Adam Smith. Frank Knight short-term horizons—that behave in ways harmful and Ronald Coase, in their seminal studies, dem- to workers, consumers, the environment, and even onstrated the institutional advantage firms have in the enterprise sector itself. Government has a role to providing cost-efficient ways of dealing with uncer- play in helping the enterprise sector achieve its po- tainty and overcoming the transaction costs inher- tential to support people’s risk management. In prac- ent in direct exchanges.9 Whereas most individuals tice, however, governments that lack the capacity, the 170 WORLD DEVELOPMENT REPORT 2014 appropriate incentives, and the correct strategy for The enterprise sector can also create opportu- public policy may undermine the flexibility and for- nities for owners of firms—and owners of capital mality of the enterprise sector that can make these more generally—to share investment risk. When the risk management channels effective. owners of capital invest in firms that are exposed to different types and levels of risk, they are able to re- duce the impact of a given negative shock on their Risk sharing investments through diversification, while still reap- The first channel through which the enterprise sec- ing the rewards of normal returns. An important tor can support people’s risk management is shar- legal risk-sharing mechanism for addressing capital ing risk among workers, among owners of firms risk is limited liability, which limits the losses for and owners of capital more generally, and between which a firm’s owners are responsible and thus helps workers and owners. Firms serve as a kind of risk overcome individuals’ natural risk aversion. Limited pool that allows workers to share the workload, fill- liability can encourage firms to take more creative ing in for an absent coworker or helping out with risks, which in turn can increase productivity and fa- unfamiliar or complex assignments. When a worker cilitate the enterprise sector’s provision of steady or falls ill, for example, another worker can assume his increasing income and employment. The develop- or her tasks temporarily, reducing the risk of unem- ment of stock markets and the exploitation of econ- ployment for the worker and helping to guarantee omies of scale have arisen in large part as a result of his or her income during the illness. More gener- limited liability. Risk sharing among firm owners ally, the complementarities between skills can occur even without formal mechanisms. of workers allows firms to respond to In places where the business climate is both positive and negative shocks For risk sharing, unfriendly, informal networks and more easily than individuals could achieving a certain size business groups often spring up to on their own, thereby stabilizing of firm is an advantage. facilitate cooperation among firm and enhancing aggregate pro- owners in the face of changing While self-employment— ductivity and income growth. commercial, economic, and po- For this type of risk sharing, size so common in many litical circumstances. is often an advantage. Smaller developing countries— Labor arrangements also per- firms tend to be more volatile. is a safety net, it is also mit risk sharing between work- A recent study found that in the a symptom of ers and firm owners. Workers can United States, which has a highly vulnerability. offer a form of insurance to firms, developed and diversified enterprise in which they agree to reductions in sector, aggregate sales growth of the me- wages or cutbacks in hours and benefits dian small firm is five times more volatile than during temporary shocks in exchange for higher that of the big publicly traded ones.11 wages in normal times. This type of risk sharing has The enterprise sectors of many developing coun- been found, for example, in manufacturing firms in tries, however, are dominated by self-employment Cameroon, Ghana, Kenya, and Zimbabwe.12 Alter- (see figure 5.1a); as a result, risk sharing among natively, firm owners willing to take chances on new workers is limited. Rates of self-employment are ventures and to accept greater profit variability can around 70 percent in South Asia and exceed 80 still offer steady wages to their workers, who are gen- percent in Sub-Saharan Africa, on average. This in- erally more concerned about the stability and pre- creases the vulnerability of the majority of workers dictability of their income. A similar dynamic applies in these regions to income shocks; a sick child, an when the source of variability is specific to the work- equipment failure, or a change in the weather could ers. For example, firms provide insurance to workers, mean the loss of a day’s income. Such high rates of in the form of a steady wage, for idiosyncratic risks self-employment also suggest that the enterprise such as illness.13 Labor contracts can facilitate these sectors in these regions are not benefiting from the arrangements by clearly defining responsibilities and specialization and increased productivity that multi- enhancing enforceability. person firms make possible. As figure 5.1b indicates, In sum, firms can serve as vehicles for risk sharing, a transition to wage employment is taking place providing insurance for shocks to individual workers in several developing countries, including Turkey, and for shocks in the production process, allowing where the share of wage employment increased 50 workers to specialize, and enabling owners of capital percent between the early 1990s and the late 2000s. to make more profitable investments. Fostering resilience and prosperity through a vibrant enterprise sector 171 Resource reallocation and innovation the capacity of the enterprise sector to respond to fu- ture shocks. Insufficient experimentation can sap the Resource reallocation and innovation is the second sector’s ability to absorb such shocks. Innovation— channel through which the enterprise sector can investing in physical and knowledge-based assets—is support risk management. Resource reallocation oc- at the heart of the ability of the enterprise sector to curs when enterprises shift resources, expand and exploit the opportunities offered by new discoveries, contract, and enter and exit markets. Reallocation particularly those opportunities that involve substan- can take place across enterprises and industries, as tial risks.17 Firms, which as entities can take on more less productive enterprises collapse and other more risks than their owners and workers individually, are productive ones are launched. This is the process of in a better position to promote the implementation creative destruction described vividly by Austrian of daring ideas that can lead to potentially produc- economist Joseph Schumpeter. The enterprise sector tive ventures. Some ventures will fail—and should be is often faced with large external shocks that can be allowed to do so. The success of the rest can provide positive or negative, and can include large and sud- both innovative processes for enterprises to absorb den changes in supply and demand, increases and shocks better and innovative products for people to decreases in input prices, advances in or obsoles- manage risk better. In the presence of risk sharing, cence of technology, and natural disasters. The sec- risk taking and tolerance for failure on a broad scale tor can absorb these shocks, and reduce the damage can be the engine of growth and poverty alleviation they cause or take advantage of opportunities, by for an entire country. reallocating resources efficiently within and across enterprises and industries to more productive areas, while still providing the income, employment, and Worker, consumer, and environmental products that facilitate people’s risk management. protection Without this process of resource reallocation, the en- The enterprise sector has the potential to support risk terprise sector would stagnate, become more volatile, management by facilitating the development and im- and experience more prolonged recessions, with dire plementation of employment standards and produc- economic and social implications. tion processes that protect workers, consumers, and Making resource allocation in China and India as the environment. Sound regulation and strong en- efficient as in the United States—provided that other forcement are crucial to bolster such protections. Firms factors such as institutional capacity were in place— that voluntarily adopt quality and environmental stan- could potentially increase total factor productivity by dards, such as those codified by the International Or- as much as 50 percent in China and 60 percent in ganization for Standardization (ISO), internalize costs India, recent research suggests.14 In the United States, of protection to their advantage, by improving their one of the world’s most dynamic economies, creative reputation or enhancing their productivity; at the same destruction accounts for more than 50 percent of time, they can also contribute to worker, consumer, productivity growth.15 Higher productivity can also and environmental protection. translate into greater job security. For instance, in Ro- Worker health and workplace safety are key com- mania, when the financial crisis hit in 2008–09, more ponents of productivity. In the best scenario, en- productive firms were less inclined to fire workers.16 terprises that recognize this connection will strive So while job turnover can carry high costs of adjust- to improve their work environment as part of their ment (which need to be addressed through inclusive strategy to maximize profits. For example, a recent social protection and other measures, as discussed meta-analysis found that workplace wellness pro- later in this chapter), rigid rules prohibiting layoffs grams reduce medical costs by $3.27 for every dol- can also carry high costs. lar spent on the program, and absenteeism costs by The enterprise sector can also support risk man- $2.73; these are benefits that accrue to both workers agement by allocating resources to the most pro- and enterprises.18 Because they can affect the bot- ductive enterprises and innovating to adjust to the tom line, reputational concerns have the potential world’s ever-changing conditions. Innovation in- to shape how enterprises treat workplace health and cludes adoption and adaptation of new technologies safety standards. and processes, especially by firms in the developing Such considerations with respect to productiv- world. Resource reallocation and innovation are the ity and reputation can encourage enterprises to de- essence of a process of experimentation and learn- velop quality standards for consumer products (box ing that can enhance productivity while improving 5.1). Standards designed to minimize defects and 172 WORLD DEVELOPMENT REPORT 2014 B O X 5 .1 Worker, consumer, and environmental protection can also be profitable business The interests of firms and those of workers, consumers, and the envi- credit.” Nestlé has embarked on similar campaigns in Brazil, Thai- ronment are not always aligned. But when they are, significant ben- land, and elsewhere. efits can arise. While consumer and environmental protection often require Nestlé’s entrance into the milk business in India is an example of public action, there are many examples of businesses launching ini- how the global supply chain can improve local conditions for work- tiatives to improve their production to benefit consumers and the ers. To establish a well-functioning milk business in the Moga dis- environment. McDonald’s Corp., for instance, has redesigned its trict of India, Nestlé had to work closely with local farmers, provid- packaging to allow consumers to access nutritional information ing financing for wells, veterinarians for sick animals, and instruction using smartphones. Firms manufacturing bottled water, such as in milk quality. This collaboration benefited Nestlé’s milk business, PepsiCo Inc. and Coca-Cola Co., have introduced bottle designs that which established a reliable supply chain. But the farmers of Moga use less plastic. Unilever, a consumer goods company, launched benefited, as well. According to Michael Porter and Mark Kramer in “Project Medusa” to increase the efficiency of the company’s water their analysis in the Harvard Business Review, “Nestlé has been able usage by reusing water evaporated during cooking for equipment to pay higher prices to farmers than those set by the government, cleaning and for cooling. and its steady biweekly payments have enabled farmers to obtain Source: WDR 2014 team based on Porter and Kramer 2006. increase product safety can raise consumer satisfac- compromised. The government has a crucial role to tion and lead to higher profits. In recent years, many play in designing and enforcing sensible regulations. firms have begun to adopt a “quality philosophy” that extends from production to identification of Flexibility and formality in the enterprise consumer needs, to product design and assurance, sector improve people’s resilience and to customer service.19 Reputational effects can also discipline firms’ treatment of the environment. At prosperity their best, social norms can discourage environmen- The shift to greater flexibility and formality can take tal mistreatment, civil society groups can watch for time, and it requires complementary reforms to environmentally harmful production practices, and strengthen institutional capacity and improve regu- consumers can become informed about the environ- lations. When state institutions are weak and regula- mental impact of the products they purchase. tions are cumbersome, there are trade-offs between At the same time, there are also many cases in flexibility and formality. When state institutions are which the enterprise sector undermines these pro- strong and regulations are sound, flexibility and for- tections or creates new risks for people, especially mality can be symbiotic. where institutions and regulations are weak. In some places, rapid economic growth has outpaced the The role and importance of flexibility development of workplace standards. Workers ex- perience both hazardous workplaces and degraded Flexibility is the capacity to adapt to changing cir- environmental conditions created by those work- cumstances. The flexibility of the enterprise sector, places. For example, residents in the Madre de Dios and of individual enterprises, goes far beyond the department in Peru, a center of the informal min- narrow ability to hire and fire workers easily. Rather, ing highlighted in the beginning of the chapter, have it extends to the capacity to reallocate resources ef- significantly higher levels of mercury than other ficiently to more productive areas to respond to Peruvians, which is likely related both to the use of short-term shocks without seriously disrupting mercury in the mining process and to the consump- operations, and to increase capacity to respond tion of fish from contaminated waters.20 Child labor to longer-term trends. Flexibility includes adjust- also remains a significant problem in many coun- ments to risk-sharing arrangements among work- tries. More than half of all child laborers—some 115 ers, among firm owners and owners of capital more million children—work in hazardous conditions, the generally, and between workers and firm owners. A International Labour Organization estimates. Even flexible enterprise sector is one in which workers’ in more developed settings, there are cases where skills are transferable, capital owners have abun- the safety and health of workers and consumers are dant options for investment, and workers and firms Fostering resilience and prosperity through a vibrant enterprise sector 173 FI G U R E 5. 2 The flexibility of the enterprise sector varies around the world a. Product market flexibilitya b. Labor market flexibiltyb 6.0 6.0 5.5 5.5 5.0 5.0 4.5 4.5 Index Index 4.0 4.0 3.5 3.5 3.0 3.0 2.5 2.5 c sia id bea ica Af and a c sia dl ean a Af and a CD ia CD ia cifi ric cifi M ibb eric ric As As lA n So a lA rib er So a OE Af OE Af ric Pa ric Pa rth ast rth ast Ca Am Ca m h ra h ra n n ut ut A d d nt No le E nt No e E ra ra an an th atin th atin Ce Ce ha ha r d ia ia Sa Sa id d d L L As As e e an an M b- b- st st Su Su pe pe d d Ea Ea an an ro ro Eu Eu Source: WDR 2014 team based on data from World Economic Forum (WEF) 2012. Note: The box plots show the minimum value, the range from the 25th percentile to the median (shaded dark), the range from the median to the 75th percentile (shaded light), and the maximum value for each region. The degree of flexibility rises with the numeri- cal value, with 1 being the least and 7 being the most flexible. Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. a. Product market flexibility corresponds to the WEF’s measure of goods market efficiency and consists of intensity of local competi- tion, extent of market dominance, effectiveness of antimonopoly policy, extent and effect of taxation, total tax rate, number of procedures and time required to start a business, agricultural policy costs, prevalence of trade barriers, trade tariffs, prevalence of foreign ownership, business impact of rules on foreign direct investment, burden of customs procedures, imports as a percentage of GDP, degree of customer orientation, and buyer sophistication. b. Labor market flexibility corresponds to the WEF’s measure of labor market efficiency and consists of cooperation in labor-employer relations, flexibility of wage determination, hiring and firing practices, redundancy costs, pay and productivity, reliance on profes- sional management, brain drain, and female participation in the labor force. are empowered to modify wages and employment ers are rigid, even small, idiosyncratic shocks can be levels potentially in their mutual interest. In a flex- very costly and threaten a firm’s survival. Finally, the ible enterprise sector, enterprises can expand and broader the opportunities in which firm owners can contract and enter and exit markets smoothly, and invest, the more diversified the risk they will bear. A can innovate to capture new opportunities. Figure shallow capital market or one with high transaction 5.2 shows two different measures of flexibility—one costs can cause investors to forgo good investment that corresponds to the efficiency of the market for opportunities. goods (panel a) and the other to the efficiency of the An enterprise sector that generates income op- market for labor (panel b). Both show that flexibility portunities even during times of economic difficulty tends to be lower in the developing world. is particularly vital in countries with weak social pro- Why is flexibility important? Flexibility enhances tection systems. In these countries, people without a each of the three channels through which the en- job cannot rely on public help and must find some terprise sector supports people’s risk management. source of income. Where the costs of formality are Flexibility facilitates the risk sharing that allows high, this income very often comes from informal workers and firm owners to respond to shocks, while work—recent research shows that in the majority providing opportunities for steady income and em- of countries, informality functions as a safety net ployment at the aggregate level. Complementary for those who are not part of the formal economy.21 skills and specialization can help firms, especially The informal sector often provides the flexibility that larger ones, take advantage of increases in demand people need for survival. or the introduction of a new technology. In contrast, A flexible enterprise sector supports the real- when labor arrangements between workers and own- location of resources within enterprises and across 174 WORLD DEVELOPMENT REPORT 2014 enterprises and industries. Flexibility means that in eling the recovery of an undistorted (flexible) econ- response to shocks, the sector can efficiently redeploy omy and the recovery of an economy in which the labor and capital to more productive enterprises and government intervenes with subsidy after a shock. more productive industries, potentially in the mutual The results are striking. The flexible economy suf- interests of workers and firm owners. As the World fered a loss of 13 percent of preshock gross domes- Development Report 2013: Jobs indicated, jobs are tic product (GDP)—84 percent of which occurred drivers of development, and all types of employment, in a single quarter—whereas the economy in which including informal jobs, can be transformational in the government intervenes suffered a loss of 36 per- improving living standards, productivity, and social cent of preshock GDP spread out over five years. A cohesion.22 The ease of dismantling collapsed busi- smaller, faster loss implies a quicker recovery. nesses and of creating new ones is crucial to ensur- An examination of recent great depressions cor- ing that new employment opportunities are created. roborates this evidence (box 5.2), as does cross- Adjustment costs during this reallocation process can country research: impediments to resource reallo- be significant, however, particularly for newly unem- cation to more productive firms reduce the pace of ployed workers. Inclusive social protection systems economic recovery. For example, by propping up need to be in place to protect the vulnerable. banks and requiring them to maintain nonperform- Research on recovery from negative shocks dem- ing loans to firms that otherwise would have failed in onstrates the harmful effects that rigidity can have. the late 1990s and early 2000s, Japan prevented their Using empirical evidence from 76 countries, a recent resources from flowing to more productive uses.23 study found that economies with policy-induced ri- Better regulation, in contrast, can promote the real- gidities such as excessive labor protections, barriers location of resources to more productive firms.24 to firm entry, burdensome bankruptcy laws, and in- Factors other than governmental regulation also dustry supports suffer deeper and more prolonged can impede flexibility. One example is gender segre- recessions (figure 5.3) than more flexible economies gation in employment, which often hampers labor do. The study substantiated this evidence by mod- market flexibility; this issue is discussed extensively in FI G U R E 5. 3 Countries with a higher regulatory burden experience more severe recessions 0.30 Correlation coefficient = 0.36 0.25 Severity of recessions 0.20 0.15 0.10 0.05 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Index of regulatory burden Industrialized Developing Source: WDR 2014 team based on data from Bergoeing, Loayza, and Repetto 2004. Note: Index of regulatory burden varies from 0 to 1, where 1 indicates the heaviest burden. Data on regulatory burden cover financial restrictions, trade barriers, firm entry costs, inefficient bankruptcy procedures, bureaucratic red tape, tax burden, and labor regula- tions for the 1990s. The severity of recessions is measured by the sum of downward output deviations from trend for each country during 1990–2000. The solid (regression) line in the figure depicts the fitted linear relationship between the y- and x-axis variables, allowing for an intercept. Fostering resilience and prosperity through a vibrant enterprise sector 175 BOX 5.2 Lessons from great depressions How should public policy respond to unusually large declines in out- and Mexico both experienced depressions in the early 1980s, but put? While the benefits and limitations of expansionary monetary productivity growth recovered, and even exceeded trend, in Chile, and fiscal policy in such cases have been examined and debated while it remained 30 percent below trend in Mexico nearly 15 years extensively, enterprise policy may play an especially important role, later. Banking and bankruptcy procedures appear to have been cen- according to comprehensive analysis of the historical evidence by tral to the different outcomes. By the early 1980s, unproductive Timothy Kehoe and Edward Prescott in their exhaustive work, Great firms in Chile were allowed to go bankrupt, with new investments Depressions of the Twentieth Century. determined by market interest rates, whereas the state-controlled In contrast to the relatively shallow and brief recessions that are banking system in Mexico continued to channel loans to some typical of business cycles, “great depressions” are unusually deep, unproductive firms through the early 1990s. Second, studies of the painful, and long.a The Great Depression of 1929–39 in the United manufacturing, construction, and mining industries suggest com- States is perhaps the best-known example, but several European petition policy may be important, insofar as more competitive countries had similar experiences in the period between the first industries tend to be more creative and productive in the aftermath and second world wars, and some recent examples have occurred in of a large output decline.b Finally, tax increases and rigid wage poli- Latin America, New Zealand, and Switzerland. cies have been shown to reduce labor input during great depres- Such experiences have often been treated by economists as sions. For example, rigid real wage policies appear to have been a exceptions to which standard macroeconomic models may not key factor in prolonging the great depression of 1928–37 in Ger- apply. Instead, Kehoe and Prescott seek to use the modern tools of many. In contrast, flexible labor regulations helped Germany escape macroeconomics (specifically, growth accounting and the general the worst of the recent crisis (see box 5.5). equilibrium growth model) to scrutinize the underlying drivers of Recent microeconomic evidence finds that, besides the types depressions in more than a dozen countries. Their findings suggest and productivity of the technology used within individual firms, that the level of productivity growth is often the dominant driver of resource allocation across firms is a crucial determinant of cross- depression dynamics, while changes to labor input are also impor- country differences in productivity.c The emerging literature in this tant in some cases. field supports the view that facilitating resource reallocation can be What policies dampen productivity growth during depressions? an essential part of bringing depressions to an end. First, not letting inefficient firms fail is crucial. For example, Chile Source: WDR 2014 team based on Kehoe and Prescott 2007. a. Kehoe and Prescott 2007 define great depressions as declines in output per capita of at least 20 percent below trend, with at least 15 percent of the decline occurring within the first decade of the depression, and annual per capita output growth remaining below trend for each year of the depression. b. Baily and Solow 2001. c. See Restuccia and Rogerson 2013 for a useful review of the recent literature. World Development Report 2012: Gender Equality and against unskilled labor.27 That said, these negative ef- Development.25 In most countries, women are more fects can be counterbalanced when cost reductions as- likely than men to participate in low-productivity sociated with innovation spur price reductions, which activities, in the informal sector, and in family em- in turn stimulate growth in aggregate terms, leading ployment. Another factor involves exclusive business to greater demand and output over time. Product in- relationships. In Sub-Saharan Africa, for example, novation, in contrast to process innovation, tends to relationships formed through business activities can stimulate domestic and foreign demand and so en- exclude unconnected communities from more prof- hance an enterprise’s demand for labor. itable investment opportunities and perpetuate exist- Flexibility is also closely related to worker, con- ing patterns of production.26 sumer, and environmental protection. In Bulgaria, Innovation and flexibility are closely linked. The for example, a firm’s capacity to change—based on more easily resources flow to entrepreneurs and measures such as leadership, trust, innovation, and firms that offer new products, better technologies, accountability—was found to be strongly positively and more efficient processes, the more quickly the associated with environmental performance.28 With- enterprise sector can generate new opportunities at out flexibility, an enterprise will be unable to adjust lower costs. Nonetheless, innovation can have adverse its products or production techniques in response to consequences for employment: for example, process the displeasure of consumers and civil society or to improvements may allow an enterprise to produce take appropriate action in the case of a defective, en- the same output with less labor, leading to cutbacks vironmentally harmful, or dangerous product. or unemployment. Even innovation that results in Although corporate responses to complaints positive (or zero) job creation might have a “skill bias” about product and worker safety are often (rightly) 176 WORLD DEVELOPMENT REPORT 2014 criticized for being insufficient or ineffective, enter- quality and safety requirements, and environmental prises do seem to be able to adjust to and, at least guidelines—all of which can be costly to implement. at times, remedy such deficiencies. One famous case In exchange, formal enterprises are entitled to en- involved the Johnson & Johnson painkiller Tylenol, hanced access to legal protection and public infra- whose market share plummeted from 37 percent structure. Formality and informality, however, are a to 7 percent after several bottles were found to have matter of degree. Some enterprises are in compliance been contaminated with cyanide. The firm removed in some but not all dimensions of formality. For ex- all Tylenol from the market and adopted and adver- ample, some enterprises that comply with registra- tised a “triple safety seal.” In less than two years, the tion procedures may substantially underreport their company had nearly regained its previous market sales and profits or only partially fulfill requirements share.29 The incident led the U.S. Food and Drug to provide benefits and workplace protection to Administration to issue regulations regarding drug workers. Moreover, there is a wide gap between de packaging, and safety shields became the norm in jure formality and de facto formality, particularly in the industry. Many enterprises operating in develop- countries with weak institutional capacity. ing countries have also responded to concerns about Formality varies widely across countries (figure product and worker safety, although these measures 5.4). The large differences in labor and production have had only limited success.30 formality across countries and regions suggest the existence of a massive labor pool yet to be mobilized The role and importance of formality in formal—and likely more productive—activities. That is especially likely to be the case in countries Enterprises are considered formal if their activities with relatively high production formality but low la- comply with laws and regulations governing taxes, bor formality. Informality is not a cause, but rather a registration, labor and workplace standards, product symptom, of underdevelopment. In general, formal FI G U R E 5.4 The formality of the enterprise sector varies around the world a. Formal production b. Formal labor a Production by formal sector, percentage of GDP, 2007 Share of labor force contributing to pension schemes, 2000s b 100 100 90 90 80 80 70 70 60 60 Percent Percent 50 50 40 40 30 30 20 20 10 10 0 0 c sia id bea ica Af and a c sia dl ean a Af and a CD ia CD ia cifi cifi ric M ibb eric ric As As lA lA n So a rib er So a OE OE Af Af ric ric Pa Pa rth ast rth ast Ca Am Ca m h h ra ra n n ut ut A d d nt nt No le E No e E ra ra an an th atin th atin Ce Ce ha ha r d ia ia Sa Sa id d d L L As As e e an an M b- b- st st Su Su pe pe d d Ea Ea an an ro ro Eu Eu Source: WDR 2014 team based on data from Schneider, Buehn, and Montenegro 2010 (panel a); World Bank Pensions (database) and World Bank World Development Indicators (database) (panel b). Note: The box plots show the minimum value, the range from the 25th percentile to the median (shaded dark), the range from the median to the 75th percentile (shaded light), and the maximum value for each region. Organisation for Economic Co-operation and Development (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. a. Values computed as 100 minus estimates on the size of shadow economies by Schneider, Buehn, and Montenegro 2010. b. Figure based on the latest data available in the 2000s. Fostering resilience and prosperity through a vibrant enterprise sector 177 employment tends to increase with the level of de- ticularly important to liberate productive resources velopment. By helping to ensure that wages are paid from an unproductive enterprise and to ensure that and contracts honored, that property will not be creditors, and potential investors in other enterprises, seized, and that financial resources are available and are protected if a business fails. Laws that protect the affordable, formality helps both workers and enter- intellectual property of established enterprises and prises to plan for the future. Compliance with labor, entrepreneurs alike are often an important incentive consumer, and environmental protections—while for innovation. Finally, formality provides the legal costly—can also lay the basis for growth, as discussed documents that are frequently necessary to access throughout this chapter. credit and other financial resources. Such access sup- Why is formality important? Like flexibility, for- ports business expansion, promotes new enterprises, mality enhances each of the three channels through and can provide a lifeline to enterprises in the face of which the enterprise sector supports people’s risk shocks. Overall, cross-country analysis suggests that management. Formal firms tend to be bigger, more an increase of one standard deviation in informality stable, and better able to provide steady work for leads to a decline of 0.7 to 1.0 percentage point in the their employees. Because formal firms have a more rate of per capita GDP growth.33 Further, countries predictable environment in which to grow and hire with high informality cannot fully capture the ben- more employees, workers are better able to share efits of globalization (box 5.3). the risk. Capital owners can invest in formal firms Formality may strengthen an enterprise’s incen- to diversify their own risk and jump-start growth for tives to provide worker, consumer, and environmental the business, thanks to the legality of contractual ar- protection. Good regulation can induce enterprises rangements. Finally, formality means that contracts to internalize the social costs of their activities. That between workers and firms are more easily enforced is particularly clear in the case of environmental rules and so less likely to be broken. Even in the absence that force firms to absorb some or all of the costs of of contracts, formality can provide both workers and polluting. It is also true of workplace protections that employers with legal recourse if a labor arrangement prohibit firms from paying very low wages when the is perceived to have been broken, reducing the risk to supply of labor is abundant. Governmental regu- both parties of noncompliance. Small firms may be lation can thus create a level playing field for firms able to rely on informal institutions to enforce con- to compete by increasing worker productivity and tracts. Such informal mechanisms become less ef- consumer satisfaction through provision of a healthy fective, however, as firms grow and the relationships workplace and environmental stewardship, rather between owners and workers become more complex. than by paying poor wages, providing poor working The stronger risk-sharing arrangements pro- conditions, or cutting costs at the expense of the en- moted by formality improve people’s level and stabil- vironment. The legal identity provided by formality ity of income, one of the most significant factors in strengthens a firm’s incentive to cultivate a reputation improving risk management. For example, average consistent with a healthy workplace, product quality, monthly earnings in Zambia generally increase with and environmental protection. formality. In Tanzania, income volatility generally Employee access to benefits and safe workplaces decreases with formality. Only 9 percent of informal tends to improve with formality. In Mozambique, the employees in francophone West Africa had a written share of paid employees with a range of employment contract with their employers.31 benefits, such as remunerated sick leave and sever- Lack of formality can inhibit risk sharing across ance payments, generally increases with the formal- firms, as well. In Zimbabwe, for instance, the risk ity of employment, as does the share of workers in that suppliers will not comply with contracts has led Zambia who use protective clothing when working.34 manufacturing firms to increase inventories and li- At the same time, formality is not a sufficient quidity reserves to protect against late delivery, non- condition for worker, consumer, and environmen- payment, or late payment.32 This imbalanced type of tal protections. These protections can be weak or risk sharing is unlikely to be efficient in a world in absent even when firms are large multinationals or which just-in-time delivery is becoming the norm. linked in a global value chain. In fact, the growth of The laws and regulations that accompany formal- global value chains has a dual effect, putting pressure ity can expedite the process of reallocating resources. on enterprises to comply with more stringent labor Well-designed and consistently applied procedures and production standards while increasing pressure are essential for resources to flow efficiently out of to reduce labor costs, even by resorting to informal collapsed enterprises and toward new ones. Bank- work. This situation results in improved employ- ruptcy law and the depth of resale markets are par- ment for some workers, but informality and instabil- 178 WORLD DEVELOPMENT REPORT 2014 BOX 5 1..2 3 The complex While relationship poverty between has declined, globalization many people aroundand world remain vulnerable to poverty the formality Globalization In a significant achievement, brings new opportunities poverty in and risks. As many developing as 20 mil- countries has after While the global crisis have all regions reduced resulted in large thelayoff shares s in the ofthe population maquiladora lion people steadily are directly declined over the past two decades. employed in the manufacturing The share of people sector liv- for that in and related live in or are vulnerable to poverty, progress has varied across industries. ing below global value chains. $2.50 a day A large has proportion dropped from 72 percent of these are women andto in 1990 rural 50 Benefi regions. Ints from globalization Europe and Central Asia, accrue more where to the formal poverty economy was already rel- a percent by migrants in poor 2010.countries. Nonetheless, a substantial proportion of people atively low, because informal substantial enterprises progress has been often do not have made the appropriate in reducing vulner- remain vulnerable In Costa to poverty, from Rica, investment with 89 percent Intel of people has translated in develop- into acceler- ability to poverty. standards and qualifi East Asia In cations and from thethe suppliers’ Pacific, the rate of poverty standpoint to com- ing countries ated formation living on less than of technical and $10 a day English in 2010, skills compared for workers. with 94 Standards pete. was cut in half from Economies with high 1990 torates 2010, informality offrom 88 percent 40 percent, are thus but often posi- percent of practice froma the in 1990. While U.S. chronic market, especially poverty declined signifi has regarding workercantly, and tioned at the 92 percent the population of lower end of thecontinues global value chain, to live and their in poverty be vul- orworkers the large share of environmental people in developing protection, were actively countries transferred to very that live local the close are nerable moreto it. Similarly, vulnerable in the shocks to global South andAsialess and Sub-Saharan likely from to benefitAfrica to poverty with economy, Intel requiring highlights the potential its suppliers for substantial and subcontractors increases in tran-to opportunities. regions, 98 percent Recent the population of research suggests lived on that less that $10has globalization a day as con- sient poverty—which meet strict worker safety have long-run can qualifi cations andconsequences sending itsfor people’s employees of 2010. to the prosperity of many countries, although sometimes tributed to health promote and livelihoods—when recycling programs peopleand areenvironmental shocks. hit by negative awareness. at the cost of rising inequality.f Following Intel’s socioeconomic contributions, the National Insur- In the long run, economies that are more open tend to have a ance Institute, Costa Rica’s autonomous insurance institution, has lower incidence of informal employment. In several countries in created the nation’s first job safety and health standard.b Asia, an increase in a country’s openness has been associated with a In Mexico, the influx of export manufacturing jobs from 1985 reduction in the incidence of its informal employment. In Mexico, to 2000 increased female formal employment nearly fivefold. reductions in tariffs after joining the North American Free Trade These jobs offered higher wages than agriculture and other alter- Agreement significantly reduced the likelihood of informality in the natives. The higher earnings increased the bargaining power of tradable sectors.g In the short run, however, the impact of globaliza- these women within their households and led to observable tion on informality may be mixed, depending on the nature of the improvement in child health: the children of women who worked globalization process and the various economic fundamentals of in formal export manufacturing were more than one standard the economy. For example, an overly rapid trade liberalization can deviation taller than the children whose mothers did not have wipe out the protected formal sectors and push their workers into their first job in manufacturing, and this effect was stronger for self-employment or informal sectors. Given the high informality in girls.c In Indonesia, during the East Asian crisis in the late 1990s, many developing countries, particularly in agriculture, retail, and the aggregate adverse effect on employment was smaller for other services, the benefits of globalization may not reach many female workers because they were concentrated in larger firms segments of the population, while shocks from the global arena can and firms that exported, which were less hard hit.d On the other negatively impact their lives. Overall, globalization exposes enter- hand, the influx of relatively low-skill jobs also induced high school prises to competition in the world market, providing incentives for students to drop out of school; for every 20 new jobs created, one productive enterprises to expand and pressures for unproductive Source: student WDR 2014 left team based school on data in Mexico. e from PovcalNet (database). More recently, the trade collapse ones to exit. a. $1.25 a day is a widely used measure of extreme poverty. However, $2.50 a day is considered a more relevant measure of extreme poverty for some regions, as Latin such WDR Source: team. and the Caribbean. The $10-a-day measure is an approximate threshold for measuring vulnerability to poverty across regions, which cor- America 2014 relates with a. UNCTAD asset holdings. The measure is based on studies suggesting that, for some regions, income of at least $10 a day is necessary to achieve the degree 2013. of economic b. World stability and resilience to shocks that characterizes middle-class households. By contrast, those living below $10 a day are vulnerable to poverty, in Bank 2006. the sense c. Atkin that they face the possibility of remaining in poverty or easily entering into poverty. See, for example, López-Calva and Ortiz-Juarez 2011; Ferreira and 2009. others 2013. d. Hallward-Driemeier, Rijkers, and Waxman 2011. e. Atkin 2012. f. Kremer and Maskin 2006. g. Aleman-Castilla 2006. ity for others. A case study of garment factories in more than 250 workers died in a fire at a garment Morocco, for example, found that some high-skilled factory in Pakistan that had been certified as meeting workers involved with the final product and with international health and safety standards.37 Indus- oversight of packaging, storage, and logistics for buy- trial disasters such as the Bhopal gas leak in 1984 and, ers had stable contracts and protections. However, more recently, the Deepwater Horizon oil spill in the factories also hired irregular workers to meet buyers’ Gulf of Mexico and the Fukushima Daiichi nuclear demands. These workers had casual contracts and disaster in Japan continue to put people and the en- were often subjected to discrimination.35 In addition, vironment at risk. because global value chains are not always linked to local value chains, the safety and quality benefits The relationship between flexibility and from industry standards that are enjoyed by global formality consumers may bypass domestic consumers.36 Furthermore, formality is not a guarantee against Burdensome laws and regulations can raise the costs negligent or criminal acts. For instance, in 2012, of formality. Complicated and lengthy business ap- Fostering resilience and prosperity through a vibrant enterprise sector 179 plication processes, onerous taxes, stringent hiring global supply chain. The quality and performance re- and firing rules, poor-quality public services such quirements (including standards and certifications) as police and courts, and a lack of educated workers for global exports are too strict to be met by the in- can discourage informal enterprises from becoming formal economy on the supply side. When interna- formal. In these cases, there is a trade-off between tional competitive pressures intensify, the informal flexibility and formality: to maintain flexibility, en- economy can actually be pushed out of the global terprises remain informal. While remaining informal value chain. As economists Rafael La Porta and An- may be optimal for individual enterprises, the enter- drei Shleifer put it recently, “Informal firms keep mil- prise sector as a whole suffers, because the advan- lions of people alive but disappear as the economy tages of formality are sacrificed. The seminal work of develops.”40 Hernando De Soto quantified the costs of formality The transition to flexibility and formality is not and the extent to which excessive regulation can lead easy. There is no single recipe for success; the tran- to abuses and foster informal activities.38 As the cost sition depends on the specific country context, in- of formality declines, the trade-off between flexibil- cluding factor endowment, history, and culture. In ity and formality also declines. Malaysia, reforms in business and labor market reg- When laws and regulations are focused, well- ulations contributed to a more flexible economy.41 designed, and consistently applied, the costs of com- In Costa Rica, relatively high education levels and plying with them are low relative to their benefits. fiscal incentives attracted large inflows of foreign In this case, formality and flexibility reinforce each direct investment that, together with advanced other. As the economy moves from high-cost formal- technology and product standards, helped cre- ity to low-cost formality, trade-offs decline and flex- ate more formal jobs and reinforce skills building. ibility and formality become symbiotic. Formality With the right steps, pace, and sequence, countries helps enterprises adapt to changing circumstances, can become flexible and formal in the longer run with laws and regulations that make it easier to rene- (figure 5.5). gotiate labor arrangements, reallocate resources, and encourage innovation, and that level the competi- How can the government help enhance tive playing field. Flexibility, in turn, makes it easier the flexibility and formality of the for enterprises to comply with laws and regulations enterprise sector? while remaining productive. Flexibility is a reasonable short-term goal. For- In the transition to formality and flexibility, the qual- mality is a longer-term objective. In the short ity and enforcement of policy and regulation are es- term, helping enterprises adjust to the world’s ever- sential. The government can play a key role in lower- changing circumstances is critical, even without for- ing the costs and increasing the benefits of formality mality. In many low-income countries, especially in and flexibility. The government’s focus should be on Sub-Saharan Africa and South Asia, even if economic improving the attractiveness of formality; penalizing growth rates are high, the formal sector cannot gen- informality would damage the safety net that infor- erate enough wage employment in the near future to mal enterprises can provide when the state is weak absorb the majority of the labor force. The informal or formality is prohibitively costly. The government sector remains the main contributor to GDP and to should also focus on reforming complementary areas employment. Self-employment and household en- at the same time. If policy reforms are to have the terprises serve as important means of survival for desired effect, they must also be tailored to specific the most vulnerable. An empirical study covering country conditions, such as institutional capacity, eight countries in Africa, for example, found that political environment, and the stage of development household enterprises were responsible for the cre- of the labor market. ation of most new nonagricultural jobs.39 Increasing The government can help in four main ways. the flexibility of the enterprise sector to create more, It can lay the basic groundwork for enterprises to and more-productive, jobs—both formal and infor- operate by strengthening property rights and reduc- mal—is crucial in helping people manage risk and ing uncertainty about government policy. It can im- reduce poverty. plement and enforce sound regulation. It can put in In the longer run, however, formality strengthens place inclusive social protection to protect workers, the contribution of the enterprise sector to making particularly the vulnerable. Finally, over time, it can people resilient to shocks. Increasingly, formality strengthen policies that improve both flexibility and is becoming a necessary condition for joining the formality in the longer run. 180 WORLD DEVELOPMENT REPORT 2014 FI G U R E 5.5 Typology of countries by the flexibility and formality of their product and labor markets More Australia Germany Netherlands flexible Austria Hong Kong Norway Belgium SAR, China Poland Canada Ireland Singapore Armenia Albania Chile Israel Sweden Azerbaijan China Costa Rica Japan Switzerland Cambodia Czech Republic Korea, Rep. United Kingdom Kazakhstan Denmark Latvia United States Peru Estonia Lithuania Finland Malaysia Product and labor market flexibility France Mauritius Benin The Gambia Morocco Botswana Georgia Sri Lanka Brazil Kyrgyz Namibia Republic Hungary Portugal Slovenia Burkina Faso Guatemala Tanzania Bulgaria South Africa Macedonia, Italy Slovak Spain Cameroon Guinea Thailand Indonesia Ukraine Republic FYR Jordan Turkey Côte d'Ivoire Lebanon Uganda Kenya Uruguay Mongolia El Salvador Madagascar Zambia Vietnam Bangladesh Ghana Pakistan Bolivia Haiti Philippines Algeria India Russian Argentina Dominican Lesotho Federation Croatia Bosnia and Honduras Senegal Herzegovina Republic Libya Trinidad and Greece Jamaica Sierra Leone Tobago Burundi Ecuador Mexico Iran, Islamic Rep. Mali Venezuela, RB Yemen, Rep. Chad Egypt, Arab Rep. Romania Nepal Zimbabwe Colombia Nicaragua Less flexible Less formal More formal Production and labor formality Source: WDR 2014 team based on data from World Bank Pensions (database); World Bank World Development Indicators (database); World Economic Forum 2012; and Schneider, Buehn, and Montenegro 2010. Note: Economies in the top row are high (above the median value) in both product market flexibility and labor market flexibility (figure 5.2); in the middle row they are high in one or the other of the two; and in the bottom row they are low (below the median value) in both flexibility indicators. Similarly, economies in the first column on the left are low in both formal production and formal labor (figure 5.4); in the middle column they are high in one of the two formality indicators; and in the last column on the right they are high in both formality indicators. Only economies with data for all four indicators are considered, and median values are calculated within this sample. Several of the policies discussed in this section Lay the groundwork for a vibrant are frequently associated with productivity and enterprise sector growth policy. This is no accident. A better invest- ment climate can improve risk management in the Some policies, such as securing property rights and enterprise sector: improving respect for rules and reducing uncertainty about government policy, are regulations and increasing flexibility can help the foundational in the sense that they lay the basic sector respond to negative shocks and take advan- groundwork for the enterprise sector to operate. In tage of opportunities. The more capable the enter- fragile and conflict-afflicted states, strengthening prise sector is of managing risk, the more produc- national institutions and improving governance to tive it can become. A government that supports the provide citizen security, justice, and jobs are crucial rule of law and transparent contracts and that does first steps to break cycles of violence.42 not act capriciously toward local and international investors can be the cornerstone for a vibrant enter- Secure property rights. Secure property rights are prise sector (see box 5.4). essential to assure investors that their risk taking Fostering resilience and prosperity through a vibrant enterprise sector 181 BOX 5.4 The resilience of the power sector during a period of civil unrest in Côte d’Ivoire was driven by protections for contracts Côte d’Ivoire is one of the countries in Sub-Saharan Africa that has a fixed price or schedule of prices. Two major civil conflicts in the pioneered private sector participation in the power sector. Since past decade significantly affected revenues and collection. Politi- 1990, a private operator (Compagnie Ivoirienne de l’Electricité, or cal fragility at the national level also made tariff adjustments next CIE) has managed the country’s power sector. As of 2006, indepen- to impossible. From late 2010, at the peak of the most recent crisis, dent power producers (IPPs) accounted for nearly two-thirds of arrears to the IPPs built up, but rather than exercise penalty clauses energy production. Severe droughts, a significant currency devalua- under the contracts or shut down production, the two IPPs instead tion immediately before the IPPs were set up, civil war, and the sus- exercised forbearance, recognizing both the underlying robust- pension of a large part of revenue from power sales for an extended ness of the contracts and the central role played by their com- period of time have all affected the sector’s performance. Rather panies in the economy. The enterprises understood that the than buckle under the combined pressures of these events, the sec- payment arrears had arisen from a temporary shortage of funds, tor has largely thrived, in part because of a combination of flexibility rather than any intention on the part of the government to default. and formality among sector participants. Indeed, the government signaled its good faith by paying all the The two main IPPs, Compagnie Ivoirienne de Production de parties in fixed proportions, according to its ability to pay. Ulti- l’Electricité (CIPREL) and Azito Energie SA (Azito), continued to mately, the IPPs understood that the sanctity of the contracts deliver electricity throughout the period of civil unrest; at times, would be honored. With the consolidation in political power employees guarded the plants around the clock—an act of resil- under a new government, this belief has been validated and the ience of which both owners and staff remain justifiably proud. Per- arrears have now been repaid. haps even more impressive has been the companies’ decision to Since the recent stabilization, both CIPREL and Azito have continue to supply power even when the government could not announced plans to expand their operations, which will increase meet the terms of its contracts. The contracts require the compa- investment and the energy supply and support economic growth nies to provide a certain amount of power each year, according to and political stability in the postconflict state. Source: Conor Healy for the WDR 2014. will yield rewards. Government nationalization and ticularly those in infrastructure and other decisions expropriation of private sector assets undermine in- that involve large sunk costs. Uncertainty increases centives to invest in a country. Countries with sound the value to investors of waiting and can lead to sig- investor protections tend to grow faster than those nificant declines in hiring, investment, and output.45 with poor investor protections; countries with weak Improving the predictability in the way the rule of enforcement of contracts are characterized by greater law is applied and the credibility of reforms is crucial macroeconomic volatility; and countries with reli- for private sector development. In Côte d’Ivoire, for able law and order have smaller informal enterprise example, the power sector thrived, even in a fragile sectors.43 environment, thanks in part to the private operator’s trust in contract durability (box 5.4). Reduce uncertainty about government policy. A stable policy environment and certainty about regulation Implement sound regulations is a key determinant of domestic as well as foreign investment (see chapter 7). In Thailand in the mid- Reform of burdensome and costly regulations that 2000s, for instance, political instability was perceived reduce enterprise sector flexibility and discourage the as a major constraint on doing business, contribut- operation of formal enterprises should be a priority. ing to a decline in business sentiment and in private Flexibility does not mean the absence of regulations; investment growth, despite improvement in several rather, sound regulations are needed to improve the other aspects of the business environment.44 When business environment and mitigate risks for enter- investors are uncertain about the future, they may prises and their owners and workers. demand higher rates of return to compensate them for the extra risk involved. They may also shorten Improve the regulations that affect the enterprise sec- their planning horizons, thus influencing their level tor. Weak or obstructive regulations can needlessly and form of investment, choice of technology, and hinder the development of the enterprise sector and willingness to train workers. Policy uncertainty can impede competition. Worldwide, some 23 percent of have dire consequences for private investments, par- registered firms consider tax administration to be a 182 WORLD DEVELOPMENT REPORT 2014 major or very severe obstacle, for example.46 More lated private sector employment by 5 percent. Those broadly, research suggests that countries that move who lost their jobs generally did not find new ones from the worst quartile of business regulations to the and left the formal labor force.54 Rigid labor market best could enjoy a 2.3 percentage point increase in regulations that reduce incentives for firms to offer annual growth.47 Several important aspects of regu- more stable jobs with enforceable contracts might lations identified in the Doing Business reports, such disproportionately hurt the most vulnerable, such as as starting a business, paying taxes, and resolving in- women and young workers. In Indonesia, for exam- solvencies, are generally more cumbersome in low- ple, the minimum wage is more binding for small income countries than in middle- and high-income firms and results in disproportionate job loss for fe- countries.48 male workers.55 Streamline business registration. An enterprise that Improve measures for resolving insolvency. Long and demonstrates compliance with the appropriate rules expensive bankruptcy procedures clutter the market and regulations should not have to wait for years to with failed firms that block opportunities for new receive a permit to operate. Many countries have enterprises. Enhanced predictability and improved successfully increased the efficiency of registration bankruptcy procedures can help facilitate respon- by establishing online procedures and linking the sible risk taking and reduce associated costs. Effec- relevant agencies with a single interface. Cross-coun- tive bankruptcy can help avoid many distortions— try research suggests that reducing the cost of regis- including, in some cases, taxpayer commitments in tration procedures to the level in the United States the form of bailouts.56 The government can stream- could increase the number of new firms by line corporate bankruptcy and encourage more than 20 percent.49 In India after rapid reuse of assets to reduce the time licensing reforms, innovation in the A better business and cost of bankruptcy and increase formal manufacturing sector in- recovery of losses for creditors and creased by roughly 5 percentage environment is investors. Reorganization may be points.50 Reducing the cost and good not only for an option for firms with proven time of registration processes productivity growth viability that are in temporary alone is not always enough to but also for resilience financial distress. The challenges promote business formalization, for the state are to balance the however. In Brazil, where obtain- in the face of protection of failed entrepre- ing a business license has been eased adverse shocks. neurs’ rights with that of creditor but still remains complicated and rights and to limit the moral hazard where the tax burden associated with for entrepreneurs to act imprudently. the registration is high, most informal firms choose to stay informal.51 Improve enforcement of regulations. Improving reg- ulatory certainty also requires strengthening the Reduce rigidities in labor market regulation. Reform- implementation and enforcement of laws. For ex- ing a single policy, such as labor market regulations, ample, when countries such as Azerbaijan and the on its own will probably not have a large impact. Yet Kyrgyz Republic were initially moving up in the when labor market regulations are too rigid, they can World Bank’s Doing Business rankings, the local pri- undermine the resilience of the economy (see spot- vate sector reported that the business environment light 5). An empirical study covering some 60 coun- had not improved much because the reforms were tries finds that moving from the 20th to the 80th per- not fully implemented. Investor Motivation Surveys centile in regulatory labor rigidity reduces the speed show a correlation between the amount of discre- of adjustment to shocks by one-third and lowers an- tion allowed in applying regulations and variance nual productivity growth by as much as 1.7 percent.52 in delays in compliance. Overall, governments with Evidence from 20 European countries shows that in less discretion in applying regulations present lower countries with high firing costs, firms are more likely investment risk. That is particularly the case where to hire temporary workers as their employment institutional capacity is weak and effective safeguards needs grow.53 In Nicaragua, a 10 percent increase in against corruption are not in place. Governments the minimum wage from 1998 to 2006 raised wages can start to address the problem by drafting laws and by 5 percent for workers whose wages were within 20 regulations with as much clarity as possible to limit percent of the minimum wage but also reduced re- discretion in interpretation. To reduce the incidence Fostering resilience and prosperity through a vibrant enterprise sector 183 of discriminatory implementation, the government ment, as an empirical study with panel data from can strengthen its accountability system, both by 97 countries reports.58 Yet flexibility can lead to increasing the transparency of regulations and by increased job turnover, which can increase the risk improving feedback mechanisms to ensure fair and of unemployment and income loss for workers in timely treatment of all entrepreneurs.57 transition. Promoting flexibility need not mean that workers who lose their jobs are completely at peril of prolonged unemployment, however. Germany and Provide inclusive social protection and Denmark provide interesting but diverse examples insurance of policies that combine labor market flexibility Policies to improve labor market flexibility and pro- with social support. In their specific contexts, the vide social protection need to be pursued in parallel. German model highlights the potential to maintain Increased flexibility improves efficiency by reallocating stable employment through internal flexibility—in resources between and within firms, but can be costly particular through adjustments to working hours for those who lose their jobs. To protect the vulner- and work-sharing—while the Danish “flexicurity” able, including those in employment transitions, the model is characterized by easy hiring and firing government needs to put in place a system to provide alongside a strong safety net and reemployment voice and inclusive social protection. In turn, an in- policies (box 5.5). clusive system that covers basic health and education needs and targets the vulnerable may also promote a Provide support for training and retraining, especially more dynamic enterprise sector (see chapter 3). for the vulnerable. Public support for training works best when enterprises and workers themselves have Build an inclusive social protection system. Flexible la- an incentive to invest in skills. To the extent that lack bor market policies can increase aggregate employ- of resources is an obstacle to training, however, the BOX 5.5 Labor market flexibility alongside social support: Examples from Germany and Denmark As mentioned at the beginning of the chapter, during the most market policies, which include job counseling and requalification, recent global recession, Germany has offered one model of how to job training, and employment with wage subsidies. Light employ- combine labor market flexibility and stable employment. Labor mar- ment protections aim to promote firms’ competitiveness by reduc- ket reforms enacted in the first half of the 2000s provided incentives ing their employment costs. However, a generous social safety net for older workers to return to work, supported job search efforts, and has high coverage and high replacement ratios: low-income groups implemented stricter monitoring of eligibility for unemployment receive up to 90 percent of the income they had been receiving benefits. In response, the share of permanent full-time jobs declined, before they were laid off. Active labor market policies help people while the share of flexible or nonstandard jobs increased. reenter the labor market. In addition, two instruments that allowed flexible work hours After the global crisis, Denmark’s GDP growth fell to minus 5.2 played a key role in stabilizing employment. Working-time accounts percent in 2009. The unemployment rate jumped from 3.3 percent enabled firms to adjust workers’ hours in response to short-term in 2008 to more than 6 percent in 2009 and hovered around 7.5 per- fluctuations in demand. This tool lowers labor costs for firms in a cent afterward. Most of the unemployed found jobs fairly quickly, downturn, helping them weather the storm without laying off work- however: 60 percent after 13 weeks, and 80 percent after 26 weeks. ers. An estimated 320,000 jobs were saved through the use of this Nearly 70 percent of Danish workers surveyed say they are fairly or arrangement. A separate short-time work arrangement was backed very confident that they can find a job if they are laid off—the high- by a government subsidy. Firms paid employees for the hours they est rate for any European Union country.b Firms are also confident worked, while the Federal Employment Agency partly compen- that they can find workers with the right skills with little cost or delay sated workers for the hours lost during the downturn. The number when the economy recovers. of short-time workers sharply increased, peaking at more than 1.5 While it is too early to say whether the German or Danish models million in May 2009. An estimated 400,000 jobs were saved through coped well with the crisis and remain fiscally affordable in the longer the use of short-time work.a run, they show that flexibility in the enterprise sector can both differ Denmark highlights another model of labor market flexibility. across countries and complement social safety nets. The choice of The Danish “flexicurity” model is characterized by easy hiring and policy tools should be based on careful analyses of the institutional, firing, compensated by a generous social safety net and active labor historical, and socioeconomic contexts unique to each economy. Source: WDR 2014 team based on Andersen and others 2011; Rinne and Zimmermann 2012. a. Boeri and Bruecker 2011. b. European Commission 2010. 184 WORLD DEVELOPMENT REPORT 2014 government can channel public funding through and the allocation of public expenditures are often public education institutions and build partnerships among the key areas in which to begin. with the private sector to develop enterprise-based skills training. The government can also provide tar- Pursue policies that improve both flexibility geted support to help unskilled workers who are un- and formality in the longer run employed or employed in informal sectors with low- paying jobs, young people transitioning from school With sound regulations and incentives for flexibility to work, and skilled workers transitioning between in place, the government can pursue policies that im- jobs to (re)integrate into the labor market. Specific prove both flexibility and formality. Policies aimed at policies may also be needed for certain groups of peo- increasing innovation and the share of highly skilled ple. The provision of affordable and quality child and workers in the labor market require investments that elderly care and the possibility of flexible working ar- typically come to fruition only in the long run. rangements, for example, are particularly important for integrating female workers into the labor market. Increase innovation by addressing constraints in access to finance. Access to finance for innovation can be a Use general revenue to finance social insurance. In de- major constraint for enterprises. Because potential veloping countries where the labor market is less for- investors may not have the information an enterprise malized, however, social insurance that is based on has about the likely success of a project, obtaining mandatory contribution from employers and em- financing for innovative projects—which are espe- ployees in the formal sector often protects only in- cially uncertain—can be difficult, especially for new siders, leaving the most vulnerable, such as informal and small enterprises. The government can address workers and the unemployed, unprotected. In many this market failure by creating a business environ- cases, labor taxes can create market distortions and ment that includes appropriate resource support so reduce firms’ incentives to offer formal and stable that the private sector is able to provide financing contracts, particular for women, youth, and other at the three stages crucial for innovative enterprises: vulnerable groups. In Mexico, for example, the close the early concept stage, when entrepreneurs need to tie between employment and social insurance provi- develop ideas into viable concepts and products; the sion has made formal employment less attractive for start-up stage, when entrepreneurs need seed fund- firms.59 It is a question for policy debate whether to ing to establish enterprises; and the growth stage, use general revenue and user fees to fund basic social when entrepreneurs need venture capital to expand. protection, such as basic health insurance and old The challenges for the state are to identify the best age pensions (see the “Focus on policy reform” at the means for offering support depending on the pre- end of this Report).60 Income taxes, property taxes, vailing business context and needs; to ensure that fi- value-added taxes, and, in resource-rich countries, nancing remains short term; to prevent the creation commodity taxes are potential sources of financing. of dependent industries; and to contain the associ- Such a shift in finance has the potential to help al- ated fiscal burden. leviate the undue burden on formal firms and move from protecting workers to protecting citizens. It can Facilitate the adoption of technology and global collabo- facilitate the government’s efforts to meet its basic ration. Creating capacities and incentives to generate function of introducing a social protection system new-to-the-world knowledge will be relatively more that provides basic services to all and supports the important in developed economies, with industries vulnerable, whether employed or not. Keeping ben- at or closer to the technology frontier. In develop- efits modest is often necessary for fiscal sustainabil- ing economies with less sophisticated technological ity and for limiting moral hazard. In less developed capabilities and less abundant resources, the adop- countries, where the needs are larger and the capacity tion of existing knowledge and its adaptation to local is more limited, fiscal sustainability and the efficiency context might be relatively more important.61 Indeed, of service delivery are likely to be more challenging. external sources of technology account for 90 percent Tailoring the design of social protection schemes to of the growth in total factor productivity in most country conditions given the stage of development developing countries.62 The government can help to of the labor market, institutional capacity, and the improve the links between research centers and uni- political environment is crucial to success. Building versities and the private sector in research and prod- institutional capacity and strengthening account- uct extension, support managerial skills training for ability to enhance the efficiency of tax collection entrepreneurs, and build the capacity to absorb new Fostering resilience and prosperity through a vibrant enterprise sector 185 technology in the general workforce. Depending on The collapse of a factory building in Bangladesh in the country’s innovative capacities, the government 2013, which killed more than 1,100 people, demon- can adopt appropriate intellectual property rights strates the need for such regulation—the building policies, which provide incentives for innovation that was constructed illegally, with poor materials, and did can be widely adopted or adapted, and facilitate tech- not comply with existing building codes. Clear guide- nology sharing between countries. Finally, the gov- lines for testing and labeling consumer products, and ernment can facilitate access to international for recalls in the case of product defects, can sources of knowledge through trade, enhance consumers’ confidence in the technology licensing, foreign direct In an ever-changing goods they buy. Environmental regula- investment, joint ventures, links world, a vibrant tions are particularly important to to the diaspora community, and enterprise sector that prevent firms from socializing the other international networks, costs of their activities. Consistent allocates resources such as those for research and implementation of regulation is development collaboration. efficiently, promotes also a concern, because interest innovation, and protects groups frequently use regulation Build the skills level of the labor workers and consumers is as a tool to limit entry and handi- force. A more skilled labor force, crucial to resilience cap competitors. Without effective whose workers can take on more and prosperity. implementation, regulation can also sophisticated tasks and are equipped become a governmental stamp of ap- with more fungible skills, is a necessary proval that does little to protect workers, condition for a flexible and formal enterprise consumers, or the environment. In these cases, sector. Education and lifelong learning are crucial regulation cannot fulfill its function of supporting for mitigating the negative impact and taking on people’s risk management. To prevent regulatory cap- the opportunities of technology innovation. In In- ture, regulation must address specific problems and dia, for example, a higher level of education of a be implemented consistently. district’s workforce is strongly linked to higher en- try rates of formal enterprises.63 Skills shortages and mismatches are among the top concerns of enter- Putting it all together prises. Worldwide, some 27 percent of registered Helping the enterprise sector remain flexible enterprises consider the lack of skilled workers to as it gradually becomes formal be one of the major constraints to doing business.64 Informational barriers are a primary contributor to In an ever-changing world, a key to resilience and this constraint—people do not know either what prosperity is a vibrant enterprise sector that pro- skills are needed or what training is available. For motes innovation; reallocates resources efficiently; example, modern management practices tend to dif- and protects workers, consumers, and the environ- fuse slowly between firms. But field experiments on ment. By promoting the rule of law and contract large Indian textile firms show that increasing aware- security, the state can provide an enabling environ- ness of the necessary management skills—gleaned ment that encourages both flexibility and formal- through free consulting on modern management ity in the enterprise sector. Not only will flexibil- practices—raised average productivity by 11 percent ity and—in the longer run—formality increase the by improving quality and efficiency and reducing resilience of workers, firm owners, and investors to inventory.65 The government can help the product negative shocks and improve their ability to take ad- market and the labor market work better by dissemi- vantage of opportunities, but these characteristics nating information about training and labor market will also help drive economic growth and alleviate outcomes. poverty. Creating such an enabling environment remains Build a regulatory framework that enhances worker, a challenge, however, because of the trade-off in de- consumer, and environmental protection. Once a veloping countries between flexibility and formality. sound basic regulatory framework is in place, govern- In the short run, the informal economy can be a way ments can gradually add targeted regulation to spur out of poverty for many of the poor and vulnerable. competition and protect workers, consumers, and In the longer run, formality strengthens the benefits the environment. Basic workplace safety standards of flexibility and ensures that the entire population can be crucial to prevent the exploitation of workers. can enjoy these benefits. Formalization based on 186 WORLD DEVELOPMENT REPORT 2014 TA B L E 5 .1 Policy priorities to improve the enterprise sector’s role in risk management POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Provide education and job-matching information Basic skills and vocational training Innovative skills training Protection Improve policy certainty and secure property rights Streamline basic regulations, Spur innovation, including technology adoption/ such as those on entry/exit adaptation and innovation at the frontier Insurance Extend social insurance coverage, possibly delinking it from work status Facilitate access to credit Facilitate access to capital markets, for both debt and equity Coping Facilitate adaptability of wages and work hours Develop insolvency mechanisms Facilitate appropriate reorganization and avoid bailouts Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing policies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably and that can be improved over time. rigid enforcement of laws and regulations is not the For this, improving legal institutions, reducing bu- solution, however: the entire enterprise sector would reaucratic discretionary power, and strengthening be worse off in a fully formal but sclerotic economy. transparency are fundamental. What is needed is a business environment that is good for all enterprises, improvement of the regu- Promote flexibility. Efficient labor and capital real- latory framework, provision of public services, and location are crucial for the enterprise sector to ad- enhanced worker skills to strengthen the advantages just to ever-changing business conditions. Adaptable of becoming formal while retaining the advantages worker skills and clear mechanisms to resolve firm of being flexible. Done at the right pace and in the insolvency are needed. right sequence, the same policy elements that focus on increasing the flexibility of the enterprise sector Provide the right incentives. Competition, stream- can contribute to increasing both flexibility and for- lined regulations, and fair enforcement can provide mality in the long run. the incentives that the enterprise sector needs to re- Table 5.1 presents public policies that can improve main flexible as it gradually becomes formal. Bailing the enterprise sector’s socially beneficial role of risk out ailing firms is seldom an advisable policy action. management. Foundational policies include focus- ing on streamlining regulations and strengthening Protect the vulnerable. A social protection system that basic public services. For countries that already have covers basic health and education needs can enhance those foundations in place, policy priorities should the flexibility of the enterprise sector and protect the move on to facilitating innovation and the entry of vulnerable, whether employed or not. The possibility efficient, productive enterprises, and easing the exit of delinking social insurance from work status (em- of inefficient, unproductive ones. ployed/unemployed, formal/informal, urban/rural) deserves further consideration. Following basic principles Keep a long-run perspective. While informality can Do not generate uncertainty or unnecessary risks. Se- act as a safety net to workers and provide flexibility curing property rights and providing a predictable to the enterprise sector in the short run, in the long political and policy environment are prerequisites run striving for formality should remain a goal for for the enterprise sector to play any beneficial role. public policy. Fostering resilience and prosperity through a vibrant enterprise sector 187 Notes 34. ILO 2009. 35. Barrientos, Gereffi, and Rossi 2011. 1. World Bank 2012. 36. Henson and Humphrey 2009. 2. Eurostat, http://epp.eurostat.ec.europa.eu/portal/page/portal/ 37. “Inspectors Certified Pakistani Factory as Safe before Dis- statistics/themes. aster,” New York Times, September 19, 2012. 3. Wölfl and Mora-Sanguinetti 2011. 38. De Soto 1989. 4. Gardner 2012. 39. Fox and Sohnesen 2012. 5. UNEP 2012. 40. La Porta and Shleifer 2008. 6. Loayza and Rigolini 2011; World Bank 2012. 41. See Ritchie 2004 and references therein for the history of 7. Maloney 2013 for the WDR 2014. reforms in Malaysia. 8. Falco and others 2012. 42. World Bank 2010. 9. Knight 1921; Coase 1937. 43. Loayza and Wada 2010. 10. For seminal papers on the topic, see Azariadis 1975 and Baily 44. World Bank 2008. 1974. 45. Bloom 2009. 11. D’Erasmo and Moscoso Boedo 2013. 46. Enterprise Surveys 2006–11, http://www.enterprisesurveys 12. Bigsten and others 2003. .org. 13. Gutierrez 2013 for the WDR 2014. 47. Djankov, McLiesh, and Ramalho 2006. 14. Hsieh and Klenow 2009. 48. World Bank 2013. 15. Caballero 2008. 49. Bartelsman, Haltiwanger, and Scarpetta 2004. 16. Merotto, Kanematsu, and Huang, forthcoming. 50. Seker 2011. 17. Dutz 2013 for the WDR 2014. 51. Bruhn and McKenzie 2013. 18. Baicker, Cutler, and Song 2010. 52. Caballero and others 2013. 19. Mitra 2008, 6. 53. Dräger and Marx 2012. 20. Gardner 2012. 54. Alaniza, Gindling, and Terrell 2011. 21. Loayza and Rigolini 2011. 55. Del Carpio, Nguyen, and Wang 2012. 22. World Bank 2012. 56. Ayotte and Skeel 2010. 23. Hoshi and Kashyap 2010. 57. Fidas and Benhassine 2013 for the WDR 2014. 24. Andrews and Cingano 2012. 58. Bernal-Verdugo, Furceri, and Guillaume 2012. 25. World Bank 2011. 59. Levy 2008. 26. Fafchamps 2001. 60. See Ribe, Robalino, and Walker 2012 for a full discussion. 27. Pianta 2006. 61. Lasagabaster 2013 for the WDR 2014. 28. Judge and Elenkov 2005. 62. Coe, Helpman, and Hoffmaister 1997. 29. Mitchell 1989. 63. Ghani 2013 for the WDR 2014. 30. Schrage 2004. 64. Enterprise Surveys 2006–11, http://www.enterprisesurveys 31. ILO 2009. .org. 32. Fafchamps, Gunning, and Oostendorp 2000. 65. Bloom and others 2013. 33. Loayza and Wada 2010. 188 WORLD DEVELOPMENT REPORT 2014 References Bruhn, Miriam, and David McKenzie. 2013. “Using Administra- tive Data to Evaluate Municipal Reforms: An Evaluation of Alaniza, Enrique, Thomas H. Gindling, and Katherine Terrell. the Impact of Minas Facil Expresso.” Policy Research Working 2011. “The Impact of Minimum Wages on Wages, Work and Paper 6368, World Bank, Washington, DC. Poverty in Nicaragua.” Labour Economics 18 (Supplement 1): Caballero, Ricardo J. 2008. “Creative Destruction.” In The New S45–S59. Palgrave Dictionary of Economics, edited by Steven N. Durlauf Aleman-Castilla, Benjamin. 2006. “The Effect of Trade Liberali- and Lawrence E. Blume. Basingstoke, U.K.: Palgrave Macmillan. zation on Informality and Wages: Evidence from Mexico.” Caballero, Ricardo J., Kevin N. Cowan, Eduardo M. R. A. Engel, Discussion Paper 763, Center for Economic Performance, and Alejandro Micco. 2013. “Effective Labor Regulation and London School of Economics and Political Science, London. Microeconomic Flexibility.” Journal of Development Economics Andersen, Torben M., Nicole Bosch, Anja Deelen, and Rob 101: 92–104. Euwals. 2011. “The Danish Flexicurity Model in the Great Coase, Ronald H. 1937. “The Nature of the Firm.” Economica Recession.” VoxEU.org, April 8. http://www.voxeu.org/article/ New Series 4 (6): 386–405. flexicurity-danish-labour-market-model-great-recession. Coe, David T., Elhanan Helpman, and Alexander W. Hoffmaister. Andrews, Dan, and Federico Cingano. 2012. “Public Policy and 1997. “North-South R&D Spillovers.” Economic Journal 107 Resource Allocation: Evidence from Firms in OECD Coun- (440): 134–49. tries.” Economics Department Working Papers 996, Organisa- D’Erasmo, Pablo N., and Hernan J. Moscoso Boedo. 2013. tion for Economic Co-operation and Development, Paris. “Intangibles and Endogenous Firm Volatility over the Busi- Atkin, David G. 2009. “Working for the Future: Female Factory ness Cycle.” Unpublished manuscript, University of Maryland, Work and Child Health in Mexico.” Unpublished manuscript, College Park, MD, and University of Virginia, Charlottes- Yale University, New Haven, CT. ville, VA. ———. 2012. “Endogenous Skill Acquisition and Export Manu- Del Carpio, Ximena, Ha Nguyen, and Liang Choon Wang. 2012. facturing in Mexico.” Working Paper 18266, National Bureau “Does the Minimum Wage Affect Employment? Evidence of Economic Research, Cambridge, MA. from the Manufacturing Sector in Indonesia.” Policy Research Ayotte, Kenneth, and David A. Skeel, Jr. 2010. “Bankruptcy or Working Paper 6147, World Bank, Washington, DC. Bailouts?” Journal of Corporation Law 35 (3): 469–98. De Soto, Hernando. 1989. The Other Path: The Economic Answer to Azariadis, Costas. 1975. “Implicit Contracts and Underemploy- Terrorism. New York: Harper & Row Publishers. ment Equilibria.” Journal of Political Economy 83 (6): 1183–202. Djankov, Simeon, Caralee McLiesh, and Rita Ramalho. 2006. Baicker, Katherine, David Cutler, and Zirui Song. 2010. “Work- “Regulation and Growth.” Economics Letters 92 (3): 395–401. place Wellness Programs Can Generate Savings.” Health Dräger, Vanessa, and Paul Marx. 2012. “Do Firms Demand Tem- Affairs 29 (2): 304–11. porary Workers When They Face Workload Fluctuation? Baily, Martin N. 1974. “Wages and Employment under Uncertain Cross-Country Firm-Level Evidence on the Conditioning Demand.” Review of Economic Studies 41 (1): 37–50. Effect of Employment Protection.” Discussion Paper 6894, Baily, Martin N., and Robert M. Solow. 2001. “International Institute for the Study of Labor (IZA), Bonn. Productivity Comparisons Built from the Firm Level.” Journal Dutz, Mark A. 2013. “Resource Reallocation and Innovation: of Economic Perspectives 15 (3): 151–72. Converting Enterprise Risks into Opportunities.” Policy Re- Barrientos, Stephanie, Gary Gereffi, and Arianna Rossi. 2011. search Working Paper 6534, World Bank, Washington DC. “Economic and Social Upgrading in Global Production Net- European Commission. 2010. Monitoring the Social Impact of works: A New Paradigm for a Changing World.” International the Crisis: Public Perceptions in the European Union (Wave 4). Labour Review 150 (3–4): 319–40. Flash Eurobarometer Series. Brussels: European Commission. Bartelsman, Eric, John C. Haltiwanger, and Stefano Scarpetta. Fafchamps, Marcel. 2001. “Networks, Communities and Markets 2004. “Microeconomic Evidence of Creative Destruction in Sub-Saharan Africa: Implications for Firm Growth and In- in Industrial and Developing Countries.” Policy Research vestment.” Journal of African Economies 10 (Supplement 2): Working Paper 3464, World Bank, Washington, DC. 109–42. Bergoeing, Raphael, Norman Loayza, and Andrea Repetto. 2004. Fafchamps, Marcel, Jan Willem Gunning, and Remco Oostendorp. “Slow Recoveries.” Journal of Development Economics 75 (2): 2000. “Inventories and Risk in African Manufacturing.” Eco- 473–506. nomic Journal 110 (466): 861–93. Bernal-Verdugo, Lorenzo E., Davide Furceri, and Dominique Falco, Paolo, William F. Maloney, Bob Rijkers, and Mauricio Guillaume. 2012. “Labor Market Flexibility and Unemploy- Sarrias. 2012. “Heterogeneity in Subjective Wellbeing: An ment: New Empirical Evidence of Static and Dynamic Ef- Application to Occupational Allocation in Africa.” Policy fects.” Working Paper 12/64, International Monetary Fund, Research Working Paper 6244, World Bank, Washington, DC. Washington, DC. Fidas, Penelope, and Najy Benhassine. 2013. “Transparency and Bigsten, Arne, Paul Collier, Stefan Dercon, Marcel Fafchamps, Access to Information in Business Regulations.” Background Bernard Gauthier, Jan Willem Gunning, Abena Oduro, Remco paper for the World Development Report 2014. Oostendorp, Cathy Pattillo, Mans Söderbom, Francis Teal, and Fox, Louise, and Thomas Sohnesen. 2012. “Household Enterprises Albert Zeufack. 2003. “Risk Sharing in Labor Markets.” World in Sub-Saharan Africa: Why They Matter for Growth, Jobs, Bank Economic Review 17 (3): 349–66. and Livelihoods.” Policy Research Working Paper 6184, Word Bloom, Nicholas. 2009. “The Impact of Uncertainty Shocks.” Bank, Washington, DC. Econometrica 77 (3): 623–85. Gardner, Elie. 2012. “Peru Battles the Golden Curse of Madre De Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, Dios.” Nature 486: 306–07. and John Roberts. 2013. “Does Management Matter? Evidence Ghani, Ejaz. 2013. “Why Do Enterprises Get Attracted to Some from India.” Quarterly Journal of Economics 128 (1): 1–51. Cities? Not to Others?” Background paper for the World De- Boeri, Tito, and Herbert Bruecker. 2011. “Short-Time Work velopment Report 2014. Benefits Revisited: Some Lessons from the Great Recession.” Gutierrez, Federico H. 2013. “Labor Contracts and Risk Sharing.” Economic Policy 26 (68): 699–765. Background paper for the World Development Report 2014. Fostering resilience and prosperity through a vibrant enterprise sector 189 Hallward-Driemeier, Mary, Bob Rijkers, and Andrew Waxman. Porter, Michael E., and Mark R. Kramer. 2006. “Strategy and Soci- 2011. “Ladies First? Firm-Level Evidence on the Labor Impacts ety: The Link between Competitive Advantage and Corporate of the East Asian Crisis.” Policy Research Working Paper 5789, Social Responsibility.” Harvard Business Review 84 (12): 78–92. World Bank, Washington, DC. Restuccia, Diego, and Richard Rogerson. 2013. “Misallocation and Henson, Spencer, and John Humphrey. 2009. “The Impacts of Productivity.” Review of Economic Dynamics 16 (1): 1–10. Private Food Safety Standards on the Food Chain and on Ribe, Helena, David A. Robalino, and Ian Walker. 2012. From Public Standard-Setting Processes.” Paper prepared for the Right to Reality: Incentives, Labor Markets, and the Challenge Joint FAO/WHO Food Standards Programme, Codex Ali- of Universal Social Protection in Latin America and the Carib- mentarius Commission, Thirty-second Session, Food and bean. Latin American Development Forum Series. Washing- Agriculture Organization, Rome, June 29–July 4. ton, DC: World Bank. Hoshi, Takeo, and Anil K. Kashyap. 2010. “Will the U.S. Bank Re- Rinne, Ulf, and Klaus F. Zimmermann. 2012. “Another Economic capitalization Succeed? Eight Lessons from Japan.” Journal of Miracle? The German Labor Market and the Great Recession.” Financial Economics 97 (3): 398–417. IZA Journal of Labor Policy 1: 3. Hsieh, Chang-Tai, and Peter J. Klenow. 2009. “Misallocation and Ritchie, Bryan K. 2004. “Politics and Economic Reform in Malay- Manufacturing TFP in China and India.” Quarterly Journal of sia.” William Davidson Institute Working Paper 655, Univer- Economics 124 (4): 1403–48. sity of Michigan Business School, Ann Arbor. ILO (International Labour Organization). 2009. The Informal Schneider, Friedrich, Andreas Buehn, and Claudio E. Montenegro. Economy in Africa: Promoting Transition to Formality: Chal- 2010. “Shadow Economies All over the World: New Estimates lenges and Strategies. Geneva: ILO. for 162 Countries from 1999 to 2007.” Policy Research Work- Judge, William Q., and Detelin Elenkov. 2005. “Organizational ing Paper 5356, World Bank, Washington, DC. Capacity for Change and Environmental Performance: An Schrage, Elliot J. 2004. “Promoting International Worker Rights Empirical Assessment of Bulgarian Firms.” Journal of Business through Private Voluntary Initiatives: Public Relations or Research 58 (7): 893–901. Public Policy?” A Report to the U.S. Department of State on Kehoe, Timothy Jerome, and Edward C. Prescott, eds. 2007. Great behalf of the University of Iowa Center for Human Rights, Depressions of the Twentieth Century. Minneapolis, MN: Fed- Iowa City. eral Reserve Bank of Minneapolis. Seker, Murat. 2011. “Effects of Licensing Reform on Firm Inno- Knight, Frank. 1921. Risk, Uncertainty, and Profit. Boston: Hough- vation.” Policy Research Working Paper 5876, World Bank, ton Mifflin Company. Washington, DC. Kremer, Michael, and Eric Maskin. 2006. “Globalization and In- UNCTAD (United Nations Conference on Trade and Develop- equality.” Working Paper 2008-0087, Weatherhead Center for ment). 2013. Global Value Chains and Development: Invest- International Affairs, Harvard University, Cambridge, MA. ment and Value Added Trade in the Global Economy: A Prelimi- La Porta, Rafael, and Andrei Shleifer. 2008. “The Unofficial Econ- nary Analysis. Geneva: United Nations. omy and Economic Development.” Brookings Papers on Eco- UNEP (United Nations Environment Programme). 2012. “Analy- nomic Activity 39 (2): 275–363. sis of Formalization Approaches in the Artisanal and Small- Lasagabaster, Esperanza. 2013. “Policy Priorities for Innovative Scale Gold Mining Sector Based on Experiences in Ecuador, Entrepreneurship.” Background paper for the World Develop- Mongolia, Peru, Tanzania and Uganda: Peru Case Study.” ment Report 2014. UNEP, Nairobi. Levy, Santiago. 2008. Good Intentions, Bad Outcomes: Social Policy, Wölfl, Anita, and Juan S. Mora-Sanguinetti. 2011. “Reforming Informality, and Economic Growth in Mexico. Washington, DC: the Labour Market in Spain.” Economics Department Work- Brookings Institution Press. ing Papers 845, Organisation for Economic Co-operation and Loayza, Norman, and Jamele Rigolini. 2011. “Informal Employ- Development, Paris. ment: Safety Net or Growth Engine?” World Development 39 World Bank. 2006. The Impact of Intel in Costa Rica: Nine Years (9): 1503–15. after the Decision to Invest. Washington, DC: World Bank and Loayza, Norman, and Tomoko Wada. 2010. “Informal Labor in the Multilateral Investment Guarantee Agency. Middle East and North Africa: Basic Measures and Determi- ———. 2008. Thailand: Investment Climate Assessment Update. nants.” Unpublished manuscript, World Bank, Washington, Washington, DC: World Bank. DC. ———. 2010. World Development Report 2011: Conflict, Security, Maloney, William. 2013. “Measuring Labor Market Risk.” Back- and Development. Washington, DC: World Bank. ground paper for the World Development Report 2014. ———. 2011. World Development Report 2012: Gender and Devel- Merotto, Dino, Kosuke Kanematsu, and Tao Huang. Forthcoming. opment. Washington, DC: World Bank. “Why Policy Makers Concerned about Jobs Need a BuDDy.” ———. 2012. World Development Report 2013: Jobs. Washington, PREM Premise Note, World Bank, Washington, DC. DC: World Bank. Mitchell, Mark L. 1989. “The Impact of External Parties on Brand- ———. 2013. Doing Business 2013: Smarter Regulations for Small Name Capital: The 1982 Tylenol Poisonings and Subsequent and Medium-Size Enterprises. Washington, DC: World Bank. Cases.” Economic Inquiry 27 (4): 601–18. ———. Pensions (database). World Bank, Washington, DC, Mitra, Amitava. 2008. Fundamentals of Quality Control and Im- http://www.worldbank.org/pensions. provement. 3rd ed. Hoboken, NJ: John Wiley & Sons. ———. World Development Indicators (database). World Bank, Pianta, Mario. 2006. “Innovation and Employment.” In The Washington, DC, http://data.worldbank.org/data-catalog/ Oxford Handbook of Innovation, edited by Jan Fagerberg, world-development-indicators. David C. Mowery, and Richard R. Nelson. New York: Oxford WEF (World Economic Forum). 2012. The Global Competitiveness University Press. Report 2012–2013. Geneva: WEF. S P OT L I G H T 5 Moving toward greater labor market flexibility: India’s uneven path About 1.5 billion people in developing countries are employed in low-productivity informal sector jobs with no social security benefits and poor working conditions. What can governments do to create better jobs, and more of them? India’s experience shows that states with more flexible labor regulations and lower costs of compliance tend to have greater labor mobility and higher productivity and employment in the formal manufacturing sector. India has one of the most rigid labor markets in the world. regulations, most other legislation has remained similar Labor market issues are governed by 45 central gov- across states. This has prompted a number of studies, ernment laws and more than 100 state statutes, many several of which are summarized in table S5.1. of which overlap. The most controversial laws regulate worker firings and closures of manufacturing firms. A stat- Learning from the experience of labor reform in India ute dating back to 1947, the Industrial Disputes Act (IDA), and elsewhere requires factories with more than 100 workers to obtain Most of the reforms in Indian states were aimed at improv- government permission to lay off employees and close ing procedures and reducing the costs of regulatory com- their operations. Such permissions are rarely granted, and pliance, rather than at enacting comprehensive structural employers face substantial fines and even a prison sen- changes. Only two states (Uttar Pradesh and Uttaranchal) tence if they fire workers illegally. Many other laws govern relaxed the threshold for applying IDA’s restrictions on work conditions, minimum wages, and benefits. Firms with worker layoffs. The more common reforms eased the re- 10 or more workers that use electric power are required to gime for use of contract labor; simplified reporting require- register under the Factories Act and to keep records and ments; limited the number of registers that need to be file regular reports on such matters as overtime work, kept; allowed shift work; and introduced self-certification wages, attendance, sick leave, and worker fines. Compli- of compliance with labor regulations to streamline or re- ance with these requirements, as well as multiple inspec- duce the number of inspections. Two other states (Haryana tions, is costly, particularly for smaller firms. and Punjab) introduced a single annual inspection for Although the regulations are meant to enhance the compliance with all labor laws, and several states now re- welfare of workers, they often have the opposite effect by quire government authorization for surprise inspections. encouraging firms to stay small and thus circumvent labor The reforms have improved the business environment and laws. To avoid compliance costs, firms reportedly break reduced transaction costs. These efforts are mostly piece- down their operations into several small separate units. meal, however; few states have addressed the vast body of In many states, firms are increasingly relying on contract administrative requirements systematically. laborers, who do not receive the same benefits as regular To create the sort of labor flexibility that promotes for- employees. Excessive protections stipulated by labor laws malization and the benefits that accompany it, India needs also discourage hiring. Large Indian firms use less labor to reduce the number of and simplify labor laws, modern- than is justifiable given the prevailing low wages. As a re- ize the IDA, and further improve administrative proce- sult of these distortions, nearly 9 in 10 Indian workers are dures. Although no single blueprint exists for how to de- employed in the informal sector and have little or no social sign and implement effective labor market reforms, several insurance or fringe benefits. This informality is associated developing countries offer examples of policy changes with markedly lower productivity. Value added per worker that increase labor market flexibility. in India’s informal manufacturing sector averages about Malaysia does not limit the duration of fixed term con- one-eighth that in the formal sector. tracts, prohibit night and weekend work, or require govern- ment permission to fire redundant workers. Labor regula- A varied regulatory burden within India by state tions protect the interests of workers but are generally not Because Indian states have the authority to amend central too burdensome for employers and thus do not discourage labor laws, the regulatory burden on enterprises varies them from hiring formal workers. The level of informality, greatly by state. Some states have made their labor laws estimated at about 35 percent of total employment, is more flexible relative to the provisions of the central acts, low relative to other developing countries. Colombia has while others have adopted additional restrictions. This reduced severance payments for employees and intro- variation has created the ideal conditions (a “natural ex- duced a wider definition of fair dismissals. To process mass periment”) to assess the economic effects of labor laws be- dismissals more quickly, it has eased the requirement for cause, while individual states have changed their labor advance notice of firings. The Czech Republic and Slovakia Spotlight 5 191 TA B L E S 5 .1 Studies of labor regulations in India Study Methodology Findings Besley and Constructed an index summarizing state-level • States with less flexible labor regulations had lower output, Burgess 2004 amendments to the Industrial Disputes Act (IDA) between productivity, investment, and employment in formal 1949 and 1992. manufacturing than they would have had if their regulations were not so rigid. • Output in informal manufacturing increased in the same states. Sharma 2009 Used the Besley-Burgess methodology to estimate • After deregulation, the number of informal establishments the impact of reforms in 1991 that removed licensing declined by 25 percent in states with more flexible labor laws, requirements to set up and expand factories. compared with states with more rigid regulations. Ahsan and Pagés Updated the Besley-Burgess index, and estimated the • States that increased labor law rigidity above the level 2009 effects of difficulties in resolving labor disputes and stipulated in the IDA experienced declines in registered restrictions on labor layoffs. manufacturing and output relative to states that did not implement such changes. • Rigid regulations for settling disputes reduced output more than the IDA provisions on layoffs. • West Bengal has lost more than 620,000 manufacturing jobs because of its restrictive regulations in these two aspects of the law alone. Andhra Pradesh and Tamil Nadu have created more than 130,000 formal manufacturing jobs, thanks to improvements in their dispute resolution regulations.a Dougherty 2008 Examined a broader range of labor regulations in 21 • The most common area of reform was contract labor; the least states. Analyzed 8 major labor legal areas. Constructed common were changes in rules governing inspections. an Organisation for Economic Co-operation and • Overall, the degree of reform was modest: no state has Development (OECD) labor reform index based on 50 introduced more than 28 reforms out of a possible 50 reform subjects of possible reform, many of which could be subjects measured. implemented by administrative procedure rather than through formal amendments to the law. • States with higher labor reform indexes had greater job turnover rates—vital for technological change and economic growth. Goldar 2011 Used the labor reform index developed by Dougherty • States with higher labor reform indexes tended to have greater employment elasticity and a higher growth of organized employment in manufacturing compared with states that introduced few reforms. a. World Bank 2010a. increased the maximum duration of fixed term contracts and reduced Goldar, Bishwanath. 2011. “Growth in Organized Manufacturing Employment in redundancy costs. Recent Years.” Economic and Political Weekly 46 (7): 20–23. Sharma, Siddharth. 2009. “Entry Regulation, Labor Laws and Informality.” Enterprise India may well choose a different path, but the overall principle Note 1, World Bank, Washington, DC. guiding reform efforts should be protecting workers—through such World Bank. 2010a. India’s Employment Challenge: Creating Jobs, Helping Workers. means as social assistance, skill building, and help finding jobs— New York: Oxford University Press for the World Bank. rather than protecting jobs. ———. 2010b. Malaysian Economic Monitor: Inclusive Growth. Washington, DC: World Bank. ———. 2012. World Development Report 2013: Jobs. Washington, DC: World Bank. Sources ———. 2013a. “Doing Business 2013. Smarter Regulations for Small and Medium Ahsan, Ahmad, and Carmen Pagés. 2009. “Are All Labor Regulations Equal? Evidence Enterprises. Economy Profile: Malaysia.” World Bank, Washington, DC. from Indian Manufacturing.” Journal of Comparative Economics 37 (1): 62–75. ———. 2013b. Doing Business 2013: Smarter Regulations for Small and Medium-Size Besley, Timothy, and Robin Burgess. 2004. “Can Labor Regulation Hinder Economic Enterprises. Washington, DC: World Bank. Performance? Evidence from India.” Quarterly Journal of Economics 119 (1): 91–134. Dougherty, Sean. 2008. “Labor Regulation and Employment Dynamics at the State Level in India.” Economics Department Working Papers 624, Organisation for Economic Co-operation and Development, Paris. Finance doesn’t need to be complex, and it can actually be helpful: an elderly man views stock prices at the Shanghai stock exchange. FPO © Qilai Shen/Panos CHAPTER 6 The role of the financial system in managing risk More financial tools, fewer financial crises The financial system to protect his wealth by spreading his can fulfill a socially business risks and opportunities and ancial system taking his company public. The ini- beneficial function of Fin risk management tial public offering on the regional Warsaw Stock Exchange was well In 1990, after the Velvet Revolu- subscribed, and he was able to di- tion ushered in an era of reform versify his wealth by buying stocks in the Czech Republic, Jan Sarkis, and bonds of other companies. Be- the son of a Greek immigrant, de- cause of his wise decisions, and the cided to start a business to produce financial system to support them, Jan bottled juices. He took out a bank loan is now a wealthy, respected man, focused and, on the advice of his local commu- on giving back to the community.1 nity, bought flood insurance and put his savings As the ups and downs of Jan Sarkis’s story il- in a bank to protect against local theft. When banks lustrate, the financial system can help people man- plunged into a crisis in 1997, Jan’s loan repayments age risk by providing them with useful financial spiked; then his savings were frozen for 14 months tools, protecting them from bad shocks, and better because his bank went bankrupt. A one-in-one- positioning them to take advantage of opportuni- hundred-year-flood in 2002 swept away his business, ties. Banks, insurance companies, brokerage houses, but, fortunately, his insurance settlement covered a stock exchanges, other financial institutions, and the majority of the losses. The twin experiences left Jan financial infrastructure (such as payment systems) with mixed feelings about the financial system. that form the financial system can collectively fulfill In 2006, the Czech National Bank, a credible this socially beneficial function of providing finan- public institution, became the integrated supervisor cial services and helping society manage risks. They of financial services, and confidence in the Czech fi- can do so by offering people market insurance (such nancial system rose. An entrepreneurial risk taker, as disaster or life insurance), self-insurance (sav- Jan took out another credit line and consolidated his ing deposits), and self-protection (safe and efficient savings in one bank account for greater efficiency. payments). People, including the poor, need not just Two years later, the global financial crisis hit. Thanks credit but a range of financial tools to manage risk to conservative supervision, the Czech banking sys- and pursue opportunity effectively and responsibly. tem was well prepared. But the Czech koruna de- However, the financial system can also hurt people preciated, pushing up the price of imported goods, if it fails to manage the risk it retains. It can generate including drinks. bad shocks that affect people directly by hindering Jan took a big risk, using his credit line to intro- access to finance or indirectly by hampering refi- duce a new sparkling drink. It was a success! Mindful nancing of enterprises, straining public finances, of past shocks and possible future ones, Jan decided and leading to loss of jobs, income, and wealth. 193 194 WORLD DEVELOPMENT REPORT 2014 Public policy thus has two important roles, both consumption, finance their own or their children’s of which help people manage risk. It can encourage education, deal with health and income shocks, im- the financial system to broaden the share of people prove nutrition, and plan for a better future, among with access to financial services (financial inclusion), other socially useful activities. In this way, the finan- giving more people more and better financial risk cial system can advance overall development and management tools. It can also encourage the finan- help create an environment of equal opportunity and cial system to better control systemic financial risk. a level playing field, including for the poor. Different Public policy should be selective, using direct inter- types of financial risk-managing tools are designed to ventions (such as subsidies and guarantees) sparingly achieve different outcomes: to avoid distorting incentives in the financial market. • Payment and foreign exchange services increase It should focus on providing adequate financial in- the security and ease of domestic transactions and frastructure (payment systems, credit history infor- international remittances, thus helping people mation) and on implementing enabling regulation self-protect. to promote greater competition and use of diverse financial tools by people. At the same time, the state • Saving instruments (bank deposits and liquid secu- needs to implement supervision of systemic risk in rities) help people smooth consumption and accu- the financial sector that is prudent but promotes de- mulate buffers for rainy days, thus providing self- velopment in the sector. Specific recommendations insurance. Sometimes, savings can be a vehicle to to achieve these ends are discussed throughout the finance unusually large expenses (investment) and chapter.2 to self-protect. This chapter focuses on the risk-managing func- tion of the formal financial system, whereas chapters • Credit (such as education loans) helps people and 3 and 4, and, to some extent, chapter 5 address in- their support systems alleviate financing con- formal financial arrangements. The formal financial straints over time and exploit opportunities with system can be defined in various ways. Here, the for- greater flexibility and resilience, thus improving mal financial system is defined as consisting of firms people’s ability to self-protect. Access to credit whose primary business activity is financial activity. in bad times, including credit they have lined up The system thus ranges from banks and insurance in advance to tap in event of shocks (contingent companies to microcredit and microinsurance firms. debt), can help people and support systems cope Financial systems perform several interconnected better. functions, including reallocating resources from sav- • Market insurance, including hedging instruments, ers to investors; monitoring managers and exerting helps people and their support systems insure corporate control; and facilitating trading, hedging, against the consequences of extreme adverse events diversifying, and pooling of risk; this chapter concen- such as death, impaired health or injury, or loss of trates on the risk management function.3 Given the income or wealth. The state can also benefit from Report’s focus on people, this chapter particularly ex- market insurance against extreme losses. amines the tension between financial inclusion and financial stability. Broader trade-offs between finan- • Debt and equity investments help people diversify cial development and stability are discussed later in wealth into a robust portfolio of instruments ac- the context of institutional reforms to improve the cording to their risk preference, thus facilitating formulation of financial sector policy. self-protection—and, if liquid, self-insurance, as well. For good risk management, people need a • Risk-taking capital (such as private equity or ven- range of financial tools ture funds) enables firms, from small ones to inter- national corporations, to take informed risks and Different financial tools serve different innovate, including through innovation-driven purposes start-ups and firm expansion. The financial system supports risk management by • Public trading of assets (commodities, securities, offering various financial tools to people and their financial derivatives) provides a mechanism for support systems (households, the community, en- discovering and determining prices that match de- terprises, the state, and even the international com- mand and supply. People then use this information munity).4 With these tools, people can smooth to make decisions about their consumption and The role of the financial system in managing risk 195 saving, business opportunities, portfolio alloca- which act as a self-insurance buffer (diagram 6.1a). tion, and strategic management of risks. Using savings to self-insure against larger, less prob- able losses can be inefficient, however. Ideally, those • Risk pricing information embedded in interest types of losses can be insured efficiently through rates, insurance premiums, and other financial credit, but only if people and their support prices is provided by the financial system, systems have access to credit in bad times thanks to its comparative advantage or can arrange for a loan that they can in this area. The financial sector Because people rely on when bad times come (con- “puts a price tag” on risk and face multiple risks, they tingent debt). Even credit can be thus helps the users of financial must employ a range of too expensive a financial tool to tools understand the cost and prepare for one-in-one-hundred- benefits of different strategies tools for their financial year events, however. For these for managing their risks. strategies to be events and losses, market insur- Each financial tool can manage effective and ance is the most efficient financial only certain risks efficiently, based reliable. tool. And even market insurers are on their frequency, intensity, and im- unable to efficiently price, retain, and pact. Self-insurance and market insur- manage losses from extreme unexpected ance help in coping with losses. Deposit ac- events, whose impact is hard to predict; in counts and electronic payment methods can make those cases, social safety nets and other solidarity payment of current (expected) expenditures more schemes can offer support. In their risk manage- efficient and secure. Remittances can also be se- ment strategy, enterprises can use risk-taking capi- curely and efficiently transferred to their receivers tal for losses larger than those efficiently covered by using electronic payments and foreign exchange contingent credit and smaller than those efficiently services (box 6.1). When expenditures jump unex- covered by market insurance. Another type of pectedly in the wake of a bad shock, the first wave hybrid financial tool is provided by religion-based of losses can be efficiently coped with by tapping (Islamic) finance. Islamic banking products such as saving deposits and liquid financial investments, partnership loans (under mudaraba or musharaka D I AG R A M 6.1 Combining financial tools improves preparation for losses of different probability and severity a. Self-insurance and market insurance tools make it b. Self-protection tools help decrease the easier to cope with losses of different severity probability of a given loss Probability of loss Probability of loss Self-protection (credit savings, electronic payments, investment diversification) Market insurance Safety Payment nets services Savings Credit Severity of loss Severity of loss Source: WDR 2014 team. Note: The diagrams depict a stylized loss distribution with fitted financial tools according to their efficiency to insure and protect against possible losses (risk) of varying frequency and intensity. The personal loss distribution function (dark red line in panels a and b) improves (shifts down and to the left) when people use financial self-protection (light red line in panel b). 196 WORLD DEVELOPMENT REPORT 2014 B OX 6.1 Better than cash: Electronic payments reduce risk and costs Cash may still be king at times, but compared with electronic pay- extend remittance services to millions of people who were previ- ments, cash payments are inefficient, can carry significant handling ously unbanked in remote, rural areas. For example, in the Philip- and transportation costs, and run the risks of theft, loss, and counter- pines, “G-Cash” and “SMART Money” serve the Filipino diaspora by feiting. The use of cash also perpetuates the shadow economy by providing remittances over their mobile money platforms.d allowing business transactions and sales to stay off official books of Electronic payments, however, involve some risks that can accounts. Greater use of electronic payments, rather than cash, could dampen consumer confidence: the risk of fraud, the risk that the save Brazil 0.7 percent of its gross domestic product (GDP) a year,a payment product issuer will go bankrupt, and the risk that records and India as much as 1.6 percent of GDP.b of the customer’s account will be corrupted or destroyed because of Individuals and small firms using electronic payments benefit problems with the issuer’s operating system. For payment products from convenient online authorizations, easier record keeping, and linked to bank accounts, the second and third risks can be mitigated the availability of dispute resolution mechanisms. By promoting by prudential and operational requirements for the bank, as well electronic payment, government policies have drastically reduced as by deposit insurance. Such arrangements may not cover non- the risk of crime and have enabled beneficiaries to keep up with bank issuers, but other mechanisms to mitigate these risks can be their financial obligations without delay. Importantly, electronic enforced. The industry actively manages security risks, and targeted payment instruments must be linked to a deposit account either at a consumer protection measures are being implemented. deposit-taking institution (bank) or in the form of e-money that can Electronic payments can help manage fraud and leakage risks in be used by banks, other financial firms, or mobile network opera- government payment programs and ultimately improve transpar- tors. For the financially unserved and underserved, electronic pay- ency and accountability. For example, in Saudi Arabia, the imple- ments are usually the first contact with formal financial services. In mentation in 2002 of the SADAD payment system, an electronic bill Pakistan, for example, more than 1.8 million branchless banking payments and settlement platform, saves the government 10–15 accounts process more than 10.4 million transactions monthly.c percent of annual revenues previously lost to human error, fraud, Electronic transfers and payments of remittances offer signif- and delay. In Brazil, a corporate card payment program (Cartão de icant benefits not only to migrants and their families but also to Pagamento do Governo Federal) has replaced the use of cash and receiving countries. Remittances from 192 million international checks for low-value procurement that is not subject to a bidding migrants (3 percent of the world’s population) totaled $501 billion in process; it has allowed agencies and cardholders to track expenses, 2011, of which $372 billion went to developing countries. Branch- while a government website openly discloses the value of transac- less banking and banking partnerships with mobile operators can tions, date, and type of merchants. Source: Maria Teresa Chimienti for the WDR 2014. a. Central Bank of Brazil, “Efficiency and Costs on Retail Payment Instruments Usage,” http://www.forodepagos.org/pdf/Custo_Eficiencia_English.pdf. b. Ehrbeck and others 2010. c. Consultative Group to Assist the Poor (CGAP), “An Overview of the G2P Payments Sector in Pakistan,” http://www.cgap.org/publications/overview-g2p- payments-sector-pakistan. d. See CGAP website, http://www.cgap.org/topics/paymentsremittances, for more information on payments and remittances. contracts) can offer entrepreneurs profit-loss risk- diversification possibilities for investors analogous to sharing qualities.5 limited-duration equity investment.6 Other financial tools decrease the chance and Because people face risks of varying frequency size of loss by enabling people to self-protect (dia- and intensity, a diversified financial risk manage- gram 6.1b). Education loans provide access to bet- ment strategy that uses a range of financial tools is ter education and thus can help decrease the chance more effective than a strategy that uses only one or of being unemployed. Housing loans ease access to too few tools. Moreover, a diversified financial strat- better housing and associated infrastructure, such as egy will be more reliable because it provides a variety sanitation, that can decrease the risk of water-borne of backups and ways to spread the financial market diseases and pandemics (box 6.2). Saving deposits risk that underlies any financial strategy. A mix of can be used to accumulate funds for large, infrequent financial tools helps people increase resilience, be- investments that increase people’s protection against cause in real life, they do not manage each risk in risks (such as a pump that cleans water). Similarly, isolation but different risks at the same time (chap- a greater variety of investment securities (stocks, ter 1, table 1.1). Moreover, these risks are more or less bonds), issued by various companies and funds with either idiosyncratic or systemic in nature. To prepare varying risk characteristics, can help people diversify for a large idiosyncratic risk (such as long-term ill- their assets and protect against loss of income and ness), market insurance (such as health insurance) wealth. Investment deposits of Islamic banks, which can be the most efficient financial risk management finance partnership loans to entrepreneurs, offer tool to use. Market insurance might be a bad tool for The role of the financial system in managing risk 197 B OX 6.2 Housing finance can improve household resilience and opportunities A house can be a lifelong investment. Housing finance allows indi- a waterproofed dwelling, and warmth or shade can all improve viduals to acquire property at an early age and spread repayments health conditions, especially among the more vulnerable young over time, as their income rises. Without proper financing, the alter- and elderly. Simple improvements like having a concrete floor can native is to spend years saving while living in unsatisfactory condi- reduce mosquito breeding grounds and thus lead to lower levels of tions, or building housing little by little at a higher cost. In old age, malaria. The availability of high-quality affordable homes enables home ownership provides security and resilience when income is families to spend a greater share of their household income on lower and would not easily cover rent payments. nutritious food, health care, and other essentials that promote good Housing finance can bring economic opportunity to house- health. Greater residential stability also reduces the stress and dis- holds. By expanding access to secured credit collateralized by hous- ruption associated with frequent or unwanted moves and provides ing, housing finance can release family wealth for other diversified a stable base for individuals with chronic illnesses and other condi- investments, unlocking the power of so-called “dead capital.”a A tions to receive needed care.c properly functioning titling system and housing finance products Improved housing contributes to safer and more resilient commu- can thus play a role in creating economic opportunity. nities. In Honduras, for instance, criminal gangs are widespread, and Housing finance improves people’s resilience and helps them lower-income communities are by far the most affected by the crimi- avoid poverty traps. To obtain a mortgage, households usually need nal activity. Improved housing through higher investment has helped to accumulate significant savings for the down payment. This reduce criminal activity and antisocial behavior. Effective housing “forced” saving alone can contribute substantially to a household’s projects have the power to change socioeconomic classification. resilience.b Through housing finance products, people have access Such projects have helped very low-income communities to achieve to better housing with better sanitary conditions, thus improving lower-middle income status and good educational levels, and other- their resilience to disease. Having utility connections, sanitation, wise marginal communities to enjoy higher security in Honduras.d Source: Simon C. Walley for the WDR 2014. a. De Soto 2000. b. Collins and others 2009, 179. c. Cohen 2007. d. RTI International 2005. protecting against a systemic risk (such as financial Savings is the most frequently used financial tool crisis) because it may fail if many insurance com- around the world, followed by insurance and credit. panies go bankrupt. At the individual level, people This pattern, however, may reflect various obstacles can prepare and cope better with systemic shocks to implementing better financial risk-managing by increasing their self-reliance (self-insurance and strategies, on both the supply and demand side. self-protection), including by using a range of suit- able financial tools. Supply-side factors that influence access In theory, it should not matter whether the financial instruments are provided by banks, microfinance Some financial systems are better than others firms, insurance companies, or capital markets as at offering access to variety long as people have access to the range of financial The range of financial tools supplied by the formal tools they need.7 In practice, however, the insti- financial system varies considerably with the stage tutional form does matter because each financial of development and personal income within a firm is licensed to provide only a specific range of country. On average, people in high-income coun- financial tools, even though several institutions can tries save through bank deposits much more than be integrated under one financial group or holding people in middle- and low-income countries (figure company. At lower levels of financial development, 6.1a). Even the poorest 40 percent of people in high- financial systems tend to be concentrated in bank- income countries (figure 6.1b) are much more likely ing; at higher levels of development, there is greater to use formal saving deposits than people in middle- diversification into capital markets, insurance com- income and low-income countries. Formal credit panies, and mutual funds (figure 6.2). Thus having a is commonly used in high- and low-income coun- financial system heavily concentrated in banks may tries; people in middle-income countries use credit constrain the provision of insurance. Similarly, the much less. The use of private health insurance across absence of capital markets, mutual funds, or broker- middle- and low-income countries differs greatly, age houses can constrain people’s options to diversify both on average and for the poorest 40 percent. wealth. 198 WORLD DEVELOPMENT REPORT 2014 F I G U R E 6 .1 The range of formal financial tools used by individuals varies by country and income a. Within-economy average b. Within-economy, income bottom 40 percent 50 50 40 40 30 30 Percent Percent 20 20 10 10 0 0 Savings Credit Health insurancea Savings Credit Health insurancea Low income Middle income High income World average Source: WDR 2014 team based on data from World Bank Global Findex (database); Demirgüç-Kunt and Klapper 2012. Note: The figure measures the percentage of adults using formal financial risk-management tools. a. Data on self-paid health insurance were not available for high-income countries. F I G U R E 6 .2 As financial systems deepen, they diversify their institutional structure Gross national income per capita in 2010 $7,355 $12,150 $26,630 $42,280 9% 6% 7% 14% 6% 5% 28% 34% 12% 29% 29% 3% 6% 8% 12% 14% 55% 49% 44% 31% Stock market Bank assets Mutual fund Insurance company Pension fund capitalization assets assets assets Source: WDR 2014 team based on data from World Bank Global Financial Development Database. Note: Size of financial system calculated for 56 countries (6 lower-middle-income, 20 upper-middle-income, and 30 high-income coun- tries) as the sum of bank assets, stock market capitalization, mutual fund assets, insurance company assets, and pension fund assets. Financial firms themselves may be constrained in nancial tools to people and their support systems, risk management by not having access to needed fi- and to efficiently absorb more risk. Likewise, pri- nancial infrastructure (electronic payment systems, vate firms may lack the capacity to assess and price credit information) or hedging tools (such as cross- certain risks (such as major natural catastrophes, currency or interest rate swaps). That, in turn, could terrorism, and epidemics) and thus focus on provid- limit their ability to offer a wider range of better fi- ing financial tools only for better-understood risks. The role of the financial system in managing risk 199 B OX 6.3 Innovative insurance mechanisms in Mongolia and Mexico Insuring against livestock mortality in Mongolia delays in funding or the reallocation of budgets intended for devel- Forty percent of Mongolia’s workforce is engaged in agriculture, opment purposes. In 1996, the government of Mexico established mainly in herding. Harsh climatic conditions periodically lead to cata- the Fund for Natural Disasters (FONDEN) to provide adequate finan- strophic losses of livestock, posing a systemic risk to herders’ liveli- cial resources for federal and state postdisaster reconstruction hoods and to Mongolia’s economy. In 2006, the government of Mon- efforts without compromising government spending already com- golia introduced an index-based livestock insurance project (IBLIP) mitted. It has evolved significantly to include broader disaster risk to provide livestock mortality insurance to herders and increase the management activities, such as funding risk assessment and risk financial resilience of Mongolia’s herders and its economy. When, in reduction and rebuilding infrastructure to higher standards. 2010, for example, a devastating winter killed nearly 22 percent of FONDEN, which supports an integrated risk financing strategy Mongolia’s livestock, the IBLIP provided $1.42 million in indemnity for the government’s contingent liabilities from disasters, has three payments to 4,706 of the 5,628 covered herders. key features. The first is a risk assessment profile. The government The insurance program uses an index based on the average has a well-defined loss-reporting mechanism that provides accurate mortality of adult livestock in each of Mongolia’s counties to deter- information about expenditures from past events. To support mine payouts. The IBLIP has proven to be an effective tool for seg- evidence-based public decision making about disaster risk manage- menting risk among herders, the domestic private insurance sector, ment, the government has invested in exposure data and an in- the government, and the international reinsurance market. It house probabilistic risk model. The second key feature is clarification ensures that each risk layer is effectively financed by the most of contingent liability. FONDEN rules and guidelines clarify how appropriate stakeholder. Participating herders retain livestock mor- total authorized resources will be determined for rapid reconstruc- tality risk of up to 6 percent. From 6 to 30 percent, a commercial tion of public infrastructure, low-income housing, and eligible natu- insurance product transfers this risk to a domestic pool of private ral environment assets, as well as how the liability will be split insurers. Above 30 percent, the government finances a social safety among federal and state governments. The third feature is risk net product. Herders who purchase the commercial insurance prod- financing. FONDEN’s integrated disaster risk financing strategy uct are automatically registered for the social safety net product at relies on a combination of risk retention instruments (self-funding, no additional cost. The government also ring-fences its fiscal expo- exceptional budget allocation) to finance recurrent expenditures, sure to extreme livestock losses by ensuring that its liability is trig- and risk transfer instruments to provide additional financing for gered only in the most extreme years, during which safety nets are immediate response (catastrophe bonds) and longer-term recon- required. Finally, by tapping international reinsurers, the IBLIP also struction (reinsurance of excessive losses) after major disasters. facilitates the transfer of livestock mortality risk out of the country. The federal and state governments spent $1.46 billion a year on disaster response from 1999 to 2011, two-thirds of which was Financing postdisaster expenditures in Mexico financed through FONDEN. The system is continuously evolving to Mexico is exposed to earthquakes, hurricanes, floods, and a wide integrate lessons learned from experience and to incorporate new variety of other geological and hydrometeorological phenomena. budgetary tools and technology to make FONDEN more effective Postdisaster recovery and reconstruction costs can jump because of and efficient. Source: Laura E. Boudreau, Daniel J. Clarke, and Olivier Mahul for the WDR 2014. When firms offer a financial product to manage a banks.9 The delivery of other financial tools, such as risk they do not fully understand, they often misprice insurance and capital market investments, needs to the product, jeopardizing their own stability and, if include clear and thorough communication between failing to pay out claims, damaging customer confi- the seller and the client to ensure that the seller un- dence as a result. For example, a private retirement derstands the expectations (risk management needs) insurance product in the Philippines was initially so of the client and that the client is aware of the proper- underpriced that the offerer nearly went bankrupt. ties of the financial tool. For instance, clearly explain- In a competitive environment possibly enhanced ing to a buyer of insurance the risk that the insurance by new entry of foreign firms, the financial industry may not cover the complete insured loss (basis risk) can increase the share of people using financial ser- is crucial for increasing take-up and renewal.10 To vices through responsible actions that account for provide financial tools that offer more complete so- risk.8 Overall, the best way for financial firms to help lutions for risk management, financial firms should advance financial inclusion is to offer simple, read- also innovate through partnerships with the state, in- ily accessible, and reliable financial tools. For exam- cluding in the area of insurance (boxes 6.3 and 6.4). ple, Kenya’s M-PESA and M-KESHO projects have The large size of the financially underserved pop- greatly broadened the use of electronic payments and ulation worldwide indicates that microfinance insti- mobile savings, and South Africa’s Mzansi accounts tutions (microcredit, microsavings, microinsurance) are now used by one in six South Africans who use can play an important role. There have been waves of 200 WORLD DEVELOPMENT REPORT 2014 B OX 6.4 Private pension insurance to confront the risk of income loss in old age Some 700 million people worldwide are over 60 years of age—a equately assessed the impact of this product on the financial con- number that is expected to increase to 1.6 billion within the next 40 ditions of its institution. A later assessment showed that two years years. A majority of these people live in developing countries, where of premiums would cover just one month of benefits, a situation there are few government income programs to support them in their that threatened the capital of the entire institution, which also old age. The governments of advanced economies have various sold other insurance products. In 1999, the retirement income types of public income support programs for people who have not activities were separated from the institution and converted into a provided adequately for their retirement. But a rising elderly popula- defined contribution scheme, with a significant loss to the original tion and a shrinking working population are threatening the fiscal contributors. viability of these programs at their existing levels. Issues of viability • In 2001, Grameen Bank in Bangladesh introduced the Grameen aside, developing countries have very few such arrangements Pension Scheme. The scheme attaches a mandatory retirement because their governments lack the capacity to finance and adminis- savings product to its loan products and requires borrowers to ter them. contribute a minimum of Tk 50 a month. Prospective borrowers are In the absence of government support, income for the aged tra- thus instructed in the need to save for old age. The bank gener- ditionally comes through the family and social networks, but these ously subsidizes the earnings on the contributions—doubling the are breaking down. Decreases in birth rates are resulting in smaller amounts contributed by individuals who make regular contribu- families, and rapid urbanization is building a distance between the tions. In turn, the bank benefits from the pension fund, which aged and their families and social networks. Increasingly, developing increases the pool of funds available to the bank to pursue its main countries are turning to private pension systems to fill the vacuum. business activity of providing microloans. Providing formal income support to the aged poses many chal- lenges for policy makers and private providers alike, including mak- • The National Jua Kali Pension Scheme in Kenya is a voluntary pen- ing people aware of the need and motivating them to save for old sion savings program developed in 2011 by the National Federation age, establishing trust in the institutions that can provide support, of Kenya Jua Kali Associations and the Retirement Benefits Author- and developing cost-effective distribution systems. These chal- ity. In the first 12 months of operation, it attracted 25,000 members. lenges have been addressed effectively in some of the very few pri- By June 2012, the quarterly contribution flows exceeded K Sh 7 mil- vate sector programs that have been implemented over the past lion ($82,000). The key features that have contributed to the early two decades: success of the scheme are strong endorsement by the supervisor and the well-respected trade organization, low distribution costs • Early unsuccessful attempts by CARD, a nongovernmental organi- (access is only through mobile phone), and an investment manager zation in the Philippines, illustrate the importance of planning and that is prepared to subsidize fund management costs. pricing retirement income products. In 1996, CARD introduced a product that provided members with P 300 ($5.45) a month from The success of any private pension initiative will require a sound their 65th birthday until death, in return for premiums of P 2.50 legal and regulatory framework, strong and capable supervisors, ($0.05) a week paid from the date of membership until retirement. good corporate governance, prudent investment practices, and This product was extremely popular. CARD, however, had not ad- cost-effective administration and distribution systems. Source: Anthony Randle for the WDR 2014. euphoria and criticism concerning microcredit and, scheme for workers in the informal economy is In- more recently, microinsurance. Microfinance needs dia’s Integrated Social Security Programme, which to assume a more realistic role in financial inclusion, insures more than 100,000 women workers and cov- taking into account business sustainability. Good ex- ers health insurance (including a maternity compo- amples exist in this respect, and it is time for others nent), life insurance, and asset insurance. As many to follow. Procredit, established in 1996, is the first credit and insurance markets in advanced econo- multinational microfinance bank; as of 2008, it was mies become saturated, multinational financial operating in 26 countries with 17,000 employees and institutions (banks and insurance companies) are $6 billion in assets. It enjoys an investment-grade looking for sustainable growth opportunities; these rating that enables it to raise long-term finance in include microfinance, which is increasingly becom- the German bond market, in addition to mobilizing ing a mainstream activity in emerging markets.11 and providing access to saving deposits locally. In de- veloping countries, microinsurance firms, requiring Demand-side factors that influence access an insurance premium as low as 50 cents, insure any- The large share of people who use only informal fi- thing from television sets to burial costs. The world’s nancial tools (savings, loans, insurance) or no finan- largest comprehensive contributory social security cial tools at all reveals a great pool of potential clients The role of the financial system in managing risk 201 The shares of formal and informal saving and borrowing change as F I G U R E 6.3 countries develop a. Savings b. Borrowing High income 45 14 42 High income 14 17 69 countries countries Middle income Middle income 8 26 67 18 13 69 countries countries Low income 11 Low income 11 33 56 18 70 countries countries 0 20 40 60 80 100 0 20 40 60 80 100 % of population % of population Saved formally Saved using other methods Formal credit Informal credit Did not save Did not borrow Source: WDR 2014 team based on data from World Bank Global Findex (database); Demirgüç-Kunt and Klapper 2012. Note: The figure shows the percentage of adults saving or borrowing any money in the past year. that could be commercially viable for the formal fi- accounts and may keep some from using banks al- nancial sector. For credit, it is probably not desirable together. People also might prefer to stay in the in- to include everybody—not everyone is creditworthy formal sector (for example, refusing to use electronic or can handle credit responsibly—but a prudential payments), or they do not understand the benefit limit may not exist for deposits and insurance.12 of using financial tools for risk management. In ad- More people use formal financial tools, such as sav- dition, low financial literacy often leads people to ings and credit, as their countries develop and their join Ponzi schemes, which frequently emerge and average income rises (figure 6.3). However, even the collapse in many developing countries (such as Al- poor use formal saving deposits that enable them bania, Nigeria, or the Philippines) and which may to make investments and better plan how to break have damaged consumer confidence in any saving out of poverty. As countries develop, informal sav- arrangements, including formal ones. ing arrangements continue to coexist with formal People are not the only ones to blame for the de- ones. Informal saving and borrowing arrangements ficient use of financial tools. Bad corporate gover- apparently have certain features that formal financial nance of financial firms contributes to low financial tools cannot provide. Completely replacing informal inclusion. Distorted incentives that focus on maxi- financial tools with formal ones may thus be neither mizing short-term profits are a particular problem: feasible nor desirable. it takes time, effort, and up-front investment to People may not use formal financial tools for broaden a client base and the range of financial tools. several reasons. The main reasons people give for Many financial firms have not taken client needs suf- not using a formal savings account are that “they ficiently into account, including their risk profile and do not have money to use it” (66 percent); “some- risk management goals, in designing and delivering one else in the family already has an account,” which financial tools. In Mexico, low-income consum- suggests there are indirect users (23 percent); “bank ers found greater price transparency at pawn shops accounts are too expensive” (24 percent); and “banks than at microfinance institutions; they also trusted are too far away” (20 percent).13 Lack of necessary department stores to hold their savings more than documentation (17 percent) and mistrust in banks they trusted banks. “They [department stores] don’t (13 percent) also discourage people from using bank give us anything, but at least they don’t take any- 202 WORLD DEVELOPMENT REPORT 2014 thing away,” some consumers noted, in reference to problems, the state should consider incorporat- the various fees and hidden charges banks levied on ing financial examples in regular public education their savings accounts. One consumer reported that curricula and partnering with the private sector to “dormancy” and other charges had reduced the value ensure proper design, implementation, and conti- of her bank savings account from 15,000 Mexican nuity of financial education programs.19 pesos to 9,000 over three years.14 Help overcome obstacles to introducing Public policy can help broaden the use of useful, innovative financial instruments. financial tools • Other direct interventions may be desirable. The Experience teaches that direct policy interventions state can help viable innovative financial products in the financial sector can have unintended conse- achieve scale and increase financial inclusion by, for quences and distort proper incentives, especially with example, introducing government-to-person (G2P) regard to shared responsibility for risk management. payments (as in India’s NREGA G2P program),20 Prime examples of possibly distortive interventions making car or mortgage insurance mandatory, or include credit subsidies and guarantees.15 Political requiring that large transactions or tax-deductible capture and lobbying have often led to misallocation expenses be made with electronic payments. In of credit within the economy, and poorly designed seeking the best solutions to advance financial in- financial literacy programs have often failed to de- clusion, the state should consider partnering with liver desired results, become quickly obsolete, or built the private sector (see boxes 6.3 and 6.4). up false confidence of consumers in their financial • Improving infrastructure is particularly impor- skills.16 Many small developing economies lack ade- tant for payment and security settlement systems, quate financial infrastructure, which cannot be viably credit information infrastructure (public credit developed by the private sector because of the small registries and credit bureaus), and collateral frame- scale of the market. Thus the infrastructure must be works (registries of movable and immovable col- developed with the participation of the state, pos- lateral, collateral appraisal, execution and sale). In sibly through private-public partnerships, to enable providing financial infrastructure, the state should greater financial inclusion and development.17 Even partner with the private sector as much as pos- where access to financial services exists, sound finan- sible to improve governance of the infrastructure cial inclusion may be endangered by a lack of con- providers, ensure timely upgrades of technology, sumer protection regarding delivery of the services.18 and encourage continuous innovation. The 2013 What can public policy do to broaden the avail- Doing Business indicators suggest that the quality ability and use of financial tools to manage risk? of collateral frameworks and the depth of credit Lessons from experience, as well as the conceptual information in developing countries are about 30 framework of this Report, suggest the following: percent and 60 percent lower, respectively, than the quality and depth in advanced economies. Further, Minimize unintended consequences of policy the state postal network could be used to house interventions. properly regulated financial agents, who could of- • Direct public interventions should be imple- fer financial tools in an easily accessible manner, mented sparingly and be carefully designed to including in hard-to-reach neighborhoods and avoid distorting incentives and undermining risk rural areas. management efforts of the financial firms and their clients. In some cases, subsidies and public Provide the right incentives, and heighten guarantees could be useful to encourage take-up confidence in financial institutions. of private insurance, thereby shifting some of the • The state should provide an enabling environ- government’s contingent liabilities to private in- ment for market development by including a legal surers: possibilities include agricultural, health, or framework for electronic payments; by requiring pension insurance (see box 6.3). the introduction of simple, low-cost bank accounts • The state should implement well-designed and for vulnerable populations such as the poor and well-targeted financial education programs to the young; and by allowing banks, nonbank finan- ensure the cost-effectiveness of its intervention cial institutions, and electronic payment providers and avoid unintended results. To overcome these to compete against each other in similar market The role of the financial system in managing risk 203 segments where appropriate. For example, the from benefiting from the advantages of a saving ac- Philippines allowed Mobile Network Operators to count (such as forgoing interest caused by delays in take on many banking operations.21 depositing money). Similarly, flexible loan sched- ules that can be readily renegotiated or forborne in • Regulatory reforms should focus on setting up an “hungry months” and prepaid when extra liquidity effective consumer protection framework that in- arrives are very useful to the poor. Finally, provid- cludes proper enforcement and dispute resolution ing some structure helps sustain self-discipline and mechanisms, such as a financial ombudsman (both commitment: for instance, through planned sav- Mexico and South Africa have established finan- ings and loan repayments schedules, supported cial ombudsmen to resolve disputes in consumer by visits from microfinance workers. Microfinance finance).22 A key goal is to instill trust in financial institutions and some mainstream financial in- institutions, including through adequate insur- stitutions could learn many useful lessons for ex- ance of retail deposits and improved quality of panding their client base and searching for new microprudential supervision. Microprudential and and sustainable business opportunities by looking business conduct regulation should cover—using at the successful programs run by nongovernmen- differentiated supervisory regimes—nonbank de- tal organizations to improve financial inclusion of posit-taking financial firms, such as saving houses, the poor and the extreme poor.26 co-ops, and credit unions; nonfinancial firms that provide credit at the point of sale; microcredit and When financial inclusion works, it promotes de- microinsurance companies; and payment and re- velopment and helps alleviate poverty. But when fi- mittance services providers.23 nancial inclusion is excessive or risks in the financial system are mismanaged, financial crises can erupt Build in information gathering and learning from with large costs to entire societies (cartoon 6.1). The impact evaluations of reforms. origins, impacts, and ways to avoid or manage finan- cial crises are discussed next. • The state should also develop data collection frame- works to continually assess gaps in financial inclu- sion and monitor and evaluate reforms. An example is Mexico’s strategy of comprehensive data collec- tion to better understand all challenges in access to finance; the data inform policy decisions, influence the business models of providers, and monitor prog- ress. In 2011, the National Households Survey of Financial Services Usage was launched—and is to be repeated every three years—to understand household motivation for using financial services, as well as barriers to greater usage.24 Enable the poor to break out of poverty traps by offering financial tools fitted to their needs. • When developing financial tools to help the poor, policy makers and microfinanciers should keep in mind the elements of reliability, convenience, flex- ibility, and structure. The great challenge of living on $2 a day is that even those $2 do not always come.25 So second best to having reliable income is having reliable financial partners and portfo- lios. Convenience and flexibility are also impor- tant because the poor need to be able to deposit and withdraw savings and take out and repay loans frequently, close to home, and without obstacles. They need flexibility in building long-term savings C A R TO O N 6. 1 Finance can help but also hurt. so that short-term difficulties do not prevent them © Matt Cartoon, The Daily Telegraph, January 18, 2009 204 WORLD DEVELOPMENT REPORT 2014 Financial crises hurt people: How can they economies. In countries in which the banking crisis be prevented? was accompanied by a sharp currency devaluation (Hungary, Ukraine), the accumulation of foreign Financial crises hurt people directly and currency debt (euros, Swiss francs) by some house- indirectly holds before the crisis may have amplified the im- Banking crises can affect people’s wealth, human pact of income shocks.30 capital, income, health, and even safety. By one Access to and the use of formal financial tools measure, the average loss of output during banking can also help people cope better with the impact of crises in the past four decades has been substantial financial crises. In the 2008 banking crises, among in both advanced countries (32.9 percent of GDP) households in Europe and Central Asia suffering in- and emerging economies (26 percent).27 The average come shocks, those that did not have a bank account loss was much smaller in low-income countries (1.6 or access to bank credit used costly coping strategies percent of real GDP), most likely because the pen- much more often (by 14 to 16 percent) than house- etration of financial services is low. Europe and Cen- holds that had such tools. tral Asia were especially affected by the 2008 wave of In banking crises, income redistribution effects banking crises; about 62 percent of households in the between the wealthy and the poor can be large. For region suffered a negative income shock mainly as example, wealthy investors tend to be better informed a result of wage reduction (job loss, lower wage, or and are able to liquidate their position first and limit lower remittances). The costly coping strategies de- their losses. Further, wealthy individuals tend to re- ployed by households to cope with this income shock ceive favorable treatment or evade controls imposed included cuts in basic consumption, health care, and during crises. Moreover, large financial transfers and education.28 opportunities for arbitrage emerge during crises, Financial shocks such as banking crises hurt which allow investors with deep pockets to purchase people through four channels: the financial system, assets at deep discounts and make large profits. labor markets, product markets, and social services. Evidence from 147 banking crises in 116 countries How does systemic risk turn into a from 1970 to 2011 suggests that the impact of bank- financial crisis? ing crises is transmitted to households most strongly through the labor market channel.29 The impact Systemic risk builds up over the financial cycle. The through the credit market seems to be less impor- financial system is naturally procyclical. Procycli- tant. Only when banking crises coincide with cur- cality can originate from the behavior of financial rency crises (large local currency depreciations) are intermediaries or from the procyclicality of the real large changes in relative prices transmitted through economy.31 Financial cycles are strongly related to the product market channel, where they affect urban business cycles. Notably, collapsing housing prices households more than rural households. As for the coincide with longer and deeper recessions, while social services channel, evidence from the Russian rapid growth in credit and increases in housing Federation suggests that the public sector cushioned prices coincide with stronger recoveries. Moreover, rather than aggravated the impact of the 1998 bank- financial cycles are highly synchronized within a ing, currency, and sovereign debt crisis. country (credit and housing price cycles) and across Macroeconomic policy, the structure and infra- countries (credit and equity cycles). Not only do the structure of the financial sector, and the design of financial and real cycles move together but banking formal safety nets play critical roles in amplifying crises can spill over to macroeconomic (sovereign or mitigating the propagation of financial crises to debt or currency) crises, such as in Malaysia in 1997, people. Household characteristics and microeco- or be triggered by a spillover from macroeconomic nomic systems also play a crucial role. Evidence crises, such as in Russia in 1998. indicates that the most important mitigators of Banking crises in Colombia (1982), Thailand income shocks transmitted to households through (1997), and Ukraine (2008) were preceded by exces- the labor market channel are diversified household sive credit growth of 40 percent, 25 percent, and 70 income, access to informal credit, and the buildup percent a year, respectively.32 Accordingly, a major of a stock of durable consumption goods. Access to concern for all countries is provision of the right informal credit was an important microeconomic amount of “equilibrium” credit: not too much and mitigator of the impact of the 2008 crisis in emerg- not too little. International standard setters have ing and developing economies but not in advanced proposed estimating equilibrium credit as the trend The role of the financial system in managing risk 205 in the credit-to-GDP ratio, obtained through statis- only later when countries reach higher income lev- tical filtering. Although such an approach can seem els. Lending concentration can also arise because of simple and transparent, its purely statistical nature an underdeveloped financial infrastructure, related- disregards fundamental changes in equilibrium party lending, pyramid ownership schemes, and credit caused by economic and financial develop- overall lack of competition in the financial system.34 ments. To strike a better balance between financial development and stability, the academic literature Interconnectedness can transform the risk from a single has proposed structural frameworks to account for institution (idiosyncratic risk) into systemic risk. On the impact of development on equilibrium credit.33 the one hand, interconnectedness of financial institu- Banking systems in developing countries can tions can have a positive impact on financial devel- incur additional exposures to systemic risk. Large opment because it promotes greater completeness of investment needs alongside short-term and often financial markets and better distribution of financial small domestic savings add to systemic risk because risks in normal times, including through innovative banks can finance only short term (generally 6 to financial instruments, or derivatives. On the other 12 months) compared with the investment financ- hand, the interconnected balance sheets of financial ing needs of the real economy (typically, 7 to 8 years, firms, through their participation in joint financial with a 2-year grace period). Using foreign savings to infrastructure, can spread a shock throughout the obtain longer-term financing can expose the banking national and even international financial system and sector to aggregate refinancing risks, and the banks sometimes amplify those shocks.35 Adverse shocks or the borrowers to foreign currency mismatches can originate from problems in one systemic institu- between their assets and liabilities (figure 6.4). An- tion or from exposure of many financial firms to a other, often neglected, systemic risk for small devel- single asset class, such as commercial real estate. The oping economies comes from lending that is con- interconnectedness and common exposures of the centrated in a small number of borrowers or sectors U.S and European banks transmitted the 2007 U.S. of the real economy. In early stages of development, subprime mortgage crisis first to Europe, and then, countries initially specialize as they open to foreign through the links between European parent banks trade, which naturally concentrates lending in fewer and their subsidiaries and branches, to emerging Eu- economic sectors. The economic structure diversifies rope. Two aspects of interconnectedness are especially F I G U R E 6.4 Banks’ aggregate refinancing risk and foreign currency mismatches can increase systemic risk in developing countries a. Europe and Central Asia’s unusually heavy b. Foreign currency mismatches on bank balance reliance on foreign savings sheets around the world 200 80 FX deposits in excess of FX loans, 180 60 Loan-to-deposit ratio, % 160 140 40 % of total loans 120 Maximum 20 100 80 0 Mean 60 –20 40 –40 20 0 –60 Minimum tin al A and be nd cifi ia ia r d ric n ric d ro ing Af ara Af t an a La cifi d ica S Pa s As Af t an ric th Am sia an So c a CI d st A rib a a Pa a an Su ica a c pe Eu erg er nt pe h h Af rth as Ca ic Sa ut rth as Am an a i Ce uro e er No e E Em As E b- No le E r dl tin E d id id M M La Source: WDR 2014 team based on data from World Bank FinStats (internal database) (panel a) and Chitu 2012 (panel b). Note: Data for panel a are as of end-2008. Regions in panel b are grouped as in Chitu 2012, with data as of end-2006. CIS = Commonwealth of Independent States. FX = foreign currency. 206 WORLD DEVELOPMENT REPORT 2014 important for developing countries: banks’ ties to the withdrawing liquid funds from Central and Eastern shadow banking system, and cross-border banking. Europe in 2009.40 In developing countries, the shadow banking sec- tor comprises financial firms focused on providing Possibility of contagion can further amplify systemic alternative sources of financing to the economy.36 risk. Contagion typically relates to the breakdown in These firms include leasing and factoring companies, confidence as systemic risk materializes; contagion credit unions, cooperatives, microfinance companies, can cause runs on bank deposits, freezes of money and pawn shops. In Thailand, the sector covers nearly and asset markets, or both. A recent example is the 40 percent of the financial system, while in Bulgaria, collapse of Lehman Brothers in 2008, which shat- its share grew rapidly by 14 percentage points from tered confidence in money market mutual funds. 2003 to 2010. The concerns about shadow banking Four days after Lehman’s bankruptcy, the U.S. gov- relate to regulatory arbitrage (lending through firms ernment was forced to announce guarantees for the with the lowest capital requirements), mostly among entire sector. In developing countries, contagion banks and nonbank credit institutions. Developing risk relates mainly to depositor confidence. Because countries need to ensure that shadow banks help pro- banks finance their long-term, illiquid assets with de- vide alternative but safe financial services, without mandable debt in the form of first-come, first-serve generating unacceptable systemic risks. In addition, deposits, depositors can consider them inherently other modes of finance such as Islamic banking (and unstable. Bank runs can occur when depositors fear insurance) have grown into systemic importance in others will withdraw before they do, leaving nothing some countries (such as Saudi Arabia, United Arab for them. Important contagion effects for develop- Emirates, and Malaysia), even though they account ing countries can also arise in the context of cross- for only about 1.5 percent of global banking assets, border banking.41 or $0.9 trillion in 2011. While Islamic banks could be less cost-effective compared with conventional Where did financial firms and past public banks, they tend to be better capitalized, have higher policies fail the most? asset quality, and be less likely to disintermediate dur- ing crises.37 Recently, bad corporate governance, distorted pri- Financial globalization, involving cross-border vate incentives, short-term horizons for profit maxi- activities of banks, has been accompanied by many mization, and coordination failures have resulted benefits, including protecting domestic economies in excessive risk taking.42 Financial firms have been from domestic shocks. From 2002 to 2012, the funds largely unsuccessful in implementing good corpo- provided by banks from advanced economies to rate governance, so the prevailing perverse incen- banks in developing countries increased from about tives, including bad compensation policies, led bank- $0.4 trillion to $1.7 trillion, translating into an av- ers (from managers to loan officers) to maximize erage inflow of $130 billion a year.38 By 2012, the short-term profits and disregard prudent risk. Bank global systemically important financial institutions managers generally lack adequate personal respon- (G-SIFIs) had 71 systemically important subsidiaries sibility for taking too much financial risk and are or branches (local SIFIs) in 43 developing countries. not held legally accountable for their bad practices. Brazil, Mexico, Peru, and Uruguay have the great- Further, financial firms, including SIFIs, ignored est systemic links to G-SIFIs, with six, four, four, their own contributions (negative externalities) to and four local SIFIs linked to G-SIFIs, respectively.39 systemic risk, and market discipline failed to enforce This growing exposure to advanced economies can consideration of these externalities. Transparency also pose potential dangers for financial stability, in- and clear disclosure of information are important to cluding exposing the domestic economy to foreign achieve proper incentives, but for the most part the shocks. For instance, in response to a negative shock financial system has not implemented these mecha- in the home country, foreign banks may decide to nisms. In this environment, investors (bond hold- curtail lending or withdraw from the host country ers and equity holders) failed to perform their basic entirely. Many countries have managed such risk by monitoring and disciplining functions to correct the requiring foreign banks to operate through an incor- incentives of financial firms. porated subsidiary with its own capital, rather than Some public policies have distorted private sector through a branch. Regional policy efforts to manage incentives for managing risk responsibly, have lacked cross-border banking risk include the Vienna Initia- a systematic approach, and have aided moral hazard tive, which prevented regional banking groups from behavior. Regulatory failures raise questions about The role of the financial system in managing risk 207 the appropriate level of government involvement in gate level. The actions of individual financial firms the operation of the financial system. Supervisors can generate negative externalities that can allow have failed to measure banks’ risks accurately or to systemic risk to build up. Moreover, monetary and set and enforce sufficient capital requirements for fiscal policies can be ineffective in managing systemic banks to be able to absorb unexpected losses reliably. risks in the financial system, especially in developing Supervisors have also failed to design and enforce countries. timely resolution of failing banks, which would limit Central banks (as in the Czech Republic, South the exposure of taxpayers to problem SIFIs.43 The Africa, and Thailand) seem to be best equipped to ineffective resolution frameworks for SIFIs have led assume the responsibility for macroprudential pol- to expectations of government bailouts. The SIFIs icy.46 First, they have an advantage in monitoring have thus tended to privatize their profits and social- macroeconomic developments. Second, central- ize their losses. Moreover, various public guarantees izing macroprudential supervision in the central and subsidies (implicit and explicit, including for bank improves coordination of crisis management lending to households) have distorted the incentives activities, especially if the central bank is also the for risk management of both banks and their clients. banking sector regulator. Third, monetary policy Most recently, the regulatory uncertainty caused by decisions undertaken by the central bank have po- the failure of governments in several developed tential implications for financial leverage (debt countries to promptly decide on, coordi- load) and risk taking. As an emerging best nate, and implement financial sector Making practice, implementation of macro- reforms held back operations of macroprudential prudential policy is being conducted the financial system and the re- by macroprudential policy com- regulators independent covery of the real economy. Too mittees—an analog to monetary much has been expected of gov- and giving them policy committees (for example, ernment regulation and super- adequate policy tools are the macroprudential committee visory capacity, in many cases. the basics for successful of the Bank of England).47 More selective policy interven- management of tions, minimizing unintended con- Choose the right indicators of systemic sequences, would be more appropri- systemic risk. risk. To assess and monitor systemic ate, in some instances. risk, the macroprudential supervisor Moreover, some public policies regu- uses analytical tools, such as stress tests, early lating systemic risk have been subject to capture warning models, and assessments of systemic im- by the financial industry.44 The observations from portance. Systemic risk assessment and monitoring the 2008 global financial crisis suggest significant need to be forward-looking, timely, and presented influence of the industry lobby on the supervision in a user-friendly way to ensure that policy makers of systemic risk, resolution of the crisis, and future act on the information received. Macroprudential regulatory reforms. The enduring challenge is to stress tests are “what if” scenario exercises to assess create mechanisms that can negate the “grabbing the resilience of the system as a whole to extreme hand” of the financial industry and politicians, while but plausible shocks.48 Early warning models and creating strong incentives for official agencies to im- assessments of systemic importance are less com- prove social welfare.45 Making regulators politically mon in developing countries that are still working and financially independent is the first step in this on building and using practical approaches to stress direction. testing. From the points of view of practicality and accountability, monitoring a selected set of simple and robust financial indicators could be preferable to What are the best-practice policies for a more complex approach involving composite in- managing systemic risk and banking crises? dicators or outputs from complex models.49 Central Pursue macroprudential policy. Macroprudential banks often publish these systemic risk assessments policy seeks to foster financial stability by managing as part of their financial stability reports to alert systemic risk and keeping it at a socially acceptable market participants, inform the public, and increase level. Such policy is needed because policy measures accountability of the macroprudential supervisors.50 focusing on the financial stability of financial insti- tutions and their actions at the individual level are Calibrate macroprudential tools to the specifics of the insufficient to foster financial stability at the aggre- country. To manage systemic risk, macroprudential 208 WORLD DEVELOPMENT REPORT 2014 TA B L E 6.1 A taxonomy of macroprudential tools Selected measures Main characteristics Country examples Aimed at borrowers Loan-to-value caps Reduces vulnerability arising from highly geared Brazil; Bulgaria; Canada; Chile; China; Colombia; borrowing Croatia; France; Hong Kong SAR, China; Hungary; India; Italy; Korea, Rep.; Malaysia; Mexico; Norway; Philippines; Poland; Romania; Singapore; Spain; Sweden; Thailand; Turkey Debt-to-income caps Reduces vulnerability arising from highly geared China; Colombia; Hong Kong SAR, China; Korea, borrowing Rep.; Poland; Romania; Serbia Aimed at financial institutions (addressing the asset side) Credit growth caps Reduces credit growth directly China, Colombia, Malaysia, Nigeria, Serbia, Singapore Foreign currency lending limits Reduces vulnerability to foreign exchange risks; Argentina, Austria, Brazil, Hungary, Poland, reduces credit growth directly Romania, Serbia, Turkey Aimed at financial institutions (addressing the liabilities side) Reserve requirements Reduces vulnerability to funding risks; reduces Brazil, Bulgaria, China, Colombia, Russian credit growth indirectly Federation Aimed at financial institutions (addressing bank buffers) Dynamic loan-loss provisioning Increases resilience and reduces credit growth Brazil, Bulgaria, Colombia, India, Mongolia, Peru, indirectly Russian Federation, Spain, Uruguay Countercyclical capital requirements Increases resilience and reduces credit growth Brazil, India indirectly Profit distribution restrictions Limits dividend payments in good times to help Argentina, Colombia, Poland, Romania, Slovak build up capital buffers in bad times Republic, Turkey Source: WDR 2014 team based on Claessens, Ghosh, and Mihet, forthcoming. Note: Countries listed in the table adopted corresponding macroprudential tools in various years from 2000 to 2010, some of them temporarily. supervisors use policy tools such as variable capital deleveraging in severe downturns. In any case, the buffers and dynamic provisioning, as well as caps on use of macroprudential tools needs to be calibrated leverage, credit growth, and the debt-to-income ra- to the specifics of a given country.54 tio (table 6.1).51 The use of macroprudential policy tools has been increasing, particularly in managing Focus on crisis preparedness for effective management systemic risk in the financial sector. Macropruden- of future financial crises. In crises, policy makers face tial tools, such as capital controls, could also be used deep uncertainty about market conditions. They more broadly in the context of macroeconomic man- must be able to mobilize expertise to decide and act, agement (chapter 7). Interestingly, emerging mar- transparently deploy legislated crisis management kets have been three to four times more likely to use tools, communicate with the public to contain un- macroprudential tools than advanced economies.52 certainty, and ensure adequate loss sharing to avoid For instance, in 2011, the Republic of Korea imposed moral hazard going forward (box 6.5). The resolu- a levy of up to 0.2 percent on bank noncore financial tion of banking crises will always be country specific liabilities to manage speculative inflows of foreign because of differences in legal framework, but reso- capital.53 Some macroprudential tools are intended lution should not compromise the bottom line of to mitigate externalities that occur in the upturn of minimizing the fiscal cost and avoiding moral hazard the financial cycle, while others are deployed to build in the future. Concerns of widespread liquidity runs buffers to mitigate any bust. For example, caps on on banks usually mean that blanket guarantees are debt-to-income and loan-to-value ratios could be given to all bank creditors. While liquidity assistance effective in reducing risk exposures in booms, while to banks needs to be provided early on, open-ended countercyclical buffers, such as additional capital and liquidity support has proven to prolong crises and reserve requirements, could help mitigate excessive could result in future macroeconomic risks (chap- The role of the financial system in managing risk 209 B OX 6.5 Preparing for a banking crisis with crisis simulation exercises To test crisis preparedness and practice using existing or proposed • Most participants—often the nation’s top decision makers—take arrangements for crisis management, the World Bank, since 2009, these exercises quite seriously, typically spending one or more has been encouraging financial policy makers to participate in finan- days fully concentrating on them. Their actions tend to be condi- cial crisis simulation exercises. During the exercise, participants tioned by whether they perceive the exercise as a (perhaps receive a stream of (generally bad) news describing the “scenario” imposed) “test” or as a (freely requested) “drill.” Ownership of the they must deal with and the tools provided by their (real or assumed) exercise by the participating public representatives is thus crucial legal, regulatory, and operational frameworks. This news arrives in for the success of the exercise. two forms: as “public information,” understood to be simultaneously • Public representatives frequently overestimate parent bank or available to all participant teams (including financial sector supervi- shareholder capacity and willingness to provide support. This sory authorities, the central bank, the finance ministry, and the overestimation then typically leads to public agencies’ inaction deposit insurer) and the market; and as “private information” from before and during the crisis, and protracted coping and recovery several fictional characters (analysts, bank inspectors, advisers, bank- from the crisis. ers, journalists, foreign authorities, politicians). Participants must share their respective pieces of information and analyses to under- • Formal arrangements to share information and coordinate action tend to spring up voluntarily and ignore possible conflicts of stand the scenario and coordinate their actions. interest. Areas for improvement are identified through exhaustive analy- sis of the exchanges that take place among the participants (typi- • Simple information sharing is much more common than joint anal- cally running into the hundreds of written memos and e-mails), as ysis of all relevant information by all responsible parties. well as between them and the fictitious characters. • Governments tend to postpone decisions on resolution of prob- While comparisons and generalizations are difficult, given the lem banks—including restructuring, recapitalization, merger, clo- small sample size and highly localized conditions (reflected both in sure, or liquidation—by implicitly or explicitly nationalizing them. the institutional identity of the participating teams and the situa- tions proposed by the fictional scenario), it is nonetheless possible • Coordination of public communications is limited, if not entirely to offer tentative impressions about the adequacy of participant absent. responses to some common challenges and about the way partici- Some of the judgment and behavior observed in the exercises pants tend to underestimate and underreact, or overestimate and reveals cognitive and behavioral failures in policy-making decisions overreact, to bad news: and actions in the face of risk, as discussed in chapter 2. Source: Aquiles A. Almansi for the WDR 2014. ter 7).55 Emergency liquidity assistance from the cen- resources to preparing such frameworks in normal tral bank should be provided only to solvent banks.56 times because crises are not likely to go away. Insolvent banks should be closed transparently to avoid moral hazard in the future. Prompt interven- Resolve failures of systematically important financial tions can reduce costs and improve efficiency.57 institutions fairly and effectively. Failing SIFIs must be resolved promptly in the view of the trade-off be- Seek private sector solutions to pass bank losses to ex- tween minimizing negative spillovers to the rest of isting shareholders, managers, and in some cases un- the financial system and minimizing future moral insured creditors first. For systemwide crises, finding hazard, while protecting taxpayers’ money (box 6.6). domestic private sector solutions could be difficult, To improve the resilience of global and national SIFIs, and reliance on foreign takeovers or government- recent proposals recommend that SIFIs hold more assisted mergers may be needed. The fiscal cost of capital and other instruments that can promptly in- banking crises averaged almost 7 percent of GDP crease their capacity to absorb losses and mitigate the during 1970–2011 (4 and 10 percent of GDP in ad- possibility of negative spillovers to the rest of the fi- vanced and developing countries, respectively). The nancial system in advance.59 To further enhance cri- two costliest banking crises occurred in Indonesia sis preparation, SIFIs should be required to prepare (1997) and Argentina (1980), with fiscal costs reach- so-called “living wills” to assist the management and ing 57 and 55 percent of GDP, respectively.58 Overall, the authorities in prompt resolution of failing SIFIs, government interventions and assistance in manag- including through partitioning and sales. Resolution ing banking crises need to be based on a sound legal could also involve the injection of public capital or framework to avoid ad hoc interventions outside the other government support, provided that sufficient existing legal framework, which can have large redis- fiscal space exists for such contingent liabilities, that tribution effects. Countries must devote time and costs to taxpayers are minimized, and that large re- 210 WORLD DEVELOPMENT REPORT 2014 B OX 6.6 Financial bailouts: “Too big to fail” versus moral hazard Domestic systemic banks are banks whose failure or severe prob- injecting public capital into failing systemic banks may be fiscally lems might generate significant negative externalities for the rest of unsustainable, counterproductive by increasing moral hazard, and the domestic financial system and the economy. While in many cases potentially inequitable by introducing large redistribution effects, banks of systemic importance can be identified ex ante using appro- benefiting shareholders or creditors of the failing bank at the priate assessment methodology, it is hard to assess ex ante which expense of taxpayers. Other resolution options, which are prefera- banks will not be systemically important in stressed market condi- ble in certain market conditions, include establishing a bridge bank tions. Thus if a bank is in trouble, policy makers first need to deter- so that the systemic part of failing bank is transferred to a new entity mine whether the bank is systemically important in current market owned by the ministry of finance and operated by the bank supervi- conditions, with a view to available legal resolution options and fiscal sor or resolution agency, with little or no permanent support of space, political economy factors, and uncertainty about possible public capital.b spillovers to the financial system and the real economy.a This box If public funds are used in systemic bank resolution, political focuses on systemic banks with insolvency problems after those economy constraints become an even more important factor banks have exhausted all possible insurance and protection mea- shaping the possibility frontier (panel a) because the legislature or sures arranged ex ante, such as capital buffers, bail-in or contingent the ministry of finance that provides public capital get involved in debt, or sale of assets. the decision making. The decision about the resolution method, Available resolution options are an important factor shaping the timing of the intervention, and its particular execution (placing possibility frontier of resolving systemic banks (panel a). Market government deposits in the bank, nationalizing it, or using the solutions are preferable at all times and may include indirect sup- bridge bank) will be influenced by politicians and the political port (mergers forced by the supervisor) or direct support from the lobby.c For systemic banks, the least-cost solution must consider government (purchase and assumption, with credit enhancement the cost to society rather than the deposit insurance fund (bank from the deposit insurance fund). However, if the market is small or resolution agency), given the likely spillover of the problem to the in distress, private sector solutions may not be available. In contrast, real economy. The possibility frontier of systemic bank resolution a. Main resolution options b. Role of deep uncertainty Risk of spillovers Risk of spillovers Mergers and acquisitions Purchase and assumption Policy maker’s risk tolerance Bridge bank Extra Open bank assistance social risk Risk of moral hazard Risk of moral hazard Source: WDR 2014 team based on Beck 2011. In times of financial stress, policy makers face the additional using public money then reinforce moral hazard and result in larger challenge of deep uncertainty about negative spillovers that can be redistribution effects, to the detriment of market discipline and tax- triggered by the closure of a systemic bank. They typically have payers (shaded area depicting extra social risk, panel b). some idea about the trade-offs, thanks to prior systemic risk and Resolution of Turkey’s 2001 systemic banking crisis through a resolvability assessments (solid line, panel b). However, if such public recapitalization program provides some elements of good assessments are not being performed, policy makers can underesti- practice in the presence of a too-important-to-fail and too-many- mate the real spillover potential, represented by the upper dashed to-fail problem. The Turkish crisis started from mounting weak- line. To complicate matters, financial firms do not fully share all pri- nesses in the banking system and collapse in investor confidence vate information. In times of stress, they can use this information after the fall of a medium-sized bank. A successful design of the asymmetry to their advantage to lobby for higher public support 2002 public recapitalization program prevented misuse of public than would be appropriate. This private information, if revealed, funds through strict eligibility criteria, an in-depth audit to transpar- could show that the spillover potential is much smaller in reality, as ently disclose capital shortfalls, and the mandatory participation of depicted by the lower dashed line. Unnecessarily large bailouts bank owners in recapitalization. Between 1997 and 2004, 21 banks (continued) The role of the financial system in managing risk 211 B OX 6.6 Financial bailouts: “Too big to fail” versus moral hazard (continued) representing about 20 percent of banking sector assets were trans- systemic importance in any market conditions, and preparing sys- ferred to the bank resolution agency, and all but one bank were temic risk assessment approaches to determine systemic impor- resolved through mergers, sales, and liquidation by 2004. The tance of banks in specific market conditions should be essential efforts of Turkish authorities to minimize the fiscal costs of the parts of any crisis preparedness efforts. In particular, recovery plans restructuring program and future moral hazard have contributed to can help increase the resilience of systemic banks and their ability greater self-reliance and self-discipline on the part of the Turkish to recover from stresses, thus indirectly enhancing overall financial banking sector, including improvements in bank risk management. stability.e To address political economy issues, decisions about This in turn underpinned the resilience of the Turkish banking sector troubled systemic banks should be broad-based, and involve the to the spillovers from the 2008 global financial crisis.d banking supervisor, resolution agency, central bank (the financial Systemic banks are likely to experience troubles in the future, stability supervisor), and ministry of finance. The right platform for especially if system-wide financial stresses emerge. Governments such decisions can be the financial stability committee (table 6.3), should therefore be well prepared to resolve troubled systemic in which all these agencies typically participate. In addition, if the banks while minimizing moral hazard and redistribution effects. impact of resolving a domestic systemic bank crosses national bor- Developing a legal framework for resolving systemic banks, pre- ders, such as the case of the Icelandic bank Kaupthing, the resolu- paring recovery and resolution plans (living wills) for banks of tion will require cross-country or regional coordination.f Source: WDR 2014 team. a. BIS 2012. b. Beck 2011. c. Brown and Dinc 2005. d. Josefsson 2006. e. BIS 2012. f. BIS 2010b. distribution effects are avoided. Another option is Resolving the tension between financial to close the SIFI and transfer its systemically impor- development and financial stability tant part to a temporary bridge bank, owned, man- aged, and then sold in a timely manner by public Important complementarities and trade-offs exist authorities.60 between boosting financial inclusion and fostering financial stability (cartoon 6.2). This section focuses Reduce regulatory uncertainty. Regulatory uncertainty on these complementarities and trade-offs, as well as can paralyze recovery from a crisis. Government in- on financial sector development and stability more terventions to manage banking crises could have large generally. repercussions for the government fiscal position and redistribution effects from taxpayers to creditors and Financial inclusion can aid stability shareholders (consider the Euro Area crisis resolution, for instance).61 For this reason, the government could Greater financial inclusion can improve the efficiency be forced to reset its tax policy and reform financial and stability of financial intermediation by making regulation as it learns about the causes of the crisis. greater and more diversified domestic savings avail- This process could involve many stakeholders and able to banks. As a result, a country’s banking system might need to be coordinated at the regional or inter- can ease its reliance on reversible foreign capital and national level. As a result, regulatory reforms could be thereby enhance its stability. Indeed, preliminary evi- protracted and their outcomes very uncertain. Banks dence suggests that a broader use of bank saving de- uncertain about how much capital and liquidity they posits made the banking systems of middle-income will need to hold will curtail their lending. Investors countries more resilient to deposit withdrawals and will hold back their projects because once they take the slowdown in deposit growth during the 2008 cri- into account the uncertainty about future taxes and sis (figure 6.5a). Similarly, the performance of loan the cost of finance, most projects will become finan- portfolios of Chilean banks suggests that aggregated cially unviable. Thus tax and regulatory reforms in re- losses on small loans present less systemic risk than sponse to crises need to be timely and decisive to ease large, infrequent, but also less predictable losses on recovery. For that to happen, improved coordination large loans.62 Thus greater financial inclusion and di- at the national, regional, and international level needs versified credit allocation may coincide with greater to be established and put into practice. stability of individual financial firms and of the entire 212 WORLD DEVELOPMENT REPORT 2014 C A R TO O N 6 .2 Trade-offs in financial sector policy pose a challenge. © John Cole/The (Scranton) Times-Tribune system. Greater financial inclusion can also enhance than $170 billion in 2008, people with low finan- financial stability indirectly by providing households cial literacy underestimated the increased burden of (and firms) with access to savings, credit, and insur- debt-servicing costs in bad times, which significantly ance tools that can bolster resilience and stability of impaired their spending capacity.65 Preliminary evi- the real economy and thus the financial system that dence suggests that excessive credit growth can im- serves it. pose heavy financial burdens on people when market If financial inclusion can enhance financial stabil- conditions deteriorate (figure 6.5b). Households that ity, can exclusion from formal financial services lead purchase the “wrong” financial tools that add to their to greater instability? Households (and small firms) risk, whether a result of their own irresponsible risk in countries with high levels of financial exclusion taking or irresponsible delivery of financial services must rely on informal financial services that can be by financial firms, jeopardize their own financial sta- poor substitutes for formal services.63 In extreme bility—and collectively, possibly the stability of the cases, informal services can increase people’s risk ex- financial system. Such risk exposures at the micro posure to shocks and be a source of instability them- level can be mitigated by an adequate level of finan- selves. For example, pyramid schemes organized as cial education and consumer protection. Financial informal savings and investment opportunities have tools with a risk profile matching that of the clients been known to trigger both political and social un- can improve outcomes in financial markets. rest and lack of confidence in the banking system.64 Stability is also impaired if the system tries to Stability is endangered when financial do more than its development permits inclusion is excessive There appears to be a limit on how much and what Inclusion of everybody in each and every financial services the financial system can provide to whom service cannot be the social objective. The U.S. sub- at a given stage of its development. This limit (a prime crisis showed that subsidized, excessive access financial-possibility frontier) is affected by many to credit, combined with tolerated predatory lending, development factors driving the provision of finan- is bad policy. Similarly, in Russia, where consumer cial services on the supply side (financial system) loans grew from about $10 billion in 2003 to more and constraining participation on the demand side The role of the financial system in managing risk 213 Prudent financial inclusion can enhance financial stability but, if F I G U R E 6.5 excessive, can weaken it a. Greater access to bank deposits can b. Excessive consumer credit can result in aid financial stability excessive indebtedness y = –0.4327 × x y = 0.2649 × x y = vulnerability of bank deposit base, 2007–10, % 50 45 SYR GAB R 2 = 0.3544 R 2 = 0.1312 40 40 GRC y = household arrears on loan, 2010, % 30 BGR LSO 35 20 PHL BWA 30 ROM 10 SYC AZE PNGGEO ISL VEN URY 25 0 UZB AGO KAZ 20 LVA –10 YEM BOL ALB IND HUN MKD BGR ARM COG TUR LKA 15 –20 ESP SVK PAK ROM CPV ITA AUT POL LTU CRI CHN IDN 10 EST –30 BTN ARG MYS NOR EGY ZAF THA PRT SVN CZE MEX TUN CHL 5 BEL –40 RUS DNK GRD –50 0 0 10 20 30 40 50 60 70 0 20 40 60 80 100 x = access to bank deposits, 2005, % x = annual growth in consumer credit, 2004–07, % Source: WDR 2014 team based on data from Han and Melecky 2013 for the WDR 2014 (panel a); European Credit Research Institute Lending to Households in Europe (database), and European Union Statistics on Income and Living Conditions (EU-SILC) Survey (panel b). Note: The solid (regression) lines in the figures depict the fitted linear relationships between the y- and x-axis variables. For panel a, the vulnerability of the bank deposit base is conditional on per capita income, bank z-score, occurrence of a banking crisis, and imple- mented explicit deposit insurance. For measurement of access to bank deposits, the composite index of access to financial services by Honohan 2008 was used. Regression results are available upon request. All middle-income countries for which data are available are included. For panel b, the countries are the EU-27 (excluding Ireland and Cyprus) plus Norway and Iceland. The arrears are on con- sumer loan repayments. The results hold if arrears are replaced by self-reported financial burden, or if the annual growth in consumer credit in 2004–07 is controlled for the size of subsequent bust in consumer credit in 2008–10. (individuals and firms).66 Following the concept In such environments, financial innovation—which of financial-possibility frontiers, countries can face can promote financial deepening and inclusion in three broad challenges. First, the frontier of a given other contexts—could pose a challenge for financial country may be low relative to its level of economic stability, especially if it becomes self-interested and development because of deficient structural factors unnecessarily complex. (such as low population density or a high degree Bank competition can be beneficial if it improves of economic informality) or nonstructural factors financial inclusion, deepens financial markets, and (such as inadequate contract enforcement or pro- generates useful innovative services at the accept- tection of property rights, or macroeconomic sta- able level of systemic risk. More intense competition bility). Second, a country’s financial system can be among banks can have positive effects on financial below its frontier because of demand constraints depth, income distribution, growth, and efficiency. (such as self-exclusion stemming from low finan- At the same time, it can also negatively affect the cial literacy and trust in banks) or supply-side con- stability of the banking system; with more pressure straints (such as lack of competition due to poor on profits, bankers have incentives to take excessive financial infrastructure or regulatory restrictions on risks.67 However, competitive lending rates reduce new products). Third, a country’s financial system entrepreneurs’ cost of borrowing and increase the can move beyond the frontier, expanding unsustain- success rate of entrepreneurs’ investments. Banks, ably through excessive investment and risk taking by in turn, experience lower default rates on their loan market participants in environments of weak super- portfolio, and the banking system as a whole enjoys vision, corporate governance, and market discipline. greater stability. The role of regulatory frameworks 214 WORLD DEVELOPMENT REPORT 2014 could be critical in shaping the tension between velopments in the sector, the use of quantifiable data bank competition and financial stability. Recent evi- to specify their forward-looking objectives is weak. dence, more attentive to systemic risk measurement, Instead of choices (which would lead to a discussion confirms that greater bank competition can be as- of trade-offs), conventional strategies tend to focus sociated with greater financial stability.68 on issues.69 Governments should instead adopt an approach that explicitly addresses policy trade-offs and begins by recognizing that the government must Policy, to succeed, must consider trade-offs make choices and that each choice has consequences. and synergies in finance The national financial sector strategy should At the level of the national government, the national clearly assign implementation of the targeted finan- financial sector strategy formulates the pol- cial development at the (identified) accept- icy for the financial sector. A well- able level of systemic risk to individual formulated strategy should set devel- Financial policy must government agencies in accord with opment targets that take into ac- their mandate. For instance, the count the systemic risk involved consider the synergies ministry of finance (or economy) in achieving them and that com- and trade-offs between could be responsible for financial municate the systemic risk ap- promoting access to development, while the central petite (tolerance) of the country more and better financial bank could be responsible for in the financial area. Preliminary the supervision of systemic risk evidence from a survey of na- tools and controlling (as in Moldova). In their financial tional financial sector strategies systemic risk in the sector strategies, most countries indicates that most strategy docu- financial sector. broadly identify the implementing ments have a clear statement of in- government agencies based on their tent, but less than half have a quantifiable overall mandates (table 6.2). Countries indicator included in their objective statements less often clearly assign specific agency responsi- (table 6.2). Although most documents refer to sys- bility for implementing measures to achieve develop- temic risk in general terms, very few refer to specific ment goals or to manage systemic risk at acceptable measures of systemic risk. With a few exceptions, the levels. Financial sector strategies should not only in- strategy implementation plans do not discuss spe- clude such assignments in the implementation plan cific trade-offs between financial development goals but should also present a mechanism through which and the management of systemic risk, even though the implementation will be coordinated, such as a many countries commit to achieving both goals standing committee. within the same strategy document. While the strate- A financial policy committee with an effective gov- gies include a rich numerical analysis of recent de- ernance structure that includes major stakeholders in TA B L E 6 .2 National financial sector strategy documents rarely consider the trade-off between financial development and stability Development objectives Clear development goals set 94 Development goals quantified 42 Tools to achieve goals identified 58 Systemic risk Risk associated with achieving goals identified 94 Systemic risk quantified 6 Tools to manage systemic risk identified 53 Trade-off Trade-off in development and systemic risk is communicated 11 Implementation plan Agencies to execute the strategy identified 92 Agencies to implement development goals assigned 64 Agencies to manage systemic risk assigned 33 Source: WDR 2014 team based on Maimbo and Melecky 2013 for the WDR 2014. Note: The table summarizes the percentage of countries meeting each requirement in a sample of 36 countries, consisting of six countries in each of the six regions: Latin America and the Caribbean, Sub-Saharan Africa, East Asia and Pacific, Europe and Central Asia, Middle East and North Africa, and South Asia. The role of the financial system in managing risk 215 TA B L E 6.3 Composition of financial stability committees in selected developing countries Country Coordination body Chair Members India Financial Stability and Minister of finance CB, MOF, regulators for securities, insurance, Development Council and pension Indonesia Financial System Stability Minister of finance CB, MOF, regulators for deposit insurance Coordination Forum and financial services Mexico Financial Systemic Stability Minister of finance CB, MOF, regulators for securities and Council banking, insurance, pension and deposit insurance Poland Financial Stability Committee Minister of finance CB, MOF, and the regulator for financial services South Africa Financial Stability Oversight CB governor and minister CB, Treasury, and the regulator for financial Committee (interim) of finance services Turkey Financial Stability Committee Deputy prime minister CB, Treasury, regulators for banking, capital markets and deposit insurance Source: WDR 2014 team based on information from the International Monetary Fund, national central banks, and ministries of finance. Note: CB = central bank. MOF = ministry of finance. financial sector policy, can improve policy coordina- of new circumstances and newly identified gaps or tion and produce balanced policies. To set compati- policy tools concerning financial development and ble and sustainable policies, a group of policy makers stability. In improving regulatory frameworks and and experts that understands the trade-offs between adopting best practices in regulation, national pol- risk and development in the financial sector should icy makers are supported by international standard be established. Many countries have established fi- setters such as the Financial Stability Board, Basel nancial stability committees to manage systemic risk Committee on Bank Supervision, International As- and crises, notably in the aftermath of the 2008 crisis sociation of Insurance Supervisors, and Interna- (table 6.3). These committees are chaired by a high- tional Organization of Securities Commissions. The level public official and include major policy makers standard setters, apart from developing guidelines in the financial area. It would be practical to extend a for best-practice regulation, provide assistance in mandate to these high-level committees to prepare a building capacity, so that national regulators advance holistic, national financial sector strategy, including their knowledge and skills to further develop their with the participation of relevant experts from aca- national financial markets. At the regional level— demia and the financial industry. Intermediate solu- closer to implementing regulatory frameworks— tions also exist. For instance in Malaysia, the central global best-practice guidelines could be elaborated bank engages with major stakeholders in financial on or turned into rules by regional standard setters. sector policy, including the ministry of finance and At the level of the European Union, these standard private sector experts, to prepare a national finan- setters would include the European Systemic Risk cial sector strategy that takes into account tradeoffs between financial (inclusion) development and sys- Board, the European Banking Authority, European temic risk in the financial system. Securities and Markets Authority, and European In- In implementing the financial sector strategy, surance and Occupational Pensions Authority. some direct policy interventions (such as provision Implementation should focus on enforcing good of financial infrastructure) could be beneficial, while corporate governance to correct the incentives of fi- others (support of state-owned banks, for example) nancial firms and the financial system to take on ex- could be controversial. Most public policies should cessive risk or pursue too much or too little financial be indirect and focused on proper regulation, be- inclusion, and to ensure that private decision mak- cause direct policy interventions can lead to unin- ing is governed by a long-term view and attention tended consequences, including large redistribution to business sustainability. Enforcing good standards effects and distorted incentives in the private sector.70 of corporate governance pertains to both the devel- Coordination of policy implementation should also opment and offering of useful, accessible, and reliable be the task of the proposed financial policy commit- financial services and to responsible risk taking that tee, which can revise the strategy periodically in light accounts for systemic externalities. The key areas 216 WORLD DEVELOPMENT REPORT 2014 to improve bank corporate governance include the an important area of policy reforms. In this regard, following:71 promoting diverse business models, including that of Islamic banking, can increase the system’s diversity • Boards should incorporate a balance of expertise along with greater financial inclusion. Furthermore, to approve and monitor the overall business strat- promoting diversity beyond the banking sector could egy of the bank, considering its long-term financial be equally important and can involve insurance com- interests, exposure to risk, and ability to manage panies and nonbanking credit institutions. risk effectively. Overall, public policy should encourage diversifi- • Senior management should ensure that bank ac- cation of financial intermediation away from banks tivities are consistent with the bank’s business into capital markets to enhance the stability of the strategy, risk appetite (tolerance), and policies ap- financial system. Recent evidence suggests that al- proved by the board. though bank lending to firms declined during the global crisis, bond financing actually increased to • Risk management, compliance, and internal audit make up much of the gap in some countries.74 Al- functions should be established, each with suffi- though banks may have practical advantage in de- cient authority, independence, resources, and ac- veloping countries, building local currency capital cess to the board. markets is desirable. Certain preconditions must • Compensation schemes should encourage an be established, involving both the government and orientation to client needs, responsible provi- the private sector, such as adequate property rights, sion of financial services and risk taking by the a legal framework, infrastructure (payment and se- bank employees, and a long-term view in business curity settlement systems), corporate governance, conduct.72 financial accounting standards, and a credible audit- ing industry. Some economies may still be far from • The board and senior management should under- establishing these preconditions, and some small stand and guide the bank’s overall structure and its economies may never generate the necessary scale. evolution, ensure that the structure is justified, and Still, firms in those countries could list on regional avoid undue complexity. or global stock exchanges, and individual investors • Disclosure requirements should enhance account- could access foreign capital markets through brokers ability of banks to depositors, creditors, and other or investment funds. Small economies with neces- clients and stakeholders; for instance, key points on sary preconditions in place and problems of small its governance structure and risk appetite should scale could consider developing local trading plat- be clearly disclosed. forms integrated with regional or global stock ex- changes—the way that Estonia, Latvia, and Lithuania In addition, external auditors should monitor integrated under the Baltic Stock Exchange. Regional compliance with any mandatory or voluntary corpo- and international initiatives have emerged to aid di- rate governance codes adopted by the financial in- versification of financial intermediation into capital dustry, and identify in their reports any gaps between markets. They include the 2003 Asian Bond Markets the existing practice and the adopted code. and Asian Bond Fund Initiatives; the 2008 Global The jury is still out on the effects of bank (fi- Emerging Markets Local Currency Bond program of nancial institutions) specialization versus diversifi- the World Bank; and the Vienna Initiative, a regional cation—for instance in lending—on financial de- public-private coordination framework for develop- velopment and stability. On the one hand, lending ment of local currency capital markets. expertise gained through specialization in certain sectors can benefit banks by enhancing their screen- ing and monitoring efficiency. On the other hand, A summary of policy recommendations diversification of lending risk across many sectors This chapter has explored the tension between finan- can enhance the stability of an individual bank by cial inclusion and stability and stressed that this ten- protecting it from correlated losses. At the system sion must be addressed when financial sector policy level, the impact of lending specialization as opposed is formulated and implemented. On the one hand, to diversification could be critical. Diversified banks excessive and reckless financial inclusion can endan- that look alike or purposely herd can actually reduce ger financial stability. On the other hand, responsible systemic stability.73 Hence, from the system’s per- financial inclusion can enhance the financial system’s spective, promoting diversity among banks could be stability directly or indirectly through greater resil- The role of the financial system in managing risk 217 TA B L E 6.4 Policy priorities to improve the financial system’s role in risk management POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Collection and analysis of data Targeted financial education IT solutions for better access to on gaps in financial inclusion financial prices System-wide collection of Financial stability reports Early warning models macroprudential data Public communication of concerns about systemic risk and steps to resolve the crisis Protection Legal frameworks and Consumer protection G2P payments Access to capital financial infrastructure market instruments Independent financial Macroprudential Crisis preparedness Crisis simulation regulators regulation frameworks exercises Corporate governance standards (for example, disclosure of ultimate controllers, risk management and internal controls, compensation policies) Insurance Legal frameworks and nsumer protection Consumer Con Compulsory insurance Fiscal insurance financial infrastructure cture infrastruc , car, mortgage) (for example, gage) mortg including PPPs Macroprudential temic risk Systemic Syst k Foreign exchange Fiscal contingent capital buffers surcharges charges surc s reserves liabilities Coping Contract enforcement cy regimes Efficient insolvency insolvenc Preserved access to o Consumer workouts and bad debt worrkouts credit protection Failing bank Emergency liquidity Blanket deposit Lending resolution assistance guarantees guarantees Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing poli- cies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably, and that can be improved over time. G2P = government to person. IT = information technology. PPPs = public-private partnerships. ience of the financial system’s clients (individuals, To broaden the availability and use of financial firms, the state). In practice, middle-income coun- tools for managing risk, public policy should focus tries face the greatest tension between allowing rapid on overcoming obstacles related to financial infra- financial inclusion and fostering financial stability. structure, the small scale of the market, and adop- Low-income countries cannot mobilize as much tion of innovative financial instruments. The state savings from households that are often constrained should promote competition among different types in their consumption. Nor are low-income countries of financial institutions and support delivery of fi- well-enough integrated into global finance to import nancial tools within efficient consumer protection large amounts of foreign savings. In contrast, finan- frameworks. To enhance management of systemic cial inclusion in high-income countries approaches risk in the financial system, public policy should fo- 90 to 100 percent, and high-income countries focus cus on establishing strong macroprudential frame- mainly on fostering financial stability. works, including crisis preparedness and resolution Table 6.4 summarizes policy recommenda- measures, that are equipped with adequate macro- tions to promote financial inclusion and enhance prudential tools, while fostering the safety and ef- financial stability in view of this tension. The rec- ficiency of financial market infrastructure. Most ommendations are grouped according to the main important, the process of public policy formulation components of effective risk management (knowl- must account for the trade-offs and synergies in fi- edge, protection, insurance, and coping) and follow nance to produce balanced policies that respect both a foundational approach from the most needed to a country’s development goals and risk appetite in advanced measures in support of risk management. the financial area. 218 WORLD DEVELOPMENT REPORT 2014 Notes from 2007 to 2011 (Laeven and Valencia 2012; Brown 2013 for the WDR 2014). 1. This story is a composite of events and recent business prac- 29. Brown 2013 for the WDR 2014. tices in the Czech Republic. 30. See Brown 2013 for the WDR 2014 for further details and 2. This chapter focuses on the financial system, not on indi- references. vidual financial institutions. Thus it does not discuss issues 31. Claessens, Ghosh, and Mihet, forthcoming; Claessens, Kose, related to microprudential supervision and deposit insurance, and Terrones 2012. for instance. 32. World Bank FinStats (internal database), calculated as the 3. Levine 1997. three-year average growth of nominal credit before the crisis. 4. For a more extensive discussion of financial inclusion, see 33. Reinhart and Rogoff 2009; BIS 2011a; Buncic and Melecky World Bank 2013. 2013a for the WDR 2014. 5. Beck, Demirgüç-Kunt, and Merrouche 2013. 34. Beck and De Jonghe 2013 for the WDR 2014. 6. Beck, Demirgüç-Kunt, and Merrouche 2013. 35. Allen and Gale 2000; Claessens, Ghosh, and Mihet, forth- 7. Levine 1997. coming. 8. The picture of competition in the financial system, based 36. Ghosh, Gonzalez del Mazo, and Ötker-Robe 2012. on available studies and data, is incomplete. Studies have fo- 37. Beck, Demirgüç-Kunt, and Merrouche 2013. cused mostly on banking; much less research has been done 38. WDR 2014 team calculations based on data from the Bank on competition in the insurance sector, capital markets (pen- for International Settlements on locational banking statistics sion funds, mutual funds, brokerage houses), and the sector of by residence (claims of Bank for International Settlements re- nonbank credit institutions and payment services providers. porting banks on banks in developing countries). For an extensive discussion of bank competition and enabling 39. WDR 2014 team calculations based on Bankscope data and public policies, see World Bank 2012b, chapter 3. See BIS the initial list of 29 G-SIFIs issued by the Financial Stability 2004 on the benefits of foreign direct investment in financial Board. Local SIFIs are defined as the top 10 banks of the do- systems. mestic financial system ranked by assets. 9. Demombynes and Thegeya 2012; Bankable Frontier Asso- 40. Empirical studies confirm that lending by subsidiaries is ciates 2009. more stable than direct cross-border lending (Allen and 10. Clarke and others 2012; Mobarak and Rosenzweig 2013. others 2011); Vienna Initiative, http://vienna-initiative.com/ 11. Green 2008. vienna-initiative-part-1/. 12. This section is focused on development. It thus does not deal 41. Laeven and Valencia 2012. Contagion can occur through de- with issues such as unmet demand for credit that results from pressed asset prices (as a result of fire sales) and information tight credit standards in recessions or crisis periods. (where one institution’s failure can cause investors to with- 13. Allen and others 2012. draw their investments in other banks in the country). Simi- 14. Collins, Jentzsch, and Mazer 2011. 15. Dowd 2009; Honohan 2010. Similarly, public deposit insur- larly, as retail runs on banks occur, cross-border wholesale and ance should strike a balance between protecting the vulner- interbank runs on solvent domestic banks can also happen able and discouraging moral hazard; it should not, for exam- because of coordination problems (Allen and others 2011). ple, distort market discipline by letting depositors ignore bank 42. BIS 2010a; Cole, Kanz, and Klapper 2012; Aebi, Sabato, and risk and simply deposit funds for the highest interest rates. Schmid 2012. 16. Braun and Raddatz 2010; Willis 2009. 43. Calomiris 2011; FSB 2010. 17. Beck 2013 for the WDR 2014. 44. Braun and Raddatz 2010; Hardy 2006. 18. OECD 2011. 45. Senior Supervisors Group 2009; Beck, Demirgüç-Kunt, and 19. See OECD, http://www.financial-education.org; World Bank, Levine 2003. http://responsiblefinance.worldbank.org; and Russia Trust 46. BIS 2011b. Fund on Financial Literacy and Education and Trust, http:// 47. BIS 2011b; Bank of England 2013. Some proposals suggest www.finlitedu.org, managed by the World Bank, for informa- that macroprudential policy committees consist of five mem- tion on designing financial education programs. bers, all from outside government and international organi- 20. In 2008, India’s National Rural Employment Guarantee Act zations: a macroeconomist, a microeconomist, a research ac- (NREGA) made more than 45 million payments to poor peo- countant, a financial engineer, and a practitioner. The board ple living in rural areas. People can receive their G2P payment should not include supervisors and regulators. Its composi- from post office saving accounts, bank accounts, and village tion should provide for objective and independent judgment. officials. 48. IMF 2012b; Melecky and Podpiera 2012. See Buncic and 21. Gupta 2013. Melecky 2013b for applications of practical stress-testing 22. Brix and McKee 2010. approaches for macroprudential policy. 23. World Bank 2012a. 49. Arnold and others 2012; Bank of England 2013. 24. World Bank 2012a, box 3. 50. Cihak and others 2012. 25. Collins and others 2009. 51. See also Lim and others 2011 for detailed country examples of 26. Faz and Breloff 2012; Hashemi and de Montesquiou 2011. the use of macroprudential tools. 27. Output losses are computed as the cumulative sum of the dif- 52. Claessens, Ghosh, and Mihet, forthcoming. ferences between actual and trend real GDP over the period 53. Shin 2010; IMF 2012a. (T, T + 3), expressed as a percentage of trend real GDP, with 54. For an example of a policy statement outlining a detailed T the starting year of the crisis. See Laeven and Valencia 2012 approach to implementing macroprudential buffers in the for more details. United Kingdom, see Bank of England 2013. Lim and others 28. Mongolia, Nigeria, and the United States are the only coun- 2011, figure 1, offers a useful conceptual framework on how to tries outside that region that experienced a banking crisis use macroprudential instruments. The role of the financial system in managing risk 219 55. Hoggarth, Reidhill, and Sinclair 2004. Bank of England. 2013. “The Financial Policy Committee’s Pow- 56. In principle, the framework for emergency liquidity assistance ers to Supplement Capital Requirements: A Draft Policy State- should allow central bank lending against a broader pool of ment.” Bank of England, London. collateral (including highly rated corporate bonds or foreign Bankable Frontier Associates. 2009. “The Mzansi Bank Account governments bonds) that is properly discounted to avoid situ- Initiative in South Africa.” Report commissioned by FinMark ations in which the central bank takes on excessive counter- Trust, Bankable Frontier Associates, Somerville, MA. party (credit) risk and needs to be recapitalized by the min- Beck, Thorsten. 2011. “Bank Failure Resolution: A Conceptual istry of finance. Framework.” In Financial Regulation at the Crossroads: Impli- 57. Dziobek and Pazarbasioglu 1997. cations for Supervision, Institutional Design and Trade, edited 58. Laeven and Valencia 2012. by Panagiotis Delimatsis and Nils Herger, 53–71. Alphen aan 59. See FSB 2010; BIS 2011a. The instruments comprise systemic den Rijn, NL: Kluwer Law International. capital surcharges, obligatory issuance of contingent capital ———. 2013. “The Supply-Demand Mismatch in Provision of obligations (CoCOs), and bail-in debt instruments. Financial Services.” Background paper for the World Develop- 60. See, for instance, BIS 2012. ment Report 2014. 61. De Grauwe 2010. Beck, Thorsten, and Olivier De Jonghe. 2013. “Lending Concen- 62. Adasme, Majnoni, and Uribe 2006. tration, and Its Implications for Systemic Risk and Public 63. Collins and others 2009; Cull, Demirgüç-Kunt, and Lyman Policy.” Background paper for the World Development Report 2012. 2014. 64. CGAP 2011. Beck, Thorsten, Asli Demirgüç-Kunt, and Ross Levine. 2003. “Bank Supervision and Corporate Finance.” Working Paper 65. Klapper, Lusardi, and Panos 2012. 9620, National Bureau of Economic Research, Cambridge, 66. Beck 2013 for the WDR 2014. MA. 67. Beck and De Jonghe 2013 for the WDR 2014. Beck, Thorsten, Asli Demirgüç-Kunt, and Ouarda Merrouche. 68. Anginer, Demirgüç-Kunt, and Zhu 2012. 2013. “Islamic vs. Conventional Banking: Business Model, 69. Lafley and others 2012. Efficiency and Stability.” Journal of Banking & Finance 37 (2): 70. World Bank 2012b. 433–47. 71. BIS 2010a; FSB 2013. BIS (Bank for International Settlements). 2004. “Foreign Direct 72. European Commission 2009. Investment in the Financial Sector of Emerging Market Econ- 73. Beck and De Jonghe 2013 for the WDR 2014. omies.” CGFS Publications 22, BIS, Basel. 74. Adrian, Colla, and Shin 2012. ———. 2010a. “Principles for Enhancing Corporate Governance.” BCBS Publications 176, BIS, Basel. ———. 2010b. “Report and Recommendations of the Cross- References Border Bank Resolution Group.” BCBS Publications 169, BIS, Adasme, Osvaldo, Giovanni Majnoni, and Myriam Uribe. 2006. Basel. “Access and Risk: Friends or Foes? Lessons from Chile.” Policy ———. 2011a. “Basel III: A Global Regulatory Framework for More Resilient Banks and Banking Systems—Revised Version Research Working Paper 4003, World Bank, Washington, DC. June 2011.” BCBS Publications 189, BIS, Basel. Adrian, Tobias, Paolo Colla, and Hyun Song Shin. 2012. “Which ———. 2011b. “Central Bank Governance and Financial Stability: Financial Frictions? Parsing the Evidence from the Financial A Report by Study Group.” Other Publications 14, BIS, Basel. Crisis of 2007–2009.” Paper prepared for the National Bu- ———. 2012. “A Framework for Dealing with Domestic Systemi- reau of Economic Research Macro Annual Conference, April cally Important Banks.” BCBS Publications 233, BIS, Basel. 20–21. Braun, Matias, and Claudio Raddatz. 2010. “Banking on Politics: Aebi, Vincent, Gabriele Sabato, and Markus Schmid. 2012. “Risk When Former High-Ranking Politicians Become Bank Direc- Management, Corporate Governance, and Bank Performance tors.” World Bank Economic Review 24 (2): 234–79. in the Financial Crisis.” Journal of Banking & Finance 36 (12): Brix, Laura, and Katharine McKee. 2010. “Consumer Protection 3213–26. Regulation in Low-Access Environments: Opportunities to Allen, Franklin, Thorsten Beck, Elena Carletti, Philip Lane, Dirk Promote Responsible Finance.” Focus Note 60, Consultative Schoenmaker, and Wolf Wagner. 2011. Cross-Border Banking Group to Assist the Poor, Washington, DC. in Europe: Implications for Financial Stability and Macroeco- Brown, Craig O., and I. Serdar Dinc. 2005. “The Politics of Bank nomic Policies. London: Centre for Economic Policy Research. Failures: Evidence from Emerging Markets.” Quarterly Journal Allen, Franklin, Asli Demirgüç-Kunt, Leora Klapper, and María of Economics 120 (4): 1413–44. Soledad Martínez Pería. 2012. “The Foundations of Financial Brown, Martin. 2013. “The Transmission of Banking Crises to Inclusion: Understanding Ownership and Use of Formal Ac- Households: Lessons from the ECA Region 2008–2012.” Back- counts.” Policy Research Working Paper 6290, World Bank, ground paper for the World Development Report 2014. Washington, DC. Buncic, Daniel, and Martin Melecky. 2013a. “Equilibrium Credit: Allen, Franklin, and Douglas Gale. 2000. “Financial Contagion.” The Reference Point for Macroprudential Supervisors.” Back- Journal of Political Economy 108 (1): 1–33. ground paper for the World Development Report 2014. Anginer, Deniz, Asli Demirgüç-Kunt, and Min Zhu. 2012. “How ———. 2013b. “Macroprudential Stress Testing of Credit Risk: Does Bank Competition Affect Systemic Stability?” Policy A Practical Approach for Policy Makers.” Journal of Financial Research Working Paper 5981, World Bank, Washington, DC. Stability 9 (3): 347–70. Arnold, Bruce, Claudio Borio, Luci Ellis, and Fariborz Moshirian. Calomiris, Charles W. 2011. “An Incentive-Robust Programme for 2012. “Systemic Risk, Macroprudential Policy Frameworks, Financial Reform.” Manchester School 79 (s2): 39–72. Monitoring Financial Systems and the Evolution of Capital CGAP (Consultative Group to Assist the Poor). 2011. “Global Adequacy.” Journal of Banking & Finance 36 (12): 3125–32. Standard-Setting Bodies and Financial Inclusion for the Poor: 220 WORLD DEVELOPMENT REPORT 2014 Towards Proportionate Standards and Guidance.” CGAP and cial Security: The Benefits of E–Payments to Indian Society.” Global Partnership for Financial Inclusion, Washington, DC. McKinsey & Company, New York. Chitu, Livia. 2012. “Was Official Dollarisation/Euroisation an European Commission. 2009. “Commission Recommendation Amplifier of the ‘Great Recession’ of 2007–2009 in Emerging of 30 April 2009 on Remuneration Policies in the Financial Economies?” Working Paper 1473, European Central Bank, Services Sector.” Commission Recommendation 2009/384/ Frankfurt. EC, European Commission, Brussels. Cihak, Martin, Sonia Munoz, Shakira Teh Sharifuddin, and Kalin Faz, Xavier, and Paul Breloff. 2012. “A Structured Approach to Tintchev. 2012. “Financial Stability Reports: What Are They Understanding the Financial Service Needs of the Poor in Good For?” Working Paper WP/12/1, International Monetary Mexico.” Brief 71306, Consultative Group to Assist the Poor, Fund, Washington, DC. Washington, DC. Claessens, Stijn, Swati R. Ghosh, and Roxana Mihet. Forthcom- FSB (Financial Stability Board). 2010. “Reducing the Moral Haz- ing. “Macro-Prudential Policies to Mitigate Financial System ard Posed by Systemically Important Financial Institutions: Vulnerabilities.” Journal of International Money and Finance. FSB Recommendations and Time Lines.” FSB, Basel. Claessens, Stijn, M. Ayhan Kose, and Marco Terrones. 2012. “How ———. 2013. “Thematic Review on Risk Governance: Peer Re- Do Business and Financial Cycles Interact?” Journal of Interna- view Report.” FSB, Basel. . tional Economics 87 (1): 178–90. Ghosh, Swati R., Ines Gonzalez del Mazo, and I nci Ötker-Robe. Clarke, Daniel J., Olivier Mahul, Kolli N. Rao, and Niraj Verma. 2012. “Chasing the Shadow: How Significant Is Shadow Bank- 2012. “Weather Based Crop Insurance in India.” Policy Re- ing in Emerging Markets?” Economic Premise 88, World search Working Paper 5985, World Bank, Washington, DC. Bank, Washington, DC. Cohen, Rebecca. 2007. “The Positive Impacts of Affordable Hous- Green, Duncan. 2008. From Poverty to Power: How Active Citi- ing on Health: A Research Summary.” Center for Housing zens and Effective States Can Change the World. Oxford, U.K.: Policy and Enterprise Commmunity Partners, Washington, Oxfam International. DC. Gupta, Sunil. 2013. “The Mobile Banking and Payment Revolu- Cole, Shawn, Martin Kanz, and Leora Klapper. 2012. “Incentiv- tion.” European Financial Review February: 3–6. Han, Rui, and Martin Melecky. 2013. “Financial Inclusion for Sta- izing Calculated Risk-Taking: Evidence from an Experiment bility: Access to Bank Deposits and the Deposit Growth during with Commercial Bank Loan Officers.” Policy Research Work- the Global Financial Crisis.” Background paper for the World ing Paper 6146, World Bank, Washington, DC. Development Report 2014. Collins, Daryl, Nicola Jentzsch, and Rafael Mazer. 2011. “Incorpo- Hardy, Daniel C. 2006. “Regulatory Capture in Banking.” Working rating Consumer Research into Consumer Protection Policy Paper WP/06/34, International Monetary Fund, Washington, Making.” Focus Note 74, Consultative Group to Assist the DC. Poor, Washington, DC. Hashemi, Syed M., and Aude de Montesquiou. 2011. “Reaching Collins, Daryl, Jonathan Morduch, Stuart Rutherford, and Or- the Poorest: Lessons from the Graduation Model.” Focus Note landa Ruthven. 2009. Portfolios of the Poor: How the World’s 69, Consultative Group to Assist the Poor, Washington, DC. Poor Live on $2 a Day. Princeton, NJ: Princeton University Hoggarth, Glenn, Jack Reidhill, and Peter Sinclair. 2004. “On the Press. Resolution of Banking Crises: Theory and Evidence.” Working Cull, Robert, Asli Demirgüç-Kunt, and Timothy Lyman. 2012. “Fi- Paper 229, Bank of England, London. nancial Inclusion and Stability: What Does Research Show?” Honohan, Patrick. 2008. “Cross-Country Variation in Household Brief 71305, Consultative Group to Assist the Poor, Washing- Access to Financial Services.” Journal of Banking & Finance 32 ton, DC. (11): 2493–500. De Grauwe, Paul. 2010. “Crisis in the Eurozone and How to Deal ———. 2010. “Partial Credit Guarantees: Principles and Practice.” with It.” Policy Brief 204, Centre for European Policy Studies, Journal of Financial Stability 6 (1): 1–9. Brussels. IMF (International Monetary Fund). 2012a. “The Interaction Demirgüç-Kunt, Asli, and Leora Klapper. 2012. “Measuring Fi- of Monetary and Macroprudential Policies—Background nancial Inclusion: The Global Findex Database.” Policy Re- Paper.” IMF, Washington, DC. search Working Paper 6025, World Bank, Washington, DC. ———. 2012b. “Macrofinancial Stress Testing: Principles and Demombynes, Gabriel, and Aaron Thegeya. 2012. “Kenya’s Practices—Background Material.” IMF, Washington, DC. Mobile Revolution and the Promise of Mobile Savings.” Policy Josefsson, Mats. 2006. “The Public Sector Recapitalization Pro- Research Working Paper 5988, World Bank, Washington, DC. gram in Turkey.” In Bank Restructuring and Resolution, edited De Soto, Hernando. 2000. The Mystery of Capital: Why Capital- by David S. Hoelscher, 369–84. New York: Palgrave Macmillan. ism Triumphs in the West and Fails Everywhere Else. New York: Klapper, Leora, Annamaria Lusardi, and Georgios Panos. 2012. Basic Books. “Financial Literacy and the Financial Crisis.” Working Paper Dowd, Kevin. 2009. “Moral Hazard and the Financial Crisis.” Cato 17930, National Bureau of Economic Research, Cambridge, Journal 29 (1): 141–66. MA. Dziobek, Claudia, and Ceyla Pazarbasioglu. 1997. “Lessons from Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises Systemic Bank Restructuring: A Survey of 24 Countries.” Database: An Update.” Working Paper WP/12/163, Interna- Working Paper WP/97/161, International Monetary Fund, tional Monetary Fund, Washington, DC. Washington, DC. Lafley, Alan G., Roger L. Martin, Jan W. Rivkin, and Nicolaj Siggel- ECRI (European Credit Research Institute). Lending to House- kow. 2012. “Bringing Science to the Art of Strategy.” Harvard holds in Europe (database). ECRI, Brussels. http://www.ecri. Business Review 90 (9): 3–12. eu/new/node/165. Levine, Ross. 1997. “Financial Development and Economic Ehrbeck, Tilman, Rajiv Lochan, Supriyo Sinha, Naveen Tahliyani, Growth: Views and Agenda.” Journal of Economic Literature 35 and Adil Zainulbhai. 2010. “Inclusive Growth and Finan- (2): 688–726. The role of the financial system in managing risk 221 Lim, Cheng Hoon, Francesco Columba, Alejo Costa, Piyabha Prepared for United States Agency for International Develop- Kongsamut, Akira Otani, Mustafa Saiyid, Torsten Wezel, and ment, Washington, DC. Xiaoyong Wu. 2011. “Macroprudential Policy: What Instru- Senior Supervisors Group. 2009. “Risk Management Lessons from ments and How to Use Them? Lessons from Country Expe- the Global Banking Crisis of 2008.” Senior Supervisors Group. riences.” Working Paper WP/11/238, International Monetary Shin, Hyun Song. 2010. “Non-Core Liabilities Tax as a Tool for Fund, Washington, DC. Prudential Regulation.” Policy Memo, February 19, Seoul. Maimbo, Samuel, and Martin Melecky. 2013. “Financial Policy Willis, Lauren. 2009. “Evidence and Ideology in Assessing the Formulation: Addressing the Tradeoff between Development Effectiveness of Financial Literacy Education.” San Diego Law and Stability.” Background paper for the World Development Review 46 (2): 415–58. Report 2014. World Bank. 2012a. “Financial Inclusion Strategies Reference Melecky, Martin, and Anca M. Podpiera. 2012. “Macropruden- Framework.” Prepared for the G-20 Mexico Presidency. Wash- tial Stress-Testing Practices of Central Banks in Central and ington, DC: World Bank. Southeastern Europe: Comparison and Challenges Ahead.” ———. 2012b. Global Financial Development Report 2013: Re- Emerging Markets Finance and Trade 48 (4): 118–43. thinking the Role of the State in Finance. Washington, DC: Mobarak, Ahmed Mushfiq, and Mark Rosenzweig. 2013. “Infor- World Bank. mal Risk Sharing, Index Insurance and Risk-Taking in Devel- oping Countries.” American Economic Review 103 (3): 375–80. ———. 2013. Global Financial Development Report 2014: Finan- OECD (Organisation for Economic Co-operation and Develop- cial Inclusion. Washington, DC: World Bank. ment). 2011. “G20 High-Level Principles on Financial Con- ———. FinStats (internal database). World Bank, Washington, sumer Protection.” OECD, Paris. DC. Reinhart, Carmen M., and Kenneth S. Rogoff. 2009. This Time ———. Global Financial Development Database. World Bank, Is Different: Eight Centuries of Financial Folly. Princeton, NJ: Washington, DC, http://data.worldbank.org/data-catalog/ Princeton University Press. global-financial-development. RTI International. 2005. “LAC Housing and Urban Upgrading As- ———. Global Findex (database). World Bank, Washington, DC, sistance Retrospective: Honduras and Ecuador, 1980–2005.” http://data.worldbank.org/data-catalog/financial_inclusion. S P OT L I G H T 6 Building resilience to global economic shocks in the Czech Republic, Peru, and Kenya While the recent global economic downturn did not spare many developing countries, they were more resilient to the 2008 global crisis than to previous crises. The East Asian countries managed systemic risk especially well, but the per- formance of several countries in Central Europe, Latin America, and Sub-Saharan Africa has also been remarkable. The experiences of three of them—the Czech Republic, Peru, and Kenya—offer two main lessons. First, pursuing macroprudential policies in good times, while continuously strengthening the domestic financial system, is key to building resilience to severe economic downturns. Second, timely countercyclical macro policies not only help man- age macroeconomic and financial cycles after crises hit but also boost preparation by building or preserving the necessary resources (fiscal space) to respond to a crisis. A difficult period of reform to safeguard against regional contagion from crisis episodes The Czech Republic started building stronger foundations in emerging markets, banks built up adequate levels of capi- for aggregate risk management following major lessons talization and sufficient levels of liquidity. learned from the 1997–98 banking crisis. In 1997 the coun- Peru liberalized foreign trade in the early 1990s, drasti- try abandoned its fixed exchange rate regime in favor of a cally reducing tariff rates and eliminating nontariff barriers. monetary policy framework based on inflation targeting. In A more favorable economic environment from 2002 to part thanks to its increased financial and political indepen- 2007 fueled economic growth. Increasing demand for the dence, the Czech National Bank managed to increase the country’s commodities (mineral ores and metals) from credibility of monetary policy and achieve greater price sta- large dynamic emerging markets in East Asia produced a bility. These achievements translated into low interest rates large positive income shock. Peru saved part of the reve- that, along with better fiscal discipline, helped the country nues from natural resources. International reserves grew to maintain a sound external position, which benefited from a the equivalent of more than 17 months of imports in 2007, trade surplus and a modest current account deficit. and the fiscal primary surplus increased. Sustained annual In addition to the strengthened policy framework, the gross domestic product (GDP) growth of nearly 7 percent Czech banking system was able to finance its lending ac- from 2002 to 2008 helped reduce poverty. In light of these tivities mainly from local deposits and extended loans to macroeconomic achievements, the international rating households in domestic currency. Thus borrowers avoided agencies upgraded Peru’s sovereign rating, paving the way unhedged exposures in foreign exchange, and the banks for major foreign investment. avoided the associated indirect credit risk. Like Peru, Kenya successfully built resilience by strength- Starting in 2006, all microprudential regulators were ening both its financial and macrofiscal systems. Although integrated under the Czech National Bank, which was al- it did not have to deal with a specific economic crisis ready the monetary authority and macroprudential super- before 2008, its economy was in trouble during the 1980s visor. Bringing microprudential and macroprudential su- and early 1990s, after experiencing two decades of high pervision under one institution enabled the Czech Republic growth. From 1991 to 1993, Kenya’s GDP growth stag- to conduct prudential supervision in a more comprehen- nated, agricultural production sharply contracted, and hy- sive manner and to better monitor how the risks from indi- perinflation flared. The government decided to implement vidual financial institutions translate into systemic risk in economic reforms to stabilize the financial sector and re- the financial sector. Integrated financial sector supervision gain sustainable growth. The banking system was strength- should also bring improvements to the coordination and ened, notably through substantial capitalization of the timeliness of policy response in future crises. banks, and access to finance for the population was im- Unlike the Czech Republic, until 2008, Peru had not been proved. Kenya also managed to decrease its public debt hit by a major economic turmoil for almost two decades. But and accumulate high international reserves (up to four until the late 1980s, the country experienced hyperinflation, months of import coverage) by adopting prudent fiscal severe macroeconomic imbalances, and massive capital policies and maintaining a healthy external position, with outflows. In the 1990s, Peru put in place key reforms to sta- strong surpluses in the service balance (mainly tourism and bilize the economy. It brought hyperinflation under control information technology) and massive inflows of foreign through explicit targets on the monetary base. Once infla- capital that compensated for the trade deficit. tion was reduced to single digits, the central bank adopted an inflation-targeting regime with a flexible exchange rate The benefits of good preparation that kept inflation in check. The tax system and the financial The relatively strong resilience of these three countries to sector were reformed. As a byproduct of these reforms and the global crisis in 2008 was the result of an arduous pro- Spotlight 6 223 cess undertaken a decade or more before the shock. Although politi- cal leaders may have been tempted to adopt procyclical measures during good economic times, these three countries understood the F I G U R E S 6 .1 Bank nonperforming loans to total gross necessity of strengthening their financial and macrofiscal systems to loans ratio, 2007–11 prepare for serious economic turmoil. This awareness proved its 12 worth when the world economy crashed. The Czech Republic demonstrated the utility of establishing an in- 10 tegrated supervisor at the national level within a strong and indepen- 8 dent central bank. Overall, the Czech government did not have to Percent undertake any major measures, and a simple relaxation of monetary 6 policy proved sufficient to ensure adequate liquidity. The adequate 4 loan-to-deposit ratio of the banking sector and low dollarization of loans through adequate pricing of foreign exchange risk by the banks 2 were also key factors in weathering the global economic shock. De- 0 spite decreased lending during the crisis period, Czech banks contin- 2007 2008 2009 2010 2011 ued to generate profits and further strengthen their capital buffers, which helped them cope with a notable increase in the share of non- Kenya Peru Czech Republic Hungary performing loans (figure S6.1). Other countries in the region, includ- Source: WDR 2014 team based on data from World Bank Global Finance ing Hungary and Ukraine, faced higher risk because of higher dollar- Development Database. ization of loans. Note: While the share of nonperforming loans fell in Kenya during the crisis Given the good conditions that Peru had created since the mid- and remained relatively low and flat in Peru, it increased significantly in the 1990s, the government was able to respond in an efficient and coun- Czech Republic; however, the country managed to stop the rise at the end of 2010. This can be considered a good performance in comparison with other tercyclical manner to sustain the national economy during the global Eastern European countries such as Hungary, where the share quadrupled crisis. The central bank injected liquidity into the financial system, in over the same period and was still increasing at the end of 2011. both local currency (nuevos soles) and U.S. dollars, to prevent a liquid- ity squeeze and a credit crunch. The monetary policy rate was low- ered to 1.25 percent, and the first package—equivalent to 3.4 percent of GDP—of a threefold stimulus plan was enacted in 2009, financed by fiscal savings. By investing in roads, housing, and hospitals; by giv- Sources ing incentives to nontraditional exporters; by supporting small and Castillo, Paul, and Daniel Barco. 2008. “Facing Up a Sudden Stop of Capital Flows: medium enterprises and farms; and by increasing expenditures in so- Policy Lessons from the 90’s Peruvian Experience.” Working Paper 2008-002, cial programs, the government aimed to sustain domestic demand, Central Reserve Bank of Peru, Lima. De Gregorio, José. 2012. “Resilience in Latin America: Lessons from Macroeconomic boost business confidence, and extend guarantees to support firms, Management and Financial Policies.” Unpublished manuscript, University of exporters, and smaller financial institutions. Peru also benefited from Chile, Santiago. a key external factor: favorable terms of trade, with a rapid recovery of De la Torre, Augusto, and Alain Ize. 2010. “Finance in Crisis: Causes, Lessons, Conse- exports to Asia (particularly China). quences, and Application to Latin America.” In The Day after Tomorrow, edited by Kenya’s demonstration of risk management is arguably even more Otaviano Canuto and Marcelo Giugale, 143–59. Washington, DC: World Bank. Ellingson, Amy. 2010. “Brazil and Peru: Economies Set to Flourish in Postrecession impressive, considering the quadruple shock it faced within a very World.” EconSouth 12 (1): 22–26. short period: postelection violence in early 2008, oil and food price IMF (International Monetary Fund). 2009. “The Implications of the Global Financial increases, catastrophic drought, and the global financial crisis. Al- Crisis for Low-Income Countries.” IMF, Washington, DC. though an increased perception of risk in the market was reflected in ———. 2010. “Peru: Staff Report for the 2010 Article IV Consultation.” IMF, Washing- ton, DC. the commercial bank lending rates, with a particularly large impact on ———. 2012. “Resilience in Emerging Market and Developing Economies: Will It the agriculture sector, the central bank successfully implemented Last?” In World Economic Outlook, 129–71. Washington, DC: IMF. countercyclical monetary policies, reducing its rate, as well as the cash Kirori, Gabriel, and Tabitha Kiriti Nganga. 2009. “Global Economic and Financial Crisis reserve ratio, to inject liquidity into the market. The banking sector and Trade: Kenya’s Experience.” Paper prepared for the Rethinking African Eco- nomic Policy in Light of the Global Economic and Financial Crisis Conference was strong enough to maintain capital adequacy ratios (19.8 percent organized by the African Economic Research Consortium. in 2009—well above the statutory requirement of 12 percent) and a Melecky, Martin. 2009. “Macroeconomic Management, Financial Sector Develop- low share of nonperforming loans. With public debt under control, ment and Crisis Resilience: Some Stylized Facts from Central and Eastern Eu- and buoyed by large international reserves, the government was able rope.” MPRA Paper 28214, University Library of Munich, Munich. to implement an ambitious fiscal stimulus program of $300 million, OECD (Organisation for Economic Co-operation and Development). 2009. Latin American Economic Outlook 2010. Paris: OECD. thereby protecting key expenditures. The stimulus boosted employ- Were, Maureen, and Samuel Tiriongo. 2012. “Central Bank’s Response to Economic ment and economic activity, notably by increasing spending on Crises from a Developing African Economy Perspective: Lessons from Kenya’s infrastructure. Experience.” Unpublished manuscript, Central Bank of Kenya, Nairobi. While many countries are still suffering from the crisis, the Czech World Bank. 2009. “Still Standing: Kenya’s Slow Recovery from a Quadruple Shock, with a Special Focus on the Food Crisis.” Kenya Economic Update, Edition 1, Republic, Peru, and Kenya, have all demonstrated an impressive abil- World Bank, Nairobi. ity to manage macrofinancial risks—offering lessons that would ben- ———. Global Financial Development Database. World Bank, Washington, DC, efit even developed countries. http://data.worldbank.org/data-catalog/global-financial-development. A stable macroeconomy reduces uncertainty and enables economic agents to focus on productive decisions rather than on trying to mitigate high risks. A pensioner copes with the 1998 financial crisis in Russia. © Anthony Suau CHAPTER 7 Managing macroeconomic risk Building stronger institutions for better policy outcomes Increasing resilience and oeconomy cent a year from 2002 to 2008). Sounder Macr promoting opportunity macroeconomic management and through sound benign conditions internationally macroeconomic policy allowed Peru to build ample li- quidity buffers and monetary and Peru has weathered a series of fiscal space, helping accommodate crises in the past 25 years. The the global shock: unlike much of first, driven by a money-financed the rest of the world, the recession fiscal expansion in the second half in Peru in 2008–09 was mild (minus of the 1980s, led to high inflation 2.1 percent from peak to trough), and and large macroeconomic imbal- the recovery was swift (two quarters) and ances. Declared ineligible for new loans strong. Peru’s GDP grew 10 percent the year from the International Monetary Fund (IMF) in after the economy hit its trough. 1986, Peru had no room for maneuver, and the new As Peru’s experience illustrates, the national gov- administration had to resort to severe monetary and ernment can play a pivotal role helping individuals fiscal tightening to bring inflation down, leading to a manage aggregate risks—domestic and interna- steep recession. The gross domestic product (GDP) tional—that they are not equipped to manage on plummeted more than 20 percent in 1990 from peak their own (cartoon 7.1). High inflation can worsen to trough. It took Peru 14 quarters to return to the income distribution, increase poverty, and lower real real GDP level it had before the crisis. wages. Unemployment rises in recessions—and even By 1997, Peru’s inflation reached single-digit more in recessions associated with financial crises.1 levels. The government continued reconstituting Aggregate risks like these can have a profound im- reserves and reducing external debt. The country pact on people, households, communities, enter- nonetheless remained vulnerable to reversals in prises, and the financial system—and each of these capital flows, which became all too apparent when agents’ capacity to manage risk. financial contagion, triggered by the East Asian The government’s conduct of macroeconomic crisis, led to a “sudden stop” in capital inflows. A policy plays a unique and pivotal role in managing recession again followed, but this one was milder risk at the national level. Macroeconomic policies (GDP fell around 4 percent from peak to trough), that are adequately designed and implemented help and the recovery was shorter (six quarters), if weak. overcome many obstacles in managing risk, includ- At the onset of the 2008–09 global crisis, Peru’s ing asymmetric information, coordination failures, economy was better equipped to withstand the un- externalities, and the provision of public goods (see precedented external shock. A credible inflation chapter 2). Thus sound macroeconomic policy pro- targeting regime kept inflation low and stable. Real motes development. A stable macroeconomic envi- GDP had grown strongly before the crisis (6.8 per- ronment and ample revenues and resources—fiscal 225 226 WORLD DEVELOPMENT REPORT 2014 in the run-up to elections or fall prey to lobby groups competing for higher spending during upswings. Coordination failures among policy makers may also lead to less than optimal allocation of policy in- struments to meet different macroeconomic targets. Problems like these can lead to uncertainty over the course of policy, which can in turn lead to greater instability and lower growth. Box 7.1 illustrates that uncertainty in global economic policy is growing. Reducing mismanagement of macroeconomic risk requires an adequate institutional framework that attains better information, improves the quality of analysis, and develops tools to help policy makers with the host of uncertainties they face. Greater pub- lic accountability for the likely costs of their policies will increase the quality of policy making. This chap- ter argues that credible, transparent, more flexible, and sustainable macroeconomic policy frameworks C A R T O O N 7.1 Systemic risks. Individuals are unequipped to manage macro- increase a country’s resilience. It highlights examples economic risks by themselves. © Jeff Parker/Florida Today of good monetary and fiscal management and offers policy recommendations suited to countries at dif- ferent levels of institutional capacity facing different space—to finance government programs and poli- constraints and opportunities. cies reduce uncertainty and enable economic agents to concentrate on productive activities rather than on trying to mitigate high risks. Gearing macroeconomic policies toward This chapter focuses on the conduct of macro- aggregate stability economic risk management—more specifically, on Using macroeconomic policies to manage monetary and fiscal policies aimed at achieving mac- economic crises and cycles roeconomic stability. To help people manage risks, monetary and fiscal policies should be credible, pre- Reducing instability and uncertainty. Macroeconomic dictable, transparent, and sustainable. That requires volatility is a source of short-term concern and an building reputation and policy space in good times impediment to achieving long-term development by keeping inflation low and stable, having exchange goals. Beyond regular business cycle fluctuations, rate arrangements that absorb shocks from the inter- volatility disrupts households’ and firms’ saving, in- national economy, and following fiscal practices that vestment, and production decisions. It reduces the generate adequate surpluses and reduce the public ability of the financial system to transform liquid debt burden. Consequently, macroeconomic policy financial instruments into long-term capital invest- makers should behave prudently during upswings. ments, as agents in the economy become reluctant Accumulating resources in good times can help fi- to enter long-term contracts. Greater output vola- nance countercyclical policies—especially programs tility—especially when accompanied by crisis epi- that protect the most vulnerable—and desired public sodes—lowers long-term growth. Increasing output investment programs. By reducing aggregate volatil- volatility by one standard deviation leads to a 1.3 ity—especially, that associated with crisis episodes— percentage point reduction in growth per capita; these policies will reduce uncertainty and help people this decline is even more sizable (2.2 percentage plan, save, and invest for the longer term. Overall, to points) during crises.3 Macroeconomic volatility also manage risk properly, policy makers must graduate worsens income inequality and poverty, as lower- from being crisis fighters to being cycle managers.2 income segments of the population are less protected That is the goal. In reality, policy makers may lack from economic downturns. Doubling aggregate expertise and institutional capacity to manage the volatility reduces the income share of the poorest economy or may have difficulty in credibly commit- quintile of the population by 2.4 percent. Moreover, ting to risk management policies. Government offi- the average increase in income inequality during cials tend to provide resources to their constituents recessions (5 percent) tends to be larger than the Managing macroeconomic risk 227 B O X 7.1 The new normal in the world economy: Heightened macroeconomic policy uncertainty in developed countries Recent policy conflict and fiscal crisis in the world’s biggest collective major financial centers, and the higher volatility of their stock mar- economies—the United States and the European Union—have gen- kets due to increased uncertainty can have a global contagion erated considerable uncertainty, leading to concerns that firms and effect. consumers may be postponing hiring and spending decisions, stall- The probability of disasters also affects people’s perceptions of ing the recovery. Uncertainty is a subjective concept, and measuring uncertainty; their decision making is disrupted by the greater likeli- it is not easy. It can be approximated with an index of economic pol- hood of disasters. Economic disasters—as defined by a peak-to- icy uncertainty that uses three groups of observable measures. The trough cumulative drop in GDP or consumption larger than 10 per- first component quantifies newspaper coverage of policy-related cent—have a mean size of 21–22 percent, an average duration of 3.5 economic uncertainty. The second measures variability of forecasts years, and an estimated probability of occurrence of 3.5 percent a of fiscal and monetary policies (as captured by the interquartile year. Disasters have the potential to destroy part of the capital stock range of 1-year-ahead forecasts of inflation, government purchases, and impair productivity; thus they are characterized by declines in and state and local government purchases). The third reflects the investment, corporate leverage, output, and employment, and number and size of tax code provisions set to expire in future years. account for part of increased risk premium in financial markets. Cali- The weight of each of these components varies according to the brated models for the United States estimate that doubling the country or region. probability of disaster reduces investment by 3.5 percent and Historically, policy uncertainty in the United States has surged unemployment by 0.8 percent. around major wars, elections, and terrorist attacks. Recently, uncer- Will policy uncertainty remain high? The prospects for a decline tainty spiked in 2008 and has remained high (panel a). A similar in U.S. policy uncertainty in the near term are not bright, largely surge has occurred in Europe since 2008; together, the two have because of the current U.S. political agenda and the polarization of contributed to increased global policy uncertainty (panel b). its political system. In the European Union, policy uncertainty will How big is the negative impact of policy uncertainty? In the remain high if concerted actions to address banking and fiscal prob- United States, policy uncertainty of the size observed on average lems at the national and supranational levels are delayed and pro- between 2006 and 2011 reduces industrial production by 2.5 per- market reforms are not undertaken in southern European countries. cent and employment by 2.4 million workers. Moreover, policy In short, economic policy uncertainty is the new normal. With this in uncertainty in the United States and the European Union spills over mind, firms and consumers should also actively manage risks—for to the rest of the world through two channels. First, these two eco- instance, by attempting to reduce exposure to the most sensitive nomic areas collectively account for more than half the world’s sectors. Such steps can help minimize the impact of international trade and outbound foreign direct investment. Second, they are policy risk. Economic policy uncertainty has increased in recent years in the United States and Europe a. United States b. Europe 300 300 Lehman collapse 250 Debt ceiling Policy uncertainty index Policy uncertainty index dispute 200 200 Fiscal LTCM cliff 150 crisis 9/11 100 100 50 0 0 1997 1999 2001 2003 2005 2007 2009 2011 90 92 94 96 98 00 02 04 06 08 10 12 19 19 19 19 19 20 20 20 20 20 20 20 European policy uncertainty Implied volatility index U.S. policy uncertainty Sources: Bloom 2013 for the WDR 2014; Barro and Ursúa 2012; Gourio 2012. Note: The index of policy uncertainty is normalized to 100 for January 2008. The data and description of these uncertainty indexes is available at http://www. policyuncertainty.com. This index may also capture macroeconomic and political uncertainty. The implied volatility index (VIX) is a measure of the implied volatility of S&P (Standard and Poor’s) 500 index options. LCTM = Long-Term Capital Management L.P. 228 WORLD DEVELOPMENT REPORT 2014 reduction during booms (0.9 percent).4 A rising tide prone to deep recessions (figure 7.1). Structural may lift all boats, as President Kennedy famously changes in advanced countries since the mid-1980s said, but a falling tide pushes them down deeper and have led to a decrease in world output volatility. Bet- longer. Macroeconomic policy will help manage ag- ter monetary frameworks that insulate the economy gregate risks—and avoid itself becoming a source of from shocks, financial innovations that reduce mar- risk—and facilitate the development of risk-sharing ket frictions, labor market flexibility, and techno- mechanisms in the economy. logical improvements in inventory management explain this reduction.5 A stable external Managing the cycle in good times and environment and improved macroeco- coping in bad times. Macroeco- nomic frameworks not only explain nomic policies help manage ag- To manage risk the lower output volatility but also gregate shocks—coming from properly, policy reduce the probability of reces- abroad, from domestic policies makers must sions among developing coun- themselves, or from systemically tries (figure 7.1). Conversely, the important domestic agents. In graduate from being unraveling of the Great Modera- reality, developing countries’ poli- crisis fighters to being tion period was also transmitted cies have been unable to contain cycle managers. to emerging markets during the boom-bust economic and financial recent crisis. Most of these coun- cycles. These countries are exposed to tries experienced a sharp slowdown. larger and more frequent external shocks, Aggressive monetary policy actions from have lower shock-absorbing capacity (including advanced countries’ central banks succeeded in less diversified economic structures, underdeveloped restoring global financial conditions. As the world financial markets, dollarized balance sheets, and economy recovered (reflected in rising commodity poor institutional quality), and are more likely to ex- prices and lower global risk aversion), the better- perience macrofinancial crises. As a result, economic prepared emerging markets were able to resume activity in developing countries is more volatile and rapid economic growth. F I G U R E 7.1 Real economic activity is more volatile and more likely to decline sharply in developing countries a. Real GDP volatility b. Proportion of years in deep recessions 0.07 0.45 Standard deviation of real GDP growth 0.40 0.06 Proportion of years in period 0.35 0.05 0.30 0.04 0.25 0.03 0.20 0.15 0.02 0.10 0.01 0.05 0 0 High Upper Lower Low High Upper Lower Low income middle middle income income middle middle income income income income income 1960–89 1990–2010 1960–89 1990–2010 Source: WDR 2014 team based on data from World Bank World Development Indicators (database). Note: Real GDP volatility is the standard deviation of GDP growth per capita. Proportion of years in deep recessions captures the share of years when per capita GDP declined more than 5 percent on a cumulative basis. Both measures have been computed for 1960–89 and 1990–2010. Managing macroeconomic risk 229 Aiding longer-term planning and development. Macro- the private sector concerning future interest rates and economic stability broadens the set of tools for long- future inflation. Inflation targeting required having term planning available to households, firms, and flexible exchange rates, so central banks could con- governments. Increased price stability in countries duct independent monetary policy. Inflation target- that have wrestled with bouts of high and volatile in- ing has been able to deliver low and stable inflation flation rates has led to the development of financial rates.10 Its success rested on the better understand- intermediaries, including local currency debt mar- ing of the monetary transmission mechanisms and kets, the revival of mortgage markets, and the emer- on three strong institutional underpinnings.11 First, gence of pension and mutual funds (see chapter 6). central banks have a clear mandate (to maintain These developments have allowed economic agents price stability) and are fully committed to achiev- to increase the average planning horizon of spending ing that goal. Second, central banks are independent and investment decisions and raise the maturity of from political interference in their decision making. domestic government debt. For instance, if Brazil had They are instrument-independent: that is, they can the long-term inflation performance of Denmark, choose and manage the instruments to achieve their the depth of its local currency bond market would be primary goal. Third, greater central bank account- almost triple its size.6 ability creates incentives to fulfill the mandate. The effectiveness of monetary policy requires the absence of fiscal dominance, reduced currency mismatches, Delivering low inflation through sound and sound domestic financial markets.12 monetary policy The institutional push toward greater transpar- High inflation distorts the saving and investment ency has strengthened the reputation of central decisions of households and firms, thus leading to banks and enhanced efficiency in the implementa- slower economic growth. Having a clearly defined tion of monetary policy. Greater institutional capac- (quantitative) nominal target helps central banks ity and flexibility to achieve the inflation target over anchor expectations about the evolution of prices. time have permitted policy makers to use monetary To stop high inflation, countries have adopted either policy instruments countercyclically. Historically, monetary aggregates or the exchange rate as their advanced countries have been able to implement nominal anchors. Choosing an exchange rate an- expansionary monetary policies during recessions. chor (hard peg) is sensible for countries with weak They have lowered policy interest rates to withstand institutions, dollarized economies, underdeveloped real shocks and stabilize output without jeopardiz- financial markets, and low-credibility central banks. ing their inflation target. In contrast, monetary au- The exchange rate anchor is easy to implement and thorities in many developing countries have acted monitor and is verifiable by market participants. It procyclically, raising policy rates during contractions has been successful in reducing inflation from his- to avoid massive capital outflows and currency de- torically high levels.7 Small countries with greater in- preciation (fear of free falling) and cutting rates in ternational trade integration tend to have hard pegs, good times to prevent surges in capital inflows and because greater exchange rate stability boosts their currency overvaluation (fear of capital inflows).13 foreign trade and investment. Targeting monetary The adoption of sound macroeconomic policies (in- aggregates provides greater exchange rate flexibility flation targeting and flexible exchange rates), public and permits central banks to pursue independent debt management strategies, and market-friendly monetary policy. However, these regimes proved to reforms (trade and financial liberalization) have be inconsistent with long-term development. Ex- helped many emerging markets sharply reduce the change rate targeters were plagued by problems of fear of free falling (figure 7.2). It thus comes as no fiscal indiscipline and monetary financing of deficits surprise that in the midst of the recent global finan- that led to a costly regime collapse.8 Money growth cial crisis, many emerging countries in Latin America targeters gradually lost the ability to anchor expec- and East Asia were able to reduce policy rates. For tations as the relationship between monetary aggre- instance, the average monetary policy rate declined gates and inflation became unstable.9 from 6.50 percent to 1.25 percent in Peru, and from Once high inflation was defeated, monetary 5.21 percent to 1.98 percent in the Republic of Korea, frameworks targeted the inflation rate to coordinate from September 2008 to September 2009. expectations. Central banks set an inflation target to The choice of exchange rate regime matters for be achieved over a specific policy horizon. Monetary its likely effects, both direct and indirect, on infla- policy announcements shaped the expectations of tion and growth. Countries with higher per capita 230 WORLD DEVELOPMENT REPORT 2014 F I G U R E 7. 2 Monetary policy has become countercyclical in some developing countries 1.0 Becoming countercyclical Always countercyclical BWA Correlation (monetary policy rate, real GDP) over 2000–09 CHL BRB LKA 0.8 ZAF COL ZMB CZE JOR BOL ARG PER KOR 0.6 IND BGD PHL MYS KWT TZA THA PAK 0.4 TTO EGY GHA VEN UGACYP TUN BGR 0.2 TUR NPL PRY MUS ISR KEN 0 BRA DZA NGA –0.2 MAR RWA URY GMB –0.4 CRI FJI CHN –0.6 SWZ –0.8 –1.0 Always procyclical Becoming procyclical –0.6 –0.4 –0.2 0 0.2 0.4 0.6 Correlation (monetary policy rate, real GDP) over 1960–99 Source: WDR 2014 team based on data from Végh and Vuletin 2012. Note: The figure compares the correlation between monetary policy rate and real GDP in 1960–99 with that in 2000–09. Both series are detrended using the Hodrick-Prescott filter. A positive (negative) correlation coefficient signals countercyclical (procyclical) monetary policy. The countries are classified as: (a) always countercyclical when the correlations are positive in both periods; (b) becoming counter- cyclical when the correlation is negative in 1960–99 and positive in 2000–10; (c) always procyclical when the correlations are negative in both periods; (d) becoming procyclical when the correlation is positive in 1960–99 and negative in 2000–10. income and deeper financial markets have benefited rates. For instance, it takes 12 months to correct half from more flexible arrangements (floating regimes), the imbalance in the current account under float- which have delivered higher growth without higher ing exchange rates, compared with 21 months under inflation.14 Floating regimes act as shock absorb- fixed rates.17 Overall, flexible exchange rate regimes ers, helping countries accommodate adverse real are needed to guarantee the long-term viability of shocks. In the presence of rigid prices, countries with an independent and sound monetary policy frame- floating rates will have smoother responses in real work. Box 7.2 illustrates the costs of not having an output to real shocks, thanks to the faster relative independent monetary policy (including a floating price adjustment that these regimes facilitate. Growth regime at the national level) within a currency union. in per capita GDP declines 38 basis points, on aver- Before the global financial crisis, the prevailing age, for countries with flexible regimes facing a 10 monetary framework in advanced countries and percent deterioration in the terms of trade, whereas some developing countries was characterized by a the decline is 83 basis points for countries with fixed fragmented approach. Inflation targeting and flexible regimes.15 Countries with floating regimes also re- exchange rates were used to achieve low and stable cover faster from natural hazards. Output growth inflation, stabilize fluctuations in output, and facili- averages 1.6 percentage points higher in countries tate external adjustment. Meanwhile, micropruden- with flexible regimes over a three-year period follow- tial regulation and bank supervision sought to pre- ing a natural hazard, but only 0.24 percent in coun- vent excessive risk taking in the financial sector. This tries with fixed rates.16 Finally, current account ad- arrangement did not account for the fact that mac- justment is faster in countries with flexible exchange roeconomic and financial cycles are tightly linked— Managing macroeconomic risk 231 B O X 7. 2 Relinquishing monetary policy flexibility: The ultimate sacrifice? Currency unions—either unilateral or multilateral arrangements— episodes of internal devaluations have occurred among high- reduce cross-border transaction costs and foster trade among their income countries since 1990, all during periods of severe recession members. In general, trade increases between 30 and 90 percent for and growing unemployment: Hong Kong SAR, China, in the early countries that join a currency union, and between 4 and 16 percent 2000s; Japan in the late 1990s and early 2000s; and Ireland in the for Euro Area countries. Joining a currency union poses a trade-off, recent crisis. Latvia, an upper-middle-income country, engineered however: a country must give up the power to have an independent an internal devaluation, but the recovery from the recent global cri- (national) monetary policy to withstand idiosyncratic asymmetric sis has been weak so far. There is compelling evidence of price and shocks. wage rigidity and lack of nominal wage declines (downward inflexi- The recent Euro Area crisis underlines the costs of not having an bility). If trading partners’ inflation does not outpace the country’s independent monetary policy (including a floating regime at the inflation, internal devaluations will be slow and costly to achieve. national level) to accommodate negative real shocks and facilitate For instance, competitiveness gains of 5 percent over three years external adjustment. Costly coping with the crisis was exacerbated would require a decade of internal devaluations in Greece. Finally, by poor preparation of the periphery countries: they had large an internal devaluation would lead to a decline in nominal GDP— external and fiscal imbalances. In the absence of monetary policy even if real economic activity increases—thus delaying the conver- space, the burden of adjustment fell on the already stressed fiscal gence to debt sustainability. policy of these countries. Some Euro Area policy makers are contemplating the option of Lack of monetary flexibility at the national level sharply limits fiscal devaluations—say, value added tax increases and payroll tax policy options for boosting growth and lifting the country out of reductions—that have the potential to raise competitiveness faster recession. The alternative, an internal devaluation (a relative price than do internal devaluations. Whether that potential can be real- decline in the absence of exchange rate devaluation), happens ized has not yet been shown. infrequently and entails a slow and painful adjustment. Only three Source: WDR 2014 team based on Rose and Stanley 2005; Micco, Stein, and Ordoñez 2003; Shambaugh 2012; Farhi, Gopinath, and Itskhoki 2011. especially in advanced economies. For instance, real reversals that disrupt economic activity. If so, they output and credit cycles of advanced countries are in should deploy macroprudential tools to reduce the the same cyclical phase (expansion or recession) 80 procyclicality of the financial system, avoid exces- percent of the time, while the likelihood of a reces- sive bank risk taking, and increase the resilience of sion in economic activity in these countries condi- systemic institutions by imposing additional capital tional on a credit crunch is 40 percent.18 The lack of a requirements.20 Movements in monetary policy rates holistic approach led central bankers to tailor policy may prove ineffective in addressing financial bubbles, actions that addressed the trade-offs between infla- as higher interest rates may have adverse unintended tion and output but did not address the buildup of consequences on output, unemployment, and volatil- financial imbalances. Yet mopping up the effects of ity. Overall, macroprudential instruments—because burst bubbles by providing unlimited liquidity and they have more direct effect on leverage than the pol- sharply reducing interest rates—the preferred policy icy rate—can give central banks more instruments to pursued in many advanced countries—can create achieve the goals of price and financial stability. For moral hazard problems. In environments with low emerging markets, controls on capital inflows to limit interest rates and excess liquidity, such as those cur- the expansion of domestic credit and prevent the in- rently experienced in advanced economies, financial crease of currency and maturity mismatches may institutions have incentives to take excessive risks enhance financial stability.21 So far, macroprudential and expand their balance sheets. Excessive leverage policies have been effective in reducing systemic risk of financial institutions in advanced countries can be in the financial sector, but their impact on stabilizing transmitted to emerging markets through surges in output fluctuations is very much an open question. capital inflows and ensuing accumulation of finan- Box 7.3 examines the ability of one tool that is popu- cial imbalances.19 lar in the developing world to eliminate excessive These developments have reignited the debate fluctuations in real GDP: legal reserve requirements. about including financial stability among central bank mandates and expanding the policy toolkit Promoting countercyclical fiscal policy to include macroprudential instruments. Central banks need to assess whether monetary and financial Countercyclical spending by the government is conditions may lead to sharp credit and asset price needed for two reasons: to transfer resources to less 232 WORLD DEVELOPMENT REPORT 2014 B O X 7. 3 Reserve requirement policy has substituted for monetary policy as a countercyclical tool in most developing countries Unlike industrial countries that tend to implement countercyclical In other words, reserve requirement policy substitutes for monetary monetary policies, many developing countries follow a procyclical policy in its countercyclical role. stance. About half of developing countries use the legal reserve Increasing the legal reserve requirement by one standard devia- requirement both as a substitute for countercyclical monetary policy tion reduces output somewhat more than increasing the monetary and as an instrument to stabilize output. In good times, policy mak- policy rate (0.39 percent versus 0.21 percent) (see figure). In princi- ers in developing countries cut policy interest rates to reduce cur- ple, both policy instruments could be used for stabilization pur- rency appreciation pressures and instead choose to increase reserve poses. In developing countries, however, monetary policy typically requirements to cool down the economy. In bad times, they increase has been used to defend a nation’s currency and contain inflation- policy interest rates to reduce depreciation pressures and decrease ary pressures, but it has responded procyclically to fluctuations reserve requirements to help the economy get out of the recession. in output. Reserve requirement policies can help stabilizing real GDP fluctuations a. Reserve requirement policy b. Monetary policy 0.4 0.4 Change in real GDP, % 0.2 Change in real GDP, % 0.2 0 0 –0.2 –0.2 –0.4 –0.4 –0.6 –0.6 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10 Quarter Quarter Source: Vuletin 2013 for the WDR 2014. Note: The figure depicts the response of real GDP to a one standard deviation increase in the legal reserve requirement rate (panel a) and to an analogous increase in the monetary policy rate (panel b). These impulse-responses were estimated from a panel vector autoregression that evaluates the impact of exogenous changes in the reserve requirement, real GDP growth, and inflation in panel a; in panel b, the monetary policy rate was used instead of the reserve requirement. The dashed lines refer to 95 percent confidence intervals. favored individuals, and to stimulate real economic East Asian crisis. Unemployment insurance in the activity in the event of shortfalls in aggregate de- Republic of Korea was extended from 12.3 percent mand. Contractions in economic activity and crisis to about 50 percent of the unemployed from 1999 episodes affect people and especially hurt those at to 2004.23 lower income levels. The global financial crisis has Countercyclical fiscal policies in advanced coun- added 53 million people to the number living below tries are triggered by automatic increases in social the $1.25-a-day poverty line, and that number is not security and welfare spending as the economy moves expected to start declining until 2015.22 Countercy- into recession—especially during crises. Historically, clical social spending and timely stimulus packages social security spending in advanced countries in- are important to protect individuals during periods creases to 13.1 percent of GDP in the year of the fi- of hardship and stimulate employment. Historically, nancial crisis, from an average of 11.4 percent before social security policies have been implemented or the crisis.24 A reduction of 1.0 percentage point in strengthened during financial crises. For instance, the growth rate of the economy is compensated by a Mexico’s social safety net for the poor, Programa de 0.36 percentage point increase in social expenditures Educación, Salud y Alimentación (PROGRESA), was in these countries, on average. Social expenditures introduced after the 1994 “Tequila” crisis. The pen- account for more than 80 percent of the overall con- sion system in the Republic of Korea was broadened tribution of fiscal policy to stabilizing output. Auto- and universal health coverage was granted after the matic movements in pensions and health spending, Managing macroeconomic risk 233 along with unemployment compensation, have been from powerful interest groups competing for public among the largest contributors.25 Increasing trans- spending. In such situations, accumulating primary fers that are well-targeted and do not distort incen- surpluses during upswings (saving for a rainy day) tives to work (say, transfers to the unemployed and can be politically costly. Fiscal resources generated the poor) have been quite effective in the United during upswings end up being captured by govern- States at stabilizing fluctuations in output. Cutting ment agencies, state-owned enterprises, provinces or transfers (by 0.6 percent of GDP) in the United states, and rent-seekers.30 Finally, voters may also seek States increases output volatility by 4 percent, the to starve governments and reduce political rents— variance of hours worked by 8 percent, and house- especially in corrupt democracies.31 26 hold consumption volatility by 35 percent. Institutional development has helped some de- In contrast, most developing countries have veloping countries escape from the trap of fiscal pro- been unwilling or unable to implement fiscal expan- cyclicality. An improvement in the quality of institu- sions during recessions.27 Their behavior arises from tions, reflected in better fiscal institutions and sound the weakness of automatic stabilizers, their procy- fiscal rules, has helped some countries graduate from clical access to world capital markets, and political fiscal policy procyclicality. More than one-third of de- economy problems. Automatic stabilizers in devel- veloping countries now follow a countercyclical fiscal oping countries are too small to have a significant policy stance (map 7.1).32 Improved fiscal outcomes smoothing effect on real economic activity: taxa- and frameworks in emerging markets have been re- tion is regressive, coverage and benefits of transfer warded in the markets by lower sovereign spreads. For programs are low, and unemployment insurance instance, the sovereign spread of Brazil declined from is almost nonexistent.28 Spending on health, 772 basis points to 145 basis points (over U.S. education, and infrastructure behaves treasuries) from end-2000 to end-2012. procyclically in good times, and it Other factors that have contributed expands faster than other types Governments to developing countries’ graduation of spending. Social spending re- may have less need from procyclicality are increases mains fairly constant during in the depth of domestic financial for discretionary downturns, rather than declin- markets and greater credibility of ing. Deploying social spending policy action if fiscal policies.33 in bad times may thus require they have flexible The effectiveness of discretion- only building up safety margins fiscal rules. ary fiscal stimulus is under intense in good times—thereby breaking debate in academic and policy cir- the cyclical pattern of boosting pub- cles. Discretionary actions to stimulate lic spending in good times.29 Procycli- consumption and hence aggregate de- cal discretionary fiscal interventions increase mand in the short run (the so-called Keynesian output volatility and hence undermine long-term multiplier) through government spending should be growth. Heavy reliance on discretion may also cre- distinguished from steps to increase productive ca- ate greater uncertainty and lead to greater instability. pacity in the short and long run (such as public in- Governments may have less need for discretionary vestment in infrastructure). Estimating these aggre- policy action if they have stronger built-in resilience gate spending multipliers (that is, the increase in GDP or large automatic fiscal stabilizers. for every dollar in additional government spending) The procyclical bias of fiscal policy in developing is not a trivial issue. The estimation should consider countries stems partly from their generally procycli- changes in government spending that are indepen- cal access to capital markets. Governments’ inability dent from economic conditions; specifically, it needs to borrow resources abroad or at home (or to borrow to isolate the effects of government spending on out- only at very high interest rates) during downturns put from reverse causality and from the influence of leads them to cut spending or raise taxes. During other forces in the economy such as natural hazards. upswings, they have access to markets and tend to The evidence consistent with this identification strat- borrow to increase public spending. Procyclical fis- egy suggests that the use of discretionary fiscal policy cal policies are also the outcome of political distor- to stimulate demand in developing countries has not tions and distributional conflicts. Governments tend been overly successful—as witnessed over the past to spend windfall revenues (stemming from rising 30 years. The (short-term) aggregate government commodity prices or higher-than-expected growth) spending multiplier in developing countries is quite during good times, when they are under pressure small: the one-year government spending multiplier 234 WORLD DEVELOPMENT REPORT 2014 M A P 7.1 Government consumption became countercyclical in more than one-third of developing countries over the past decade Always countercyclical Becoming countercyclical Becoming procyclical Always procyclical Missing data Source: WDR 2014 team estimations based on Frankel, Végh, and Vuletin 2013 methodology. Map number: IBRD 40099. Note: The map shows the evolution of the cyclical stance of fiscal policy from 1960–99 to 2000–12. The cyclical stance is measured in a regression of the (Hodrick- Prescott) cyclical component of general government consumption expenditure on its own lagged value, and the cyclical component of real GDP. The sign of the coefficient on the cyclical component of real GDP indicates whether government consumption expenditure is procyclical (positive sign) or countercyclical (nega- tive sign). The coefficient on the cyclical component of real GDP was estimated separately for the periods 1960–99 and 2000–12. Then, countries are classified as always countercyclical (in both periods); becoming countercyclical (only countercyclical in 2000–12); becoming procyclical (only procyclical in 2000–12); and always procyclical (in both periods). The likely endogeneity of the cyclical component of real GDP was controlled for by using as instruments the (current and lagged value of the) cyclical component of real GDP of the country’s main trading partners and international oil prices, as well as the lagged value of the country’s own cyclical component of real GDP. fluctuates between 0.5 and 0.7, on average.34 A recent est rate is unresponsive to increases in government survey for the United States suggests that one-year spending—especially when interest rates are near government spending multipliers are somewhere zero. If government spending goes up for 12 quarters between 0.5 and 1.5, although, there is diversity in but the nominal interest rate does not vary, the im- the methods used to identify exogenous fiscal policy pact multiplier is 1.6.37 changes.35 Box 7.4 reviews a few contrasting experi- The effectiveness of government expenditure in ences in industrial and developing countries. building productive capacity goes beyond the ho- The magnitude of the aggregate spending mul- rizon of output impact multipliers. Public invest- tiplier, however, depends on the country’s initial ment projects, especially infrastructure projects, can conditions—and the evidence comes mostly from have lasting positive effects on GDP, investment, and developed countries. Multipliers tend to be stronger productivity—especially when the economy’s stock in recessions than in booms. The one-year govern- of infrastructure capital is relatively low.38 The evi- ment spending multiplier is about 0.5 in recessions dence shows that while the short-run impact of out- and booms; however, the response soon falls below put to government investment is 0.6 in developing zero in expansions, while it rises steadily to 2.5 after countries, its cumulative impact rises to a long-run five years during recessions.36 Aggregate multipliers value of 1.6.39 Public infrastructure projects require are less than 1.0 when central banks can adjust inter- coordination among different levels of government, est rates in response to macroeconomic conditions. and they undergo an extensive planning, bidding, However, they are much larger if the nominal inter- contracting, construction, and evaluation process. Managing macroeconomic risk 235 B O X 7. 4 Fiscal stimulus: The good, the bad, and the ugly China. China’s response to the 1997–98 East Asian financial crisis is all, purchases peaked at 0.2 percent of GDP and infrastructure spend- viewed as a successful countercyclical intervention by a developing ing at 0.05 percent of GDP in the third quarter of 2010. Finally, grants country in the face of an external shock. The stimulus was timely, provided to state and local governments to start infrastructure proj- coincided with a growth slowdown, worked primarily through ects and purchase goods and services were shifted toward reducing increases in infrastructure spending, and reversed at a good time. state and local net borrowing. In sum, the federal government stimu- Several factors account for this success. First, China’s public debt was lus compensated for collapsing state expenditures; the net increase very low before the crisis, giving China more scope to engage in defi- in government spending was negligible. cit spending. Second, sustained growth and China’s prohibition on Argentina. The risks of a mistimed expansionary change in fiscal subnational government borrowing ensured that credit-constrained policy are illustrated by Argentina in the mid-1990s. In the midst of local governments had strong need for financing for infrastructure the post-Tequila boom (1996–98), Argentina undertook an expan- projects. Third, the stimulus, although modest at 1 percent of GDP, sionary fiscal policy that further fueled the boom. This expansion was leveraged into larger investment spending. The central govern- was motivated by revenue shortfalls, partly linked to pension reform, ment participated in infrastructure projects only by cofinancing whose adverse short-term draws on public revenue were not offset them; private banks provided the rest of the financing once the gov- by increases in other taxes or reductions in spending. As a result, ernment provided loan guarantees. Finally, the fiscal expansion was the ratio of public debt to GDP kept rising despite the economic implemented during a period when growth was still very rapid. bonanza. When the 1998 Russian financial crisis erupted, triggering United States. The American Recovery and Reinvestment Act (enacted sudden stops in capital inflows among emerging markets, Argenti- in February 2009) failed to stimulate the U.S. economy from 2009 to na’s fiscal authorities had no room for maneuvering to mitigate the 2011. Temporary transfers to raise personal disposable incomes were recession. Instead, they had to engage in a severe fiscal contraction not translated into higher consumption, as recipients used these that deepened the slump and eventually led to the collapse of the resources to pay off debt and save. The amount of resources allotted monetary policy regime. to federal government purchases and services was quite small. Over- Sources: Kraay and Servén 2013 for the WDR 2014; Taylor 2011. Public infrastructure stimulus may not automati- well-functioning and already-tested projects and fi- cally translate into commensurate increases in the nancing preappraised and “shovel-ready” new proj- supply of infrastructure services because of limited ects, before embarking on untried public spending or low-quality projects in the pipeline and inefficien- projects that risk becoming “white elephants.” cies in the selection and implementation of these Fiscal expansions should focus on growth- projects. The disconnect between spending and asset enhancing spending programs or on areas where accumulation is particularly acute when governance expenditures are reversible. Such expansions will and fiscal institutions are weak, as is the case of many not jeopardize long-run fiscal and debt sustainabil- developing countries. ity. Policy makers should concentrate on projects that act as automatic stabilizers. Examples include means-tested social benefit programs that expand Elements of sound fiscal policy expansions during downturns, as more people fall below eligi- Fiscal expansions need to be credibly and sustain- bility thresholds, and then contract as the economy ably financed. Only those developing countries with recovers. Similarly, workfare programs that clearly strong fiscal positions and large reserve stocks (such pay below-market wages will attract participants in as Chile, China, Malaysia, and Turkey) can afford downturns but will not be appealing once the econ- to finance fiscal expansions. This point underscores omy recovers. The risks of unsustainable accumula- the importance of building up fiscal buffers in good tion of public debt are also reduced to the extent that times so that they are available in bad times. increases in spending occur in areas such as infra- Fiscal expansions should be timely but not structure, where costs may be recovered through fu- rushed. Timely action is a challenge in developing ture user fees. Overall, strengthening automatic sta- countries, where data quality and fiscal institutions bilizers—or designing programs that resemble them are often weak. Serious risk can arise from rushing to for use during recessions—are sustainable ways to expand public spending without adequate oversight conduct countercyclical policies. institutions and capacity to appraise new projects in Historical experience in developing countries place. Policy makers should first consider expanding challenges the notion that expansionary fiscal policy 236 WORLD DEVELOPMENT REPORT 2014 is effective. That does not mean that expansionary tingent liabilities. Explicit contingent liabilities stem policy cannot play a role in mitigating the effects of from the government’s need to meet the terms of crises, however. It does mean that recommendations contracts and regulations; examples include credit for countercyclical fiscal measures should incorpo- guarantees and public-private partnerships in in- rate the sobering lessons from past experience. Two frastructure. Implicit contingent liabilities are taken priorities should be considered in the use of expan- by the government on the basis of political commit- sionary fiscal policy. First, social safety nets should be ments (financial bailouts) or humanitarian grounds strengthened to help the most vulnerable and those (disaster relief), or provision of public goods (envi- most affected by the crisis to cope, especially in areas ronmental clean-up), for instance. Table 7.1 sum- where short-term coping mechanisms can have se- marizes public policy actions to prevent or deal with vere long-term impact, such as cutbacks in children’s recession and budgetary surprises (see also the dis- food consumption or education (see chapter 3). Sec- cussion on contingent liabilities later in this chapter). ond, government spending should focus on areas that are likely to contribute to long-term growth, Creating space to cope with downturns and such as infrastructure. build resilience Using fiscal rules appropriately. Fiscal rules have Generating sustainable fiscal resources to emerged in response to fiscal profligacy and to cor- finance stabilization policies and rect distorted incentives and contain pressures to long-term social programs overspend in good times. They typically impose year-by-year numerical limits on debt, expenditure, Making fiscal room to maneuver to cope with revenue, or budget balances. However, during the shocks and unexpected obligations recent financial crisis, these annual numerical tar- Funding for stabilization policies and long-term so- gets did not facilitate adjustment to adverse shocks, cial programs is limited by the ability of the govern- shifted expenditure composition away from social ment to save and borrow resources. In this context, and investment spending, and created incentives in creating space for policy actions requires assessing countries with large imbalances to erode transpar- the sustainability of public debt, the nature and tim- ency through the use of creative accounting.40 Rules ing of desired expenditures, the responsiveness of should recognize that fiscal sustainability is an inter- public revenues to economic activity, the exposure to temporal concept and allow for temporary deficits fiscal risks, and the government’s capacity to repay its accompanied by subsequent offsetting surpluses. debt. In the latter case, debt defaults and the ensuing Fiscal rules targeting budget balances along the cycle credit downgrades reduce the country’s creditworthi- provide flexibility to respond to shocks and meet the ness and deepen the downturn in economic activity. sustainability criteria. These rules are far from be- ing a panacea, however. Their credibility and effec- Preparing to cope with macroeconomic disasters. Se- tiveness rest on their design, adequate institutional vere macroeconomic contractions are typically ac- capacity, clear operational procedures, and effective companied by declines in public revenue collection communication strategies (see the “Focus on policy and the call for social expenditure increases. In the reform” at the end of this Report). past 50 years, middle-income countries have been Medium-term expenditure frameworks (MTEFs) in recession 14 to 16 percent of the time, and low- constitute another option to implement forward- income countries, a staggering 27 percent of the time. looking multiyear budget planning. Under these Meanwhile, industrial countries have spent 7 percent frameworks, spending may not exceed expected rev- of the time in sharp recession, as defined by cumu- enues and is allocated through medium-term sector lative declines in real GDP per capita of more than strategies. MTEFs combine top-down approaches 5 percent. The timely and appropriate response to a to allocate aggregate resources to spending agencies, collapse in aggregate demand will be limited by the bottom-up determination of the resource needs of health of the government fiscal position. spending agencies, and assessment of the links be- tween funding and results. Currently, more than Preparing to cope with budgetary surprises. Govern- two-thirds of all countries have adopted MTEFs.41 ments need to safeguard fiscal space to manage im- plicit obligations (such as social security programs) Managing assets prudently. To weather adverse or obligations arising from explicit or implicit con- shocks, some countries have accumulated reserves Managing macroeconomic risk 237 TA B L E 7.1 Policies to prevent or cope with fiscal risks Policies regarding asset management Policies regarding liability management Dealing with recessions/macroeconomic disasters Structural budget rules and sovereign wealth Fiscal consolidation: debts and deficits reduction; sound debt funds to safeguard savings in good times management strategies (reduce interest rate, currency, and refinancing risks); hedge against currency and interest rates Effective tax administration Reprioritization of expenditures Saving part of donor assistance (especially for low-income countries) Integration of risk analysis and mitigation (especially for natural disasters) into a public investment frameworks; market insurance for Commodity price hedging (for resource-rich future costs/damages of public investment countries) Social security financing: encourage personal responsibility through Sound governance for state-owned enterprises setting up privately managed, fully funded contribution funds to mitigate fluctuations of revenues Increased transparency and disclosure of budget expenditures that are legally binding in the long term; reduce earmarking Dealing with budgetary surprises Transparent tax frameworks to mitigate volatility Establishment of a fiscal risk management unit in the ministry of fiscal revenues due to tax competition of finance to manage contingent liabilities International diversification of assets to Guarantees: Build adequate incentives through risk sharing; accommodate changes in global market sentiment valuing and provisioning Well-funded deposit insurance funds Increase transparency and build credibility to avoid legal claims against the government Adequate microprudential and macroprudential supervision to mitigate the need for financial bail-outs Adequate foreign exchange reserves for uncalled capital Rules on the subnational government borrowing operations to reduce the risk of bail-outs Preparation and implementation of sovereign disaster risk financing strategies, including retention instruments (such as reserves and contingency budgets) and risk transfer mechanisms (such as parametric insurance and catastrophe bonds) Adequate infrastructure and early warning systems for natural disasters Environmental norms to prevent residual damages associated with pollution Source: WDR 2014 team. while others have set up sovereign wealth funds in average.43 International reserves as a share of GDP response to commodity price booms (such as oil- nearly tripled in upper-middle-income countries exporting countries) or large export-led booms in over the past decade (to 30.8 percent of GDP in 2010 economic activity (as in China). Reserve accumula- from 10.9 percent in 2000). Safer assets (reserves and tion can be used to limit exchange rate volatility and debt) have driven the accumulation of foreign assets cushion aggregate domestic spending during current in emerging market economies, as opposed to ad- account reversals.42 Holding reserves can be costly, vanced countries, which accumulated riskier assets however, because their return is lower than the in- (equity and foreign direct investment). Compared terest rate offered on government debt. The cost of relative to the rest of the world, emerging markets carrying reserves for the median emerging market are now in a net creditor position in safe assets, while was around 0.5 percent of GDP from 2001–09, on they are in a net debtor position in riskier assets.44 238 WORLD DEVELOPMENT REPORT 2014 The net accumulation of foreign assets and persis- sue debt in their own currency, thus reducing their tent current account surpluses led to the emergence dependence on external funding and exposure to and growth of sovereign wealth funds (SWFs). Ac- exchange rate risk. Second, public bond issuances cording to their source of revenue, they can be clas- have gradually shifted from short-term and floating sified as commodity SWFs, funded by revenues from debt to fixed rate debt—thus reducing the exposure commodity exports (such as the Government Pen- to interest rate fluctuations. Third, governments have sion Fund of Norway, and Saudi Arabian Monetary raised the average life of their portfolio by issuing Authority Foreign Holdings); and noncommodity long-term, fixed-rate instruments. Established issuers SWFs, funded by transferring assets from interna- such as Brazil, Colombia, Jamaica, Mexico, Peru, Tur- tional reserves, government budget surpluses, and key, and Uruguay have been issuing 30-year bonds privatization revenues (such as China’s SAFE Invest- since 2006. These longer maturities have allowed debt ment Company and Singapore’s Temasek Holdings). managers to reduce the risk of refinancing. Finally, SWFs have multiple goals, including stabilizing the stable domestic macroeconomic environment government revenue, managing intergenerational and financial market reforms have allowed govern- savings and pension liabilities, and making long-term ments to diversify their funding sources. Domestic investments. SWFs need legitimacy and credibility institutional investors are playing a larger role, with to protect their capital from depletion by the gov- pension funds and insurance companies increasing ernment or the current generation.45 Together, they their demand for government bonds. Foreign inves- manage more than $5 trillion in assets (with oil- and tors have increased their role as well, especially in gas-related SWFs accounting for nearly 60 percent of countries that have (or have recently obtained) in- the total), compared to world international reserves vestment grade status.49 of $11 trillion and worldwide GDP of $71 trillion in 2012.46 Countries have enacted laws and created Enhancing the scope for public debt management with institutions to set up management principles and market insurance. Market instruments help govern- investment policies for their funds. Procedures gov- ments secure the funds needed to deal with the after- erning the funding, withdrawal, and spending of the math of large negative shocks, such as fluctuations capital should be tailored to the specific fund’s goals. in interest rates, exchange rates, and commodity Stabilization and saving funds are typically made prices, and reversals of capital inflows. Some gov- up of excess (commodity or government) revenues. ernments have issued explicit state-contingent debt Fund withdrawal is sometimes flexible; however, it to hedge against some of these risks—for instance, requires an investment mandate to minimize unex- government debt indexed to GDP, exports, or ex- pected resource demands from the government.47 An port commodity prices (such as copper prices). The SWF’s spending plans should be part of a coherent hedging potential of these instruments has not yet policy framework, need to be flexible, and—if nec- been fully realized, however. State-contingent secu- essary—be able to be used to meet unexpected and rities, already traded in international markets, can large adverse shocks. For instance, Timor-Leste’s provide additional insurance, helping governments Petroleum Fund has invested in the country’s elec- build portfolios with countercyclical returns. For in- tricity grid and transportation networks. Chile has stance, the risk of reversals in capital flows (sudden drawn upon its SWF to help rebuild areas damaged stops)—driven by episodes of international flight by the 2011 earthquake (box 7.5).48 Greater account- to quality—is strongly correlated with jumps in the ability increases the credibility and effectiveness of Chicago Board Options Exchange Market Volatility these funds. All of them submit reports to the gov- Index (VIX), which measures anticipated volatility ernment on a regular basis. For instance, the Kuwait in the Standard & Poor’s 500 index. A strategy that Investment Authority has an independent board that “shorts” VIX-linked contracts may provide a good reports to the Council of Ministers. hedge against sudden-stop events. Similarly, market instruments are available to hedge against commod- Adopting sounder public debt management strategies. ity price fluctuations, and the welfare gains from do- Along with sharp debt reduction, public debt man- ing so are potentially large.50 While these measures agers in emerging markets have engineered a major offer some potential insurance against shocks to ex- shift in the risk and maturity profile of government ternal funding, their effectiveness relies on the pres- debt portfolios. First, the development of local cur- ence of deep-pocketed and informed creditors who rency bond markets has allowed governments to is- are willing to take on emerging market risks. That Managing macroeconomic risk 239 B O X 7. 5 Managing commodity revenues in Chile: An example of sound institution building and management of public resources Chile is the world’s largest producer of copper, accounting for 43 per- also established two sovereign wealth funds (SWFs). The law con- cent of world exports in 2010. Copper is an important source of gov- templated greater disclosure and transparency in the conduct of fis- ernment revenues: about half the income from copper is public, in cal policy. New government administrations are required to publish the form of tax revenues and profits from Corporación Nacional del their fiscal policy framework for their four-year term and to issue an Cobre de Chile—a state-owned enterprise that controls about one- annual report on the financial state of the government, its fiscal sus- third of the country’s copper production. How does Chile manage tainability, and its macrofinancial implications, along with an esti- this natural wealth? The answer lies in the increasing institutionaliza- mate of the CAB. In turn, the calculation of the CAB target requires tion of prudent fiscal policies for better risk management. an annual estimation of government contingent liabilities. In 2001, Chile became the only country besides Norway to cor- To finance the government’s future pension liabilities, a Pension rect for cyclical influences of the business cycle and the price of Reserve Fund was created. In good times, fiscal surpluses in excess its main commodity export goods. A fiscal rule that commits the of the structural target (and after contribution to the pension fund) government to a target level for a cyclically adjusted balance are channeled to the new Economic and Social Stabilization Fund. (CAB) aims at saving during high-revenue periods and deploying In bad times, resources are withdrawn from the stabilization fund resources in bad times, over and above the saving or dissaving tar- to finance budget deficits (including payments into the pension get reflected by the CAB. The fiscal rule provides a predictable path fund). for fiscal policy and has reduced uncertainty about fiscal revenues The international investment of the resources held in the two associated with copper prices. While the rule is a very sound mea- SWFs can be undertaken directly by the treasury or outsourced to sure, its workings are far from perfect. Chile has yet to establish an the Central Bank of Chile—or to private fund managers hired by the ex ante escape clause from the rule (that would apply under pre- central bank. The law also created a new independent committee, specified conditions) or ex post sanctions for violating the rule and the Advisory Financial Committee for Fiscal Responsibility Funds, ensuring corrections. which provides nonbinding recommendations to the ministry of Chile also made fiscal policy a cornerstone for managing finance on fund investment policies and regulations, and publishes resource revenues. The Fiscal Responsibility Law enacted in 2006 an annual report on the financial performance of the SWFs. Finan- (Law 20128) provided an institutional framework that strengthened cial statements from the SWFs are audited by an independent inter- the link between the fiscal rule and use of government savings. It national agency. Source: Fuentes 2013 for the WDR 2014. may be a problem, particularly when shocks to exter- sheet when an increase in the international reserves nal funding often originate as shocks to the creditor’s requires debt to be issued in foreign currency. In ability to lend.51 Canada, the country with the greatest degree of in- tegration, the management of both assets and liabili- Implementing a management framework for sovereign ties is assigned to one agency or ministry, which del- assets and liabilities. Sound risk management by the egates responsibilities for day-to-day management government requires the effective implementation of (for example, to the central bank) and coordinates a sovereign asset-liability management framework. the borrowing and investing programs.52 This framework requires enhanced coordination by the various governmental institutions that control Managing macroeconomic contingent and manage specific sovereign financial assets and liabilities liabilities: for example, coordination between the treasury and the central bank, when the latter is- Managing public liabilities entails not only ad- sues debt or holds windfall revenues (in the case of dressing current obligations but also focusing on commodity-exporting countries). In practice, the contingent liabilities, thus reflecting the increased management of assets and liabilities is rarely coor- awareness of the ability of these liabilities to impair dinated. Partial coordination efforts integrate the fiscal sustainability. For instance, calls on govern- management of some (but not all) balance sheet ment guarantees will trigger budgetary obligations. items. For instance, Finland, Greece, and Turkey have integrated management of the net position on cen- Controlling social security expenditures. In many tral government debt and cash reserves. In Hungary, countries, social security benefits (either publicly or the central bank prepares a consolidated balance privately provided) are implicit public guarantees. 240 WORLD DEVELOPMENT REPORT 2014 To the extent that they are politically binding (and governments need to develop fiscal frameworks to at times unavoidable), these benefits are ultimately better assess and mitigate risks associated with con- a government obligation. Disclosure of long-term tractual obligations. They should develop methods budgetary pressures associated with social security to project the costs, evaluate the merits of taking on and demographic trends permits countries to man- these liabilities, and declare the conditions under age the associated risks better. For instance, Australia, which the government will meet these obligations. New Zealand, the United Kingdom, and the United For instance, Australia and Canada have developed States have published stand-alone, long-term fiscal principles to regulate the participation of the govern- sustainability reports. European Union countries, ment as a guarantor in loan operations—including along with Brazil and Japan, report long-term fiscal the identification, pricing, coverage, and evaluation outlooks on pension and social security spending.53 of the risk. In the case of public-private partnerships The fiscal costs of social security can be mitigated in infrastructure, the governments of Colombia and by transforming these implicit open-ended guaran- South Africa have established frameworks to ensure tees into explicit but limited ones. 54 For example, the proper risk taking and the allocation of risks. The 1981 pension reform in Chile introduced privately risk allocation under these schemes is reflected in na- managed individual retirement accounts. A similar tional or international legislation.57 reform was adopted by several countries in Latin America and Eastern Europe from 1981 to 2004. Enhancing fiscal policy decision making through trans- Nevertheless, the 2008 crisis brought up shortcom- parency and disclosure. By enhancing the quality ings in the effectiveness of the model, including high of information on fiscal risks, transparency builds fees in private accounts, distributional effects, and support for prudent fiscal policies, promotes better political interference. In March 2008, Chile enacted policy actions, and leads to better risk mitigation. a comprehensive pension reform law that addressed Disclosure strengthens confidence and credibility in critical policy areas (related worker coverage, gender public sector accounts and in the sustainability of fis- equity, pension adequacy, and administrative fees) cal policy. Credibility, in turn, reduces sovereign bor- and set up a basic universal pension as a supplement rowing costs and improves the government’s access to the individual accounts system. Under the new law, to international capital markets. Greater fiscal trans- the Sistema de Pensiones Solidarias has been added parency is positively associated with improvements to the existing mandatory individual accounts to in- in a country’s credit rating: on average, credit spreads crease coverage. It also introduced a noncontributory decline 11 percent when governments choose to be- basic solidarity pension (Pensión Básica Solidaria).55 come more transparent.58 In best practice, some governments (Australia, Brazil, Chile, Colombia, In- Developing sound frameworks for explicit contingent donesia, New Zealand, and Pakistan) have published liabilities. Optimal design of contingent contracts statements of fiscal risks to their balance sheets and may reduce moral hazard by either the beneficiary hence to their policy stance. Sometimes, however, full or the guaranteed party. It is necessary to develop disclosure of government obligations may lead to public-private risk-sharing mechanisms so that the moral hazard. For instance, reporting some implicit guaranteed party or the beneficiary bears some risk. contingent liabilities may lead some agents to take For instance, coverage ratios of credit guarantees on excessive risk under the impression that the gov- should provide incentives for lenders to properly as- ernment may step in to cover any losses. Moreover, sess and monitor borrowers. Most practitioners ar- information that may endanger the government in gue that lenders should retain a significant part of the the event of litigation should not be revealed. Box 7.6 risk, from 30 to 40 percent. In practice, the median highlights the experience of Colombia in disclosing guarantee covers 80 percent of the loan, while some fiscal risks. schemes offer guarantee to lenders up to 100 per- cent.56 Governments should also consider other risk- Maintaining adequate regulatory and crisis resolution sharing mechanisms, such as termination clauses that frameworks to protect against financial bailouts. Gov- allow them to close arrangements when the instru- ernment bailout of the financial sector can both be ment is no longer needed, requirements to post col- costly and impair the sustainability of its financial lateral or to have an ownership stake, and measures accounts. The median fiscal cost of financial system to share the upside potential along with the downside bailouts in 87 crisis episodes from 1970 to 2011 was risks. Rather than implementing ad hoc mechanisms, approximately 7 percent of GDP (4 percent of GDP Managing macroeconomic risk 241 B O X 7. 6 Disclosing fiscal risks in Colombia: A path to greater transparency and credibility in risk management Colombia’s resilience to the global economic turbulence is partly ects. It uses statistical models to estimate the probability of the attributed to its sound fiscal institutions. The gradual consolidation occurrence and financial impact of risks. For instance, contingent lia- of the fiscal framework over the past 15 years was strengthened with bilities arising from PPP contracts were estimated at 0.27 percent of the enactment of the Responsibility and Transparency Law (Law 819) GDP in 2012. The creation of the National Infrastructure Agency and in 2003. This law established the Medium-Term Fiscal Framework new laws governing PPPs have also improved the technical and (MTFF), a tool for macrofiscal programming that set high standards financial structuring of investment projects and reduced incentives for fiscal policy transparency and risk management. to renegotiate contracts. The MTFF must contain a 10-year fiscal programming framework Assessing contingent liabilities related to lawsuits against the state. and an assessment of the main fiscal risks and the fiscal impact of These liabilities are numerous and have growing fiscal impact: they economic decisions adopted the previous year. This assessment is consolidated yearly in a public document submitted to Congress represent an estimated 71.1 percent of GDP for 2011–21. To assess before the budget bill is discussed. The law requires the central gov- them, a probability tree is used to account for all stages of litigation. ernment to analyze the country’s macroeconomic and fiscal perfor- It is based on historical information of similar actions in similar juris- mance and its macroeconomic framework over a 10-year horizon, dictions and on qualitative analysis. The growing fiscal impact of numerical targets for the nonfinancial public sector primary balance these liabilities is partly explained by the low quality of public and its future financial plans, the fiscal cost of laws enacted the pre- defense. The recent creation of the National Agency for the Legal vious year, and the fiscal impact of quasi-fiscal operations (from the Defense of the State is a step in the right direction to control these central bank and deposit insurance agency). costs. Evaluating nonexplicit debt and contingent liabilities. Law 448 of 1998 Determining contingent liabilities from guarantees in public credit oper- issued the first regulations for managing budgetary contingencies, ations. Assessing these contingencies is based on the estimation of giving the ministry of finance (MoF) responsibility to approve and solvency probability curves. Estimates suggest they amount to 0.22 monitor their assessment, and created a hedging mechanism, the percent of GDP for 2011–21. Contingency Fund of State Entities. In 1999, for the first time, the Dealing with contingent liabilities related to disasters. Earthquakes and National Planning Department estimated the present value of pen- floods have had high fiscal impact in Colombia.a The government sion liabilities and built an “extended balance sheet” for the public has strengthened its institutional and financial capacity to deal more sector, including an assessment of nonexplicit debt and contingent efficiently with natural hazards by creating a national coordinating liabilities. In 2003, Law 819 required the MTFF to include an annual body in 1985, called the Prevention of and Attention to Disasters Sys- assessment of nonexplicit public debt associated with pension and tem; conducting impact evaluation studies of the potential costs and severance liabilities, and a valuation of contingent liabilities related the financial capacity of the state; and creating a financial fund to to state guarantees in public-private partnership (PPP) projects, loan recover and adapt infrastructure affected recently by floods. Cur- guarantees, and lawsuits against the state. For instance, the MTFF rently, the MoF is designing a disaster risk financing strategy, which estimated the net present value of pension liabilities at 114 percent includes enhancing its management of the possible budgetary of GDP in 2012. impacts of disasters through instruments such as contingent credit Identifying contingent liabilities from guarantees for public-private part- and parametric reinsurance, and improving insurance of public nerships in infrastructure. The MoF identifies, manages, and monitors assets through risk pooling and standard insurance requirements for risks affecting the expected financial results of infrastructure proj- concession contracts. Source: Salazar 2013 for the WDR 2014. a. For example, damages from the 1999 earthquake in the coffee region totaled $1.6 billion. Floods from 2010 to 2012 caused around $4.5 billion in damage to transport infrastructure and the agricultural sector. for industrial countries, and 10 percent for devel- sound entities to monitor their financial position. oping countries). Recent banking crises have been In addition, the government needs to formulate ad- among the costliest in terms of government obliga- equate frameworks for crisis resolution that provide tions: the fiscal costs of the bailout in Iceland and clear expectations of ex post risk sharing. Ireland have exceeded 40 percent of GDP so far.59 Protection against the risk associated with finan- Designing a comprehensive strategy to manage liabili- cial bailouts requires both an adequate institutional ties associated with natural hazards. Moving from framework that regulates the behavior of financial postdisaster coping to proactive budget planning intermediaries (through limits on risk taking) and would help increase the financial and fiscal resil- 242 WORLD DEVELOPMENT REPORT 2014 ience of countries to natural hazards.60 The Disaster Provide the right incentives Risk Finance Insurance (DRFI) program proposes a sovereign disaster risk financing framework built on To improve the predictability and credibility of policy a risk-layering, bottom-up approach to help coun- responses, increase the autonomy and accountability tries formulate budget plans before natural hazards of monetary policy makers. One of the major insti- occur and reduce volatile and open-ended fiscal tutional achievements in monetary policy has been exposures to contingent liabilities associated with to shield the central bank’s policy-making decisions them. Government can effectively manage high from political interference and providing the mone- probability, low-impact events (such as localized tary authority with independence to create policy in- floods, storms, or landslides) through risk retention struments to achieve its goals. However, with greater tools like annual budget allocation and domestic re- autonomy comes greater responsibility. Advanced serves. Mexico has an annual budget allocation of countries and some emerging market economies $800 million for its National Fund for Natural Disas- have made great strides in achieving greater trans- ters. Intermediate layers of risk can be addressed with parency in monetary policy making. Currently, cen- contingent credit lines or budget reallocation, such tral banks must explain their policy frameworks, de- as World Bank development policy loans with a ca- scribe the ways they intend to reach their goals, and tastrophe deferred drawdown option (CAT-DDO). provide information about the models built to for- Costa Rica, Guatemala, Colombia, and the Philip- mulate economic policy analysis. The disclosure of pines drew down funds from their CAT-DDO after this information improves the capacity of economic natural disasters. Financial market instruments— agents to anticipate monetary policy decisions and traditional and parametric insurance, and alterna- understand the central bank’s decision-making pro- tive risk transfer mechanisms, particularly catastro- cess. Monetary policy decisions are more predict- phe bonds—are better suited to manage liabilities able when communication from the central bank is associated with low frequency, high-impact events. timely, clear, collegial, and tailored to its audience. For instance, Mexico issued the first CAT-bonds in 2006. Other risk-pooling mechanisms include re- Promote flexibility gional disaster insurance facilities established in the Caribbean and in the Pacific Islands (see chapter 8). Create the right incentives for better fiscal policy mak- Governments can also acquire commercial insurance ing by shifting toward flexible rules within an adequate that will cover the costs of environmental cleanup of institutional framework. Political authorities in many properties under their control that exceed their bud- countries have granted the monetary authority inde- geted resources. The international community— pendence, a very precise mandate to guide monetary multilateral agencies, more specifically—also pro- policy, and incentives to be more transparent and vide contingent instruments and emergency loans to accountable. These changes would also benefit fis- finance disaster risk (see chapter 8). cal policy arrangements—although to different de- grees. The goal of fiscal policy is to achieve long-term budgetary discipline while allowing for flexibility Putting it all together: What to avoid and to pursue short-term countercyclical actions. Suc- what to do to improve risk management cess requires rules that do not constrain short-term at the macroeconomic level flexibility. These flexible rules, in turn, require the development of supporting institutions. In this con- Do not generate uncertainty or greater risks text, independent fiscal agencies can help inform, Maintain the predictability of economic policies and evaluate, and implement rules-based fiscal policies. focus on sustainability. Governments should try to implement time-consistent, predictable, and sus- Build the foundation for long-term risk tainable policies. Otherwise, their policy actions may management create a vicious cycle that leads to counterproductive procyclical responses later. Policy-induced volatility Enhance fiscal policy credibility by creating indepen- raises aggregate instability and reduces growth. An dent fiscal agencies. Independent monetary policy increase of one standard deviation in fiscal policy committees have greatly improved monetary policy volatility, for example, reduces long-term economic making. Similarly, independent fiscal councils could growth by 0.74 percent a year.61 be created to monitor fiscal discipline and restrain Managing macroeconomic risk 243 policy makers from spending sprees in good times, line ministries, and a supreme auditing institution. thus helping overcome agency and common pool In addition, the design and application of a sover- problems and fostering coordination (see chapter eign asset-liability management framework entail 2). Unlike monetary policy, there is less consensus coordination among institutions controlling re- on the goals of fiscal policy (such as the appropriate sources and generating obligations.65 level of sustainable debt), and fiscal authorities have a wider set of instruments in their toolkit. Full del- Protect the vulnerable egation of policy-making decisions is implausible, given the redistributive nature of fiscal policy. How- Protect the vulnerable from the distributional conse- ever, government officials can empower councils to quences of shocks or of the policies themselves. Macro- conduct and monitor some budget procedures.62 economic policies have distributional consequences. For instance, fiscal councils can correct overly opti- Lower-income groups tend to be more affected by mistic official forecasts of budget and GDP growth external shocks and macroeconomic imbalances. by providing independent (legally binding) ones Higher-income groups tend to be more strongly af- (as in Chile and the United Kingdom) or by audit- fected by financial crises. Macroeconomic misman- ing them (as in Sweden). They can increase rule agement tends to disproportionately increase the flexibility by defining ex ante contingencies that unemployment rate among low-income households trigger escape clauses to the rule (Switzerland), and the young. Wealth shocks associated with provide positive analysis and normative equity price busts tend to disproportion- assessments of policies (Belgium and ately affect those aged 26–35 years (the the Netherlands), and identify rule The increasing most leveraged age group). Deeper deviations associated with bad complexity of financial markets tend to amplify policies. Greater accountability macroeconomic shocks, thus strengthening the strengthens the reputation of management case for tighter macroprudential these councils.63 However, fiscal regulation. More stringent regu- councils are not a panacea. They necessitates continuous lations in product markets tend to are subject to problems of time strengthening of have an adverse impact on young inconsistency, capture, and lack of institutional capacity. people and poorer segments of so- legitimacy (see the “Focus on policy ciety. The negative consequences of reform” at the end of this Report for macroeconomic shocks and policies more detail on their optimal design). In on the less favored segments of society can countries with weak institutions and capacity, a be alleviated by institutions and policies that fa- good foundation starts with more comprehensive cilitate risk sharing. Positive measures include social fiscal frameworks—including top-down approaches protection (such as unemployment benefits) and to budget planning and cooperative bargaining that policies that facilitate resource reallocation (such as impose binding budget constraints and put a pre- more flexible entry and exit of firms, more flexible mium on fiscal policy transparency.64 business regulation, trade openness, and prudent fis- cal policies).66 Ensure adequate institutional capacity to carry out macroeconomic policies and address implementa- Keep a long-run focus tion problems. The increasing complexity of mac- roeconomic management necessitates continu- Policy makers should be proactive and keep their ous strengthening of institutional capacity, which sights on long-term development. Macroeconomic should be supported by qualified staff. For instance, risk management should concentrate on managing the greater sophistication of monetary policy re- the cycle prudently and on developing systematic, gimes and fiscal rules and the management of con- credible, and sustainable policy responses. Forward- tingent liabilities demand rethinking and enhancing looking policy makers should avoid making hasty institutional arrangements. Effective policy imple- decisions during crises and instead focus on build- mentation requires a high level of institutional co- ing the resilience of individuals, households, com- ordination. Monitoring and managing fiscal risks munities, the enterprise sector, the financial system, associated with contingent liabilities may call for co- and the economy as a whole. They need to build the ordination among various risk management units, proper economic institutions that can create incen- 244 WORLD DEVELOPMENT REPORT 2014 TA B L E 7. 2 Policy priorities to improve risk management at the macroeconomic level POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge Data ata collection and dissemination Da quality Improve quality of data transparency Monetary policy tran nsparency Disclosure of fiscal risks Protection Central bank independence Inflation targeting Flexible exchange rate regime Build stronger fiscal frameworks/institutions Debt/deficit reduction Insurance monetary Countercyclical monetary policy; policy y; mech hanisms; contingent bonds Hedging mechanisms; accumulattion reserve accumulation omatic stabilizers Design better automatic auto ers stabilize Strengthen ngthen automatic Stren utomatic stabilizers and au Countercyclical so social discretionary retionary s discr social ocial spending ocial spending Coping international nal financial Support from internation credit edit lines Contingent cre institutions Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing poli- cies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably and that can be improved over time. tives to invest and thus foster long-term develop- public programs (notably, investments in human ment. Table 7.2 summarizes government policies for capital and infrastructure), and devising policies to risk management, arrayed as priorities for countries take advantage of trade and financial integration in at different levels of institutional development and the global economy will create the right incentives to capacity. Overall, safeguarding macroeconomic sta- deliver sustained growth without undermining the bility, building reserve funds to finance desirable management of aggregate volatility. Managing macroeconomic risk 245 Notes 33. Calderón and Schmidt-Hebbel 2008. 34. Kraay 2012; Barro and Redlick 2011. 1. The increase in the unemployment rate relative to precrisis 35. Ramey 2011. levels tends to be between 2 and 4.5 percentage points higher 36. Auerbach and Gorodnichenko 2012. than in a regular recession. See Calvo, Coricelli, and Ottonello 37. Christiano, Eichenbaum, and Rebelo 2011. 2012. 38. Calderón, Moral-Benito, and Servén 2013. 2. Other macroeconomic policies, including trade and financial 39. Ilzetzki, Mendoza, and Végh 2013. openness, also permit the national government to stabilize 40. Fatás and Mihov 2010; Schaechter and others 2012. the economy by sharing risks internationally. Although im- 41. World Bank 2013 portant in the policy toolkit, their discussion goes beyond the 42. Aizenman, Edwards, and Riera-Crichton 2011. scope of this chapter. 43. IMF 2011. 3. Hnatkovska and Loayza 2005. 44. Didier, Hevia, and Schmukler 2012. 4. Calderón and Levy Yeyati 2009. 45. Aizenman and Glick 2009 document that countries with 5. Galí and Gambetti 2009. SWFs have superior governance to that of other developing 6. Burger and Warnock 2006. countries, including better government effectiveness, regula- 7. Levy Yeyati and Sturzenegger 2003. tory quality, and control of corruption. The transparency and 8. Tornell and Velasco 2000. credibility of SWFs tend to reflect their national governance 9. Mishkin and Savastano 2001. benchmark. 10. Mishkin and Schmidt-Hebbel 2007. 46. Sovereign Wealth Fund Institute, SWF Rankings http://www 11. Mishkin 2004. .swfinstitute.org/fund-rankings/. 12. Fiscal dominance refers to the situation where monetary 47. Das and others 2009. policy is used to alleviate problems of fiscal sustainability 48. Ang 2012. through the monetization of public debt—and with little re- 49. Anderson, Caputo Silva, and Velandia-Rubiano 2011. gard for the inflationary consequences. 50. Borensztein, Jeanne, and Sandri 2013. 13. Végh and Vuletin 2012. 51. Fostel and Geanakoplos 2008. 14. Husain, Mody, and Rogoff 2005. 52. Das and others 2012. 15. Edwards and Levy Yeyati 2005. 53. Cebotari and others 2009. 16. Ramcharan 2007. 54. Cebotari 2008. 17. Ghosh, Qureshi, and Tsangarides 2013. 55. Kritzer 2008. 18. Claessens, Kose, and Terrones 2012. 56. Beck, Klapper, and Mendoza 2010. 19. Bruno and Shin 2013. 57. Schick 2002. 20. De Gregorio 2012; Canuto and Cavallari 2013. 58. Glennerster and Shin 2008. 21. Committee on International Economic Policy and Reform 59. Laeven and Valencia 2012 estimated the fiscal cost of bail- 2011. outs as gross fiscal outlays related to the restructuring of the 22. Chen and Ravallion 2010. financial sector. The estimate includes fiscal costs related to 23. Keum and others 2006. bank recapitalizations and excludes asset purchases and direct 24. Prasad and Gerecke 2010. liquidity assistance from the treasury. 25. Darby and Melitz 2008. 60. This paragraph draws heavily from GFDRR 2012. 26. McKay and Reis 2013. 61. Fatás and Mihov 2013. 27. Frankel, Végh, and Vuletin 2013. 62. Calmfors and Wren-Lewis 2011. 28. Suescún 2007. 63. Wyplosz 2013. 29. Arze del Granado, Gupta, and Hajdenberg 2010. 64. Dabla-Norris and others 2010. 30. Tornell and Lane 1999. 65. Das and others 2012. 31. Alesina, Campante, and Tabellini 2008. 66. Ahrend, Arnold, and Moeser 2011. 32. Frankel, Végh, and Vuletin 2013. 246 WORLD DEVELOPMENT REPORT 2014 References Calmfors, Lars, and Simon Wren-Lewis. 2011. “What Should Fiscal Councils Do?” Economic Policy 26 (68): 649–95. Ahrend, Rudiger, Jens Arnold, and Charlotte Moeser. 2011. “The Calvo, Guillermo A., Fabrizio Coricelli, and Pablo Ottonello. 2012. Sharing of Macroeconomic Risk: Who Loses (and Gains) from “The Labor Market Consequences of Financial Crises with or Macroeconomic Shocks.” Economics Department Working without Inflation: Jobless and Wageless Recoveries.” Working Paper 877, OECD (Organisation for Economic Co-operation Paper 18480, National Bureau of Economic Research, Cam- and Development), Paris. bridge, MA. Aizenman, Joshua, Sebastian Edwards, and Daniel Riera- Canuto, Otaviano, and Matheus Cavallari. 2013. “Monetary Policy Crichton. 2011. “Adjustment Patterns to Commodity Terms of and Macroprudential Regulation Whither Emerging Markets.” Trade Shocks: The Role of Exchange Rate and International Policy Research Working Paper 6310, World Bank, Washing- Reserves Policies.” Working Paper 17692, National Bureau of ton, DC. Economic Research, Cambridge, MA. Cebotari, Aliona. 2008. “Contingent Liabilities: Issues and Prac- Aizenman, Joshua, and Reuven Glick. 2009. “Sovereign Wealth tice.” Working Paper WP/08/245, International Monetary Funds: Stylized Facts about Their Determinants and Gover- Fund, Washington, DC. nance.” International Finance 12 (3): 351–86. Cebotari, Aliona, Jeffrey Davis, Lusine Lusinyan, Amine Mati, Alesina, Alberto, Filipe R. Campante, and Guido Tabellini. 2008. Paolo Mauro, Murray Petrie, and Ricardo Velloso. 2009. “Fis- “Why Is Fiscal Policy Often Procyclical?” Journal of the Euro- cal Risks: Sources, Disclosure, and Management.” Departmen- pean Economic Association 6 (5): 1006–36. tal Paper 09/01, International Monetary Fund, Fiscal Affairs Anderson, Phillip R. D., Anderson Caputo Silva, and Antonio Ve- Department, Washington, DC. landia-Rubiano. 2011. “Public Debt Management in Emerg- Chen, Shaohua, and Martin Ravallion. 2010. “The Developing ing Market Economies: Has This Time Been Different?” Policy World Is Poorer Than We Thought, but No Less Successful in Research Working Paper 5399, World Bank, Washington, DC. the Fight against Poverty.” Quarterly Journal of Economics 125 Ang, Andrew. 2012. “The Four Benchmarks of Sovereign Wealth (4): 1577–625. Funds.” In Sovereign Wealth Funds and Long-Term Investing, Christiano, Lawrence J., Martin S. Eichenbaum, and Sergio Re- edited by Patrick Bolton, Frederic Samama, and Joseph E. belo. 2011. “When Is the Government Spending Multiplier Stiglitz, 94–105. New York: Columbia University Press. Large?” Journal of Political Economy 119 (1): 78–121. Arze del Granado, Javier, Sanjeev Gupta, and Alejandro Hajden- Claessens, Stijn, M. Ayhan Kose, and Marco E. Terrones. 2012. berg. 2010. “Is Social Spending Procyclical?” Working Paper “How Do Business and Financial Cycles Interact?” Journal of WP/10/234, International Monetary Fund, Washington, DC. International Economics 87 (1): 178–90. Auerbach, Alan J., and Yuriy Gorodnichenko. 2012. “Measuring Committee on International Economic Policy and Reform. 2011. Rethinking Central Banking. Washington, DC: Brookings. the Output Responses to Fiscal Policy.” American Economic Dabla-Norris, Era, Richard Allen, Luis-Felipe Zanna, Tej Prakash, Journal: Economic Policy 4 (2): 1–27. Eteri Kvintradze, Victor Duarte Lledo, and Irene Yackovlev. Barro, Robert J., and Charles J. Redlick. 2011. “Macroeconomic 2010. “Budget Institutions and Fiscal Performance in Low- Effects from Government Purchases and Taxes.” Quarterly Income Countries.” Working Paper WP/10/80, International Journal of Economics 126 (1): 51–102. Monetary Fund, Washington, DC. Barro, Robert J., and José F. Ursúa. 2012. “Rare Macroeconomic Darby, Julia, and Jacques Melitz. 2008. “Social Spending and Disasters.” Annual Review of Economics 4 (1): 83–109. Automatic Stabilizers in the OECD.” Economic Policy 23 (56): Beck, Thorsten, Leora F. Klapper, and Juan Carlos Mendoza. 2010. 715–56. “The Typology of Partial Credit Guarantee Funds around the Das, Udaibir S., Yinqiu Lu, Christian Mulder, and Amadou N. R. World.” Journal of Financial Stability 6 (1): 10–25. Sy. 2009. “Setting Up a Sovereign Wealth Fund: Some Policy Bloom, Nicholas. 2013. “Has Economic Policy Uncertainty Slowed and Operational Considerations.” Working Paper WP/09/179, Down the World Economy?” Background paper for the World International Monetary Fund, Washington, DC. Development Report 2014. Das, Udaibir S., Yinqiu Lu, Michael G. Papaioannou, and Iva Borensztein, Eduardo, Olivier Jeanne, and Damiano Sandri. 2013. Petrova. 2012. “Sovereign Risk and Asset and Liability Man- “Macro-Hedging for Commodity Exporters.” Journal of Devel- agement Conceptual Issues.” Working Paper WP/12/241, opment Economics 101 (2): 105–16. International Monetary Fund, Washington, DC. Bruno, Valentina, and Hyun Song Shin. 2013. “Capital Flows and De Gregorio, José. 2012. “Price and Financial Stability in Modern the Risk-Taking Channel of Monetary Policy.” Working Paper Central Banking.” Economía 13 (1): 1–11. 18942, National Bureau of Economic Research, Cambridge, Didier, Tatiana, Constantino Hevia, and Sergio L. Schmukler. MA. 2012. “How Resilient and Countercyclical Were Emerging Burger, John D., and Francis E. Warnock. 2006. “Local Currency Economies during the Global Financial Crisis?” Journal of Bond Markets.” IMF Staff Papers 53 (1): 133–46. International Money and Finance 31 (8): 2052–77. Calderón, César, and Eduardo Levy Yeyati. 2009. “Zooming in Edwards, Sebastian, and Eduardo Levy Yeyati. 2005. “Flexible Ex- from Aggregate Volatility to Income Distribution.” Policy Re- change Rates as Shock Absorbers.” European Economic Review search Working Paper 4895, World Bank, Washington, DC. 49 (8): 2079–105. Calderón, César, Enrique Moral-Benito, and Luis Servén. Forth- Farhi, Emmanuel, Gita Gopinath, and Oleg Itskhoki. 2011. coming. “Is Infrastructure Capital Productive? A Dynamic “Fiscal Devaluations.” Working Paper 17662, National Bureau Heterogeneous Approach.” Journal of Applied Econometrics. of Economic Research, Cambridge, MA. Calderón, César, and Klaus Schmidt-Hebbel. 2008. “Business Cy- Fatás, Antonio, and Ilian Mihov. 2010. “The Euro and Fiscal cles and Fiscal Policies: The Role of Institutions and Financial Policy.” In Europe and the Euro, edited by Alberto Alesina and Markets.” Working Paper 481, Central Bank of Chile, Santiago, Francesco Giavazzi, 287–324. Chicago: University of Chicago Chile. Press. Managing macroeconomic risk 247 ———. 2013. “Policy Volatility, Institutions and Economic Mishkin, Frederic S. 2004. “Can Inflation Targeting Work in Growth.” Review of Economics and Statistics 95 (2): 362–76. Emerging Market Countries?” Working Paper 10646, National Fostel, Ana, and John Geanakoplos. 2008. “Leverage Cycles and Bureau of Economic Research, Cambridge, MA. the Anxious Economy.” American Economic Review 98 (4): Mishkin, Frederic S., and Miguel A. Savastano. 2001. “Monetary 1211–44. Policy Strategies for Latin America.” Journal of Development Frankel, Jeffrey A., Carlos A. Végh, and Guillermo Vuletin. 2013. Economics 66 (2): 415–44. “On Graduation from Fiscal Procyclicality.” Journal of Devel- Mishkin, Frederic S., and Klaus Schmidt-Hebbel. 2007. “Does In- opment Economics 100 (1): 32–47. flation Targeting Make a Difference?” Working Paper 12876, Fuentes, Rodrigo. 2013. “Sovereign Wealth Funds: The Case of National Bureau of Economic Research, Cambridge, MA. Chile.” Background paper for the World Development Report Prasad, Naren, and Megan Gerecke. 2010. “Social Security Spend- 2014. ing in Times of Crisis.” Global Social Policy 10 (2): 1–30. Galí, Jordi, and Luca Gambetti. 2009. “On the Sources of the Great Ramcharan, Rodney. 2007. “Does the Exchange Rate Regime Mat- Moderation.” American Economic Journal: Macroeconomics 1 ter for Real Shocks? Evidence from Windstorms and Earth- (1): 26–57. quakes.” Journal of International Economics 73 (1): 31–47. GFDRR (Global Facility for Disaster Reduction and Recovery). Ramey, Valerie A. 2011. “Can Government Purchases Stimulate 2012. “Disaster Risk Financing and Insurance Program.” the Economy?” Journal of Economic Literature 49 (3): 673–85. World Bank and GFDRR, Washington, DC. Rose, Andrew, and Tom D. Stanley. 2005. “A Meta-Analysis of the Ghosh, Atish R., Mahvash S. Qureshi, and Charalambos G. Tsan- Effect of Common Currencies on International Trade.” Jour- garides. 2013. “Is the Exchange Rate Regime Really Irrelevant nal of Economic Surveys 19 (3): 347–65. for External Adjustment?” Economics Letters 118 (1): 104–09. Salazar, Natalia. 2013. “Fiscal Risk Management for Development: Glennerster, Rachel, and Yongseok Shin. 2008. “Does Transpar- The Case of Colombia.” Background paper for the World ency Pay?” IMF Staff Papers 55 (1): 183–209. Development Report 2014. Gourio, François. 2012. “Disaster Risk and Business Cycles.” Amer- Schaechter, Andrea, Tidiane Kinda, Nina Budina, and Anke ican Economic Review 102 (6): 2734–66. Weber. 2012. “Fiscal Rules in Response to the Crisis toward Hnatkovska, Viktoria, and Norman Loayza. 2005. “Volatility and the ‘Next Generation’ Rules: A New Dataset.” Working Paper WP/12/187, International Monetary Fund, Washington, DC. Growth.” In Managing Economic Volatility and Crises: A Prac- Schick, Allen. 2002. “Budgeting for Fiscal Risk.” In Government titioner’s Guide, edited by Joshua Aizenman and Brian Pinto, at Risk: Contingent Liabilities and Fiscal Risk, edited by Hana 65–100. Cambridge, UK: Cambridge University Press. Polackova Brixi and Allen Schick, 79–97. New York: Oxford Husain, Aasim M., Ashoka Mody, and Kenneth S. Rogoff. 2005. University Press. “Exchange Rate Regime Durability and Performance in De- Shambaugh, Jay 2012. “The Euro’s Three Crises.” Brookings Papers veloping Versus Advanced Economies.” Journal of Monetary on Economic Activity 1: 157–211. Economics 52 (1): 35–68. Suescún, Rodrigo. 2007. “The Size and Effectiveness of Automatic Ilzetzki, Ethan, Enrique G. Mendoza, and Carlos A. Végh. 2013. Fiscal Stabilizers in Latin America.” In Fiscal Policy Stabiliza- “How Big (Small?) Are Fiscal Multipliers?” Journal of Mon- tion and Growth: Prudence or Abstinence?, edited by Guillermo etary Economics 60 (2): 239–54. E. Perry, Luis Servén, and Rodrigo Suescún, 75–120. Washing- IMF (International Monetary Fund). 2011. “Assessing Reserve ton, DC: World Bank. Adequacy.” IMF, Washington, DC. Taylor, John B. 2011. “An Empirical Analysis of the Revival of Keum, Jaeho, Jiyeon Chang, Deok-Soon Hwang, Dong-Heon Fiscal Activism in the 2000s.” Journal of Economic Literature Kim, Jooseop Kim, Byung-Hee Lee, Kyu-Yong Lee, Seong-Jae 49 (3): 686–702. Park, and Kil-sang Yoo. 2006. Employment Insurance in Korea: Tornell, Aaron, and Philip R. Lane. 1999. “The Voracity Effect.” The First Ten Years. Seoul: Korea Labor Institute. American Economic Review 89 (1): 22–46. Kraay, Aart. 2012. “How Large Is the Government Spending Multi- Tornell, Aaron, and Andrés Velasco. 2000. “Fixed versus Flexible plier? Evidence from Lending by Official Creditors.” Quarterly Exchange Rates: Which Provide More Fiscal Discipline?” Jour- Journal of Economics 127 (2): 829–87. nal of Monetary Economics 45 (2): 399–436. Kraay, Aart, and Luis Servén. 2013. “Fiscal Policy as a Tool for Végh, Carlos A., and Guillermo Vuletin. 2012. “Overcoming the Stabilization in Developing Countries.” Background paper for Fear of Free Falling: Monetary Policy Graduation in Emerg- the World Development Report 2014. ing Markets.” Working Paper 18175, National Bureau of Eco- Kritzer, Barbara E. 2008. “Chile’s Next Generation Pension Re- nomic Research, Cambridge, MA. form.” Social Security Bulletin 68 (2): 69–84. Vuletin, Guillermo. 2013. “Effects and Role of Macroprudential Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises Policy.” Background paper for the World Development Report Database: An Update.” Working Paper WP/12/163, Interna- 2014. tional Monetary Fund, Washington, DC. World Bank. 2013. Beyond the Annual Budget: Global Experience Levy Yeyati, Eduardo, and Federico Sturzenegger. 2003. “To Float with Medium-Term Expenditure Frameworks. Washington, or to Fix: Evidence on the Impact of Exchange Rate Regimes DC: World Bank. on Growth.” American Economic Review 93 (4): 1173–93. ———. World Development Indicators (database). World Bank, McKay, Alisdair, and Ricardo Reis. 2013. “The Role of Automatic Washington, DC., http://data.worldbank.org/data-catalog/ Stabilizers in the U.S. Business Cycle.” Working Paper 19000, world-development-indicators. National Bureau of Economic Research, Cambridge, MA. Wyplosz, Charles. 2013. “Fiscal Rules: Theoretical Issues and His- Micco, Alejandro, Ernesto Stein, and Guillermo Ordoñez. 2003. torical Experiences.” In Fiscal Policy after the Financial Crisis, “The Currency Union Effect on Trade: Early Evidence from edited by Alberto Alesina and Francesco Giavazzi, 495–525. EMU.” Economic Policy 18 (37): 315–59. Chicago: University of Chicago Press for NBER. S P OT L I G H T 7 Diseases without borders: Managing the risk of pandemics Pandemics such as influenza (flu), AIDS, plague, and smallpox have caused episodes of overwhelming misery and economic and social disruptions throughout history. Today, a pathogen that originates in a chicken flock or a goat herd in a remote village in Asia or Africa and is then transmitted to humans can reach major cities on all continents within 36 hours. Because everyone is vulnerable, management of pandemic risk is the quintessential global public good that can yield benefits for all but can be supplied only through collective action. Any country’s efforts to reduce the risk are of limited benefit unless all other countries take supportive measures. Sources of pandemics and development implications on—a failure of public health service delivery—AIDS Pandemics do not start in a vacuum; their onset is shaped spread unchecked for decades. The costs of this manmade by human action. A staggering 2.3 billion infections by zoo- delay are still rising. In contrast, prompt public health ac- notic (animal-borne) pathogens afflict people in develop- tion to isolate infected people helped stop the SARS out- ing countries every year. Some 75 percent of pathogens break. Contagion is far less likely to take off in countries capable of causing human disease are now animal-borne. that detect disease early and implement effective control This is a major concern because health, nutrition, and food measures promptly. To date, no mechanism ensures the and income security all decline when livestock and people strengthening of veterinary and human public health sys- are diseased. The poorest, often living close to livestock or tems in countries that are unable to detect and control dis- wild animals, are most vulnerable. This disease burden per- eases, although such “weak links” put all countries at risk. sists because of weak veterinary and human public health systems that fail to detect diseases and allow them to Mitigating impacts of a pandemic spread. Adding to the risks, livestock numbers are projected Contingency planning, and periodic simulation exercises to grow very quickly in developing countries. Some patho- by governments, firms, and communities, as part of disaster gens spread not just across species but also through trade risk preparedness, can mitigate impacts. Health sector and travel across borders and continents. Even worse, some plans can help cope with surges of patients. Networked in- become capable of easy human-to-human spread and thus dustries like power, transport, finance, and food distribu- have great impact, like AIDS, flu, or severe acute respiratory tion can avoid major disruptions when the main firms have syndrome (SARS). Any country’s failure to stop contagion business continuity plans. Likewise, security and other early at its animal source can cause a pandemic. A severe flu government services need operational continuity plans in pandemic could more than double the total burden of dis- the event of high worker absenteeism. Communications in- ease. Moreover, economic activity would suffer from worker country and across borders are vital, as the differing degrees absenteeism, cascading service disruptions, and human of SARS contagion within Canada clearly demonstrated. reactions to fear and rumors, which can spread faster than Advance planning for truthful, complete, coordinated, and the disease itself. Much of the economic costs would result timely communications about the disease and government from avoidance behaviors; these costs could account for as responses can reduce uncertainty and rumors. Too few gov- much as 60 percent of total economic costs. Poor countries, ernments, communities, and firms make and test contin- especially fragile and conflict-affected states, may be least gency plans for complex disasters (including pandemics), able to cope. despite evidence that these activities are highly beneficial. In short, pandemics are an undermanaged risk. Pan- Preventing a pandemic demic prevention and preparedness tend to be sidelined, To stop contagion, it is essential to act early, at the source, especially in the health sector, where the responsibility of- and quickly. Early warning requires cooperation from farm- ten rests. Health authorities focus on immediate problems ers and communities. If farmers who report disease are and do not readily work with veterinary authorities to pre- punished by having their livestock destroyed without com- vent diseases of tomorrow or coordinate societywide pre- pensation, they will hide disease from the authorities. The paredness. Why such neglect? The economic and social main cause of pandemic risk is low capacity of public vet- impacts of contagion are often ignored, so the total risk is erinary and human health systems. Bringing them up to underestimated. Recent experience shows how wide this meet minimum international standards requires only mod- gap can be. The 2003 SARS outbreak, which killed about 10 est resources: $3.4 billion a year for all developing coun- percent of the 8,000 people it infected, caused $54 billion tries, compared with the current level of barely $450 mil- in economic damage (mostly canceled travel, lost retail lion. The expected annual benefits of robust systems are at trade, and associated cross-border economic shocks). A se- least $37 billion, more than 10 times the costs. Because vere flu pandemic could cost 4.8 percent of global gross public health authorities failed to detect the disease early domestic product (GDP), or more than $3 trillion, trigger- Spotlight 7 249 B O X S 7.1 An emergency response to a top global catastrophic risk How H5N1 avian flu galvanized the international community Why an emergency response? Two goals: How did it work? • Control H5N1 avian flu at its source in poultry to reduce pandemic • Initiated by the United States and the European Commission, risk to humans and the world economy. the International Partnership on Avian and Pandemic Influenzas • Prepare all countries to cope with a pandemic. engaged all countries. • After launch at the United Nations (UN) General Assembly in 2005, Results? Notable achievements, but risks remain: political support was galvanized at five ministerial conferences in • Largest global public health program to date reduced risk through 2006–10. prevention and preparedness; assistance was delivered quickly to • A UN System coordinator worked with the World Bank to support over 100 developing countries. the World Health Organization (WHO), the World Organisation for • H5N1 avian flu was controlled in most of the 63 countries in Africa, Animal Health (OIE), the Food and Agriculture Organization of the Asia, and Europe where it appeared but still circulates in a dozen United Nations (FAO), and others to develop a common strategy countries. Preventing renewed spread of this virus is technically rapidly. possible and cost-effective, yet most of the required investments • Financing of $3.9 billion in 2005–10 helped over 100 developing in veterinary and human public health systems are unfunded. countries. No new fund was created; instead, the World Bank mon- • Preparedness for pandemics was boosted, as evidenced by itored financing gaps in recipient countries and organizations. responses to the 2009 H1N1 flu pandemic. • Partners agreed on a strategy to reduce health risks at the animal- • The onset of the financial crisis in 2008 and decline in media atten- human-environment interface (One Health), steered by the WHO- tion sidelined pandemic prevention, leaving weaknesses in veteri- OIE-FAO tripartite. nary and public health systems unaddressed and undermining • The United Nations and partners founded a network for prepared- the sustainability of investments made. ness, the Towards a Safer World Initiative, but sustainable funding and other support remain uncertain. Source: WDR 2014 team. ing a global recession. The international community has not yet ex- ternational community rapidly mobilized and deployed resources for pressed its demand for pandemic risk reduction by adopting explicit zoonotic disease control and pandemic preparedness, but the effort goals but instead resorts to emergency responses that contain the dissipated soon after the threat left the headlines. threat only temporarily (box S7.1). Because risk governance is not Current and future generations would benefit if the international backed by resources and authority, numerous weak links persist in community set a goal to reduce pandemic risk. The goal would em- global defenses against contagion. Fearing trade and travel restric- power international organizations to raise risk awareness and moti- tions, countries may hide diseases, facilitating their spread; incentives vate prevention and preparedness; provide relevant knowledge, ca- for compliance with disease reporting are weak or absent. Govern- pacity building, and technical assistance to developing countries; ments with resources try to protect their own populations by spend- assess the performance of national veterinary and human public ing large amounts on ex post measures (vaccines, masks, and antivi- health systems and their collaboration; and mobilize resources for rals), although prevention at the source of the threat has much higher strengthening these systems. benefit-cost ratios. In particular, the promising One Health approach Sources to reducing disease risks through systematic collaboration between Barrett, Scott. 2007. Why Cooperate? The Incentive to Supply Global Public Goods. animal and human health services is underfunded. New York: Oxford University Press. Burns, Andrew, Dominique van der Mensbrugghe, and Hans Timmer. 2008. “Evaluat- Promising precedents ing the Economic Consequences of Avian Influenza.” World Bank, Washing- ton, DC. The international community has already eradicated two devastating Jonas, Olga. 2013. “Pandemic Risk.” Background paper for the World Development scourges: smallpox in 1979, and rinderpest (cattle plague) in 2011. Report 2014. Smallpox killed as many as 500 million people in the 20th century Mariner, Jeffrey C., James A. House, Charles A. Mebus, Albert E. Sollod, Dickens alone. Rinderpest, with its high fatality rate, decimated herds and Chibeu, Bryony A. Jones, Peter L. Roeder, Berhanu Admassu, and Gijs G. M. van ’t Klooster. 2012. “Rinderpest Eradication: Appropriate Technology and Social economies for centuries and catalyzed the founding of the World Or- Innovations.” Science 337 (6100): 1309–12. ganisation for Animal Health (OIE) in 1924. Intergovernmental coop- The SARS Commission. 2006. “The Story of SARS.” In The SARS Commission Final eration, science-based disease control strategies, mass vaccination, Report: Spring of Fear, vol. 2, 38–429. Ontario: Commission to Investigate the In- and surveillance were among the elements behind these successful troduction and Spread of SARS in Ontario. campaigns. The disease risk was reduced to zero. The benefits are last- UNSIC (United Nations System Influenza Coordination) and World Bank. 2010. Animal and Pandemic Influenza: A Framework for Sustaining Momentum. Fifth ing and already outweigh the control costs many times over. Interna- Global Progress Report. Bangkok: UNSIC. tional coordination and strong public health agencies broke the chain World Bank. 2012. “People, Pathogens, and Our Planet: The Economics of One of transmission of SARS. Faced with the H5N1 avian flu threat, the in- Health.” World Bank, Washington, DC. Warming Arctic waters: a risk with global impact. Scaling up risk management requires a cohesive international community that works collectively and has the capacity to mobilize resources and establish mechanisms to enforce agreements. © Ira Meyer/National Geographic CHAPTER 8 The role of the international community When risks exceed national capacity tional communi erna ty Global problems call for Int national community that takes on the global players task of delivering it.2 Global pub- lic goods benefit all countries and Unmanaged risk does not respect populations, but they are likely to boundaries. Once triggered, pan- yield the greatest benefit to those demics and financial crises can countries whose weak infrastruc- circle rapidly around an increas- tures and limited access to coping ingly interconnected globe. Con- tools make them less equipped to flicts can quickly spill over into deal with the adverse consequences neighboring countries. Droughts, of these kinds of risks. floods, and violent storms can dev- This chapter looks at the circum- astate an area, a country, or an entire re- stances in which the international commu- gion. Left unmitigated, climate change is likely nity—defined here as a collection of organizations to intensify all these risks. Moreover, each of these of global cooperation, providers of development fi- risks is capable of reversing gains in development nance and expertise, global standard setters, policy and jeopardizing the well-being of generations. makers, global charities, other nongovernmental or- The increasing interconnectedness of the world— ganizations, global media, and the scientific commu- through trade, communications, travel, information, nity—has a role to play in helping people and their and finance—has made possible the rapid economic governments manage risk and pursue development growth that has helped reduce poverty and open opportunities (diagram 8.1). The chapter explores opportunities for the developing world.1 But that what the international community can do that other same interconnectedness also magnifies the poten- economic and social systems cannot—and what it tial impact of these global risks and complicates their should (or should not) do to scale up collective ef- management. forts to manage risk. The chapter does not intend No one country or agent acting alone can deal ef- to address all possible risks at the global scale, but fectively with a risk that crosses a national border. instead aims to illustrate the common factors that Clearly, risks that spread across and affect multiple enhance or undermine the effectiveness of actions by countries or generations merit international atten- the international community. It does so by focusing tion. But international action is also justified when a on five areas of risk: disasters, global financial crises, country-specific shock is simply too large for a coun- environmental risks, pandemics, and risks associated try to resolve on its own, even when its implications with fragile and conflict-affected states (FCSs). do not go beyond national boundaries. Managing In so doing, the chapter underscores a key mes- these kinds of risks becomes a global public good, sage of this Report: risk management requires shared whose benefits also transcend boundaries, providing responsibility and actions by various economic and a central rationale for collective action by an inter- social systems, from households to the international 251 252 WORLD DEVELOPMENT REPORT 2014 tively by facilitating information sharing, devoting more resources to capacity building, and protecting D I A G R A M 8 .1 The agents of the international the most vulnerable. The international community community must have the capacity to mobilize resources and establish mechanisms to enforce agreements, even when some countries are unwilling to cooperate, by Global using its various tools to realign incentives around cooperation basic, common goals. organizations Global standard- setters, regulators, International What circumstances call for action by the international financial international community and why? courts institutions Countries face many risks that may overwhelm national capacity—some country-specific, others shared; some simple, others more complex; some stemming from natural causes, and others gener- Global ated by actions of other actors. Regardless of the type Global civil society and cause, some risks are just too big for countries to media (NGOs) handle alone. Beyond the national level, the interna- Global tional community provides assistance, expertise, and scientific collaboration to better manage these risks. community Severe economic and humanitarian crises Source: WDR 2014 team. Crises and disasters can put severe strains on people Note: NGO = nongovernmental organization. and the systems that support them. The interna- tional community can support people where na- tional authorities, usually the first line of support, community. The international community is called are overwhelmed by a negative shock that can expose on when managing risks requires efforts and public economies to significant volatility, distress public re- goods that go beyond the capacity of national eco- sources, disrupt access to markets, and retard prog- nomic and social systems. It can strengthen national ress in development, as well as where states might be efforts to manage risks that cross borders or genera- cut off from access to international resources.3 The tions or that produce outcomes exceeding the capac- international community’s risk-sharing tools can be ity of a country to manage alone. Its interventions particularly useful for lower-income countries that aim to generate and disseminate global knowledge are disproportionately affected by economic risks and expertise that improve risk awareness and the and disasters (as measured in lives lost and damages capacity to assess and manage risk, set rules and stan- relative to economic size), given their greater expo- dards to make negative outcomes less likely or costly, sure to such shocks, weak institutional capacity, and and strengthen coping capacity through quick mobi- limited access to insurance, credit markets, and other lization of global resources. tools that can mitigate their effects (map 8.1; figure These are lofty goals. In practice, the interna- 8.1). Even in developed countries, prolonged periods tional community has not been very effective in of uncertainty and weak economic activity following managing risks that transcend boundaries—but it crises can reverse years of economic and development can do better. Too often knowledge does not result advances and change people’s lives dramatically.4 The in effective action. Less emphasis is put on preparing sovereign and financial crisis in the Euro Area and the for risk than on responding to it after the fact. Some- subprime mortgage crisis in the United States appear times, too much weight is placed on avoiding risk to have undone 7 to 12 years of economic progress in rather than on managing it to seize development several countries hardest hit by the crisis, resulting in opportunities. Diverging national interests and risk unprecedented levels of unemployment and loss of management capacities undermine cooperation economic and social well-being.5 and the effectiveness of global efforts. Scaling up People living in fragile and conflict-affected states risk management requires a cohesive international face these risks on an ongoing basis. More than 1.5 community that enables its actors to work collec- billion people live in these states, in environments The role of the international community 253 M A P 8 .1 Developing countries in general have suffered higher mortality associated with disasters Most affected Least affected Missing data Source: WDR 2014 team based on data from EM-DAT OFDA/CRED International Disaster Database and World Bank World Development Indicators (database). Map number: IBRD 40100. Note: The map depicts the extent to which countries experience deaths from disasters, with the number of deaths scaled by population, averaged over the period 1990–2011. Countries are divided into equally sized categories from the most affected to the least affected. F I G U R E 8 .1 Damages from a disaster can exceed a country’s annual GDP in developing countries 25 Up to 11% of GDP Damages, average 1988–2011, US$ billions 20 Up to 14% 15 of GDP 10 5 Up to 127% of GDP Up to 365% of GDP 0 High income Upper middle Low income and Small island states countries income countries lower middle income countries Source: WDR 2014 team based on data from EM-DAT OFDA/CRED International Disaster Database and World Bank World Development Indicators (database). Note: GDP = gross domestic product. 254 WORLD DEVELOPMENT REPORT 2014 typically characterized by corruption and weak- can provide free passage to pathogens that cause in- nesses in governance and institutional capacity, with fectious diseases, some of which can travel around the minimal access to functioning market mechanisms world in less than 36 hours.7 Indeed, in recent years, or governments that can help them manage risk— greater mobility of people and goods has contrib- much less protect them from additional risks. These uted to the spread of zoonotic diseases that originate people make up 15 percent of the world population, in animals but then transfer to humans. The H5N1 but they represent nearly one-third of people in ex- (avian) flu killed 59 percent of the confirmed human treme poverty, one-third of the HIV-related deaths in cases in Asia and the Middle East during 2003–13. poor countries, one-third of people lacking access to The H1N1 (swine) flu killed an estimated 151,700– clean water, one-third of children who do not com- 575,400 people during the first year the virus cir- plete primary education, and half of all child deaths.6 culated (2009).8 AIDS, which also originated in State fragility and violent conflict pose significant animals, continues to destroy lives globally, although risks not only to citizens but to global and regional improved treatment has slowed fatalities since 2004; security in an interconnected world. By improving still, an estimated 1.7 million people died from AIDS incomes, economic prospects, and the environment in 2011 alone, and another 2.5 million contracted the for health, security, and education, sustained and disease. At the same time, globalization and scientific well-targeted engagement by the international com- advances have improved understanding of many munity can help reduce social and economic tensions pathogens, including how they can be detected and that inflame conflict, and instead create an environ- diagnosed rapidly to enable disease control. Global- ment that nurtures development opportunities. ization also supports greater collaboration among scientists and public health officials and enables the media to inform people even in remote areas of risks Risks that recognize no boundaries (see spotlight 7 on managing pandemics). Some risks have implications that cross geographi- Global efforts are also essential where risks may cal borders—or even generations. The consequences evolve slowly, with few immediately visible implica- of the risks taken today may not be visible for many tions. HIV/AIDS was not detected until well after years. Given their complexity, individual risk man- it had been established in populations around the agement actions may be insufficient, or even made world. Climate change risk is another example that ineffective by others’ actions. The international com- has been building slowly and nearly invisibly for munity can facilitate risk sharing across countries generations (box 8.1). Climate extremes such as heat and generations in cases where managing risks col- waves and heavy precipitation have been increasing lectively encourages complementarities across indi- for the past 50 years and are expected to worsen as vidual actions and enhances their impact. atmospheric concentrations of greenhouse gas emis- Global financial and economic crises are clear ex- sions reach unprecedented levels, with potentially amples of cases where risks may transcend national catastrophic and irreversible consequences.9 While borders. As the links intensify, problems originating all countries are vulnerable to the effects of climate in one country can introduce turmoil and under- change, developing countries are disproportionately mine development elsewhere. International integra- affected because they have the least capacity to pre- tion therefore presents a double-edged sword for risk pare and cope; three-quarters of the people killed management: it can create more opportunities for in disasters in the past two decades lived in lower- international risk sharing and help countries diver- income countries and small island states.10 The sify idiosyncratic shocks, but it can also generate new global community has a responsibility to help the types of risk through economic and financial conta- world’s most vulnerable people and provide broader gion. The ongoing global financial crisis that origi- and longer-term perspectives to tackle these risks. nated in advanced countries, for instance, has damp- Risk management actions by one country (or a ened growth through close economic and financial generation) may also create additional risks and un- linkages across countries and led to a slowdown in dermine stability and development efforts of others. the progress toward meeting Millennium Develop- For example, national policies to promote growth ment Goals (MDGs); 50 million more people fell and escape poverty risk may create growing pressures into extreme poverty in 2009 alone and 64 million on shared resources (such as oceans, waterways, fish additional people had become poor by 2010. stocks, and the atmosphere), resulting in degrada- Health risks can also cross national boundaries. tion of resources that hurt other countries and future Increased air travel and trade in goods and services generations (the so-called tragedy of the commons). The role of the international community 255 B O X 8 .1 Climate change and implications for development Climate change is the rise in Earth’s temperature associated with threat to development for both current and future generations; the increased atmospheric concentrations of heat-trapping green- estimated cumulative cost from damage to health, food security, house gases such as carbon dioxide (CO2). The steady rise, dating and the physical environment ranges from $2 trillion to $4 trillion by from the Industrial Revolution, has been attributed largely to 2030 depending on the climate scenario. human activities, including the burning of fossil fuels and defores- Mitigating climate change is a prime example of a global public tation. New research into climate change suggests that Earth is good that requires collective action. Collective action is needed warmer today than at any time during the past 11,300 years as because while each country prefers that others supply the good greenhouse gas concentrations have continued to rise: the con- (free-riding on others), each also recognizes that if everyone centration of the main greenhouse gas, CO2, rose from its pre- depended on others to supply the good, the result would be bad for industrial level of 278 parts per million (ppm) to a daily average of everyone, suggesting that there is an advantage in collective provi- 400 ppm in May 2013, approaching the 450 ppm threshold that cor- sion. Climate change mitigation faces several important obstacles. responds to a likely increase in Earth’s temperature of more than First, despite improved confidence in climate models, significant 2oC—the warming level that the international community commit- scientific uncertainty remains on the critical warming thresholds ted itself to avoid because of its potentially catastrophic and irre- (so-called tipping points) and on the magnitude of climate change versible consequences (see panel a). effects. Second, climate change effects are not uniform across coun- The effects of climate change are already visible in widespread tries, creating diverging incentives for action. The absence of a melting of Arctic glaciers, rising sea levels, and higher frequency global authority to enforce cooperation across nations undermines and severity of extreme weather events and natural hazards (see collective efforts, combined with the free-riding problems, as each panel b). If concentrations of greenhouse gases continue unabated, country hopes that others will bear the cost of climate change miti- a warming of more than 4oC could occur as early as the 2060s, with gation. Third, short-termism and different valuations of ecosystems, large-scale impacts on human and ecological systems—including, biodiversity, and loss of life breed inaction and pass the risk to future heightened risk of inundation of coastal areas, spread of infectious generations. Despite general consensus that it is a serious threat, diseases, declining water and food security, destruction of habitats and decades of debate and negotiations notwithstanding, climate for many species, and adverse social and economic consequences change risks are likely to grow until these challenges are effectively of large displaced populations. Climate change is hence a serious addressed. a. Rising temperature and CO2 concentrations b. A changing pattern of natural disasters 0.7 420 100 4.1 2.6 Othera Base period for anomalies: 1951–80 8.9 Earthquakes 0.6 90 27.8 11.6 Epidemics 0.5 390 80 70 Degrees, Celsius 0.4 3.2 60 ppm 0.3 360 Other Percent 24.9 48.2 weather- 50 related 0.2 eventsb 40 0.1 330 30 0.0 20 40.0 –0.1 300 28.8 Storms 1959 1969 1979 1989 1999 2009 10 Temperature anomalies, 5-year average 0 CO2 emission concentrations (right axis) 1903–62 1963–2012 Source: WDR 2014 team based on Aldy, Orszag, and Stiglitz 2001; Barrett 2003, 2007, 2008; Cole 2007; DARA International 2012; IPCC 2007; Jacoby, Rabassa, and Skoufias 2011 (for loss estimates); Lenton and others 2008; Marcott and others 2013; Mercer 2011; Stern 2007; World Bank 2009, 2012c; and data from EM-DAT OFDA/CRED International Disaster Database; NASA Goddard Institute for Space Studies Surface Temperature Analysis (database); and Scripps Institution of Oceanography, Atmospheric CO2 Concentration at Mauna Loa Observatory, Hawaii (database). Note: CO2 = carbon dioxide. a. “Other” refers to volcanoes, insect infestations, and complex disasters. b. “Other weather-related events” refers to floods, droughts, extreme temperatures, and wildfires. 256 WORLD DEVELOPMENT REPORT 2014 Dams to control water levels and retain water can af- global community can step in when risks cannot be fect water security for millions of downstream users resolved effectively by individual countries or such in neighboring countries. In each of these examples, regional bodies. Regional or international courts countries acting in their own interest obtain imme- of justice, for instance, may deal with cross-border diate gain from their actions, while losses from the disputes that cannot be resolved because of mutu- impact of adverse consequences are not felt imme- ally exclusive demands. Treatment of these issues de- diately. If all countries try to safeguard their own in- serves more extensive discussion and analyses than terests, individual actions can collectively cause large are possible within the scope of this Report. damages to all involved, in some cases with irrevers- ible consequences. How does the international community Similar beggar-thy-neighbor policies and col- enhance risk management? lective action failures are observed in international finance and trade. National measures to protect the The international community supports national ef- domestic financial system by ring-fencing affiliates of forts to manage risks by addressing some of the key cross-border banks may reduce contagion risks and obstacles to effective management of risks that go be- fiscal costs of a failing foreign bank, but they may also yond national capacity: information gaps, limited ac- weaken the resilience of the home country financial cess to markets and resources, externalities imposed system, raise the cost of capital and liquidity in both by actions of other actors, and cognitive and behav- home and host countries, and limit the ability of ioral biases. Members of the international commu- banks to manage funding risks. Fear of ring-fencing nity contribute to strengthening the key pillars of may induce global banks to pull out of other host risk management defined in chapter 1: knowledge, countries, hurting those with less developed finan- protection, insurance, and coping (diagram 8.2). cial markets. Similarly, history points to how inter- national trade collapsed when many countries intro- Generating and disseminating global duced beggar-thy-neighbor trade policies during the knowledge and expertise Great Depression.11 More recently, increased export barriers by exporters and reduced import tariffs by Lack of relevant knowledge is a key obstacle to ef- importers during the food price crisis of 2008 caused fective risk management. Knowledge deficiencies the world price of grain to jump, forcing other coun- become more formidable as risks grow in intensity tries to adopt similar measures. These uncoordinated and complexity and as the uncertainties about their actions turned out to be completely unsuccessful in sources, drivers, and potential impacts deepen. Lack- protecting the poor against the food shock—close ing knowledge, countries or individuals may con- to half of the increase in the world price of rice is tribute to, or overlook, environmental risks; spread, estimated to have come from countries’ attempts to or fail to protect against, communicable diseases; insulate themselves from higher rice prices.12 or take excessive risks in search of high returns. In Not all risks that exceed national borders are truly this context, knowledge becomes a global public global, however. Some risks, such as armed conflict good that contributes, or limits damage, to develop- between neighboring countries, may affect only ment. The international community plays an impor- a few countries, as may disputes over natural re- tant role in supplying it. sources, such as those arising from management of International financial institutions (IFIs)—such waterways. Such risks may be more appropriately or as the International Monetary Fund (IMF), the efficiently managed by bilateral or regional institu- World Bank, the Bank for International Settlements, tions that provide appropriate forums, frameworks, and the Organisation for Economic Co-operation and incentives for addressing the risks. The subsid- and Development (OECD)—as well as other coor- iarity principle may suggest that the risk should be dination agencies that accumulate extensive country handled by the lowest level of authority capable of information and experience, can provide a broad, addressing the matter effectively, before it becomes impartial knowledge base for countries and indi- a regional or global problem. Regional economic viduals to draw upon to help bridge gaps between communities are important layers of support in global objectives and national policies on a range of Africa, Asia, and the Caribbean, dealing with risks issues. Global knowledge and expertise can act as a and creating opportunities through cooperation guide and tool to raise awareness, especially where and development actions in areas such as trade, national agents fail to recognize far-reaching and energy, industry, security, and environment. The longer-term implications of their actions, for them- The role of the international community 257 D IAG R AM 8.2 Role of the international community Responding to severe Circumstances Addressing cross-national and humanitarian or economic crises cross-generational risks for country-specific risks Informational Cognitive or Access to Obstacles behavioral Externalities markets and gaps gaps resources Risk management Knowledge Protection Insurance Coping elements • Global knowledge • Rules or • Ex ante pooling of • Rapid mobilization Tools of the and expertise regulations global resources of global resources international • Capacity building • Platform for policy • Facilitation of • Other international community • Information dialogue regional resource risk-sharing tools campaigns • Contingency pools • Scientific research planning Source: WDR 2014 team. selves or others. Clear and prompt communication Improving protection through global rules, of the knowledge is crucial, if it is to succeed in af- capacity building, and coordination fecting behavior. The international community offers a range of Accumulating knowledge about the drivers and po- tools to fill knowledge gaps. From IFIs and coordi- tential effects of risks is necessary but not sufficient nation agencies to the scientific community, think to encourage appropriate risk management action. tanks, media, and civil society, international organi- Design and implementation of rules, regulations, zations collect, review, analyze, synthesize, and dis- standards, and frameworks for collective action can seminate information and research findings on eco- provide incentives and guidance to better manage a nomic, financial, health, environmental, safety, and range of risks that affect multiple nations and gen- other risks that have a bearing on development and erations. Some examples include the global financial stability, and publish cross-country information, re- sector reforms to strengthen the financial infrastruc- search, and policy analyses that can help assess risks ture and create a more resilient financial system fol- in an increasingly interconnected world. The inter- lowing the 2008 financial crisis;14 the Millennium national community can also provide platforms for Development Goals to reduce poverty and address knowledge exchange and publish periodic risk as- a range of risk affecting development;15 the United sessments or information at the global level on a va- Nations effort to encourage preparedness for pan- riety of risks. These knowledge platforms can play an demics and incorporate pandemic response plans important role in bringing national interests closer into disaster risk management (the Toward a Safer to establish shared goals. With capacity-building World Initiative);16 and A New Deal for Engagement technical assistance provided by IFIs, national au- in Fragile States.17 Several other examples, including thorities can design monitoring and early warning those for managing environmental risks, are out- systems and contingency-planning frameworks for lined more fully in table 8.1. timely detection of problems and effective responses Technical capacity to implement rules, regula- to them.13 Global media can play a crucial role in tions, and standards is necessary for such efforts to disseminating alerts generated by these detection succeed. Weak capacity in the veterinary and human systems. public health systems in developing countries, for ex- 258 WORLD DEVELOPMENT REPORT 2014 TA B L E 8 .1 Examples of global actions and factors underlying their impact Goals and results Underlying reasons for success or failure Kyoto Protocol (1997) Goals: Failed to attract broad-based support, ensure compliance (in the absence of an effective Reduce greenhouse gas emissions of 38 industrial enforcement mechanism), and make parties take substantial actions; some of the largest countries as a confidence-building step to reach the emitters either did not participate (industrial countries such as the United States and goal of the United Nations (UN) Framework Canada) or were not required to cut emissions (middle-income countries such as China Convention on Climate Change (UNFCCC) to stabilize and India). greenhouse gas concentrations at a level that would Bundled together targets for several greenhouse gases to achieve cost-effectiveness, prevent dangerous interference with the climate. but at the expense of lowering emission reduction targets. Required a 5.2% cut, on average, in the emissions of Diverging incentives and interests (no clear self-enforcing common goal): industrial countries below their 1990 levels between 2008 and 2012. • Perception that an individual country is too small to make a change. • Climate change does not affect all countries the same way; some benefit from it in Results: the short run, while some are hurt more than others. These differences create Took effect in February 2005 when the two varying views about benefits and costs of action to mitigate climate change. conditions of ratification were met (ratification by 55 • Competing domestic policy imperatives, including political factors, and short-term nations and ratification by nations that produce 55% economic considerations; nonparticipation (particularly by developing countries) to of the emissions). By April 2006, 141 countries had avoid hurting growth. ratified the protocol. Many countries did not meet • Free-rider problems with costly steps to mitigate climate change. their targets, however, and actually increased emissions, resulting in a global rise from 1990 levels. Montreal Protocol (1987) Goals: Broad participation: First treaty to reach universal ratification (197 UN nations). It started Protect ozone layer by banning ozone-depleting with 24 signatories and the European Economic Community in 1987, and was eventually chemicals (ODCs). signed by many, including developing countries. Results: Addressed the problem by chemicals (source), not timetable (targets). Emissions of most depleting substances have been Cost-effective substitutes for ODCs already existed. brought under control; signs indicate that the ozone layer will recover within the next 100 years; Negotiations included civil society and scientists to overcome informational barriers; high developed countries have reduced their production, degree of scientific consensus and evidence provided credibility. consumption, and emission of chemicals controlled Right incentives (and common interests): by the protocol by 99%; developing countries by 72% • Wide recognition that ozone depletion has serious, quickly visible consequences and reductions are continuing. (health issues such as cancer). Some increase in some of the chemicals authorized • Created strong incentives to participate and comply: the treaty set out reasonable for short-term substitution. plans for implementation with appropriate support coupled with trade restrictions— bans on trade between parties and nonparties in ozone-depleting substances and products containing the substances—to spur compliance. • Recognized importance of developing new technologies using nondepleting alternatives and providing access to developing countries. • Set up a multilateral fund to provide incremental funding to developing countries for transitioning to phase out harmful substances; provided institutional support (a key motivation for the participation of developing countries in the Protocol). Smallpox eradication campaign (1967–79) Goals: Broad cooperation achieved. Eradication of a pandemic disease that killed 300 Strong leadership and commitment from the World Health Organization (WHO), backed million–500 million people. by political commitment from governments. Financial and technical assistance from developed to developing countries that lacked resources and capacity to eradicate alone. (continued) The role of the international community 259 TA B L E 8 .1 Examples of global actions and factors underlying their impact (continued) Goals and results Underlying reasons for success or failure Results: Right incentives that were self-enforcing (no formal enforcement by WHO was needed): WHO declared the world free of smallpox in May • Costs (about $300 million worldwide) were negligible, compared with benefits: the 1980. United States got back its entire contribution in 26 days (in health costs saved)—a First disease eradicated by human effort. benefit-cost ratio of over 400:1. • The disease affected every country (ease of spread with trade and movement of Seen as a unique achievement in the history of people) with direct consequences; eradication succeeded because smallpox was international cooperation. eliminated everywhere. Strong U.S. support (monetary and technical) and other support from U.S. Centers for Disease Control. Scientific research showing feasibility of eradication; technical breakthrough (with a new type of needle) lowered the cost of vaccination. Surveillance and containment strategy: strong focus on preventing the disease from spreading by seeking and monitoring new cases. Controlling HIV/AIDS Goals: Global cooperation has had some success: Eradication of the disease. • In 2010, number of people on antiretroviral treatment in low- and middle-income Results: countries reached 47% of the need, up from 39%. Progress in treatment research has decreased the • Number of health facilities in these countries has risen significantly. number of people dying from HIV/AIDS. • About 35% of pregnant women living with HIV in these countries receive care, up from 7% in 2005. The number of new cases of HIV/AIDS has been decreasing since its peak in the late 1990s. Treatment 2.0 launched in 2010, aiming at higher efficiency: simplified, more affordable diagnosis and treatment; and integrated, decentralized HIV service delivery. But the number of people living with HIV is still rising. Despite availability of prevention, there is a problem of incentives: • HIV takes 5–10 years to manifest; people with low life expectancy may not protect themselves and may spread the disease. • Promotion of treatment is a double-edge sword—treatment may create an externality and lower the incentive to protect. • Limited access to information and protection in low-income countries. Hyogo Framework for Action (HFA) (2005–15) Goals: Provide governments with a common set of terms, approach, and platform, facilitating Reduce the impact of natural hazards by making cooperation at the international level. prevention a priority at all levels under the The structure of the HFA (organized by expected outcome, strategic goals, and priorities coordination of UN International Strategy for Disaster for actions) and guidance on its implementation supported development of comparable Reduction. framework at regional and national levels. Results: • Good communication and rising public awareness (various organizations have Increasing number of reporting countries. launched global campaigns). Increase in disaster reduction and recovery activities • Establishment of a scientific and technical committee. at all levels. More progress needed in preparation. • Progress in implementation (creation of a special representative of the UN Secretary- General for implementation of the HFA). • Broad participation/acceptance of the framework. High and increasing damages caused by natural hazards provide strong incentives to reduce the risk. Consequences are concrete and immediate. Source: WDR 2014 team based on United Nations Framework Convention on Climate Change 1998; Barrett 2006, 2008; UNEP 2007; Stern 2007; Rae 2012; Center for Global Development, “Case 1: Eradicating Smallpox,” http://www.cgdev.org/doc/millions/MS_case_1.pdf; World Health Organization, http://www.who.int; UNISDR 2006, 2007; OECD and G20 2012. 260 WORLD DEVELOPMENT REPORT 2014 ample, has undermined implementation of the Inter- eration among sovereign nations are discussed later national Health Regulations (2005) on surveillance, in the chapter. control, and reporting of contagious pathogens. The international community could help countries build Mobilizing global resources for preparation, implementation capacity. It could also make periodic mitigation, coping, and recovery assessments of whether national policies and prac- tices have high potential for cross-border spillovers. Countries’ efforts to prepare for risk notwithstand- These discussions could focus on areas where the lack ing, crises and disasters do happen, and when they of implementation capacity undermines a country’s do, significant resources are spent on coping with ability to conform to global agreements and manage their consequences and recovery. The international risk effectively. community has a range of risk-sharing tools to help The international community can provide tech- countries deal with extreme (tail-risk) scenarios such nical assistance to support initiatives designed to as disasters, both before and after the event (box 8.2). protect against various risks. It can support capac- ity building to strengthen governance; to build early Offering support for coping. In their most typical warning and monitoring systems for infectious dis- form, international risk-sharing solutions involve eases, crises, and disasters; and to design proactive direct ex post support from bilateral or multilateral crisis and disaster management strategies that reduce creditors or private organizations. A key driver of this the need for costly coping measures after the fact. The support is the need for timely action to mitigate a IFIs can support the development of markets for debt sudden shortage of resources (monetary or human) and reserve management and hedging instruments following a severe crisis or a disaster. Direct interna- to manage financial risks, particularly where small tional interventions can be justified when resources or segmented markets can block efficient private sec- to protect vulnerable populations are unavailable tor solutions to risk and prevent the pooling of risk from capital markets, self-insurance, or functioning across markets. Specific risk management strategies communities and governments or when the risk that include developing alternative risk-financing tools distress and contagion will escalate to other countries such as catastrophe bonds that transfer the risk of is high. Examples include the financial stabilization a disaster to markets by allowing the issuer to forgo packages arranged for several Euro Area states, the repayment of the bond principal if a major disaster liquidity provisions to unclog international finan- occurs.18 Weather hedges are another example of an cial markets during the global financial crisis, the instrument that transfers the risk to financial mar- support that 36 donors provided to more than 100 kets; these hedges are based on an underlying weather developing countries to control the H5N1 avian flu index, with payments triggered by prespecified ad- and prepare for a possible pandemic during 2005–10, verse weather events. and direct humanitarian help to people in FCSs. An important role for the international commu- Several international community actors play a nity lies in facilitating the collective action and co- role in coping. The IMF, the World Bank, and other operation necessary to supply global public goods. IFIs, as part of their mandates, pool risk across coun- By providing a platform for policy dialogue and tries and lend to countries experiencing actual or po- coordination among sovereign states (key building tential external funding pressure as countries work to blocks of the international community), the inter- restore stability or sustain development spending in national community can promote implementation the wake of a crisis and correct underlying problems. of agreed rules and regulations that reduce global Remittances from immediate or extended family risks, as well as cooperation that improves develop- members abroad provide risk pooling at the family ment outcomes. Such cooperation could facilitate level, allowing for more direct and timely relief in the further liberalization of international trade and presence of adverse domestic shocks. Remittances capital flows; support strong, sustainable, and in- and kinship support are among the traditional cop- clusive growth, or engage with FCSs on a sustain- ing mechanisms for FCSs, especially when effective able basis; and take a balanced approach to risks and government support is not available. Civil societ- opportunities. Cooperation can also limit potential ies, including global nongovernmental organiza- externalities and inconsistencies in implementa- tions, combine in-kind transfers with foreign onsite tion that could jeopardize outcomes in a tightly managerial services to deal with local bottlenecks. integrated and interconnected world. Some of the International investors also boost domestic capacity challenges associated with securing effective coop- through portfolio and direct investment flows. The role of the international community 261 BOX 8.2 International support for disaster risk management Disaster and climate risk management are increasingly recognized • Risk identification. By quantifying risks and anticipating the as key priorities for development, and many actors are working to potential negative impacts of natural hazards on society and shift the focus from ex post response toward preparation and pre- the economy, disaster and climate risk assessments can help vention at all levels of government. At the international level, the UN governments, communities, and individuals make informed Office for Disaster Risk Reduction coordinates efforts across the decisions about managing risk. UN system and tracks progress toward the implementation of the • Risk reduction. Anticipatory action can reduce existing risks Hyogo Framework for Action (HFA) to make prevention of natural and prevent the creation of new risks. disasters a priority. The scientific community, civil society, and inter- • Preparedness. Technical assistance and financing of climate national financial institutions, as well as regional intergovernmental services help establish early warning of extreme events. They organizations that foster regional disaster-risk-reduction coopera- also increase climate-modeling capacity to design effective tion, support the efforts for HFA implementation. adaptation policies. In 2006, the Global Facility for Disaster Reduction and Recovery • Financial protection. Advisory services on disaster risk financ- (GFDRR) was created to forge a global partnership to intensify sup- ing and insurance help protect governments, businesses, and port for mainstreaming disaster risk management into national households from the economic burden of disasters; increase development planning. Housed within the World Bank, the GFDRR the state’s financial capacity to respond to emergencies; pro- has helped the Bank move from a reactive approach to a more strate- mote deeper insurance markets at regional and sovereign lev- gic, long-term approach focused on reducing risk. Bank disaster- els; and support social protection strategies for the poorest. related financing doubled from 1984–2006 to 2007–11. The share of • Resilient recovery and reconstruction. The Bank supports funding by the International Development Association for climate country-led Post-Disaster Needs Assessments, which esti- adaptation rose from 9 percent to 16 percent, and for climate mitiga- mate the impact on people, including development needs, tion rose from 5 percent to 16 percent from fiscal 2011 to fiscal 2012.a and economic losses following a disaster. The estimates The Bank supports disaster resilience in developing countries provide the basis for planning recovery and reconstruction through a five-pillar approach: efforts. Source: Robert Reid for the WDR 2014. a. IDA 2012. Providing insurance mechanisms. Besides emergency because of the stigma effect of seeking financial help. assistance, international risk-sharing mechanisms Finding an efficient design for global safety nets has include insurance that pools risk and transfers re- also been a challenge, given the difficult trade-off sources from good to bad times. IFIs such as the between limiting moral hazard and preventing li- Multilateral Investment Guarantee Agency offer po- quidity crises from turning into insolvency. Instead, litical risk insurance to reassure foreign investors and stronger links among emerging economies triggered promote investment flows to countries as part of ex interest in regional reserve pooling and swap lines ante risk management. The IFIs also provide emer- to serve as insurance; however, these schemes are of gency disaster response tools by creating a range of limited lending power and effectiveness in dealing products countries can access with great flexibility with covariate liquidity shocks.21 and speed, such as the World Bank’s Immediate Re- sponse Mechanism and Catastrophe Deferred Draw- Facilitating regional insurance. Besides its more di- down Option (CAT DDO) instrument.19 Moreover, rect engagement, the international community can the ongoing financial crisis facilitated the creation of also play a more indirect catalyzing and technical insurance tools for countries experiencing volatility role by helping countries in a particular region pool and instability despite relatively strong fundamentals, resources that they can use in an emergency. Such such as the IMF’s Flexible Credit Line or the Bank’s mechanisms bode well for the principle of shared Development Policy Loan with DDO.20 responsibility in managing risk and enhance coun- These tools are also intended to reduce the de- tries’ capacity to jointly access international markets mand for self-insurance through excessive reserve ac- at a lower premium than they could obtain individu- cumulation—a factor that contributed to global im- ally. These facilities are particularly helpful for small balances as external account deficits of systemically states where private markets are nonexistent, small, important economies widened because of higher segmented, poorly functioning, or unaffordable to demand for reserve currencies. Countries have been the most vulnerable, and where access to credit, in- reluctant to use some of these tools, however, in part surance, and reinsurance markets is limited. Three 262 WORLD DEVELOPMENT REPORT 2014 BOX 8.3 Regional risk-sharing solutions: Promoting financial resilience to disaster risks Innovative disaster risk financing and insurance (DRFI) solutions are Caribbean countries are members of the facility, which provides being developed by international financial institutions, in partner- them with immediate liquidity in case of a major hurricane or ship with donors and other members of the international commu- earthquake. Members pay an annual premium depending on their nity. These tools are particularly important for developing countries risk exposure. The CCRIF uses risk pooling (through joint reserves with high exposure to natural hazards, but limited resources, finan- and lower reinsurance rates) and shared operating costs to provide cial capacity, and access to cheap credit and insurance markets. The coverage at a significantly lower cost than each country would pay regional risk-pooling mechanisms discussed below illustrate four key acting separately. With technical assistance from the World Bank, roles the international community can play in advancing DRFI solu- the facility tackles the technical, actuarial, legal, fiduciary, and tions: convening power; promotion of public goods that permit the financial engineering aspects of designing and implementing an development of risk market infrastructure; technical assistance and independent, sustainable facility. The CCRIF, which is funded by specialized expertise; and provision of initial seed capital, contingent participating countries and donors, has provided immediate liquid- loans, and credit enhancements. ity funding to governments on eight occasions since its launch in 2007. Increasing access to catastrophe insurance in Southeast Europe Financing a proactive approach in the Pacific The Southeast Europe and the Caucasus Catastrophe Risk Insurance Transitioning from relying on humanitarian aid and other ex post Facility (SEEC-CRIF) was launched in 2009 to support development resources to more efficient ex ante DRFI requires investment in of a catastrophe and weather risk insurance market for the region. disaster risk assessment and financing tools—as well as funding to The initiative simultaneously addresses three bottlenecks of market implement the solution (such as paying insurance premiums). The development: risk market infrastructure, regulatory framework, Pacific Catastrophe Risk Assessment and Financing Initiative has and government policy. It provides pivotal public goods, including invested in the development of probabilistic catastrophe risk mod- country-specific catastrophe risk models and a web-based insurance els acceptable to the international reinsurance market. The facility underwriting platform to facilitate the sale of reliable, cost-efficient provides technical assistance on implementation and advises mem- catastrophe insurance products. The CRIF also helps participating bers on financial solutions to reduce their exposure and to improve countries incorporate risk awareness, knowledge, and skills related financial and economic planning (insurance, donations, reserves, to climate change and disasters into their development policies. The and contingencies). For the sovereign catastrophe-risk insurance World Bank supports the CRIF with technical assistance and loans to pilot launched as part of the initiative in 2013, the Japanese govern- the facility and member governments. ment funded the first and part of the second year’s premium for five participating countries. Providing technical assistance to launch state-of-the-art risk-pooling in the Caribbean The Caribbean Catastrophe Risk Insurance Facility (CCRIF) is the first-ever multicountry risk pool to insure against disasters. Sixteen Source: Laura Boudreau, Hannah Yi, and Olivier Mahul for the WDR 2014. examples of regional facilities, designed with sup- ing by global systemically important banks and for port from the international community, are outlined policy makers to use in assessing systemic risks.22 The in box 8.3. World Organisation for Animal Health has evaluated public veterinary systems in more than 100 countries for their ability to detect and control diseases and re- How effective is the international duce contagion risks. The Intergovernmental Panel community in resolving global risks? on Climate Change brings together scientists period- The international community has made significant ically to review research from around the world and progress in addressing risks through knowledge update and fine-tune assessments on the drivers and tools. It has put great effort into data collection and consequences of climate change; the UN Framework risk analysis to improve assessment of risks and has Convention on Climate Change monitors trends in developed innovative tools and databases to analyze greenhouse gas emissions to inform policy analyses risk from adverse natural hazards (box 8.4). It has and discussions at national and international levels. worked to reduce data and information gaps. In a IFIs monitor and analyze a wealth of economic, fi- recent joint effort, for example, the Bank for Inter- nancial, environmental, and developmental data and national Settlements, the Financial Stability Board, trends that help inform national policies. and the IMF developed a common data template Significant efforts have also been made to apply for markets to use in monitoring excessive risk tak- this knowledge. The international community has put The role of the international community 263 BOX 8.4 Global efforts to provide tools and databases for assessing disaster risk By anticipating and quantifying potential damages from natural haz- by-building data will be available for a selected number of areas, ards, disaster and climate risk assessments can help communities, and the number of areas will increase over time. companies, and governments make more informed decisions, such Probabilistic risk-modeling techniques are now increasingly used as where and how to build safer schools, how to insure farmers to evaluate uncertainty inherent in complex systems, including nat- against drought, and how to protect coastal cities against rising sea ural events. Probabilistic risk modeling is also being coupled with levels. climate change models to assess the likelihood and severity of Estimates of potential exposure of physical assets and popula- future hazards, over the time horizons needed for decision making tions to risk are necessary to develop any risk reduction strategy, in sectors such as urban planning.a A free platform, CAPRA (Central as well as for effective emergency response and crisis management American Probabilistic Risk Assessment), has been developed to use in general. Although the most detailed exposure data are available a probabilistic methodology to visualize, quantify, and track sources primarily in high-income countries, international actors are work- of risk resulting from a range of hazards in Central America and is ing with developing countries to build their own asset exposure being rolled out in other regions. inventories. For example, the Pacific Catastrophe Risk Assessment Data sharing and open systems promote transparency and and Financing Initiative has created the largest-ever collection of accountability and enlist a wide range of participants in the chal- geospatial information on disaster risks available for Pacific Island lenge of building resilience. For example, the Open Data for Resil- countries, with quantification of potential disaster losses from ience Initiative uses free and open-source software from eight lead- earthquakes, tsunamis, and tropical cyclones. Resulting exposure, ing international organizations and data providers to enable people hazard, and risk maps and data are shared with policy makers and and institutions to collaborate on building drought resilience in the public. the Sahel. Similarly, InaSAFE is a free and open-source software that The Global Earthquake Model is a global collaborative effort to produces natural hazard impact scenarios, providing a simple yet pool knowledge and provide people with tools and resources to rigorous way to combine data from scientists, local governments, assess earthquake risk anywhere in the world. The goal is to provide and communities to assess likely effects of future disaster events. a global exposure database by the end of 2013 that contains aggre- The tool was piloted by the city of Jakarta for emergency planning gate information on population and residential buildings. Building- during the 2012 flood season. Sources: Robert Reid for the WDR 2014. a. Ranger and others 2011 in place rules and standards to encourage responsible the onset of the global financial crisis, economies and risk management behavior. It has convened experts, financial systems of advanced and developing coun- national and global policy makers, and standard set- tries remain vulnerable to the risk of renewed ten- ters around the world to solve global problems. It has sions, as some underlying economic and structural made progress in using risk assessments to generate weaknesses remain unresolved. Negative feedback and communicate predictions and warnings of natu- loops across banking, sovereign, and real risks and ral hazards. Improved access to global media and the competing macroeconomic priorities complicate Internet has allowed rapid sharing of disease intel- policy responses. Progress remains limited in ar- ligence and scientific research on disease control, en- resting climate change, despite substantial available vironmental risks, and financial risks, among others. knowledge and emphasis on the dangers of inaction. Early warning systems have been developed for many Progress in preventing and preparing for pandemic types of hazards, helping to reduce the number of risk is limited, even as costly zoonotic disease out- deaths from disasters. Enhanced monitoring of eco- breaks continue to occur (including the recent out- nomic, financial, social, geopolitical, environmental, breaks of H7N9 and the coronavirus). The majority and technological risks is being used to assess low- of the Millennium Development Goals are not ex- probability, high-impact risks to the global system pected to be met by fragile states by the 2015 target and to push for risk-mitigating policies, including date (box 8.5); by that time, these countries also are those that would require international coopera- expected to account for half of the world’s poor.24 tion.23 Global resources have been used when coun- The inability to move forward more aggressively to tries faced mounting difficulties that also risked spill- deal with these risks is costly, taking already scarce ing over to others. resources away from development efforts, in some But overall effectiveness has been limited. In cases slowing or reversing hard-won development particular, the international community as a whole gains and imposing huge costs on future generations. could have been more forceful in addressing some of Several common elements play a role in this the key risks that cross boundaries. Five years after poor performance. Insufficient access to available 264 WORLD DEVELOPMENT REPORT 2014 BOX 8.5 Well-managed risks can unleash development opportunities in fragile and conflict-affected states International engagement in fragile and conflict-affected states flict to continue or resume. In an interconnected world, such con- (FCSs) poses considerable risks for donors and implementing part- flicts have significant economic and social costs that reach beyond ners, who must contend with high levels of insecurity, political insta- national borders. bility, weak institutions, and the failure of basic state functions that typically characterize such states. These characteristics undermine High degree of risk aversion the social and economic support systems through which the interna- Yet within the donor community and its implementation partners, tional community provides support (including the state and local the emphasis tends to be on avoiding risk, both in where and how communities and institutions). In these complex and fast-changing donors engage, and within the organizational cultures of donors. environments, outcomes are hard to foresee and control, and the Aid flows to poor and fragile countries are volatile and unpredict- possibility of returning to violent conflict is always present. People’s able. High degrees of uncertainty and information gaps can also ability to manage risks is seriously constrained, and the conse- lead to overly pessimistic perception of risks and unrealistic expec- quences of a risk materializing are often a matter of life and death. tations for what aid can achieve in short time frames. Concerns At the same time, international engagement in these high-risk about corruption discourage donor and investor engagement. environments can make particularly important contributions to Tighter reporting and accountability requirements and less reliance development. Because of the low starting point, effective interna- on local initiative reduce the speed, flexibility, and innovation that tional assistance can achieve more in these transitional contexts are key to taking advantage of short-lived opportunities in these than in most other situations.a Where state-society relations are fast-changing situations. These shortfalls are manifested by a lack of renegotiated and state institutions redefined, international engage- progress in attaining the Millennium Development Goals (MDGs) in ment, including aid, has the potential to provide critical catalytic many FCSs (figure) and in a large number of countries returning to and transformative support. Moreover, the risks of not engaging conflict. The absence of active engagement with and support of can be high—both for the countries themselves and for the interna- development over the long term prevents the building of national tional community, if the lack of international assistance allows con- capacity. Fragile and conflict-affected states have made slow progress toward the MDGs 100 Progress toward goal to date, % 80 60 40 20 0 Reduction Primary Gender parity Gender parity Gender parity Reduction in Improved Improved in extreme completion (primary and (primary) (secondary) child mortality access to access to poverty rate secondary) under five safe water sanitation Middle income countries Low income countries Fragile situations Source: WDR 2014 team based on World Bank and IMF 2010. Some change in approach is under way Deal for Engagement in Fragile States, agreed in 2011 in Busan, FCSs and development partners are now concluding that appropri- Republic of Korea, recognized that the risk of not engaging can out- ate risk taking is essential for improved outcomes and that a better weigh most risks of engagement. It emphasized the need for joint balance must be struck between risk and opportunity. The 2011 WDR assessments of the specific risks and context-specific, joint donor on conflict suggested that poorly designed and rushed donor risk-mitigation strategies. The g7-plus group of states (those affected responses can exacerbate the significant risks in engaging in FCSs, by conflict and now in transition to the next stage of development) and that risk-opportunity assessments should be used more fre- and its development partners have committed to support develop- quently to see how aid itself might be a risk mitigation measure ing countries’ efforts to strengthen core institutions and policies by through its impact on local systems and capacities. The 2011 WDR aiming to manage, rather than avoid, risk, and minimizing the risk of recommended greater monitoring of government-executed pro- reverting to conflict through joint efforts of donors and fragile states. grams, risk sharing through pooled funds, and proactive planning of Balancing risks and opportunities requires a parallel focus on risk contingencies based on risk-opportunity assessments. The New contextual, programmatic, and institutional risks and collective The role of the international community 265 BOX 8.5 Well-managed risks can unleash development opportunities in fragile and conflict-affected states (continued) approaches. Recent research by the OECD provided policy recom- risks and respond flexibly. Greater honesty and transparency mendations for donors to match aspiration with action and move about exposure to all risks is needed between donors and from risk aversion to balancing risk and opportunity: those they fund, with an explicit focus on building local capacities and a commitment to collectively manage asso- • Establish institutional cultures that encourage appropriate risk ciated risks.b taking in FCSs. This would involve setting incentives for staff and implementing partners to consider risks in relation to The broader international community could support donors’ opportunity and communicate openly about why some risks efforts to manage the risk of engagement in FCSs: in FCSs are worth taking, along with devising specific risk- • It could help donors identify, better understand, and monitor management frameworks. risks specific to the FCS, notably by developing a set of indica- • Agree on realistic objectives and frameworks for measuring tors to monitor frequently. The New Deal Fragility Assess- results. Donors and partners should establish such frame- ments will likely use a common set of indicators that can be works for measuring results in complex environments, tai- applied flexibly to take country context and local needs into lored to specific circumstances of FCSs. account. Better access to such knowledge could inform deci- sion making and prioritization, alleviate risk aversion by the • Simplify procedures for the release and delivery of aid. This step donor community, and identify the local capacities and insti- would facilitate rapid and flexible responses and transfer of tutions that need strengthening. funds to take advantage of the narrow windows of opportu- nity to influence the course of events. • Contingency plans, early warning, and crisis and disaster man- agement systems could be developed to mitigate the • Establish a common framework for understanding and assessing extreme risks facing FCSs, in partnership with multilateral risk that ensures focus on people and contextual risks. By con- institutions. Targeted capacity-building assistance could ducting fragility assessments, several New Deal pilot countries strengthen implementation deficiencies, improve institu- have sought to develop a joint understanding with develop- tional capacity, and reduce corruption risks. Development ment partners and civil society of the causes, features, and agencies, civil society, and the media could partner in the drivers of fragility, and of sources of resilience that can form a effort, making wrongdoing costly. Coordinated donor in- basis for dialogue and joint risk mitigation strategies. volvement in reconstruction efforts proved useful after the • Identify options to share risks and maximize collective impact by 2004 tsunami in Aceh. pooling efforts and funding. Joint efforts can reduce individual • Both donors and FCSs could do risk pooling with the interna- actors’ exposures to political and reputational risks and dilute tional community’s help (involving bilateral and multilateral the risk of program failure. Options for sharing risks include partners). The positive experiences with regional risk-pooling pooled funds, joint guidelines, and mutual accountability facilities could provide useful guidance. The international frameworks. community could help mobilize multilateral donor funding • Understand and facilitate the role of multilateral organizations and design harmonized proactive risk-management frame- as “risk-pooling mechanisms.” Donors need to give imple- works and by providing political risk insurance in multilateral menting partners the necessary scope and means to take platforms. Source: WDR 2014 team based on Laura Mazal, Diane Koester, and Sophie Walker for the WDR 2014; Asbjorn Wee for the WDR 2014; OECD 2011a, 2011c; OECD Development Assistance Committee 2012; Fengler and Kharas 2011; World Bank 2011; and World Bank and IMF 2010. a. Transitional situations may include transitioning from conflict to peace or violence to security. b. Ongoing case studies by the OECD and the U.K. Department for International Development highlight a number of interesting donor approaches and innovative practices on risk management already being used in FCSs. In Nepal, conflict-sensitive programming is used as a risk management practice that has been main- streamed across donor operations. Specialized risk management units help pool resources in addressing security, fiduciary, and other risks encountered in operational work. In Somalia, a UN Risk Management Unit was set up to manage fiduciary risk and monitor implementing partners, where limited access and freedom of movement hamper the ability to undertake direct field monitoring. knowledge, resources, and capacity hamper the ac- tive enforcement mechanisms undermine interna- cumulation and use of this knowledge to take ap- tional cooperation. propriate risk management action. In some cases, more emphasis is put on avoiding risk than on Problems in formulating and transforming taking well-informed risk and managing it (as with knowledge to action the international community’s engagement with fragile states). Political economy constraints and Continued gaps in information constrain knowledge lack of proper incentives, accountability, and effec- and action in some areas. Information asymmetries 266 WORLD DEVELOPMENT REPORT 2014 continue to hamper global efforts to effectively man- instance, in the past five years, insurance covered less age risks in the financial sector. Despite the efforts of than 20 percent of total disaster losses in developing the global community, some key information needed countries, on average, compared to about 60 percent to identify a buildup of systemic financial risks re- in North America, according to SwissRe. Shortfalls mains unavailable to markets and policy makers. In in funding the cost of mitigation of, and adaptation countries where financial systems are regionally or to, climate change have been an obstacle to reducing globally interconnected, the lack of exposure data greenhouse gas emissions in developing countries across institutions hinders identification of emerg- and to reaching agreements in global negotiations.28 ing risks and undermines the usefulness of early Although climate change risk and loss of biodiversity warning systems to trigger appropriate action. Early and resources are global problems, vulnerability as warning systems for natural hazards are also of lim- well as efforts to mitigate or adapt to these risks are lo- ited use when information is imprecise and is not cal, and constrained by national and local authorities’ communicated quickly or clearly, as experienced in capacity to implement necessary corrective actions. some developing countries.25 Similarly, infectious Cognitive, behavioral, and political economy fac- disease controls are often undermined by weak com- tors also get in the way of translating available infor- munication between public health authorities and mation into actionable knowledge. Despite wide- within the public; delays in detection and diagnosis spread availability of information on the evidence of caused by information gaps lead to late and more drivers of climate change and other environmental costly control measures. Failure to translate scientific risks, disasters, or the possibility of yet another pan- knowledge for use by local practitioners also limits demic, individuals, communities, and governments the appreciation of, and response to, various risks. continue to overlook their potential exposure to While extensive data are available on environmental what they view as rare or distant events, underesti- risks, they remain scattered and lagged, with lim- mate the potential cost, and fail to insure or otherwise ited systematic dissemination of the key messages to protect themselves (and others). Similarly, small- summon national or global action. probability, high-impact risks are often ignored in Information asymmetries also affect the interna- the face of short-term challenges, resulting in under- tional community’s ability to engage effectively in investment in preventive steps. A recent global survey fragile and conflict-affected states, thus undermining on climate change found, for instance, a clear indica- their ability to support people’s risk management. tion of short-sighted attitudes to climate change risk Corruption and political risks, which typically char- and greater attention to what is seen as more press- acterize these states, undermine the competitiveness ing and urgent matters facing the world (figure 8.2). and investment appeal of their economies.26 Insuf- These responses suggest that a non-negligible part ficient information about the extent of corruption, of the global population discount the future heavily political risks, and local authorities’ implementation and place a lower value on adverse consequences of will and capacity adds to donor risk aversion, reduces climate change, such as loss of biodiversity and in- their willingness to engage, and may focus their at- creased frequency and cost of disasters. tention on attaining results that, while safe, may not Deep uncertainty adds to the behavioral and help these states strengthen national systems and ad- cognitive biases. For example, the push to mitigate dress peace-building needs.27 Missed opportunities climate change risks is undermined by uncertainty for engagement (through market access or develop- about the benefits and costs of taking such action ment aid), in turn, raise the risk that political transi- and by a lack of consensus on the critical thresholds tion will threaten prosperity and social cohesion and (tipping points) for greenhouse gas concentrations could create a vicious circle of fragility, poverty, de- beyond which small changes in Earth’s temperature spair, continued corruption, and conflict, with costs could have catastrophic consequences. The absence of that can have broader cross-border implications, as scientific consensus on these thresholds undermines observed in the Middle East and Africa. In such cir- incentives for international cooperation. Recent ex- cumstances, risk of inaction can be very high. perimental research suggests that if this threshold Many tools designed by the international com- could be identified with certainty, and if the rela- munity to manage complex risks elude countries tive cost of avoiding it were low, the fear of crossing with limited resources and capacity to understand it could reduce the free-riding behavior of countries and implement them. Insufficient resources and ca- and induce them to join in the needed collective ac- pacity make it difficult for countries to access and tion to avoid catastrophe.29 Deep uncertainty may apply available information and knowledge, afford also undermine the ability to assess complex macro- insurance, reinsure, and take preventive actions. For financial risks. The difficulty of anticipating the com- The role of the international community 267 FIGURE 8.2 A 2011 survey highlights divergent national interests and short-sighted views on climate change a. Those concerned about climate change b. Reasons for not being concerned North 50 Warmer temperatures 12 America are good for me Europe 68 Not yet a problem 14 Global Will not affect me 18 69 average in my lifetime Technologies will 23 Asia Pacific 72 take care of it Middle East Humans do 37 80 and Africa not cause it Latin Other more serious 48 90 America and urgent problems 0 20 40 60 80 100 0 10 20 30 40 50 60 % of respondents % of respondentsa Source: WDR 2014 team based on data from survey responses of 51 countries in Nielsen Company 2011. a. Percentages equal more than 100 percent because respondents could choose all choices that apply. plex feedback loops between financial, sovereign, and encouraging excessive risk taking by financial institu- real sector risks, for instance, has contributed both to tions perceived as too important to fail, and reducing the severity of the ongoing global financial crisis and their incentives to self-insure by holding capital and to the challenges in resolving it. liquidity in line with the risks taken (see chapter 6). Despite the high benefit-cost ratios of better prep- aration (see chapter 1), evidence suggests that more Perverse incentives that discourage effective emphasis is placed on ex post risk management. This risk management emphasis is evident in donor financing for disasters: Emphasis on ex post risk management creates moral of the total development assistance allocated for hazard. Ready availability of help to recover from a disaster-related activities between 1980 and 2009, disaster or a crisis may encourage public and private only 3.6 percent ($3.3 billion) was devoted to preven- agents to be less cautious in taking risk or in protect- tion and preparedness (figure 8.3).32 By contrast, the ing or insuring against it. The ex post availability estimated economic losses from disasters over the past of disaster aid (given the reluctance to deny help to 30 years amount to $3.5 trillion—with a record $380 those who have not taken sufficient prevention mea- billion in 2012. Financial crises also divert resources sures—the Samaritan’s dilemma) may, in some cases, from growth and development: the cost of direct sup- weaken incentives of governments to invest in warn- port from national governments to financial institu- ing systems or enforce strict zoning and building tions during the 1990 crisis ranged from less than 5 regulations in disaster-prone areas, or for individuals percent of gross domestic product (GDP) in Sweden to insure or avoid settling in such areas, when other to more than 55 percent in Indonesia. Since 2008, the options are available.30 For governments (or donors), cost of direct support and government guarantees to the political reward for well-funded and costly hazard the financial system ranged from about 10 percent of prevention may be seen as small compared with the GDP in the United States to more than 50 percent gain from an efficient ex post response.31 Similarly, in Ireland.33 Realization of these contingent liabilities sustained investments in public health systems to reduces the fiscal room available for social spending prevent pandemics from developing may be crowded (see chapter 7), and makes it more likely that national out by funding for mitigation programs, such as governments will call for international support when stockpiling of medications. In finance, the absence future problems hit. In the health area, the total cost of effective cross-border resolution regimes to deal of major zoonotic disease outbreaks reached an esti- with failing systemic banks and national govern- mated $80 billion over 2007–09, compared with an ments’ tendency to rescue them create moral hazard, annual estimated cost of $1.9 billion–$3.4 billion to 268 WORLD DEVELOPMENT REPORT 2014 FIGURE 8.3 Disaster-related donor assistance has focused more on coping than preparedness and prevention 30 25 20 US$ billions 15 10 5 0 1980–84 1985–89 1990–94 1995–99 2000–04 2005–09 Emergency response Reconstruction relief Prevention and preparedness Source: WDR 2014 team based on data from AidData Aid Activity (database). build and operate One Health approaches to prevent taining pandemics) require global collective action and control these diseases.34 by sovereign nations. Collective action depends, first, Myopia about risks may also reduce the perceived on recognition of shared interests. If there is no per- urgency for action, while creating tendencies to pass ceived commonality of interests, cooperation is un- the risk, and the associated cost of today’s inaction, likely. Cooperation also fails if there is no “global au- on to others. The costs of climate change mitigation thority” that can assess global risks and exert coercive not borne by today’s generation will be passed to fu- sanctions on sovereign countries that fail to take ture generations when mitigation will likely be more agreed-upon actions.37 Without explicit enforcement costly and possibly too late to have the intended ef- mechanisms, international agreements to provide a fect. A 2007 estimate by the Stern Review placed the global public good must rely on voluntary participa- cost of unmitigated climate change at a per- tion, which works only if the incentives are manent annualized loss of 5–20 percent “right” or a “common goal” has been rec- of global output by 2050, compared International ognized.38 That is, multilateral coop- with a cost of 1 percent to stabilize eration works best when national carbon emissions.35 Fears that cooperation works interests are well aligned, or when other countries will impose trade best when national impediments arising from vested and travel restrictions may also interests are well interests or other domestic policy dampen a government’s willing- aligned or when priorities are not overriding. ness to share information on the Reducing greenhouse gas outbreak of a disease, increasing domestic policy emissions to mitigate climate the eventual cost of stopping it.36 imperatives are not change and prevent its catastrophic Imprudent government spending overriding. consequences is a perfect illustra- raises the debt burden of future genera- tion of the challenges facing collective tions. And short-sighted domestic political action. Climate change affects countries considerations create incentives to delay tough policy and regions—and even populations within a given measures to resolve a crisis, compounding the cost of country—unevenly, benefiting some and hurt- an eventual resolution for all countries involved. ing some more than others. Continued uncertainty Finally, divergent national interests undermine about the level of climate change thresholds, the per- international cooperation and create incentives for ceived unevenness of climate change effects across inaction in the absence of agreed common goals and nations, and competing domestic policy impera- standards that are enforceable. Global public goods tives create diverging incentives for taking mitiga- (such as controlling climate change, arresting exploi- tion action. As a result, countries have been unable tation of natural resources, and curbing loss of biodi- to forge a lasting agreement ratified by all nations, versity; restoring global financial stability; or con- let alone a mechanism to enforce it.39 In contrast, The role of the international community 269 in two successful examples of international coop- The lack of international assistance leaves se- eration—smallpox eradication and protection of the vere risks to people unaddressed and increases the ozone layer—common interests helped remove bar- eventual cost of engagement. Preventing states from riers to collective action: everybody was vulnerable falling into conflict can be more cost-effective than to the highly damaging and quickly visible health responding once they have failed: studies have esti- consequences (see table 8.1). The looming threat of mated that each dollar spent on conflict prevention a nuclear war, with devastating consequences for the can generate, on average, savings of $4 to the inter- world, also spurred 189 nations to sign the Nuclear national community.42 Delayed response can also be Non-Proliferation Treaty in 1968, which helped con- very costly in terms of human lives, as demonstrated tain the spread of nuclear weapons (to fewer than 10 by recent events in Somalia, where a famine took countries), although long-term viability of the treaty many lives during 2010–11 despite 11 months of re- remains vulnerable to the presence of several nations peated early warnings, with opportunities for early with the capacity to build nuclear weapons.40 intervention missed because of perceived political Diverging incentives and collective action traps risk.43 In an increasingly interconnected world, the also play a role in the slow progress in resolving cost of such inaction goes beyond national borders, the ongoing global economic and financial crisis.41 resulting in increased refugee populations, spread For example, the Basel III framework designed to of communicable diseases, crime, conflict, economic strengthen the soundness of the global financial sys- losses, and growing pressure on public goods (such tem following the ongoing financial crisis has faced as water, sanitation, education, housing, and health challenges in its formulation and implementation. services) in neighboring countries that absorb af- The desire of individual nations to protect their fected populations.44 One study estimated that shar- banking systems has led to divergent views among ing a border with a fragile state can reduce a coun- advanced countries and between advanced and de- try’s economic growth by 0.4 percent annually.45 veloping countries on the stringency of the new stan- dards and pace of their implementation; some coun- Policy implications and takeaways tries have unilaterally introduced stricter national regulations as a result, in effect creating regulatory The international community has made remarkable gaps. Similarly, the efforts of the Group of 20 worked progress in providing a range of tools for effective well at the start of the financial crisis, when country risk management, but much more needs to be done leaders supported expansionary policies to restore to forge consensus on risks that transcend national financial stability and counter economic downturn. and generational borders. In a world with a tight net- Continued cooperation has become more challeng- work of interconnections, “global problems require ing as expansionary policies in advanced countries global solutions,” but in the absence of an effective have stimulated large capital inflows to emerging global risk governance mechanism with an interna- market countries and complicated their macroeco- tional body that has appropriate accountability and nomic management. enforcement powers over sovereign nations, the in- Diverging national interests also contribute to ternational architecture necessary to provide the slow progress in resolving the problems facing fragile global public goods and address global risks has not and conflict-affected states. Concerns about whether kept pace with the connectivity that glues the world the resources devoted to FCSs are used effectively together and the complexities such connectivity have made donors less inclined to engage, reducing creates.46 the effectiveness of aid in many fragile states. On the The limited progress made in managing global one hand, expectations have risen that aid could help risks has put into doubt the ability of the interna- achieve peace-building and state-building objectives. tional community to foster collective action among a On the other hand, applying the same reporting and large number of nations with diverging interests, ca- accountability requirements as in more stable envi- pacity constraints, and incentives to free ride on the ronments and requiring rapid and visible results of- actions of others. This collective inaction poses sig- ten leave limited room for flexibility and innovation, nificant challenges to the goals the international undermining the effectiveness of engagement. While community aims to safeguard, from eliminating donors have been stressing since 2011 that they have poverty to restoring peace, building resilience and a common interest in ensuring successful engage- prosperity, and achieving a more equitable distribu- ment with fragile states, they have struggled to adapt tion of income around the world. their systems for implementation and control to ef- Does this mean the world should give up on the fectively meet these challenges. goal of attaining global solutions and turn its back 270 WORLD DEVELOPMENT REPORT 2014 TA B L E 8 . 2 Policy priorities to improve risk management at the international community level POLICIES TO SUPPORT RISK MANAGEMENT FOUNDATIONAL ADVANCED Knowledge data Improve data quality and availability Eliminate minate information gaps on financial Elim ins stitutions and exposures institutions Inten nsify scientific r Intensify research, esearch, improve improv ve knowledge on global risks, and step up inf formation/edu ucation information/education campaigns to rais campaigns se risk awareness on raise n importance of preparation A on basic RM tools, Provide TA tools, EWSs, Advisory Ad continge ency planning, dvisory on EWSs, contingency , debt/ contingenncy planning, m contingency arket/institutio onal market/institutional reserve res serve management, hedging hedg instrumen nts ging instruments developmment, communication, governance development, governa ance Protection Des sign t Design targeted argeted global rules, regulations, re egulations, standards, rds, standar collaboration ollaboration through platforms and ensure co atforms for policy dialogue pla disaster Financing for dis saster prevention d n and Facilitate ilitate implementa Faci implementation mitigation/adaptation, gation/adaptation, ation of mitig preparedness; m itigation and adaptation; mitigation ad dapta ation; contingency planning mechanisms, g mechanism ms, EWSs anning mechanisms/EWSs contingency planning pla mechanis EWSs sms/E sic nutrition, education Vaccination, basic bas educ on programs, catio bsidies/financing of Subsidies/fi Sub of R&D technology transfer, nsfer, peacekeeping e tran ffort eff Insurance nt elements Contingent credit lines with grant gran elemen nts es including Global Safety Net Contingent credit lines line Facilitate regional reserve erve pool and catastrophe i rese insurance nsurance mechanisms Coping emergency Humanitarian, emergenc cy response, and Technical echnical support for e Te mergency response and emergency f (e.g., reconstruction relief ., food, shelter, health) (e.g. reconstruction econstruction re Stabilization and targeted geted targ d development financing nanc cing Emergency mergency liquidity/swap Em wap lines liquidity/sw Source: WDR 2014 team. Note: The table presents a sequencing of policies based on the guidance of chapter 2 for establishing policy priorities: be realistic in designing poli- cies tailored to the institutional capacity of the country, and build a strong foundation that addresses the most critical obstacles sustainably and that can be improved over time. EWSs = early warning systems. R&D = research and development. RM = risk management. TA = technical assistance. on globalization, relying, instead, on individual, na- Successful international cooperation requires a tional actions to address the complex risks that have cohesive international community where national been collectively created? Individual and national interests are well aligned. It requires an international actions are, of course, essential for any international community that has the capacity to mobilize re- action to be taken, but more ambitious and coor- sources and to establish mechanisms that can enforce dinated efforts are necessary to change the course agreements, even when not all countries are willing and ensure that the whole is greater than the sum to cooperate. That capacity, in turn, rests on the in- of its individual parts. Moving away from global ternational community’s ability to realign incentives cooperative solutions would be costly for develop- around shared goals and to attract participation of ment, especially for developing countries and the major players capable of achieving progress. The poor that have benefited the most from improved international community can scale up risk manage- access to credit and foreign investment flows facili- ment to the extent it can devise innovative mecha- tated by globalization. The international commu- nisms that have a better chance of securing coopera- nity has much to lose by failing to cooperate. Taking tion with appropriate combinations of knowledge, advantage of the positive steps at the local, national, protection, insurance, and coping tools (table 8.2). and individual levels, and building on the lessons learned from the successful examples of interna- When incentives are well aligned: Pursue tional cooperation, the international community proactive and well-coordinated interventions should therefore strive to preserve the gains from globalization and continue its efforts to find the International cooperation works best when incen- right tools, incentives, and institutions to achieve tives are well aligned with a clear course of action. In international cooperation. this case, scaling up risk management requires pro- The role of the international community 271 active, well-coordinated interventions by the inter- communication, and disease-control systems; and national community. For global risks such as finan- developing financial markets for catastrophe-risk cial crises or pandemics, the risk of rapid spillover in financing to facilitate private sector risk solutions. a tightly interconnected world helps align national • Ease resource constraints. Financial support from interests that call for well-coordinated national ac- the international community could augment na- tions to contain risks at the source. The effectiveness tional resources by facilitating and supporting of these actions rests critically on prompt sharing of regional risk-pooling solutions. Financing should information and resources, effective coordination of focus on areas that matter the most and on people actions, and appropriate capacity and infrastructure most vulnerable to shocks. The international as- to monitor, identify, and prevent problems from sistance in cooperation with local and national arising and spreading beyond national borders. authorities to rebuild infrastructure and establish Knowledge is fundamental to broadening per- early warning systems in Indonesia after the 2004 spectives and addressing the problems when they tsunami focused on reducing vulnerabilities to emerge. Access to knowledge is therefore the first step future disasters, whereas slow progress in restor- in boosting risk management capacity. Greater ef- ing infrastructure and access to sanitation, treated forts are particularly needed to do the following: water, and health care following the 2010 earth- • Narrow existing information gaps and address quake in Haiti added to vulnerabilities, including cognitive and behavioral biases. The international to deadly disease outbreaks like cholera (box 8.6). community could increase its own dissemination • Provide appropriate incentives for preparation and and communication of data and analysis and fa- limit moral hazard. Taking into account the cilitate sharing of information and best degree of self-insurance and protection practices, particularly for countries The international and making financing contingent on with limited access to information. adequate risk management could More systematic, frequent, and community should help limit moral hazard. Donor targeted dissemination through strive to preserve the aid to low-income countries knowledge platforms and in- gains from globalization and FCSs can be combined with formation campaigns can help and find the right targeted technical assistance to build longer-term perspectives tools, incentives, and reduce vulnerability to future on rare, high-impact, or distant institutions to achieve shocks and strengthen institu- risks, raising awareness of the tional and governance capacity dangers of inaction. global cooperation. and processes. National and interna- • Reduce the degree of uncertainty about tional platforms can be strengthened to specific risks facing the global system. More assure investors, unleashing capital neces- resources should be devoted to consolidating and sary for growth and rebuilding. disseminating scientific research that can expand knowledge and reduce uncertainty. Knowledge When incentives are not well aligned: Use of the likelihood and nature of complex risks can incremental approaches to global solutions heighten the ability to assess risks and the need for collective action. As the successful global cam- When major sovereigns are not fully engaged—that paigns to eradicate smallpox and protect the ozone is, where progress on fostering collective action has layer demonstrate, partnership with the scientific been limited—new ways of thinking about interna- community and civil society can prompt effective tional cooperation are necessary. Where the conse- action.47 quences of inaction are potentially catastrophic and irreversible, as with climate change, loss of biodiver- International community efforts should focus on sity, or exhaustion of scarce natural resources, lack of providing greater resources for capacity building and full scientific certainty about the dangerous thresh- risk management actions: olds or tipping points should not be used as a reason • Support capacity building for risk management. The for postponing action (cartoon 8.1). On the contrary, international community can further intensify ef- preventive action should be taken in the face of un- forts to assist countries where capacity constraints certainty.48 For these risks, progress can still be made continue to undermine effective risk management. outside a multilateral treaty with full participation.49 The efforts could focus on the capacity to design The international community could embrace in- contingency plans and early-warning, monitoring, cremental deals and actions by an initially small group 272 WORLD DEVELOPMENT REPORT 2014 BOX 8.6 A tale of two disasters The international community typically provides valuable resources earthquake-resistant homes and implement projects to reduce vul- when countries are hit by a massive shock. How those resources are nerability to disasters. used to support national efforts has an important bearing on the Thousands of miles away from Aceh, a similar disaster hit Haiti, results attained. another very poor, fragile island country, where a powerful earth- Aceh Province, a remote region of Indonesia then struggling quake in 2010 left around 230,000 dead and 1.5 million homeless, with conflict, bore the brunt of a powerful earthquake and a massive after poorly constructed homes and infrastructure collapsed. The tsunami that swept over the Indian Ocean in December 2004. More global community rushed to the scene, bringing supplies and vast than 100,000 people in Aceh were killed and over 500,000 were left sums of money. In the midst of this global mobilization, a second homeless. With the financial impact estimated at 97 percent of disaster hit Haiti, when deficient control measures and infrastructure Aceh’s gross domestic product, a special multidonor fund was cre- led to a massive cholera outbreak, killing nearly 8,000—the cholera ated, pooling contributions from 15 countries and organizations to came from a faulty sanitation system at a base of peacekeeping coordinate resources to support the national efforts and govern- troops from a cholera-infected region in South Asia. Only a small ment reconstruction strategy. The funds helped communities to part of the massive foreign aid reached the government because of rebuild houses, local infrastructure, ports, and lost businesses; to donor concerns about the funds’ mismanagement by weak institu- offer scholarships to poor children; and to establish disaster warn- tions and corruption. As of mid-2013, some 350,000 people remain in ing and response systems in hazard-prone areas. Global efforts temporary housing, with little access to sanitation, piped or treated focused on helping local communities and the government to build water, waste management, health care, or education. Source: WDR 2014 team based on Larrimore and Sharkey 2013 and Global Facility for Disaster Reduction and Recovery, “Sendai dialogue: Resilience stories,” https://www.gfdrr.org/node/1308. of participants, while maintaining global collective The board could provide valuable inputs to the coali- action with full participation as the ultimate goal. tion of the willing on the specific issues that need ur- If it can demonstrate benefits from action, the in- gent attention, and offer credibility and legitimacy to cremental approach can serve as a building block to the coalition’s efforts. Further details and background global deals. Countries, international organizations, on the incremental approach are provided in the and specialized entities could form a “coalition of the “Focus on policy reform” at the end of the Report. willing” to coordinate, advocate, and take prompt ac- The crucial step in this incremental approach is tion to address the risks, while creating incentives for finding a common goal around which like-minded others to join, converging over time on a global deal participants can work to realign national interests and with full participation. The coalition should include incentives to examine complex issues and take con- the actors that contribute the most to the problem crete actions. A number of global or regional agree- and those most affected by it, as well as engaging the ments have been reached through such incremental scientific community, civil society, and media. After approaches that started from smaller-scale initiatives global climate change negotiations made only lim- to address a pressing problem of common inter- ited progress in 2009 and 2010, calls for such coali- est (including the Montreal Protocol, Nuclear Non- tions have increased, especially in Europe.50 Proliferation Treaty, European Union, and World The international community could continue to Trade Organization). Once that goal is found, the ways have a crucial role in this setting by developing ap- to achieve it are no different from the essential ele- proaches to tackle the problem collectively; providing ments of reaching a global deal with full participation: platforms for policy discussion; monitoring, report- ing, and aggregating actions; and anchoring them to • Improved access to knowledge and advocacy. The in- existing global frameworks to demonstrate that in- ternational community could do more to facilitate cremental steps and global deals are connected and information sharing and offer longer-term per- heading in the right direction. One way to achieve this spectives through focused information campaigns is by establishing an international risk board that, as and open, transparent knowledge platforms where with the IPCC, would work with scientific and expert individual actions (or inactions) impose serious communities around the world, as well as relevant externalities on others. Such platforms and con- international financial and knowledge institutions, vincing evidence from the scientific community to pool all available knowledge to identify, assess, and (including through an international risk board) manage the major risks that cross national and gen- can help bring diverging views closer, creating a erational boundaries in the near and longer term.51 greater sense of urgency for collective action. Such The role of the international community 273 C A R T O O N 8 .1 Delayed action for climate change can have irreversible consequences. © Kevin Kallaugher/The Economist knowledge was crucial in the success of the Mon- reduced fuel subsidies to encourage more environ- treal Protocol and smallpox eradication. mentally friendly energy options,54 or trade restric- tions to encourage participation and compliance • Financial and technological incentives. These in- with agreements). Subsidies could reward com- centives could help lower participation costs and panies that undertake research to develop green encourage other countries to join the coalition— technologies. Carbon taxes and markets and other particularly developing countries that may be the incentives to limit emissions are being introduced most affected but have the least ability to cope. For in many places in recent years, including in China climate change or loss of biodiversity, for example, and several U.S. states (see the “Focus on policy technology transfers from developed countries reform” at the end of the Report). Larger-scale and could stimulate more environmentally friendly in- coordinated efforts would be needed, however, to dustries and induce the use of cleaner technologies make a material difference and avoid economic and investments in research and development to distortions. devise methods to support climate change mitiga- tion and adaptation and protection of scarce natu- The incremental approach discussed here is not ral resources.52 For example, developed countries without risks and is clearly a second best to a global made a collective commitment to provide new and solution with full cooperation, in effect formalizing additional resources for climate adaptation and free riding by those outside the coalition. There is mitigation in the 2009 and 2010 climate negotia- also no guarantee that the incremental actions will tions, but scaling up funding requires substantial succeed in scaling up efforts and participation to full efforts to mobilize existing and new sources of global action. But the alternative of waiting until an finance.53 International cooperation benefited acceptable deal is reached and all the uncertainties greatly from such transfers in the eradication of resolved is also not viable, if the irreversible con- smallpox and the protection of the ozone layer. sequences of inaction on key global risks are to be avoided. The international community therefore has • Positive and negative financial incentives. These a crucial responsibility to take and support the steps incentives can also help internalize the cost of ex- necessary to protect the world’s vulnerable popula- ternalities created by individual actions (incentives tions and its future generations from the costly and include carbon taxes, cap-and-trade mechanisms, irreversible consequences of today’s inaction. 274 WORLD DEVELOPMENT REPORT 2014 Notes a DDO, a contingent credit line that allows the borrower country to rapidly meet its financing requirements follow- 1. Ghemawat and Altman 2012; IMF 2011. ing a shortfall in resources due to adverse economic events 2. Kaul 2003; Stiglitz 1999; World Bank 2007. (see the experience of Indonesia, which used the DPL DDO 3. In the latter case, where the scope for risk management is in the midst of the 2008–09 financial crisis to support ongo- potentially high, humanitarian instruments may be used to ing access to international capital markets at favorable terms reach out to individuals and communities directly to deal thereby sending a strong positive signal to international and with extreme human and financial costs of recurrent conflicts domestic markets about its economic strength). Detailed (OECD Development Assistance Committee 2012). discussions on experiences with DDO instruments are pro- 4. Crises can result in sharp output losses, increased debt, large vided in http://www.managingclimaterisk.org/document/ fiscal costs, and average recovery time of two to three years; CC_WB.pdf and http://treasury.worldbank.org/web/docu see Laeven and Valencia 2012. ments/DDO_MajorTermsConditions_July12013.pdf. 5. Economist 2012; Calvo 2013 for the WDR 2014; Ötker-Robe 21. Kawai and Lombardi 2012. and Podpiera 2013 for the WDR 2014. 22. Ötker-Robe and others 2011; IMF and FSB 2011. 6. OECD 2013; World Bank 2013. 23. International Monetary Fund and Financial Stability Board, 7. Also see Jonas 2013 for the WDR 2014. IMF-FSB Early Warning Exercises 2012, http://www.imf.org; 8. Centers for Disease Control and Prevention, “First Global World Economic Forum, http://www3.weforum.org/docs/ Estimates of 2009 H1N1 Pandemic Mortality,” http://www WEF_GlobalRisks_Report_2013.pdf; OECD 2011b. .cdc.gov/flu; World Health Organization Facts Sheet, http:// 24. Ötker-Robe 2013; OECD 2013; World Bank 2013. www.who.int; U.S. Department of Health and Human Ser- 25. UNISDR 2006. vices, Pandemic Flu History, http://www.flu.gov. Pandemics 26. See World Bank, World Governance Indicators data- can also have significant direct and indirect economic costs. base for corruption rankings, http://info.worldbank.org/ 9. IPCC 2012. About three-fourths of the total number of di- governance/wgi/index.asp; and World Economic Forum sasters since 1903 have taken place in the past three decades, Global Competitiveness Index, http://www.weforum.org/ when the Earth’s temperature started to rise rapidly. issues/global-competitiveness. 10. Ghesquiere and others 2012. 27. Evidence suggests that corruption can lower a countries’ eco- 11. Kindleberger 1973. nomic growth as much as 0.5–1.0 percent a year and decrease 12. Anderson, Ivanic, and Martin 2013. investments in health and education systems, leading to an 13. Early warning systems are an effective tool for preparation increase in infant mortality, poverty, and inequality. Standard against risks. After the tsunami in 2004, for example, coun- & Poor’s estimates that in countries with corruption, inves- tries in the Indian Ocean region invested in tsunami warning tors, including donors, have a 50–100 percent chance of losing systems. Similar systems have been set up for storms in Ban- their investments within five years (up to $30 billion in aid to gladesh and Cuba and are connected to a web of public shel- Africa has ended up in foreign bank accounts; see Transpar- ters. Subbiah, Bildan, and Narasimhan (2008) calculate very ency International, http://www.transparency.org/). high benefit-cost ratios for systems that warn of storm-related 28. See, for example, World Bank 2009; Aldy, Orszag, and Stiglitz floods in Bangladesh. Also see World Bank and UN 2010. 2001; Barrett 2003. 14. BCBS 2009, 2011; FSB 2010; IMF and FSB 2011. 29. Barrett and Dannenberg 2012. 15. UN General Assembly 2000. 30. For example, Nicaragua declined to pursue a weather- 16. United Nations 2011, Towards a Safer World initiative. indexing program after it had been priced in the global re- 17. OECD 2011a. insurance market, citing, among other things, international 18. Examples from the World Bank include development of a assistance following Hurricane Mitch in 1998 as an indication platform for a multicountry, multiperil catastrophe bond (the of dependable alternatives (World Bank and UN 2010). MultiCat Program, in collaboration with Mexico) that trans- 31. Aizenman and Ötker-Robe 2013 for the WDR 2014; Noy fers risk to private investors and allows pooling of multiple 2012. risks to take advantage of diversification benefits; intermedia- 32. Ghesquiere and others 2012. tion services to help Malawi protect against the risk of severe 33. IMF 2012. drought; and advisory services to help Turkey establish na- 34. World Bank 2012a. tional catastrophe insurance pool for earthquakes; see Mahul 35. World Bank 2009. and Cummins 2009; Mahul and Ghesquiere 2010. 36. Brahmbhatt and Dutta 2008 argue that efforts to avoid infec- 19. For example, since December 2011, the World Bank Immedi- tion through reduced travel or trade account for 60 percent of ate Response Mechanism has allowed low-income countries the economic costs during a pandemic. to rapidly access a portion of their undisbursed investment 37. Aldy, Orszag, and Stiglitz 2001; Banerjee 2012; Barrett 2007, project balances to mitigate the impact of natural disasters or 2008; Dellink and Finus 2012; Kaul 2003; Stern 2007. economic shocks on vulnerable groups and to protect criti- 38. Stern 2007; Kaul 2003. cal development spending. Similarly, the Bank’s Development 39. See Warner 2013, who discusses these challenges in detail. Policy Loan (DPL) with CAT DDO has a contingent credit 40. Campbell, Einhorn, and Reiss 2004; Fitzpatrick 2009. line that provides immediate liquidity to IBRD countries in 41. Sheng 2013. the aftermath of a natural disaster; http://www.gfdrr.org/sites/ 42. Collier and Hoeffler 2004; Chalmers 2004. gfdrr.org/files/documents/DRFI_CatDDO_ProductNote_ 43. Bailey 2013. Jan11.pdf. 44. See Jim Yong Kim, “Spillover from Syria: Helping a Neighbor 20. Examples are the IMF’s Flexible Credit Line, which allows Cope,” The Guardian Online—Global Development, July 29, qualified countries to draw on the credit line at any time 2013. within a specified window, and the World Bank’s DPL with 45. DFID 2005. The role of the international community 275 46. Goldin 2013; Hale 2011; Lagarde 2012. Barrett, Scott, and Astrid Dannenberg. 2012. “Climate Negotia- 47. See, among others, Barrett 2003, 2007, 2008; Barrett and tions under Scientific Uncertainty.” PNAS 109 (43): 17372–76. Danneberg 2012; and Stern 2007. BCBS (Basel Committee on Banking Supervision). 2009. 48. See, for example, the United Nations Rio Declaration from the “Strengthening the Resilience of the Banking Sector.” Consul- 1992 United Nations Earth Summit. http://www.unesco.org/ tative Document, Bank for International Settlements, Basel. education/nfsunesco/pdf/RIO_E.PDF. ———. 2011. “Basel III: A Global Regulatory Framework for 49. See also Jim Yong Kim, “Make Climate Change a Prior- More Resilient Banks and Banking Systems.” Bank for Inter- ity,” Washington Post Opinions, January 24, 2013, http:// national Settlements, Basel. www.washingtonpost.com/opinions/make-climate-change- Brahmbhatt, Milan, and Arindam Dutta. 2008. “Economic Effects a-priority/2013/01/24/6c5c2b66-65b1-11e2-9e1b-07 during Outbreaks of Infectious Disease.” World Bank Research db1d2ccd5b_story.html. Digest 2 (2): 7. 50. Falkner, Stephan, and Vogler 2010; Goldin 2013; and Hale Calvo, Sara. 2013. “Financial Crises, Social Impact, and Risk Man- 2011. See also http://www.euractiv.com/climate-environment/ agement: Lessons and Challenges.” Background paper for the europe-looks-coalition-willing-d-news-508909. World Development Report 2014. 51. See a related proposal by the German Institute for Develop- Campbell, Kurt M., Robert J. Einhorn, and Mitchell B. Reiss. 2004. ment 2009. The Nuclear Tipping Point: Why States Reconsider Their Nu- 52. See, for example, Barrett 2003; The Royal Society 2009; World clear Choices. Washington, DC: Brookings Institution Press. Bank 2009. Caravani, Alice, Smita Nakhooda, Charlene Watson, and Liane 53. The agreements required provision of $30 billion in Fast Start Schalatek. 2012. “The Global Climate Finance Architecture.” Finance for 2010–12 and $100 billion a year by 2020. While Climate Finance Fundamentals 2, http://www.climatefunds the $30 billion goal is close to being realized, the second goal update.org. is falling short of the needs, in part because of the fiscal prob- Chalmers, Malcolm. 2004. “Spending to Save? An Analysis of the lems in advanced countries. World Bank 2009, 2012b, 2012c; Cost Effectiveness of Conflict Prevention.” Paper prepared for Caravani and others 2012; Schalatek and others 2012a, 2012b. the Bottom Billion Conference organized by the Centre for the 54. Ending fuel subsidies globally could lead to a 5 percent fall Study of African Economies, Oxford University, Oxford, U.K., in emissions by 2020. See Jim Yong Kim, “Make Climate June 27–29. Change a Priority,” Washington Post Opinions, January 24, 2013, Cole, Daniel H. 2007. “Climate Change and Collective Action.” http://www.washingtonpost.com/opinions/make-climate- Indiana University, Bloomington, IN. change-a-priority/2013/01/24/6c5c2b66-65b1-11e2-9e1b-07 Collier, Paul, and Anke Hoeffler. 2004. “Aid, Policy and Growth db1d2ccd5b_story.html. in Post-Conflict Societies.” European Economic Review 48 (5): 1125–45. DARA International. 2012. Climate Vulnerability Monitor: A Guide to References the Cold Calculus of a Hot Planet. Madrid: DARA International. AidData. AidData by Activity (database). AidData, Washington, Dellink, Rob, and Michael Finus. 2012. “Uncertainty and Climate DC, http://aiddata.org/content/index/data-search. Treaties: Does Ignorance Pay?” Resources and Energy Econom- . ics 34 (4): 565–84. Aizenman, Joshua, and Inci Ötker-Robe. 2013. “Managing Risk for Development: International Risk Sharing Tools.” Background DFID (Department for International Development). 2005. “Why paper for the World Development Report 2014. We Need to Work More Effectively in Fragile States.” DFID, Aldy, Joseph E., Peter R. Orszag, and Joseph E. Stiglitz. 2001. London. “Climate Change: An Agenda for Global Collective Action.” Economist. 2012. “Lost Economic Time: The Proust Index.” Febru- Paper prepared for the Timing of Climate Change Policies ary 25. Conference organized by Center for Climate and Energy Solu- Falkner, Robert, Hannes Stephan, and John Vogler. 2010. “Interna- tions, Washington, DC, October 10–12. tional Climate Policy after Copenhagen: Towards a ‘Building Anderson, Kym, Maros Ivanic, and Will Martin. 2013. “Food Price Blocks’ Approach.” Global Policy 1 (3): 252–62. Spikes, Price Insulation, and Poverty.” Policy Research Work- Fengler, Wolfgang, and Homi J. Kharas. 2011. “Delivering Aid Dif- ing Paper 6535, World Bank, Washington, DC. ferently: Lessons from the Field.” Economic Premise 49. World Bailey, Rob. 2013. “Managing Famine Risk: Linking Early Warning Bank, Washington, DC. to Early Action.” Chatham House, London. Fitzpatrick, Mark. 2009. “Successes and Failures of NPT or World Banerjee, Subhabrata Bobby. 2012. “A Climate for Change? Criti- without NPT?” Paper presented at a seminar on Nuclear Non- cal Reflections on the Durban United Nations Climate Change Proliferation: Challenges & Opportunities, Webster Univer- Conference.” Organization Studies 33 (12): 1761–86. sity, St. Louis, MO, October 29. Barrett, Scott. 2003. “Creating Incentives for Cooperation: Stra- FSB (Financial Stability Board). 2010. “Progress since the Wash- tegic Choices.” In Providing Global Public Goods: Managing ington Summit in the Implementation of the G20 Recom- Globalization, edited by Inge Kaul, Pedro Conceição, Katell Le mendations for Strengthening Financial Stability.” Report to Goulven, and Ronald U. Mendoza, 308–28. New York: Oxford G20 Leaders, FSB, Basel. University Press. German Institute for Development. 2009. “Globalization at the ———. 2006. “The Smallpox Eradication Game.” Public Choice Crossroad: An “International Panel on Systemic Risks in the 130 (1/2): 179–207. Global Economy” is Needed.” Briefing Paper 6/2009. German ———. 2007. Why Cooperate? The Incentive to Supply Global Institute for Development, Bonn. Public Goods. New York: Oxford University Press. Ghemawat, Pankaj, and Steven A. Altman. 2012. “DHL Global ———. 2008. “Climate Treaties and the Imperative of Enforce- Connectedness Index 2012: Analyzing Global Flows and Their ment.” Oxford Review of Economic Policy 24 (2): 239–58. Power to Increase Prosperity.” Deutsche Post DHL, Bonn. 276 WORLD DEVELOPMENT REPORT 2014 Ghesquiere, Francis, Prashant, Robert Reid, Jan Kellett, Shyam Mercer. 2011. “Climate Change Scenarios: Implications for Strate- KC, and Jack Campbell. 2012. “The Sendai Report: Managing gic Asset Allocation.” Mercer, New York. Disaster Risks for a Resilient Future.” Paper prepared for the NASA (National Aeronautics and Space Administration). God- Sendai Dialogue organized by World Bank, Sendai, Japan. dard Institute for Space Studies Surface Temperature Analy- Goldin, Ian. 2013. Divided Nations: Why Global Governance Is sis (GISTEMP) (database), NASA, New York, http://data.giss Failing and What We Can Do about It. New York: Oxford Uni- .nasa.gov/gistemp/. versity Press. Nielsen Company. 2011. “Sustainable Efforts & Environmental Hale, Thomas. 2011. “A Climate Coalition of the Willing.” Wash- Concerns around the World.” A Nielsen Report. Nielsen, New ington Quarterly Winter 34 (1): 89–101. York. IDA (International Development Association). 2012. “Implemen- Noy, Ilan. 2012. “Investing in Early Warning Systems: A Global tation and Results Progress Report.” IDA16 Mid-Term Review, Fund for Disaster Preparedness.” Paper prepared for the World Bank, Washington, DC. Copenhagen Consensus Project organized by Copenhagen IMF (International Monetary Fund). 2011. “Mapping Cross- Consensus Center, Copenhagen, March. Border Financial Linkages: A Supporting Case for Global Fi- OECD (Organisation for Economic Co-operation and Develop- nancial Safety Nets.” IMF, Washington, DC. ment) 2011a. “A New Deal for Engagement in Fragile States.” ———. 2012. “Taking Stock: A Progress Report on Fiscal Adjust- Paper prepared for the International Dialogue on Peacebuild- ment.” Fiscal Monitor, October, IMF, Washington, DC. ing and Statebuilding organized by OECD, November 29– IMF and FSB. 2011. “The Financial Crisis and Information Gaps.” December 1. Implementation Progress Report to the G20, IMF and FSB, ———. 2011b. Future Global Shocks: Improving Risk Governance. Washington, DC. OECD Reviews of Risk Management Policies. Paris: OECD. IPCC (Intergovernmental Panel on Climate Change). 2007. “Cli- ———. 2011c. Managing Risks in Fragile and Transitional Con- mate Change 2007—Mitigation of Climate Change: Contri- texts. The Price of Success? Paris: OECD. bution of Working Group III to the Fourth Assessment Report ———. 2012. Fragile States 2013. Resource Flows and Trends in a of the Intergovernmental Panel on Climate Change.” Cam- Shifting World. Paris: OECD. bridge University Press, Cambridge, U.K. ———. 2013. Fragile States: Resource Flows and Trends. Conflict ———. 2012. Managing the Risks of Extreme Events and Disasters and Fragility Series. Paris: OECD. to Advance Climate Change Adaptation. Special Report of the OECD Development Assistance Committee (DAC). 2012. Inter- IPCC. Cambridge, U.K.: Cambridge University Press. national Support to Post-Conflict Transition: Rethinking Policy, Jacoby, Hanan, Mariano Rabassa, and Emmanuel Skoufias. 2011. Changing Practice. DAC Guidelines and Reference Series. “Distributional Implications of Climate Change in India.” Paris: OECD. Policy Research Working Paper 5623, World Bank, Washing- OECD and G20 (Group of 20). 2012. Disaster Risk Assessment ton, DC. and Risk Financing: A G20/OECD Methodological Framework. Jonas, Olga. 2013. “Pandemic Risk.” Background paper for the Paris: OECD. World Development Report 2014. . Ötker-Robe, Inci, 2013. “Seizing Opportunities under Extreme Kaul, Inge. 2003. Providing Global Public Goods: Managing Global- Risks: Fragile and Conflict-Affected States.” Let’s Talk Devel- ization. New York: Oxford University Press. opment (blog), May 17. http://blogs.worldbank.org/develop Kawai, Masahiro, and Domenico Lombardi. 2012. “Financial Re- menttalk/seizing-opportunities-under-extreme-risks-fragile- gionalism.” Finance and Development 49 (3): 23–25. Kindleberger, Charles P. 1973. The World in Depression: 1929– and-conflict-affected-states. . 1939. Berkeley, CA: University of California Press. Ötker-Robe, Inci, Aditya Narain, Anna Ilyina, and Jay Surti. Laeven, Luc, and Fabian Valencia. 2012. “Systemic Banking Crises 2011. “The Too-Important-to-Fail Conundrum: Impossible Database: An Update.” Working Paper WP/12/163, Interna- to Ignore and Difficult to Resolve.” Staff Discussion Note tional Monetary Fund, Washington, DC. SDN/11/12, International Monetary Fund, Washington, DC. . Lagarde, Christine. 2012. “Fragmentation Risks.” Finance and Ötker-Robe, Inci, and Anca Podpiera. 2013. “Social Impact of Fi- Development 49 (3): 26–27. nancial Crises.” Background paper for the World Development Larrimore, J. T., and Brielle Sharkey. 2013. “Haiti Continues to Report 2014. Struggle Three Years after the Earthquake.” Council on Hemi- Rae, Ian. 2012. “Saving the Ozone Layer: Why the Montreal Pro- spheric Affairs, Washington, DC, January 18, http://www.coha tocol Worked.” The Conversation, September 10, http://the .org/haiti-continues-to-struggle-three-years-after-the-earth conversation.com/saving-the-ozone-layer-why-the-montreal- quake/. protocol-worked-9249. Lenton, Timothy M., Hermann Held, Elmar Kriegler, Jim W. Ranger, Nicola, Stephane Hallegatte, Sumana Bhattacharya, Mur- Hall, Wolfgang Lucht, Stefan Rahmstorf, and Hans Joachim thy Bachu, Satya Priya, K. Dhore, Farhat Rafique, P. Mathur, Schellnhuber. 2008. “Tipping Elements in the Earth’s Climate Nicolas Naville, Fanny Henriet, Celine Herweijer, Sanjib Pohit, System.” PNAS 105 (6): 1786–93. and Jan Corfee-Morlot. 2011. “An Assessment of the Potential Mahul, Olivier, and J. David Cummins. 2009. Catastrophe Risk Impact of Climate Change on Flood Risk in Mumbai.” Climate Financing in Developing Countries: Principles for Public Inter- Change 104 (1): 139–67. vention. Washington, DC: World Bank. The Royal Society. 2009. “Geoengineering the Climate: Science, Mahul, Olivier, and Francis Ghesquiere. 2010. “Financial Protec- Governance and Uncertainty.” The Royal Society, London. tion of the State against Natural Disasters: A Primer.” Policy Schalatek, Liane, Smita Nakhooda, Sam Barnard, and Alice Cara- Research Working Paper 5429, World Bank, Washington, DC. vani. 2012a. “Adaptation Finance.” Climate Finance Funda- Marcott, Shaun A., Jeremy D. Shakun, Peter U. Clark, and Alan C. mentals 3, http://www.climatefundsupdate.org. Mix. 2013. “A Reconstruction of Regional and Global Temper- ———. 2012b. “Mitigation Finance.” Climate Finance Funda- ature for the Past 11,300 Years.” Science 339 (6124): 1198–201. mentals 4, http://www.climatefundsupdate.org. The role of the international community 277 Scripps Institution of Oceanography. Atmospheric CO2 Con- UN General Assembly. 2000. “Resolution Adopted by the Gen- centration at Mauna Loa Observatory (Hawaii) (database), eral Assembly: United Nations Millennium Declaration.” Scripps Institution of Oceanography, La Jolla, CA, http:// A/RES/55/2, United Nations, New York, September 18. scrippsco2.ucsd.edu/data/atmospheric_co2.html. Université Catholique de Louvain. EM-DAT: The OFDA/CRED Sheng, Andrew. 2013. “The Contradictions of System Stability: International Disaster Database. Université Catholique de One Asian View.” Speech at Thirteenth IMF/World Bank/ Louvain, Brussels, http://www.emdat.be. Federal Reserve Board Annual International Conference on Warner, Koko. 2013. “Climate Change and Global Warming: The Policy Challenges for the Financial Sector, Washington, DC, Role of the International Community.” Background paper for June. the World Development Report 2014. Stern, Nicholas. 2007. The Economics of Climate Change: The Stern World Bank. 2007. “Global Public Goods: A Framework for the Review. Cambridge, UK: Cambridge University Press. Role of the World Bank.” Paper prepared for the Development Stiglitz, Joseph E. 1999. “Knowledge as a Public Good.” In Global Commitee Meeting, Washington, DC, October 21. Public Goods: International Cooperation in the 21st Century, ———. 2009. World Development Report 2010: Development and edited by Isabelle Grunberg, Inge Kaul, and Marc A. Stern, Climate Change. Washington, DC: World Bank. 308–25. New York: Oxford University Press. ———. 2011. World Development Report 2011: Conflict, Security, Subbiah, A., Lolita Bildan, and Ramraj Narasimhan. 2008. “Assess- and Development. Washington, DC: World Bank. ment of the Economics of Early Warning Systems for Disas- ———. 2012a. “People, Pathogens and Our Planet: The Econom- ter Risk Reduction.” Background paper submitted to Global ics of One Health.” World Bank, Washington, DC. Facility for Disaster Reduction and Recovery, Washington, DC. ———. 2012b. “A Strategic Framework for the World Bank UNEP (United Nation Environment Programme). 2007. “Mon- Group: Completion Report FY09–11.” Development and treal Protocol: A Success in the Making.” UNEP, Nairobi. Climate Change, World Bank, Washington, DC. UNISDR (United Nations International Strategy for Disaster Re- ———. 2012c. “Turn Down the Heat: Why a 4oC Warmer World duction). 2006. “Global Survey of Early Warning Systems: An Must Be Avoided.” World Bank, Washington, DC. Assessment of Capacities, Gaps and Opportunities Towards ———. 2013. “Stop Conflict, Reduce Fragility and End Poverty: Building a Comprehensive Global Early Warning System for Doing Things Differently in Fragile and Conflict-Affected All Natural Hazards.” UNISDR, Geneva. Situations.” World Bank, Washington, DC. ———. 2007. “Hyogo Framework for Action 2005–2015: Build- ———. World Development Indicators (database). World Bank, ing the Resilience of Nations and Communities to Disasters.” Washington, DC, http://data.worldbank.org/data-catalog/ UNISDR, Geneva. world-development-indicators. United Nations. 2011. “Beyond Pandemics: A Whole-of-Society World Bank and IMF. 2010. Global Monitoring Report: The MDGs Approach to Disaster Preparedness.” Towards a Safer World after the Crisis. Washington, DC: World Bank. initiative. September. World Bank and UN. 2010. Natural Hazards, UnNatural Disasters: United Nations Framework Convention on Climate Change. The Economics of Effective Prevention. Washington, DC: World 1998. “Kyoto Protocol.” United Nations, New York. Bank. FOCUS ON POLICY REFORM Mainstreaming risk management into the development agenda: Selected institutional reforms The discussion that follows expands on four selected areas where fundamental institutional reforms for better risk management are needed. The four areas cover integrated risk management, fiscal and financial risk management, social insurance and work status, and multinational approaches to address global risks. Why these four? They are by no means exhaustive, but they do represent a framework through which many specific recommendations provided in the Report could be implemented. They discuss innovative solutions to long-standing problems in developing countries. They cut across risks and social systems, using a holistic approach to risk management. Along with the other recommendations in the Report, they can contribute to mainstreaming risk management into the develop- ment agenda. The four reforms may require substantial changes in the way national governments develop and implement their general plans. Reform 1. Establish a national risk board to assess risks—what the Dutch and U.K. authorities call risk man- and manage risks in an integrated way agement “capabilities.” Important trade-offs and synergies also exist across What is the problem? Governments and public agencies of- risks or across scales. For instance, providing a public retire- ten manage specific risks in an isolated manner, which can ment scheme reduces risks for individuals but may increase lead to ineffective formulation and implementation of risk aggregate fiscal risks. And a watershed that covers multiple management strategies. For example, while the ministry of municipalities can be managed effectively only in a coordi- finance can create and regulate a health care insurance sys- nated manner. An integrated and multistakeholder ap- tem to better manage health risks, the usefulness of that proach helps deal with these trade-offs and reduces the system depends on the availability of competent health likelihood of simply transferring a risk of one type to a risk care providers, which is a responsibility of the health minis- of another type (such as from idiosyncratic risk to systemic try. Shifting from one energy source to another (coal to gas risk) or from one agent to another. or nuclear) may reduce one kind of pollution but increase Involving more stakeholders (policy makers, industry other pollutants or security risks. experts, and academics) in the process of designing a na- Such “risk-risk trade-offs” and coordination problems tional risk management strategy also makes the process often arise from narrow decisions by risk managers with more transparent and less prone to political capture and restricted perimeters of responsibility.1 Ideally, broader introduces natural accountability mechanisms. All too of- analyses can help risk managers develop “risk-superior so- ten, risks that evolve over long time horizons and the lack lutions” that reduce multiple risks at the same time.2 Look- of clear indicators of success for risk management limit the ing at risks in an integrated manner helps define policy accountability of decision makers for their risk manage- priorities and avoids overspending on managing one risk ment choices. This problem can be addressed, in part, by an while neglecting others, helping to achieve a good balance independent and multistakeholder entity that analyzes between preparation for low-probability but high-impact and publishes assessments of risk management practices events (such as earthquakes) and less spectacular risks within a country and that makes expert and policy-relevant (such as truck accidents) that are more prevalent and are recommendations. also costly to society. Managing individual risks entails both trade-offs and What is the solution? A national risk board should be created synergies. A multistakeholder approach to national risk to provide integrated risk management at the national management helps identify and capture synergies across level. This recommendation builds on analogous proposals, risks: for instance, developing the capacity to evacuate including the national Council of Risk Analysts proposed by populations while taking into account the constraints of Graham and Wiener,3 and the World Economic Forum’s pro- available crisis management infrastructure. National risk posal to establish a country risk officer4—similar to the po- assessments undertaken in the Netherlands and the sition of chief risk officer that has been created in many United Kingdom aim explicitly at identifying investments multinational companies, notably financial corporations. that increase the ability to anticipate and manage multiple The board’s expertise should cover the areas of military, Overview Focus on policy reform 279 security, and terrorism risk; economic risk; environmental, health, and could either be an advisory body or have powers to implement rec- technological risk; and social risk. It should also consider the actions ommendations, or a combination of both. It could consist entirely of undertaken by other countries, multinational firms, and the global experts or policy makers or a combination of both. There are trade- community. offs among these design choices, which are illustrated in diagram A national risk board can be set up as a standing (permanent) F1.1. For instance, a board of experts with powers to implement policy committee and should have powers to issue “act-or-explain” recom- could lack legitimacy, especially if it were to implement policies with mendations directed at the relevant authorities responsible for policy significant redistribution effects (such as raising taxes to cover disas- implementation. That is, government agencies and local authorities ter insurance premiums).6 In contrast, a board of experts issuing only would have to act on the board’s recommendations or explain why nonbinding recommendations could lack relevance to policy making they have decided to discard them. or be unable to influence actual decisions. If the board comprises only The board should analyze risks and risk management policies and policy makers and issues nonbinding recommendations, it could lack practices, including synergies and trade-offs across risks or across en- credibility. Finally, if a board has implementation powers but consists tities; define priorities in risk management; and make recommenda- only of policy makers, it could lack expertise and be vulnerable to po- tions for appropriate policies to pursue. Many countries already have litical capture. To avoid becoming a powerless body, the board should regular national risk assessments conducted by multistakeholder have sufficient visibility: its chair should be a highly visible policy teams involving various ministries and often including representa- maker, and its annual meeting should be chaired by the head of gov- tives of the private sector and civil society. The Netherlands, Singa- ernment. The board should be held accountable by having to publish pore, the United Kingdom, and the United States have undertaken its recommendations, by issuing annual reports with policy priorities such assessments, and other countries, such as Morocco, are working and their analytical substantiation, and by being subjected to annual to set up a national assessment process. But this process is usually hearings in front of a legislative committee. carried out by a temporary, ad hoc group that exists only while the The appropriate institutional design will depend on the country assessment is taking place. Moreover, the political relevance and political and institutional context. For instance, rather than establish- accountability of such ad hoc groups generally have been weak. ing an independent government agency, Jamaica, Mexico, and Mo- Some countries go beyond risk assessments. Some have created rocco are considering placing the integrated risk management func- multiministry bodies in charge of information exchange and coordi- tion within the government structure. Such an institutional design nation for risk management, but these bodies usually deal with a spe- may be practical in countries with an effective and independent civil cific risk—most often natural disasters, as in Peru. Few countries actu- service, with the national risk board members appointed as expert ally have an integrated risk management agency that deals with technocrats with guaranteed positions for periods that extend be- multiple risks. yond a political cycle. However, any institutional design should seek One country that does is Singapore, which has a framework, the to balance legitimacy, relevance, credibility, and expertise (depicted Whole-of-Government Integrated Risk Management approach, dedi- as the balanced region in diagram F1.1). cated to avoiding silo effects within the government and to managing risks in an integrated manner.5 The institutional umbrella of the frame- work is the Strategy Committee, which is charged with steering and D I A G R A M F1.1 Balancing the trade-offs in the reviewing the implementation of the framework. The committee, institutional design of a national risk board which meets quarterly, comprises permanent secretaries from various Independent ministries across government and is chaired by the Head of Civil Ser- experts vice. In addition, the Homefront Crisis Management system includes a Lower Lower ministerial committee chaired by the Minister of Domestic Affairs, relevance legitimacy which is responsible for crisis management. It is supported by the Homefront Crisis Executive Group, which comprises senior represen- tatives from ministries and government agencies. This multirisk ap- proach is complemented by more sectoral agencies, such as the Na- tional Security Coordination Secretariat, which focuses on national Balanced Advisory Implementation security issues. Singapore’s institutional arrangement for integrated region role role risk management involves a great deal of specialization and a com- plex coordination process that has evolved over time. For developing countries, a simpler, consolidated arrangement that involves less specificity and specialization in the institutional de- Lower Lower sign and more explicit and robust coordination mechanisms might be credibility expertise desirable. The proposed National Risk Board takes into account such considerations. Policy makers How can it be implemented? The board needs to have the required Source: WDR 2014 team. expertise, be credible and relevant, and have sufficient legitimacy. It 280 WORLD DEVELOPMENT REPORT 2014 Reform 2. Create independent fiscal and financial agencies to How can it be implemented? Fiscal authorities have adopted quantita- promote sustainable policies tive limits on deficits, spending, debt, or some combination, to con- tain fiscal profligacy. However, these numerical limits have restricted Establish fiscal councils to promote fiscal sustainability countercyclical responses during downturns and have led politicians What is the problem? Very few developing countries have been able to to circumvent them through the use of creative accounting, such as conduct countercyclical fiscal policies.7 Rather than saving during Stability and Growth Pact rules in the European Union.13 Rather than good times, policy makers typically increase government spending, imposing rigid numerical limits, fiscal authorities should focus on us- run budget deficits, and accumulate debt. Over the past five decades, ing flexible procedural rules that target the structural budget balance government spending has behaved procyclically in more than 90 and provide a blueprint to achieve this target over time. Targeting percent of developing countries; in sharp contrast, it has been structural budget balances—as is done in Chile and Norway—can de- countercyclical in 80 percent of industrial countries.8 Procyclical fiscal liver fiscal discipline and endow policy makers with flexibility to con- policies have increased output volatility and hindered long-term duct countercyclical policies. Before the crisis, in 2007, strong eco- growth throughout the developing world.9 nomic performance and sharp increases in the prices of oil and copper Two main factors explain this procyclical bias in developing coun- allowed Chile and Norway—through their rules—to amass a signifi- tries. First, limited access to world capital markets during recessions cant amount of public savings. The general government primary sur- forces governments to raise taxes and cut spending in bad times. pluses that year were 11.8 of gross domestic product (GDP) in Chile Second, political economy considerations—including distributional and 15.7 percent in Norway, providing a comfortable cushion for conflicts and information asymmetries—prevent governments from countercyclical policies following the crisis. acting prudently during upswings. Competition among multiple Currently, more than 40 percent of advanced countries and about power blocs for greater revenue windfalls leads to overspending and 20 percent of emerging market have a national fiscal rule targeting overprovision of some public goods.10 Voters’ perception that their the structural budget balance.14 However, the effectiveness of these governments are rent-seeking leads to increasing popular pressure to rules rests upon their credibility and flexibility: they may lack credibil- lower taxes and increase spending in good times.11 ity if not accompanied by budget transparency and clear operational By contrast, monetary authorities in several developing countries guidance or if they are overly ambitious or unrealistic. Defining a have succeeded in adopting a credible, predictable, and sustainable structural budget balance rule can create monitoring and communi- regime in the form of inflation targeting. Several developed and de- cation problems. Moreover, fiscal rules cannot anticipate every possi- veloping countries have maintained low and stable inflation, thanks ble contingency. Their flexibility could be enhanced through the de- to monetary frameworks that benefit from a clear mandate, indepen- sign and incorporation of escape clauses that would take into account dence from political interference, and accountability for policy mak- extreme events (crises, disasters).15 Fiscal councils can help identify ers’ actions. A greater institutional push toward transparent monetary the events that trigger escape clauses and decide on the treatment of frameworks has provided central banks the flexibility to conduct cumulative deviations.16 countercyclical policies without jeopardizing inflationary goals. There Fiscal councils can shield some budget procedures from political is need for similar credible, predictable, and sustainable frameworks pressure, thereby containing the government’s incentives to over- for fiscal policy. spend. Overspending and lack of budget discipline can be traced, in part, to overly optimistic government forecasts.17 Fiscal councils can What is the solution? The codification of flexible fiscal rules in legisla- produce official forecasts for GDP growth and government budgetary tion, along with the operation of autonomous fiscal councils, has items. The U.K. Treasury (ministry of finance), for instance, has dele- the potential to restrain policy makers from spending sprees in nor- gated such forecasts to the Office for Budget Responsibility. Forecast- mal times and to allow for additional (spending) stimulus in crisis ing contains its own risks, however. Forecasting errors in uncertain times. Given the redistributive nature of fiscal policy, full delegation of environments can threaten the credibility of the council.18 The accu- policy making to these councils is unrealistic. Fiscal councils can racy of the council’s real GDP growth and budget forecasts will be nonetheless shape incentives more effectively than can a process that reduced by the greater volatility associated with higher economic un- simply and mechanically follows numerical limits on budgetary ag- certainty. Councils will have to be held accountable for incorrect gregates. The councils should have a clear mandate, autonomy to predictions. operationalize budget procedures, and the power to monitor compli- By providing independent analysis of fiscal plans and executed ance with the fiscal rule. Fiscal councils should hold policy makers policies, councils raise voters’ awareness of the consequences of pol- accountable for their actions and be accountable for their advice and icy actions. For instance, the Netherlands Bureau for Economic Policy recommendations. To put fiscal councils in place and uphold their Analysis (CPB) evaluates whether government policies threaten fiscal powers, broad consensus needs to be built to implement these insti- sustainability. Councils can also evaluate the cost of electoral plat- tutional reforms and encourage policy makers to deliver viable coun- forms plans and coalition agreements after elections.19 Finally, fiscal tercyclical actions. Severe crises may provide that opportunity—that councils can hold policy makers accountable for choices made re- has been the case in the European Union with the new Fiscal Com- garding the cyclical operation of the rule and define clear legal sanc- pact Treaty and “Two-Pack” regulation proposal.12 However, establish- tions before the fact for noncompliance.20 ing these councils requires strong institutional underpinnings. In Governments have incentives to dismiss the advice of fiscal coun- countries with weak governance and capacity, transparent and com- cils. Councils can be dismantled if their critique of the government is prehensive fiscal frameworks (including top-down approaches to too severe or if they are formed without adequate political consen- budgeting) would provide a good foundation for more institution sus—as was the case in Hungary. Fiscal councils need legitimacy, as building in the future. well as budgetary and political independence, to work effectively and Focus on policy reform 281 to avoid political capture. So far, countries have not granted political bly the microprudential regulator and the business conduct regulator, autonomy to fiscal councils. Councils have had to rely on informal in- to implement measures to foster financial stability.26 The FPC is likely dependence acquired through the buildup of reputation over time. to receive more direct tools to fulfill its statutory responsibility.27 Councils with the largest degree of informal independence are the oldest ones—Denmark’s economic council, the Netherlands’ CPB, and Reform 3. For debate: Should access to social insurance be tied the U.S. Congressional Budget Office.21 Limited resources and budget to work status? dependence on governmental offices can reduce the councils’ quality of work—as has happened in Canada and Sweden.22 Social insurance (including pensions and health insurance) protects The council board members should be recruited competitively. people’s income and consumption in the face of potentially devastat- Reputational costs of bad performance would act as a disciplining de- ing shocks such as illness or life-cycle transitions such as old age.28 vice. Nonetheless, members’ idiosyncrasies or dismal performance This is particularly true for the most vulnerable segments of the pop- can affect the work of the entire council. Regular evaluations are war- ulation, which lack the resources and access to financial markets ranted to hold council members accountable, including testifying on to accumulate savings and purchase private insurance products. A a regular basis before the legislative body and continuous evaluation good social insurance system is one that is inclusive, that protects by international peer councils or expert groups.23 people equitably, that is fiscally sustainable in the long term, and that minimizes disincentives to work, save, and participate in the for- Put in place independent macroprudential supervisors for financial mal economy. stability What is the problem? Many countries have established so-called con- What is the problem? The main difficulties for the financial system are tributory social insurance systems, financed by mandatory payroll managing systemic risk (stemming from negative externalities and taxes levied on employers and contributions paid by employees. In herding behavior among individual financial firms) and avoiding reg- economies with high levels of formality, this system has been success- ulatory capture by politicians and the financial industry (chapter 6). ful in providing insurance to most people. By contrast, in countries with large shares of self-employed and agricultural workers, contribu- What is the solution? The solution is to delegate the oversight of finan- tory systems cover only a minority of the population. The traditional cial stability to an independent macroprudential committee, possibly approach thus ends up excluding many workers—mostly those who under the central bank. In a number of emerging market countries, are low-income, are self-employed, or work in agriculture. including the Czech Republic, South Africa, and Thailand, the respon- To narrow the coverage gap, a growing number of countries have sibility for financial stability oversight already has been given to the introduced noncontributory insurance, where benefits are financed central bank, while in many others, central banks have implicitly taken by general revenues (figure F1.1a). For example, 13 countries in Latin on this responsibility. Central banks seem to be best equipped to as- America and the Caribbean have both noncontributory and contribu- sume the statutory responsibility for macroprudential policy.24 tory systems. The introduction of noncontributory systems has helped The macroprudential committee should include selected policy increase coverage, reducing catastrophic health expenditures and stakeholders and independent experts, following the successful ex- curbing poverty among the elderly. In fact, aside from the former so- ample of monetary policy committees. It should use selected indica- cialist countries in Eastern Europe and Central Asia, only in those de- tors of systemic risk to detect excessive acceleration or concentration veloping countries with large noncontributory systems are more than of indebtedness in the financial sector or the real economy. To man- half the households with elderly members in the poorest 40 percent age any emerging excess, the committee would be directly equipped of the population covered (figure F1.1b).29 with macroprudential tools or with the ability to recommend actions However, combining contributory and noncontributory systems is to other regulators on an act-or-explain basis. The committee should particularly challenging. For workers and employers at the margin of be accountable to the legislative body. the formal sector, participating in a mandatory contributory system is not worthwhile. Meanwhile, combined with other factors (such as How can it be implemented? A possible role model for other countries, minimum wages), the additional labor cost levied by the payroll tax including developing ones, is the United Kingdom’s macropruden- for mandatory contributory systems discourages employers from hir- tial committee—the Financial Policy Committee, or FPC. The FPC is ing formally—or hiring at all—particularly for low-skill jobs. Thus if the chaired by the central bank governor and includes deputy governors benefits of contributing to social insurance are uncertain and the en- for financial stability, monetary policy, and prudential regulation; the forcement of mandated payments is weak, having these parallel sys- director of financial stability; the chief executive of the Financial Con- tems may undermine both the incentives for employers to hire for- duct Authority (business conduct regulator); four independent experts; mally and for employees to seek formal employment. Evidence from and a representative of the U.K. Treasury, who has no voting rights. Chile, Colombia, and Mexico shows that the interplay of contributory The FPC has the statutory responsibility to identify, monitor, and and noncontributory systems has led to declines in formal employ- take actions to remove or reduce systemic financial risk, with the view ment, and there is widespread evidence that smaller, informal firms to protecting and enhancing the resilience of the U.K. financial system. tend to be less productive and pay lower wages.30 For workers who It uses a set of systemic risk indicators to identify and monitor systemic move between formal and informal jobs or in and out of the labor risk.25 Since mid-2011, it has been equipped with direct powers to ad- force, replacement rates tend to be low, or in some cases they might just the capital requirements that banks must hold (the macropruden- not be eligible to receive benefits at all.31 Moreover, workers in coun- tial buffer) to mitigate systemic risk. It can also issue act-or-explain tries with rapidly aging populations make contributions toward in- recommendations to other policy makers in the financial sector, nota- creasingly uncertain benefits—all of which increases their perception 282 WORLD DEVELOPMENT REPORT 2014 F I G U R E F1.1 Noncontributory pension programs have expanded coverage in developing countries, especially for the poorest a. Regional coverage b. Country-level coverage 100 HUN SVK 15,000 HRV LTU 80 POL % of population 60+ with GNI per capita, PPP, 2005 pension coverage COL TUR ARG 60 10,000 VEN MUS BGR DOM SRB BRA BLR MNE CRI 40 THA BIH ECU PER ALB SLV UKR 5,000 MDV GTM JOR ARM BTN LKA PRY MAR GEO BOL MNG HNDNICPHL 20 PAK IND VNM MDA LAO TLS NGA KGZ AFG KHM TJK RWANPL GHA 0 MWI 0 0 20 40 60 80 100 l A nd CD c ia be nd ia r d a ric Af an tin Pa s As sia th Am ific S o an ha i c a La nd st A ra a rib a a OE Af nt pe Su rth st Coverage of poorest 40% of households h Ca ic ut a n a Ea Ce uro e er No le E ra with elderly members, % E d Sa id M b- Contributory Noncontributory Sources: WDR 2014 team based on data from World Bank Pensions (database), United Nations 2009 (panel a); and Evans, forthcoming (panel b). Note: For panel a, coverage rates are for total regional populations; years vary between 2001 and 2012. Organisation for Economic Co-operation and Develop- ment (OECD) countries in the figure are high-income countries that have been members of the OECD for at least 40 years. All other countries are grouped into geographic regions. For panel b, years vary between 2003 and 2010. Countries marked in blue have noncontributory programs. GNI = gross national income. PPP = purchasing power parity. of contributions as a pure tax on labor, especially in the presence of in a fiscally sustainable manner. This is particularly true for countries parallel noncontributory systems. Finally, the rapid aging process tak- where the old-age dependency ratio is growing rapidly (figure F1.2). ing place in many countries is threatening the fiscal sustainability of In practice, many developing countries would be able to provide only contributory systems, forcing governments to transfer additional a minimum level of benefits, possibly to only a targeted population. resources.32 Thus countries would need to consider their long-term fiscal capacity in relation to their future commitments to decide what level of cover- What could be done? One potential solution is to provide basic bene- age and benefits would be appropriate. Countries might also choose fits using general revenues, instead of labor taxes. For health care, user different ways to raise the necessary revenue. Some countries would fees could also be levied. The provision of basic benefits would make have to introduce new taxes or raise existing ones; in other cases, they social insurance similar to other basic public services and recognize its may be able to reform spending items like energy subsidies or use level of priority in public spending. Funding basic social insurance resource-based revenues where available. through general revenues would make the insurance more inclusive Noncontributory schemes provide crucial protection for the poor. by breaking the traditional eligibility condition linked to work status. However, if the benefits that can be sustainably offered by noncon- Moreover, it could limit the distortions in the labor market, to the ex- tributory systems are too basic, additional contributions to health and tent that general revenues are collected in a less distortionary way.33 pension systems may be necessary. If contributory and noncontribu- Advanced countries such as Australia, New Zealand, and the tory systems do coexist, policy makers should design both systems in United Kingdom rely mostly on universal basic pensions and provi- a way that avoids creating distortions in the labor market. In some sion of health care, whereas developing countries such as Mauritius contexts, that implies reforming contributory systems to make contri- and South Africa rely mostly on noncontributory systems for pen- butions voluntary or reducing the mandatory contribution rates. In all sions.34 Several low- and middle-income countries have also begun to cases, contributory systems should provide benefits that are clearly offer universal access to health insurance, starting with the poor. linked to contributions. Incentives to save can have a significant im- China, India, Thailand, Turkey, and Vietnam are a few examples. In all pact, as well, with examples including automatic enrollment, match- these cases, benefits do not depend on labor taxes and therefore are ing contributions, simplifying processes, and lowering information accessible to people in the informal sector. barriers through financial literacy. New Zealand’s KiwiSaver scheme is an interesting example of an automatic enrollment program (with an How would it work? While provision of universal benefits is desirable, “opt-out” option) that has increased retirement savings for about half not all countries are in a position to provide them at adequate levels the population.35 Focus on policy reform 283 F I G U R E F1. 2 Increasing coverage will require higher levels of spending in countries with aging populations a. Pension spending b. Old-age dependency ratio 200 Correlation = 0.64 200 Correlation = 0.57 Population 60+ covered Population 60+ covered 150 150 100 100 50 50 0 0 0 5 10 15 20 0 10 20 30 Pension spending, % of GDP Old-age (64+) dependency ratio East Asia and Pacific Latin America and the Caribbean South Asia Europe and Central Asia Middle East and North Africa Sub-Saharan Africa Sources: WDR 2014 team based on data from World Bank Pensions (database), United Nations 2009, and World Bank World Development Indicators (database). Note: Years vary between 2001 and 2012. Coverage of the 60+ population might exceed 100 percent in some countries because people younger than 60 are eligible for pensions or because coverage includes disability or death benefits. In panel b, the old-age dependency ratio is calculated as the ratio of people 64+ to the working-age population. GDP = gross domestic product. Reform 4. For the international community, embrace not seen since 2 million to 4 million years ago). This approaches the incremental approaches that can increase traction toward 450 ppm threshold level that corresponds to a likely increase in tem- global solutions perature larger than 2oC degree—the warming level the international community has committed to avoid.36 What is the problem? Globalization has contributed to rapid economic In the absence of a global deal, many unilateral climate action growth and to reduced poverty around the world. But it has also plans to limit greenhouse gases have been put forward in recent years made economic, social, and ecological systems more interdepen- by private actors, civil society groups, and municipal and subnational dent, generating gains from collaboration, while also increasing the governments (including in China and several U.S. states). Several coun- prevalence of cross-border risks, such as climate change, loss of bio- tries have introduced measures, including incentives that can limit car- diversity, overuse of natural resources, global financial crises, and bon emissions (table F1.1). These unilateral actions are welcome, but pandemics. Containing global risks requires timely, proactive, and more ambitious and coordinated national and international efforts are concerted action because no country acting alone can manage them needed to make a material difference and to ensure that the overall effectively and achieve the scale required to address them compre- effort is greater than the sum of its individual parts. Yet some useful hensively. Unfortunately, in the absence of an effective global risk international actions, including cooperation to develop and share governance mechanism led by an international body that has appro- technologies and improvement in existing financial instruments, have priate accountability and enforcement powers over sovereign na- been postponed in the expectation that they will be part of a soon-to- tions, the international architecture necessary to address global risks be-signed global agreement, reflecting differing views on who is re- has not kept pace with the complexities arising from increased global sponsible and incentives to free-ride on potential actions by others connectivity. and to wait for new, equitable, financing instruments. The limited progress to reach a global deal in some areas, in turn, has cast into doubt the ability to foster collective action among a large What is the solution? For certain global risks such as climate change or number of nations with diverging interests, capacity constraints, and biodiversity loss, preserving collective action with full participation is incentives to free ride. Global negotiations to secure agreements with the ultimate goal. In the interim, however, the international commu- full participation have stalled—most spectacularly for climate change, nity is increasingly embracing incremental approaches that can as atmospheric concentrations of greenhouse gases have continued increase traction toward global solutions. When incentives are to rise, with potentially catastrophic and irreversible consequences. misaligned, major sovereigns are not fully engaged, and the conse- The concentration of the main greenhouse gas, carbon dioxide (CO2), quences of inaction are disastrous, progress can still be made outside rose from its preindustrial level of 278 parts per million (ppm) to more a full-participation multilateral treaty. Incremental deals and actions than 390 ppm as of May 2013 (hitting 399.91 ppm in Hawaii—a level by an initially small group of participants can serve as building blocks 284 WORLD DEVELOPMENT REPORT 2014 TA B L E F1.1 National policy measures limiting carbon emissions CO2 emissions evolution and shares Measures included in KPMG Green Tax Index Share Change Material CO2 emissions of CO2 in CO2 Carbon Green resource Pollution in metric tons emissions emissions and Renewable vehicles efficiency control and per capita in in 2009 since 1990 Energy climate Green energy and Water and waste ecosystem Country 2009 (%) (%) efficiency change innovations and fuels buildings efficiency management protection Australia 18.38 1.33 39.28 X X X X X X X United States 17.28 17.62 8.61 X X X X X X X X Canada 15.24 1.71 14.19 X X X X X Russian Federation 11.09 5.24 –26.23* X X Korea, Rep. 10.36 1.69 106.27 X X X X X X Netherlands 10.26 0.56 3.36 X X X X X X South Africa 10.12 1.66 49.62 X X X X X X X X Finland 10.03 0.18 3.52 X X X X X X Belgium 9.60 0.34 –4.50 X X X X X X X Ireland 9.34 0.14 32.59 X X X X X X Germany*** 8.97 2.44 –21.08** X X X X X Japan 8.63 3.66 0.59 X X X X X X United Kingdom 7.68 1.58 –16.77 X X X X X X X X Singapore 6.39 0.11 –32.05 X X X X X X X Spain 6.28 0.96 31.69 X X X X X X China 5.77 25.56 212.39 X X X X X X X France 5.61 1.21 –8.94 X X X X X X Argentina*** 4.36 0.58 55.15 X X Mexico 3.98 1.48 41.93 X X X X X Brazil 1.90 1.22 75.76 X X X X India 1.64 6.58 186.63 X X X X X Sources: WDR 2014 team based on data from KPMG Green Tax Index (database) (for national or subnational policy measures limiting carbon emissions), which analyzes the 21 largest economies of the world, and the World Bank World Development Indicators (database) (for CO2 emissions data, as reported in the Carbon Dioxide Information Analysis Center, Environmental Science Division, Oak Ridge National Laboratory, Tennessee). Note: X indicates the existence of a tax-related or non-tax-related measure in place at national or subnational levels. Energy efficiency refers to measures encouraging the purchase of energy efficient equipment (excluding measures specific to green vehicles or buildings). Carbon and climate change refers to penalties on high emissions (such as carbon taxes, emission trading systems or cap and trade mechanisms, and carbon sequestration incentives and penalties). Green innovation includes incentives for research and development for green technologies. Renewable energy and fuels refers to use of tax codes to encourage the production or use of renewable or alterna- tive fuels and/or penalizing the use of fossil fuels. Green vehicle and buildings refers to tax incentives to reduce the energy consumption of buildings, increase their water efficiency and sustainability of building materials, and the purchase, lease and use of greener (fuel efficient, hybrid, electric) vehicles (excluding tax penalties and incentives related to fuels, which are included under renewable energy and fuels). Water efficiency includes use of taxes to encourage corporations to conserve and recycle water supplies. Material resource efficiency and waste management include use of taxes to promote conservation of material resources, reduction of waste, and recycling of waste materials. Pollution control and ecosystem protection include incentives to purchase equipment to reduce the pollution generated by the company or to encourage businesses to rehabilitate contaminated lands. * 1991 value; ** 1992 value. *** Measures recorded by KPMG may not include some of the recent initiatives. For example, in Germany, recent information suggests addi- tional measures in the areas of waste management, pollution control, and biodiversity protection (see German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU), “The Energy Concept and its Accelerated Implementation,” http://www.bmu.de/en/topics/climate-energy/transformation-of-the-energy- system/resolutions-and-measures/). In Argentina, recent information also suggests programs that limit carbon emissions, including energy efficiency, green buildings, waste management, pollution control, and ecosystem protection. CO2 = carbon dioxide. for global deals, by demonstrating benefits from action. The goal is to governments (including the U.S. state of California) and countries (Aus- align incentives around a common goal in a group of like-minded tralia, China, Japan, New Zealand) are using the lessons from Europe’s participants that can examine complex issues and take concrete ac- Emissions Trading System introduced to meet emission commitments tion. Over time, more public and private actors would be attracted cost-effectively. Growing coalitions of more than 30 developed and to join the incremental approaches to build convergence toward a developing countries include the Partnership of Market Readiness and global deal. the Climate, working on solutions to carbon pricing,38 and the Clean The incremental approach has precedents. The Montreal Protocol Air Coalition of the United Nations Environment Programme, catalyz- to protect the ozone layer was signed by 24 countries in 1987, but re- ing rapid reductions in short-lived climate pollutants.39 ceived universal ratification during the 1990s through the combined efforts of governments, international organizations, nongovernmen- How can it be implemented? Countries, international organizations, tal organizations, and scientists, who presented and disseminated and private sector entities of the international community can form convincing evidence on the need for urgent action.37 The 1968 Nuclear “coalitions of the willing” (or even better, “coalitions of the working”) Non-Proliferation Treaty was helped by the earlier Limited Test Ban that could coordinate, advocate, and take action on some compo- Treaty that expanded from 3 to 119 signatories from 1963 to 1992, set- nents of elusive global risks, such as climate change and the loss of ting a precedent for future arms negotiations. Several subnational biodiversity.40 The coalitions should engage the scientific community, Focus on policy reform 285 civil society, and media, and rely on information and peer pressure to of the causes, dynamics, and consequences of key systemic risks that demonstrate positive action and leadership, while inducing partici- pose threats to development. It could also analyze the interactions pants to comply and nonparticipants to join in a “race to the top.” In- and prioritize across various risks and systematically bring its analysis ternational organizations can continue to contribute by offering ways to the attention of policy makers and the international community. In to develop approaches to tackle the problem collectively, by provid- so doing, it could provide valuable inputs to the coalition of the will- ing platforms for policy discussion and by monitoring, reporting, and ing on the specific issues that require urgent attention and offer cred- aggregating the actions to ensure that incremental steps are heading ibility and legitimacy to the coalition’s efforts. in the right direction. For legitimacy and fairness, these coalitions must include actors Notes that contribute most to the problem, as well as those most affected by 1. Graham and Wiener 1995, chapter 1. it. They should start with specific, concrete actions that can set the 2. Graham and Wiener 1995, chapter 11. 3. Graham and Wiener 1995, chapter 11, 257–60. momentum for subsequent steps. The coalitions should create incen- 4. WEF 2007. tives for others to join over time to bring the coalition to a global scale. 5. OECD 2009. Steps could include, as appropriate, promoting technological change 6. However, in other cases when such a governmental body works with indirect that lowers participation costs (such as supporting cheaper ways to policy tools, such as the monetary policy, an institutional design along these cut emissions by providing subsidies or funding for green technolo- lines could be preferable. 7. Kaminsky, Reinhart, and Végh 2005. gies, or backing technology transfers to developing countries). 8. Frankel, Végh, and Vuletin 2013. Granted, there are risks with this approach—not least because it is 9. Aghion and Marinescu 2008; Woo 2009. a “second best” solution that in effect allows free-riding by those out- 10. Tornell and Lane 1999. side the coalition. Incremental deals may fail to scale up efforts suffi- 11. Alesina, Campante, and Tabellini 2008. ciently—or worse, reduce the urgency of global cooperation. More- 12. Barnes, Davidsson, and Rawdanowicz 2012. 13. Kumar and others 2009. over, for certain risks, global collective action is still the only viable 14. Schaechter and others 2012. approach, given the rapid spillover risks in a tightly interconnected 15. Wyplosz 2013. world. For example, once a pandemic is under way, no individual 16. Debrun, Hauner, and Kumar 2009. country or region can unilaterally protect itself without global coop- 17. Frankel 2011 shows that the authorities overestimate the persistence of booms eration that enables information to be shared and assists countries and underestimate that of recessions. 18. Wren-Lewis 2010. lacking the capacity to detect and contain the contagion.41 Smallpox 19. Bos and Teulings 2010. was declared eradicated in 1979 because it was eliminated in every 20. Debrun 2011. country through global cooperation; if the disease had persisted in 21. Debrun, Hauner, and Kumar 2009. only one nation, all others would remain vulnerable.42 Resolving 22. Calmfors and Wren-Lewis 2011; Page 2010. 23. Lane 2010. global financial crises in a highly connected world also requires global 24. BIS 2011; Bank of England 2013; IMF 2011; Nier and others 2011. cooperation, with well-coordinated policy responses and information 25. Bank of England 2013. sharing; uncoordinated actions are unable to prevent contagion and 26. Financial Services Act 2012. block movement of activities to less well-regulated and more- 27. Bank of England 2013. protected locations that retain systemic risk. 28. The term social insurance typically includes pensions, health, and unemployment insurance. Because most developing countries with social insurance offer only These limitations notwithstanding, the alternative—to wait until pensions and health benefits, the discussion in this section focuses on these two a universally acceptable deal is reached—is not viable for global risks areas. such as climate change, biodiversity loss, or overuse of natural re- 29. Evans, forthcoming; Levy and Schady 2013. sources, if the irreversible consequences of inaction are to be avoided. 30. Levy and Schady 2013; Pagés-Serra 2010; ILO 2009; La Porta and Shleifer 2008. That is especially the case for those who have done the least to cause 31. Ribe, Robalino, and Walker 2012. 32. This is particularly true for “pay-as-you-go” systems, in which the current labor the problem but will suffer some of the worst consequences. The in- force finances the benefits provided to current beneficiaries. ternational community therefore has the moral responsibility to take 33. Frölich and others, forthcoming. and support action on behalf of the vulnerable populations of today 34. Holzmann, Robalino, and Takayama 2009. and the future. It should buttress the incremental approach with stra- 35. Hinz and others 2013. tegic thinking about which specific issues to tackle urgently first, 36. World Bank 2012. 37. UNEP 2007. while anchoring its actions to existing global frameworks to demon- 38. See ongoing efforts to establish a carbon price at http://www.worldbank.org/ strate that incremental and global deals are connected.43 en/news/opinion/2013/05/16/tackling-climate-change-robust-carbon-price. One way to achieve this is to anchor the efforts of the coalition to 39. For further details, see http://www.unep.org/ccac/. the goals of the current global initiatives under the auspices of United 40. See discussions in Falkner, Stephan, and Vogler 2010, Hale 2011, and Goldin Nations. Alternatively, an international risk board, similar to the na- 2013; and increased calls for such coalitions in Europe following the limited progress made in global negotiations in Copenhagen (2009) and Cancun (2010) tional risk board proposed in Reform 1, could be established, in the at http://www.euractiv.com/climate-environment/europe-looks-coalition- form of an international panel on global systemic risks.44 The panel willing-d-news-508909. could invite the scientific and expert community around the world to 41. Goldin 2013; Jonas 2013. pool all available knowledge to identify, assess, and manage the ma- 42. Barrett 2006. 43. Falkner, Stephan, and Vogler 2010. jor global risks that cross national and generational boundaries in the 44. Bodies of this sort already exist for specific risks (such as the Intergovernmental near and longer term. Through its long-term orientation, interdisci- Panel on Climate Change, or the Intergovernmental Platform on Biodiversity and plinary nature, and the participation of global experts, the board Ecosystem Services) but none that consider multiple global risks in a systematic could focus on providing credible, reliable, and impartial assessments way. See a related proposal by the German Institute for Development 2009. 286 WORLD DEVELOPMENT REPORT 2014 References IMF (International Monetary Fund). 2011. “Macroprudential Policy: An Organizing Aghion, Philippe, and Ioana Marinescu. 2008. “Cyclical Budgetary Policy and Eco- Framework.” IMF, Washington, DC. nomic Growth: What Do We Learn from OECD Panel Data?” In NBER Macroeco- Jonas, Olga. 2013. “Pandemic Risk.” Background paper for the World Development nomics Annual 2007, vol. 22, edited by Daron Acemoglu, Kenneth Rogoff, and Report 2014. Michael Woodford, 251–78. Chicago: University of Chicago Press. Kaminsky, Graciela L., Carmen M. Reinhart, and Carlos A. Végh. 2005. “When It Rains, Alesina, Alberto, Filipe R. Campante, and Guido Tabellini. 2008. “Why Is Fiscal Policy It Pours: Procyclical Capital Flows and Macroeconomic Policies.” In NBER Macro- Often Procyclical?” Journal of the European Economic Association 6 (5): 1006–36. economics Annual 2004, vol. 19, edited by Mark Gertler and Kenneth Rogoff, Bank of England. 2013. “The Financial Policy Committee’s Powers to Supplement 11–53. Cambridge, MA: MIT Press. Capital Requirements: A Draft Policy Statement.” Bank of England, London. KPMG. Green Tax Index (database). KPMG, Amstelveen. http://www.kpmg.com/ Barnes, Sebastian, David Davidsson, and Lukasz Rawdanowicz. 2012. “Europe’s New global/en/issuesandinsights/articlespublications/green-tax/pages/default.aspx. Fiscal Rules.” Economics Department Working Paper 972, Organisation for Eco- Kumar, Manmohan, Emanuele Baldacci, Andrea Schaechter, Carlos Caceres, Dae- nomic Co-operation and Development, Paris. haeng Kim, Xavier Debrun, Julio Escolano, Jiri Jonas, Philippe Karam, Irina Yaka- Barrett, Scott. 2006. “The Smallpox Eradication Game.” Public Choice 130 (1–2): dina, and Robert Zymek. 2009. “Fiscal Rules—Anchoring Expectations for Sus- 179–207. tainable Public Finances.” International Monetary Fund, Washington, DC. BIS (Bank for International Settlements). 2011. “Central Bank Governance and Finan- Lane, Philip. 2010. “Some Lessons for Fiscal Policy from the Financial Crisis.” IIIS Discus- cial Stability: A Report by Study Group.” BIS, Basel. sion Paper 334, Trinity College, Dublin, Ireland. Bos, Frits, and Coen Teulings. 2010. “CPB and Dutch Fiscal Policy in View of the Finan- La Porta, Rafael, and Andrei Shleifer. 2008. “The Unofficial Economy and Economic Development.” Brookings Papers on Economic Activity 39 (2): 275–363. cial Crisis and Ageing.” Document 218, Bureau for Economic Policy Analysis (CPB), Levy, Santiago, and Norbert Schady. 2013. “Latin America’s Social Policy Challenge: the Netherlands. Education, Social Insurance, Redistribution.” Journal of Economic Perspectives 27 Calmfors, Lars, and Simon Wren-Lewis. 2011. “What Should Fiscal Councils Do?” (2): 193–218. Economic Policy 26 (68): 649–95. Nier, Erlend W., Jacek Osinski, Luis I. Jacome, and Pamela Madrid. 2011. “Towards Debrun, Xavier. 2011. “Democratic Accountability, Deficit Bias, and Independent Effective Macroprudential Policy Frameworks: An Assessment of Stylized Institu- Fiscal Agencies.” Working Paper WP/11/173, International Montetary Fund, tional Models.” Working Paper WP/11/250, International Monetary Fund, Wash- Washington, DC. ington, DC. Debrun, Xavier, David Hauner, and Manmohan Kumar. 2009. “Independent Fiscal OECD (Organisation for Economic Co-operation and Development). 2009. “Innova- Agencies.” Journal of Economic Surveys 23 (1): 44–81. tion in Country Risk Mangement.” OECD Studies in Risk Management, OECD, Evans, Brooks. Forthcoming. “An Examination of Pension Coverage of Elderly in the Paris. Developing World.” Social Protection and Labor Policy Note, World Bank, Wash- Page, Kevin. 2010. “Lessons from Canada.” Paper presented to the Conference on ington, DC. Independent Fiscal Institutions, Fiscal Council Republic of Hungary, Budapest. Falkner, Robert, Hannes Stephan, and John Vogler. 2010. “International Climate Policy Pagés-Serra, Carmen, ed. 2010. The Age of Productivity: Transforming Economies from after Copenhagen: Towards a ‘Building Blocks’ Approach.” Global Policy 1 (3): the Bottom Up. Washington, DC: Inter-American Development Bank. 252–62. Ribe, Helena, David A. Robalino, and Ian Walker. 2012. From Right to Reality: Incentives, Financial Services Act. 2012. Chapter 21. London: TSO (The Stationery Office). Labor Markets, and the Challenge of Universal Social Protection in Latin America Frankel, Jeffrey A. 2011. “Over-Optimism in Forecasts by Official Budget Agencies and and the Caribbean. Washington, DC: World Bank. Its Implications.” Oxford Review of Economic Policy 27 (4): 536–62. Schaechter, Andrea, Tidiane Kinda, Nina Budina, and Anke Weber. 2012. “Fiscal Rules Frankel, Jeffrey A., Carlos A. Végh, and Guillermo Vuletin. 2013. “On Graduation from in Response to the Crisis toward the ‘Next Generation’ Rules: A New Dataset.” Fiscal Procyclicality.” Journal of Development Economics 100 (1): 32–47. Working Paper WP/12/187, International Monetary Fund, Washington, DC. Frölich, Markus, David Kaplan, Carmen Pagés, Jamele Rigolini, and David Robalino. Tornell, Aaron, and Philip R. Lane. 1999. “The Voracity Effect.” American Economic Forthcoming. Social Insurance and Labor Markets: How to Protect Workers While Review 89 (1): 22–46. Creating Good Jobs. Washington, DC: World Bank, Inter-American Development UNEP (United Nations Environment Programme). 2007. “A Success in the Making: The Bank, and Institute for the Study of Labor. Montreal Protocol on Substances that Deplete the Ozone Layer.” UNEP, Nairobi. German Institute for Development. 2009. “Globalization at the Crossroad: An “Inter- United Nations. 2009. “World Population Prospects: The 2008 Revision.” Department national Panel on Systemic Risks in the Global Economy” Is Needed.” Briefing Pa- of Economic and Social Affairs Population Division, United Nations, New York. per 6/2009, Bonn. http://www.die-gdi.de/CMS-Homepage/openwebcms3.nsf/ WEF (World Economic Forum). 2007. “Global Risks 2007: A Global Risk Network (ynDK_ contentByKey)/ANES-7T5J6U/$FILE/BP%206.2009.pdf. Report.” WEF, Geneva. Goldin, Ian. 2013. Divided Nations: Why Global Governance Is Failing and What We Can Woo, Jaejoon. 2009. “Why Do More Polarized Countries Run More Procyclical Fiscal Do about It. Oxford: Oxford University Press. Policy?” Review of Economics and Statistics 91 (4): 850–70. Graham, John D., and Jonathan B. Wiener. 1995. Risk Versus Risk: Tradeoffs in Protecting World Bank. 2012. “Turn Down the Heat: Why a 4ºC Warmer World Must Be Avoided.” Health and the Environment. Cambridge, MA: Harvard University Press. World Bank, Washington, DC. Hale, Thomas. 2011. “A Climate Coalition of the Willing.” Washington Quarterly 34 (1): ———. Pensions (database). World Bank, Washington, DC. http://www.worldbank 89–101. .org/pensions. Hinz, Richard, Robert Holzmann, David Tuesta, and Noriyuki Takayama. 2013. Match- ———. World Development Indicators (database). World Bank, Washington, DC. ing Contributions for Pensions: A Review of International Experience. Washington, http://data.worldbank.org/data-catalog/world-development-indicators. DC: World Bank. Wren-Lewis, Simon. 2010. “Macroeconomic Policy in Light of the Credit Crunch: The Holzmann, Robert, David A. Robalino, and Noriyuki Takayama. 2009. Closing the Cov- Return of Countercyclical Fiscal Policy?” Oxford Review of Economic Policy 26 (1): erage Gap: The Role of Social Pensions and Other Retirement Income Transfers. 71–86. Washington, DC: World Bank. Wyplosz, Charles. 2013. “Fiscal Rules: Theoretical Issues and Historical Experiences.” ILO (International Labour Organization). 2009. The Informal Economy in Africa: In Fiscal Policy after the Financial Crisis, edited by Alberto Alesina and Francesco Promoting Transition to Formality: Challenges and Strategies. Geneva: ILO. Giavazzi. Chicago: University of Chicago Press. Appendixes Abbreviations and data notes Abbreviations AIDS Acquired Immunodeficiency Syndrome BIS Bank for International Settlements CAB cyclically adjusted balance CAT DDO Catastrophe Deferred Drawdown Option CCRIF Caribbean Catastrophe Risk Insurance Facility CCT conditional cash transfer CGAP Consultative Group to Assist the Poor CRIF Catastrophe Risk Insurance Facility DRFI Disaster Risk Financing and Insurance DRM disaster risk management EU European Union FAO Food and Agriculture Organization of the United Nations FCS fragile and conflict-affected state FDI foreign direct investment FONDEN Fund for Natural Disasters FX foreign exchange/currency G2P government-to-person GDP gross domestic product GEF Global Environmental Facility GFDRR Global Facility for Disaster Reduction and Recovery GFSN Global Financial Safety Net G-SIFI global systemically important financial institution GNI gross national income HFA Hyogo Framework for Action HIV Human Immunodeficiency Virus IBRD International Bank for Reconstruction and Development IFI international financial institution IMF International Monetary Fund IPP independent power producer ISO International Organization for Standardization MDG Millennium Development Goal MTEF medium-term expenditure framework MTFF medium-term fiscal plan NGO nongovernmental organization NRA national risk assessment ODC ozone depleting chemical OECD Organisation for Economic Co-operation and Development OIE Office International des Epizooties, or World Organisation for Animal Health ppm parts per million 289 290 WORLD DEVELOPMENT REPORT 2014 PPP public-private partnership SARS severe acute respiratory syndrome SEEC-CRIF Southeast Europe and the Caucasus Catastrophe Risk Insurance Facility SIFI systemically important financial institution S&P Standard & Poor’s SWF sovereign wealth fund UN United Nations UNEP United Nations Environment Programme VIX Volatility Index WDI World Development Indicators WDR World Development Report WEF World Economic Forum WHO World Health Organization Data notes The use of the term countries to refer to economies countries are grouped into geographic regions. implies no judgment by the World Bank about Countries with less than 0.5 million population as the legal or other status of territory. The term of 2010 are not included in the sample for analytical developing countries includes low- and middle- purposes. income economies and thus may include economies in transition from central planning, as a matter of Income groupings are based on World Bank income convenience. Dollar figures are current U.S. dollars, classifications as of July 1, 2012, based on 2011 gross unless otherwise specified. Billion means 1,000 national income per capita—except for the tables in million; trillion means 1,000 billion. the selected indicators, which were prepared based on the income classifications as of July 1, 2013. See For regional comparisons, figures in this Report use the selected indicators section for more details. the following country groupings: Organisation for Economic Co-operation and Development (OECD), Maps East Asia and Pacific (EAP), Europe and Central The maps numbered IBRD 40097, 40098, 40099, Asia (ECA), Latin America and Caribbean (LAC), and 40100 were produced by the Map Design Unit Middle East and North Africa (MENA), South Asia of the World Bank. (SAR), and Sub-Saharan Africa (SSA). The OECD group refers to high-income countries that have The boundaries, colors, nominations, and any other been members of the OECD for at least 40 years information shown on those maps do not imply, on (Australia, Austria, Belgium, Canada, Denmark, the part of the World Bank Group, any judgment on Finland, France, Germany, Greece, Ireland, Italy, the legal status of any territory, or any endorsement Japan, Luxembourg, Netherlands, New Zealand, or acceptance of such boundaries. Taiwan, China, Norway, Portugal, Spain, Sweden, Switzerland, receives the same ranking as China. United Kingdom, and United States). All other Background papers . Aizenman, Joshua, and Inci Ötker-Robe. 2013. “Manag- Fidas, Penelope D., and Najy Benhassine. 2013. “Trans- ing Risk for Development: International Risk Sharing parency and Access to Information in Business Tools.” Regulations.” Anderson, Phillip R. D. 2013. “Risk Market Transfer.” Foa, Roberto. 2013. “Household Risk Preparation Indices: Ashwill, Maximillian, and Rasmus Heltberg. 2013. “Is Construction and Diagnostics.” There a Community-Level Adaptation Deficit?” Fuentes, Rodrigo. 2013. “Sovereign Wealth Funds: The Ashwill, Maximillian, and Andrew Norton. 2013. “Rights Case of Chile.” and Social Action for Risk Management: Reflections on Genicot, Garance, and Ethan Ligon. 2013. “Risk and the Global, National and Local Roles and Responses.” Extent of Insurance.” Auriol, Emmanuelle. 2013. “Taxation Base and Barriers to Gereffi, Gary, and Xubei Luo. 2013. “Risks and Opportuni- Entrepreneurship in Developing Countries.” ties of Participation in Global Value Chains.” Baliki, Ghassan. 2013. “Crime and Victimization.” Ghani, Ejaz. 2013a. “Are Manufacturing Sectors Moving Beck, Thorsten. 2013. “The Supply-Demand Mismatch in Out of Large Cities?” Provision of Financial Services.” ———. 2013b. “Who Creates Jobs in India?” Beck, Thorsten, and Olivier De Jonghe. 2013. “Lending ———. 2013c. “Why Do Enterprises Get Attracted to Concentration, and its Implications for Bank Perfor- Some Cities? Not to Others?” mance and Systemic Risk: Global Evidence.” Gooptu, Sudarshan, and Auguste T. Kouame. 2013. “The Bloom, Nicholas. 2013. “Has Economic Policy Uncertainty ‘4-3-2 Framework’: A Framework for Dealing with the Slowed Down the World Economy?” Debt-Related Risks of Highly Indebted Small States.” Brown, Julia K., Ahmed Mushfiq Mobarak, and Tetyana V. Grinbaum, Mikael, and Susan T. Jackson. 2013. “Man- Zelenska. 2013. “Barriers to Adoption of Products and aging Risk for Development: Expert Workshop Technologies That Aid Risk Management in Develop- on Local Violence in Developing Countries.” ing Countries.” Gutierrez, Federico H. 2013. “Labor Contracts and Risk Brown, Martin. 2013. “The Transmission of Banking Sharing.” Crises to Households: Lessons from the ECA Region Hallegatte, Stéphane. 2013. “An Exploration of the 2008–2012.” Link between Development, Economic Growth, Buncic, Daniel, and Martin Melecky. 2013. “Equilibrium and Natural Risk.” Credit: The Reference Point for Macroprudential Han, Rui, and Martin Melecky. 2013. “Financial Inclusion Supervisors.” for Financial Stability: Access to Bank Deposits and the Cáceres-Delpiano, Julio. 2013. “Literature Review: Family Deposits Growth in the Global Financial Crisis.” Formation and Fertility as Risk Coping Mechanisms.” Heltberg, Rasmus, Ana María Oviedo, and Faiyaz Taluk- Calvo, Sara Guerschanik. 2013. “Financial Crises, dar. 2013. “What Are the Sources of Risk and How Do Social Impact, and Risk Management: Lessons and People Cope? Insights from Household Surveys in 15 Challenges.” Countries.” de la Fuente, Alejandro, Eduardo Ortiz-Juárez, and Carlos Inglehart, Ronald. 2013. “Results from the Latest Wave of Rodriguez Castelan. 2013. “Living on the Edge: Vulner- the World Values Survey.” ability to Poverty and Public Transfers in Mexico.” Jonas, Olga B. 2013. “Pandemic Risk.” Dutz, Mark A. 2013. “Resource Reallocation and Innova- Kelman, Ilan. 2012. “Disaster Mitigation Is Cost Effective.” tion: Converting Enterprise Risks into Opportunities.” Khokhar, Tariq. 2013. “Leveraging New Technology Eden, Maya. 2013. “Managing Contingent Liabilities Using for Data-Driven Risk Mitigation and Management: Market Risk Transfer Instruments: A Review of the Selected Examples and Summaries.” Literature.” 291 292 WORLD DEVELOPMENT REPORT 2014 Kraay, Aart, and Luis Servén. 2013. “Fiscal Policy as a Tool Premand, Patrick. 2013. “From Risk Coping to Risk Man- for Stabilization in Developing Countries.” agement: Productive Safety Nets in Africa.” Lakhani, Sadaf. 2013. “Forced Displacement: Moving from Salazar, Natalia. 2013. “Fiscal Risk Management for Devel- Managing Risk to Facilitating Opportunity.” opment: The Case of Colombia.” Lambert, Sylvie, and Philippe de Vreyer. 2013. “Household Teal, Francis J. 2013. “The Enterprise Sector: Providing Risk Management in Senegal.” Employment and Sharing Risk.” Lasagabaster, Esperanza. 2013. “Policy Priorities for Inno- van Aalst, Maarten, Tom Mitchell, Jan Kellett, and Florence vative Entrepreneurship.” Pichon. 2013. “Incentives in Disaster Risk Management Maimbo, Samuel, and Martin Melecky. 2013. “Financial and Humanitarian Response.” Policy Formulation: Addressing the Tradeoff between Vuletin, Guillermo. 2013. “Effects and Role of Macropru- Development and Stability.” dential Policy.” Maloney, William F. 2013. “Measuring Labor Market Risk.” Warner, Koko. 2013. “Climate Change and Global Warm- Moscoso Boedo, Hernan J. 2013. “Enterprise Sector Risk.” ing: The Role of the International Community.” . Ötker-Robe, Inci, and Anca Podpiera. 2013. “Social Impact Wethli, Kyla. 2013. “Benefit-Cost Analysis for Risk Man- of Financial Crises: Evidence from the Recent Global agement: Summary of Selected Examples.” Financial Crisis.” Zhu, Nong, and Xubei Luo, 2013. “What Drives Productiv- Oviedo, Ana María, and Harry Moroz. 2013. “The Impacts ity Volatility of Chinese Industrial Firms?” of Risk.” Petesch, Patti. 2013. “How Communities Manage Risks of Crime and Violence.” Selected indicators General notes economies with sparse data or with populations of less than 3 million. The term country, used inter- Tables 1 to 10 include 75 indicators relevant to the changeably with economy, does not imply political management of risk in the context of development. independence but refers to any territory for which These indicators summarize the level and dimen- authorities report separate social or economic sta- sions of risk facing each country, together with the tistics. Data are shown for economies as they were capacity to manage risk along several different social, constituted in 2012. Unless otherwise noted, data economic, financial, and environmental dimensions. for China do not include data for Hong Kong SAR, Definitions of each indicator are provided in the China; Macao SAR, China; or Taiwan, China. Data technical notes that follow the tables. for Indonesia include Timor-Leste through 1999. Data for Serbia include Montenegro through 2005 Sources and exclude Kosovo from 1999 onward. Data for The indicators in the tables come from a vari- Sudan include South Sudan unless otherwise noted. ety of authoritative sources, including the World Bank, other international organizations such as the Classification of economies and summary United Nations and International Monetary Fund, measures government agencies, member country statistical publications, research institutes, and peer-reviewed For operational and analytical purposes, the World academic papers. The source for each indicator is Bank’s main criterion for classifying economies is identified in the technical notes. gross national income (GNI) per capita. Based on its Many of the indicators in this section have been GNI per capita, every economy is classified as low included in the World Bank’s Open Data catalog income, middle income (subdivided into lower- and are available for download. The World Bank’s middle and upper-middle), or high income. Income Open Data terms of use allow users to use these classifications are set each year on July 1. These offi- data freely, subject to a limited set of conditions. We cial analytical classifications are fixed during the encourage readers to access the data at http://data World Bank’s fiscal year (ending on June 30); thus catalog.worldbank.org and to review the terms of countries remain in the categories in which they are use at http://data.worldbank.org/summary-terms- classified irrespective of any revisions to their per of-use. Several of the indicators available in the data capita income data. Low-income economies are catalog were obtained from research papers with the those with a GNI per capita of $1,035 or less in permission of the authors. These indicators in par- 2012. Middle-income economies are those with a ticular were developed for specific research studies GNI per capita of more than $1,035 but less than and may not be suitable for other purposes. Neither $12,616. Lower-middle-income and upper-middle- the authors nor any institutions with which they are income economies are separated at a GNI per capita affiliated make any warranties with respect to the data of $4,086. High-income economies are those with a and shall not be liable in connection with the data’s GNI per capita of $12,616 or more. When changes use under any circumstances. in classification are made, aggregates based on the new income classifications are recalculated for all past periods to ensure that a consistent time series is Country coverage maintained. Tables 1 and 3 to 8 include data for 133 countries. Summary measures are either totals (indicated Table 2 includes key development indicators for 81 by a t if the aggregates include estimates for miss- 293 294 WORLD DEVELOPMENT REPORT 2014 ing data and nonreporting countries, or by an s 0 or 0.0 means zero or small enough that the for simple sums of the data available), weighted number would round to zero at the displayed averages (w), unweighted averages (u), or me- number of decimal places. dian values (m) calculated for groups of econo- – in dates in column headings, as in 2003–12, mies. Data for economies not appearing in refers to a time period that spans years. Unless the tables have been included in the summary mea- otherwise noted in the column heading, these sures, where data are available; otherwise it is as- indicators are sums of annual values for the sumed they follow the trend of reporting economies. time period shown. Where missing data accounts for a third or more of the overall estimate, however, the group measure is $ means current U.S. dollars unless otherwise reported as not available. noted. Data in italics are for a year or period other than that Symbols specified in the column heading. .. means that data are not available or that aggre- gates cannot be calculated because of missing data in the years shown. Selected indicators 295 Classification of economies by region and income, FY2014 Low income Guatemala Kazakhstan Slovak Republic Afghanistan Honduras Lebanon Spain Bangladesh India Libya Sweden Benin Indonesia Malaysia Switzerland Burkina Faso Lao PDR Mexico United Arab Emirates Burundi Mauritania Panama United Kingdom Cambodia Moldova Peru United States Central African Republic Morocco Romania Uruguay Chad Nicaragua Serbia Congo, Dem. Rep. Nigeria South Africa Selected island states Eritrea Pakistan Thailand (Tables 2 and 8 only) Ethiopia Papua New Guinea Tunisia Antigua and Barbuda Guinea Paraguay Turkey Bahamas Haiti Philippines Turkmenistan Barbados Kenya Senegal Venezuela, RB Bermuda Kyrgyz Republic Sri Lanka Comoros Liberia Sudan High income Dominica Madagascar Syrian Arab Republic Australia Dominican Rep Malawi Ukraine Austria Fiji Mali Uzbekistan Belgium Grenada Mozambique Vietnam Canada Haiti Myanmar West Bank and Gaza Chile Jamaica Nepal Yemen, Rep. Hong Kong SAR, China Maldives Niger Zambia Croatia Marshall Is Rwanda Czech Republic Micronesia Fed States Sierra Leone Upper middle income Denmark Philippines Somalia Albania Finland Samoa South Sudan Algeria France Seychelles Tajikistan Angola Germany Solomon Is Tanzania Argentina Greece St Kitts and Nevis Togo Azerbaijan Ireland St Lucia Uganda Belarus Israel St Vincent and The Zimbabwe Bosnia and Herzegovina Italy Grenadines Brazil Japan Trinidad and Tobago Lower middle income Bulgaria Korea, Rep. Armenia China Lithuania Bolivia Colombia Netherlands Cameroon Costa Rica New Zealand Congo, Rep. Dominican Republic Norway Côte d’Ivoire Ecuador Poland Egypt, Arab Rep. Hungary Portugal El Salvador Iran, Islamic Rep. Russian Federation Georgia Iraq Saudi Arabia Ghana Jordan Singapore 296 WORLD DEVELOPMENT REPORT 2014 TABLE 1 Key indicators of development Population age Gross national Gross national Gross Life expectancy Adult Population composition  incomea income, PPPb domestic at birth literacy Average % $ $ product rate annual Density ages $ per $ per per capita  Years, Years, % ages 15 Millions growth, %  per sq. km  0–14  billions  capita  billions  capita  growth, %  male  female and older  2012 2000–12 2012  2012  2012  2012  2012  2012  2012  2011  2011  2005–11c  Afghanistan 30 3.1 46 47 16.6 570 40.7d 1,400d 4.4 49 49 .. Albania 3 –0.4 115 21 12.9 4,090 29.7 9,390 0.5 74 80 96 d Algeria 38 1.6 16 27 155.1 4,110 285.0 7,550d 0.6 72 75 73 Angola 21 3.4 17 48 95.4 4,580 114.3 5,490 3.5 50 53 70 Argentina 41 0.9 15 24 .. ..e .. Ѡ .. 72 80 98 Armenia 3 –0.3 104 20 11.1 3,720 20.8 6,990 7.0 71 77 100 Australia 23 1.4 3 19 1,351.2 59,570 982.2 43,300 1.8 80 84 .. Austria 8 0.5 103 15 407.6 48,160 373.2 44,100 0.4 78 84 .. Azerbaijan 9 1.2 112 22 56.3 6,050 87.5 9,410 3.1 68 74 100 Bangladesh 155 1.3 1,188 31 129.2 840 319.9 2,070 5.1 68 70 57 Belarus 9 –0.5 47 15 61.8 6,530 143.9 15,210 1.6 65 77 100 Belgium 11 0.7 368 17 501.3 44,990 447.6 40,170 –1.1 78 83 .. Benin 10 3.1 89 43 7.5 750 15.8 1,570 2.6 54 58 42 Bolivia 10 1.8 10 35 23.3 2,220 52.1 4,960 3.5 64 69 91 Bosnia and Herzegovina 4 0.0 75 16 17.8 4,650 36.0 9,380 –0.6 73 78 98 Brazil 199 1.1 23 25 2,311.1 11,630 2,328.8 11,720 0.0 70 77 90 Bulgaria 7 –0.9 67 14 50.2 6,870 112.4 15,390 1.4 71 78 98 Burkina Faso 16 2.9 60 46 10.9 670 24.9 1,510 6.9 54 56 29 Burundi 10 3.2 384 44 2.4 240 5.5 560 0.7 49 52 67 Cambodia 15 1.6 84 31 13.0 880 35.1 2,360 5.4 62 64 74 Cameroon 22 2.6 46 43 25.4 1,170 50.3 2,320 2.1 51 53 71 Canada 35 1.0 4 16 1,777.9 50,970 1,483.6 42,530 0.6 79 83 .. Central African Republic 5 1.8 7 40 2.2 490 3.9 860 2.1 47 50 56 Chad 12 3.4 10 49 9.3 740 16.4 1,320 1.9 48 51 34 Chile 17 1.0 23 21 249.5 14,280 372.1 21,310 4.6 76 82 99 China 1,351 0.6 145 18 7,748.9 5,740 12,435.4 9,210 7.3 72 75 94 Hong Kong SAR, China 7 0.6 6,866 12 261.6 36,560 379.6 53,050 0.3 80 87 .. Colombia 48 1.5 43 28 333.6 6,990 482.2 10,110 2.6 70 77 93 Congo, Dem. Rep. 66 2.8 29 45 14.8 220 24.5 370 4.3 47 50 67 Congo, Rep. 4 2.7 13 42 11.1 2,550 15.2 3,510 1.1 56 59 .. Costa Rica 5 1.7 94 24 42.0 8,740 60.5d 12,590d 3.6 77 82 96 Côte d’Ivoire 20 1.7 62 41 24.2 1,220 38.8 1,960 7.0 54 57 56 Croatia 4 –0.3 76 15 56.7 13,290 84.3 19,760 –1.7 74 80 99 Czech Republic 11 0.2 136 15 190.6 18,130 259.8 24,710 –1.5 75 81 .. Denmark 6 0.4 132 18 334.1 59,770 242.3 43,340 –0.8 78 82 .. Dominican Republic 10 1.4 213 31 56.2 5,470 101.0d 9,820d 2.6 71 76 90 Ecuador 15 1.8 62 30 80.5 5,190 148.5 9,590 3.3 73 79 92 Egypt, Arab Rep. 81 1.7 81 31 241.8 3,000 536.3 6,640 0.5 71 75 72 El Salvador 6 0.5 304 31 22.5 3,580 42.8d 6,790d 1.0 67 77 84 Eritrea 6 3.7 61 43 2.8 450 3.4d 560d 3.6 59 64 68 Ethiopia 92 2.7 92 43 37.4 410 104.2 1,140 5.7 58 61 39 Finland 5 0.4 18 16 254.1 46,940 209.2 38,630 –0.7 77 84 .. France 66 0.6 120 18 2,742.9 41,750 2,412.6 36,720 –0.5 78 85 .. Georgia 5f 0.2f 65f 18 14.8f 3,280f 26.4 f 5,860f 5.3f 70 77 100 Germany 82 0.0 235 13 3,603.9 44,010 3,430.1 41,890 0.6 78 83 .. Ghana 25 2.5 111 39 39.3 1,550 49.2 1,940 5.6 63 65 67 Greece 11 0.3 88 15 262.4 23,260 287.2 25,460 –6.2 79 83 97 d Guatemala 15 2.5 141 41 47.0 3,120 74.8 4,960d 0.4 68 75 75 Guinea 11 2.2 47 42 5.3 460 11.3 980 1.3 53 56 41 Haiti 10 1.4 369 35 7.7 760 12.6d 1,240d 1.4 61 63 49 Honduras 8 2.0 71 36 16.4 2,070 30.9d 3,890d 1.4 71 75 85 Hungary 10 –0.2 110 15 123.2 12,390 205.9 20,710 –1.4 71 79 99 India 1,237 1.4 416 29 1,890.4 1,530 4,749.2 3,840 1.9 64 67 63 Indonesia 247 1.4 136 29 844.0 3,420 1,188.0 4,810 4.9 68 71 93 Iran, Islamic Rep. 76 1.2 47 24 .. ..e .. .. .. 71 75 85 Iraq 33 2.6 75 41 191.2 5,870 140.2 4,300 5.7 66 72 78 Ireland 5 1.6 67 22 178.8 38,970 164.6 35,870 0.7 78 83 .. Israel 8 1.9 365 28 224.7 28,930 218.0 28,070 2.8 80 84 .. Italy 61 0.6 207 14 2,061.3 33,840 2,002.3 32,870 –2.7 80 85 99 Japan 128 0.0 350 13 6,105.8 47,870 4,629.7 36,290 2.1 79 86 .. Jordan 6 2.3 71 34 29.9 4,720 38.8 6,130 0.6 72 75 93 Kazakhstan 17 1.0 6 25 163.5 9,730 200.7 11,950 3.5 64 74 100 Kenya 43 2.7 76 42 36.2 840 76.1 1,760 1.5 56 58 87 Korea, Rep. 50 0.5 515 15 1,133.8 22,670 1,548.7 30,970 1.6 78 84 .. Kyrgyz Republic 6 1.1 29 30 5.5 990 12.6 2,260 –2.1 66 74 99 Lao PDR 7 1.7 29 36 8.4 1,260 18.1 2,730 6.1 66 69 73 Lebanon 4 2.6 433 22 40.7 9,190 63.7 14,400 0.4 70 75 90 Liberia 4 3.1 44 43 1.6 370 2.5 600 7.9 56 58 61 Libya 6 1.4 3 29 .. ..e .. .. .. 72 78 89 Lithuania 3 –1.3 48 15 41.3 13,850 67.9 22,760 5.3 68 79 100 Madagascar 22 2.9 38 43 9.7 430 21.2 950 0.3 65 68 64 Malawi 16 2.8 169 45 5.0 320 13.9 880 –1.0 54 54 75 Malaysia 29 1.8 89 27 286.4 9,800 483.2 16,530 3.9 72 77 93 Mali 15 3.1 12 47 9.8 660 17.2 1,160 –4.1 50 52 31 Mauritania 4 2.8 4 40 4.2 1,110 9.6 2,520 4.9 57 60 58 Mexico 121 1.3 62 29 1,176.9 9,740 2,015.8 16,680 2.6 75 79 93 Moldova 4g –0.2g 108g 17 7.4g 2,070g 13.1 g 3,690g –0.8g 66 73 99 Morocco 33 1.0 73 28 97.1h 2,940h 166.6h 5,040h 1.2h 70 74 56 Selected indicators 297 TABLE 1 Key indicators of development (continued) Population age Gross national Gross national Gross Life expectancy Adult Population composition  incomea income, PPPb domestic at birth literacy Average % $ $ product rate annual Density ages $ per $ per per capita  Years, Years, % ages 15 Millions growth, %  per sq. km  0–14  billions  capita  billions  capita  growth, %  male  female and older  2012 2000–12 2012  2012  2012  2012  2012  2012  2012  2011  2011  2005–11c  Mozambique 25 2.7 32 45 12.8 510 25.7 1,020 4.7 49 51 56 Myanmar 53 0.7 81 25 .. ..i .. .. .. 63 67 92 Nepal 27 1.4 192 36 19.2 700 41.1 1,500 3.4 68 70 60 Netherlands 17 0.4 497 17 809.1 48,250 731.5 43,620 –1.4 79 83 .. New Zealand 4 1.2 17 20 134.9 30,620 132.0 29,960 2.3 79 83 .. Nicaragua 6 1.3 50 33 9.9 1,650 23.7d 3,960d 3.7 71 77 78 Niger 17 3.7 14 50 6.4 370 11.2 650 7.0 54 55 29 Nigeria 169 2.6 185 44 241.1 1,430 409.1 2,420 3.6 51 53 61 Norway 5 0.9 16 19 496.2 98,860 336.1 66,960 1.7 79 84 .. Pakistan 179 1.8 232 34 225.4 1,260 543.6 3,030 2.4 65 66 55 Panama 4 1.8 51 29 37.7 9,910 67.8d 17,830d 8.9 74 79 94 d Papua New Guinea 7 2.4 16 38 12.8 1,790 19.9 2,780d 5.7 61 65 61 Paraguay 7 1.9 17 33 22.0 3,290 37.5 5,610 –2.9 70 75 94 Peru 30 1.2 23 29 176.5 5,880 306.9 10,240 5.0 71 77 90 Philippines 97 1.8 324 35 238.7 2,470 425.2 4,400 4.8 66 72 95 Poland 39 0.0 127 15 488.3 12,670 816.0 21,170 1.9 73 81 100 Portugal 11 0.2 115 15 216.6 20,580 260.7 24,770 –3.0 78 84 95 Romania 21 –0.4 93 15 179.6 8,420 347.8 16,310 4.0 71 78 98 Russian Federation 144 –0.2 9 15 1,822.7 12,700 3,260.6 22,720 3.0 63 75 100 Rwanda 11 2.6 464 44 6.2 560 13.9 1,250 5.0 54 57 71 Saudi Arabia 28 2.8 13 30 500.5 18,030 694.4 25,010 4.8 73 75 87 Senegal 14 2.8 71 44 14.2 1,040 26.3 1,920 0.7 58 60 50 Serbia 7 –0.3 83 16 38.1 5,280 80.8 11,180 –1.2 72 77 98 Sierra Leone 6 3.1 83 42 3.5 580 8.1 1,360 13.0 47 48 42 Singapore 5 2.3 7,589 16 250.8 47,210 324.6 61,100 –1.1 80 84 96 Slovak Republic 5 0.0 113 15 92.9 17,170 134.0 24,770 1.8 72 80 .. Somalia 10 2.7 16 47 .. ..i .. .. .. 50 53 .. South Africa 51 1.3 42 30 389.8 7,610 572.6 11,190 1.3 52 53 89 South Sudan 11 4.1 .. 42 7.0 650 .. .. –57.7 .. .. .. Spain 46 1.1 93 15 1,391.4 30,110 1,493.8 32,320 –1.5 79 85 98 Sri Lanka 20 0.5 324 25 59.3 2,920 124.5 6,120 9.2 72 78 91 j j Sudan 37 2.4 16 41j 53.8j 1,450j 75.3j 2,030j 0.6j 60 63 71 Sweden 10 0.6 23 17 535.0 56,210 420.1 44,150 0.0 80 84 .. Switzerland 8 0.9 200 15 661.6 82,730 449.8 56,240 –0.1 81 85 .. Syrian Arab Republic 22 2.6 122 35 56.3 2,610 116.5 5,200 0.8 74 77 83 Tajikistan 8 2.2 57 36 6.9 860 17.8 2,220 5.4 64 71 100 Tanzania 48 2.8 54 45 26.7 570k 73.6 k 1,590k 3.7k 57 59 73 Thailand 67 0.6 131 18 347.9 5,210 630.0 9,430 6.1 71 78 94 Togo 7 2.6 122 42 3.3 500 6.1 920 2.9 56 59 57 Tunisia 11 1.0 69 23 44.8 4,150 100.9 9,360 2.6 73 77 78 Turkey 74 1.3 96 26 801.1 10,830 1,345.7 18,190 0.9 72 76 91 Turkmenistan 5 1.2 11 29 28.7 5,550 49.9d 9,640d 9.7 61 69 100 Uganda 36 3.4 182 49 16.0 440 41.4 1,140 0.0 53 55 73 Ukraine 46 –0.6 79 14 159.6 3,500 332.5 7,290 0.4 66 76 100 United Arab Emirates 9 9.3 110 14 321.7 36,040 378.3 42,380 –0.8 76 78 90 United Kingdom 63 0.6 261 18 2,418.5 38,250 2,331.9 36,880 –0.5 79 83 .. United States 314 0.9 34 20 15,734.6 50,120 15,887.6 50,610 1.5 76 81 .. Uruguay 3 0.2 19 22 45.9 13,510 52.9 15,570 3.6 73 80 98 d Uzbekistan 30 1.6 70 29 51.3 1,720 111.6 3,750d 6.6 65 71 99 Venezuela, RB 30 1.7 34 29 373.5 12,470 393.0 13,120 3.9 71 77 96 Vietnam 89 1.1 286 23 124.1 1,400 305.6 3,440 3.9 73 77 93 West Bank and Gaza 4 2.7 672 41 .. ..l .. .. .. 71 75 95 Yemen, Rep. 24 2.6 45 41 26.0 1,110 53.7 2,310 –2.2 64 67 64 Zambia 14 2.8 19 47 19.1 1,350 22.8 1,620 4.0 49 49 71 Zimbabwe 14 0.8 35 40 9.3 680 .. .. 2.2 52 50 92 World 7,046s 1.2w 54w 26w 70,571.6t 10,015w 85,463.2t 12,129w 1.0w 68w 72w 84w Low income 846 2.2 56 39 494.1 584 1,173.7 1,387 3.6 58 61 61 Middle income 4,898 1.2 77 27 21,396.9 4,369 35,469.4 7,242 3.8 67 71 83 Lower middle income 2,507 1.6 122 31 4,706.2 1,877 9,808.1 3,912 2.5 64 68 71 Upper middle income 2,391 0.8 56 22 16,704.9 6,987 25,679.4 10,741 4.5 71 75 94 Low & middle income 5,744 1.3 73 29 21,902.7 3,813 36,624.4 6,376 3.6 66 70 80 High income 1,302 0.6 25 17 48,952.3 37,595 49,167.5 37,760 0.7 76 82 .. a. Calculated using the World Bank Atlas method. b. PPP = purchasing power parity; see the technical notes. c. Data are for the most recent year available. d. The estimate is based on regression; others are extrapolated from the 2005 International Comparison Program benchmark estimates. e. Estimated to be upper middle income ($4,086–$12,615). f. Excludes Abkhazia and South Ossetia. g. Excludes Transnistria. h. Includes Former Spanish Sahara. i. Estimated to be low income ($1,035 or less). j. Excludes South Sudan. k. Covers mainland Tanzania only. l. Estimated to be lower middle income ($1,036–$4,085). 298 WORLD DEVELOPMENT REPORT 2014 TABLE 2 Key indicators of development for other economies Population Population age composition Gross national incomea Average annual Density Thousands growth, % per sq. km % ages 0–14 $ millions $ per capita 2012 2000–12 2012 2012 2012 2012 American Samoa 55 –0.4 276 .. .. ..d Andorra 78 1.5 167 .. .. ..e Antigua and Barbuda 89 1.1 202 25 1,126 12,640 Aruba 102 1.0 569 20 .. ..e Bahamas, The 372 1.9 37 22 7,795 21,280 Bahrain 1,318 5.7 1,734 20 20,084 16,050 Barbados 283 0.5 659 19 .. ..e Belize 324 2.6 14 34 1,322 4,180 Bermuda 65 0.4 1,296 .. 6,903 106,920 Bhutan 742 2.3 19 29 1,797 2,420 Botswana 2,004 1.1 4 34 15,477 7,720 Brunei Darussalam 412 1.8 78 26 .. ..e Cape Verde 494 0.9 123 30 1,882 3,810 Cayman Islands 58 2.7 240 .. .. ..e Channel Islands 161 0.7 849 15 .. ..e Comoros 718 2.6 386 42 605 840 Cuba 11,271 0.1 106 17 .. ..d Curacao 152 1.1 342 20 .. ..e Cyprus 1,129 1.5 122 17 22,708g 26,000g Djibouti 860 1.4 37 34 .. ..h Dominica 72 0.2 96 .. 463 6,460 Equatorial Guinea 736 2.9 26 39 9,983 13,560 Estonia 1,339 –0.2 32 16 21,200 15,830 Faeroe Islands 50 0.5 35 .. .. ..e Fiji 875 0.6 48 29 3,675 4,200 French Polynesia 274 1.2 75 23 .. ..e Gabon 1,633 2.4 6 38 16,438 10,070 Gambia, The 1,791 3.1 177 46 912 510 Greenland 57 0.1 0i .. .. ..e Grenada 105 0.3 310 27 750 7,110 Guam 163 0.4 302 27 .. ..e Guinea–Bissau 1,664 2.2 59 42 916 550 Guyana 795 0.6 4 37 2,710 3,410 Iceland 320 1.1 3 21 12,393 38,710 Isle of Man 85 0.9 150 .. .. ..e Jamaica 2,712 0.4 250 28 13,929 5,140 Kiribati 101 1.6 124 32 228 2,260 Korea, Dem. Rep. 24,763 0.7 206 22 .. ..j Kosovo 1,806 0.5 166 27 6,576 3,640 Kuwait 3,250 4.4 182 25 133,824 44,730 Latvia 2,025 –1.3 33 15 28,725 14,180 Lesotho 2,052 0.8 68 37 2,823 1,380 Liechtenstein 37 0.9 229 .. .. ..e Luxembourg 531 1.6 205 17 40,898 76,960 Macao SAR, China 557 2.1 19,885 12 30,440 55,720 Macedonia, FYR 2,106 0.2 83 17 9,877 4,690 Maldives 338 1.8 1,128 29 1,947 5,750 Malta 418 0.8 1,307 15 8,268 19,760 Marshall Islands 53 0.1 292 .. 217 4,140 Mauritius 1,291 0.7 636 20 11,063 8,570 Micronesia, Fed. Sts. 103 –0.3 148 36 342 3,310 Monaco 38 1.3 18,790 .. .. ..e Mongolia 2,796 1.3 2 27 8,844 3,160 Montenegro 621 0.1 46 19 4,309 6,940 Namibia 2,259 1.5 3 37 12,813 5,670 New Caledonia 258 1.6 14 23 .. ..e Northern Mariana Islands 53 –2.1 116 .. .. ..e Oman 3,314 3.4 11 24 53,598 19,120 Palau 21 0.7 45 .. 205 9,860 Puerto Rico 3,667 –0.3 413 20 66,002 18,000 Qatar 2,051 10.3 177 13 150,427 78,720 Samoa 189 0.7 67 38 608 3,220 San Marino 31 1.2 521 .. .. ..e São Tomé and Príncipe 188 2.5 196 42 249 1,320 Seychelles 88 0.7 191 22 1,022 11,640 Sint Maarten (Dutch part) 39 2.1 1,150 .. .. ..e Slovenia 2,058 0.3 102 14 46,737 22,710 Solomon Islands 550 2.4 20 40 620 1,130 St. Kitts and Nevis 54 1.4 206 .. 714 13,330 St. Lucia 181 1.2 297 24 1,181 6,530 St. Martin (French part) 31 0.7 569 .. .. ..e St. Vincent and the Grenadines 109 0.1 280 26 698 6,380 Suriname 535 1.1 3 28 4,534 8,480 Swaziland 1,231 1.2 72 38 3,518 2,860 Timor–Leste 1,210 2.9 81 46 4,447 3,670 Tonga 105 0.6 146 37 445 4,240 Trinidad and Tobago 1,337 0.4 261 21 19,258 14,400 Turks and Caicos Islands 32 4.5 34 .. .. ..e Tuvalu 10 0.4 329 .. 60 6,070 Vanuatu 247 2.4 20 37 762 3,080 Virgin Islands (U.S.) 105 –0.3 301 21 .. ..e a. Calculated using the World Bank Atlas method. f. The estimate is based on regression; others are h. Estimated to be lower middle income ($1,036–$4,085). b. PPP = purchasing power parity; see the technical notes. extrapolated from the 2005 International Comparison i. Less than 0.5. c. Data are for the most recent year available. Program benchmark estimates. j. Estimated to be low income ($1,035 or less). d. Estimated to be upper middle income ($4,086–$12,615). g. Data are for the area controlled by the government of e. Estimated to be high income ($12,616 or more). Cyprus. Selected indicators 299 Gross national income, PPPb Gross domestic product per capita Life expectancy at birth Adult literacy rate $ millions $ per capita % growth Years, male Years, female % ages 15 and older 2012 2012 2012 2011 2011 2005–11c American Samoa .. .. .. .. .. .. Andorra .. .. .. .. .. .. Antigua and Barbuda 1,715f 19,260f 1.3 .. .. 99 Aruba .. .. .. 73 78 97 Bahamas, The 10,895f 29,740f 0.3 72 79 .. Bahrain 26,802 21,420 –0.5 75 76 92 Barbados .. .. .. 74 80 .. Belize 2,175f 6,880f –0.5 75 78 .. Bermuda .. .. –2.0 77 82 .. Bhutan 4,678 6,310 7.6 65 69 53 Botswana 33,114 16,520 5.2 54 52 84 Brunei Darussalam .. .. 0.7 76 80 95 Cape Verde 2,144 4,340 3.5 70 78 84 Cayman Islands .. .. .. .. .. 99 Channel Islands .. .. .. 78 82 .. Comoros 882 1,230 0.5 60 62 75 Cuba .. .. 2.1 77 81 100 Curacao .. .. .. 72 80 .. Cyprus 25,671g 29,400g –4.9g 77 82 98 Djibouti .. .. 3.2 56 59 .. Dominica 874f 12,190f –1.8 .. .. .. Equatorial Guinea 13,901 18,880 –0.3 50 52 94 Estonia 29,511 22,030 3.3 71 81 100 Faeroe Islands .. .. .. 79 85 .. Fiji 4,265 4,880 1.4 67 72 .. French Polynesia .. .. .. 73 78 .. Gabon 23,328 14,290 3.6 62 64 88 Gambia, The 3,327 1,860 2.7 57 60 50 Greenland .. .. .. 68 73 .. Grenada 1,087f 10,300f –1.2 74 77 .. Guam .. .. .. 74 79 .. Guinea–Bissau 1,981 1,190 –3.8 47 50 54 Guyana 2,703f 3,400f 4.2 67 73 .. Iceland 10,832 33,840 1.3 81 84 .. Isle of Man .. .. .. .. .. .. Jamaica .. .. –0.5 71 76 87 Kiribati 341f 3,380f 0.9 .. .. .. Korea, Dem. Rep. .. .. .. 66 72 100 Kosovo .. .. 2.9 68 72 .. Kuwait 147,287 49,230 3.6 74 76 94 Latvia 42,567 21,020 7.3 69 79 100 Lesotho 4,528 2,210 2.8 49 47 90 Liechtenstein .. .. .. .. .. .. Luxembourg 34,646 65,190 –2.2 79 84 .. Macao SAR, China 37,533 68,710 7.9 79 83 93 Macedonia, FYR 24,354 11,570 –0.3 73 77 97 Maldives 2,602 7,690 1.4 76 78 98 Malta 11,291 26,990 0.6 80 84 92 Marshall Islands .. .. 1.8 .. .. .. Mauritius 20,425 15,820 2.7 70 77 89 Micronesia, Fed. Sts. 423f 4,090f 1.4 68 70 .. Monaco .. .. .. .. .. .. Mongolia 14,265 5,100 10.6 65 73 97 Montenegro 8,654 13,930 0.4 72 77 98 Namibia 16,880 7,470 3.0 62 63 89 New Caledonia .. .. .. 73 80 96 Northern Mariana Islands .. .. .. .. .. .. Oman 71,696 25,580 –2.2 71 76 87 Palau 356f 17,150f 4.5 .. .. .. Puerto Rico .. .. 1.3 75 83 90 Qatar 161,789 84,670 8.8 79 78 96 Samoa 807f 4,270f 0.4 70 76 99 San Marino .. .. .. 80 86 .. São Tomé and Príncipe 349 1,850 1.3 63 66 89 Seychelles 2,262f 25,760f 2.5 70 77 92 Sint Maarten (Dutch part) .. .. .. 73 78 .. Slovenia 56,072 27,240 –2.6 77 83 100 Solomon Islands 1,192f 2,170f 1.7 66 69 .. St. Kitts and Nevis 926f 17,280f –2.2 .. .. .. St. Lucia 1,993f 11,020f –3.9 72 77 .. St. Martin (French part) .. .. .. .. .. .. St. Vincent and the Grenadines 1,182f 10,810f 1.5 70 74 .. Suriname 4,541f 8,500f 3.5 67 74 95 Swaziland 5,958 4,840 –3.0 49 48 87 Timor–Leste 7,761f 6,410f 5.5 62 63 58 Tonga 540f 5,140f 0.4 69 75 99 Trinidad and Tobago 29,957f 22,400f 0.9 67 74 99 Turks and Caicos Islands .. .. .. .. .. .. Tuvalu .. .. 1.0 .. .. .. Vanuatu 1,112f 4,500f 0.0 69 73 83 Virgin Islands (U.S.) .. .. .. 76 82 .. 300 WORLD DEVELOPMENT REPORT 2014 TABLE 3 Selected risk indicators Incidence of natural hazards Large recessions (droughts, earthquakes, floods, storms) Incidence of epidemics Adult mortality rate Homicide rate Per 1,000 Per 1,000 Per 100,000 Years in recession Total events Total epidemics male female people   1991–2000 2001–10 1993–2002 2003–12 1993–2002 2003–12 2007–11a 2007–11a 2010 Afghanistan 0 0 26 62 18 3 407 374 2.4 Albania 3 0 6 6 2 0 93 45 4.0 Algeria 4 0 17 29 1 0 123 99 1.5 Angola 3 0 9 22 7 9 382 333 19.0 Argentina 2 2 25 20 0 1 158 72 3.4 Armenia 0 1 4 1 0 0 160 78 1.4 Australia 0 0 47 46 0 1 81 47 1.0 Austria 0 0 11 8 0 0 100 48 0.6 Azerbaijan 0 0 8 5 0 0 178 68 2.2 Bangladesh 0 0 80 55 10 4 161 134 2.7 Belarus 0 0 4 1 2 0 334 112 4.9 Belgium 0 0 13 9 0 0 107 61 1.7 Benin 0 0 6 8 11 8 327 271 15.1 Bolivia 0 0 12 15 4 4 221 164 8.9 Bosnia and Herzegovina 0 0 3 9 1 0 132 68 1.5 Brazil 2 0 30 47 7 3 214 112 21.0 Bulgaria 5 1 6 15 0 0 197 88 2.0 Burkina Faso 0 0 5 11 7 9 296 246 18.0 Burundi 6 0 5 27 10 2 410 373 21.7 Cambodia 0 0 11 11 4 3 258 217 .. Cameroon 4 0 6 7 10 6 404 372 19.7 Canada 2 0 16 33 2 1 92 55 1.6 Central African Republic 4 4 6 13 6 2 456 422 29.3 Chad 3 4 7 13 5 13 368 313 15.8 Chile 0 0 17 15 0 0 121 56 3.2 China 0 0 188 234 3 3 135 86 1.0 Hong Kong SAR, China 1 0 10 8 0 1 72 36 0.5 Colombia 2 0 31 39 1 1 191 88 33.4 Congo, Dem. Rep. 10 1 8 19 31 34 405 351 21.7 Congo, Rep. 5 0 4 7 5 11 330 295 30.8 Costa Rica 0 0 16 21 1 0 109 57 11.3 Côte d’Ivoire 6 3 1 4 6 6 366 337 56.9 Croatia 0 2 3 7 0 0 136 55 1.4 Czech Republic 0 1 7 11 0 0 138 63 1.7 Denmark 0 2 2 3 0 0 107 65 0.9 Dominican Republic 1 0 9 30 2 4 197 129 24.9 Ecuador 1 0 16 13 6 2 159 83 18.2 Egypt, Arab Rep. 0 0 9 4 0 2 138 83 1.2 El Salvador 0 0 15 17 4 2 278 118 64.7 Eritrea 3 7 3 4 0 0 338 255 17.8 Ethiopia 3 2 24 28 9 7 298 252 25.5 Finland 3 2 .. .. .. .. 123 56 2.2 France 0 0 37 21 0 1 116 54 1.1 Georgia 4 0 7 8 0 0 175 66 4.3 Germany 0 1 20 22 1 1 101 54 0.8 Ghana 0 0 5 8 8 5 250 220 15.7 Greece 0 0 21 13 0 0 99 45 1.5 Guatemala 0 0 18 23 3 0 223 120 41.4 Guinea 4 0 4 7 5 7 347 298 22.5 Haiti 1 5 15 48 0 4 260 231 6.9 Honduras 0 0 20 24 4 2 162 113 82.1 Hungary 3 1 9 9 0 0 229 99 1.3 India 0 0 85 133 28 8 251 164 3.4 Indonesia 2 0 68 119 11 7 199 163 8.1 Iran, Islamic Rep. 0 0 59 41 1 0 162 75 3.0 Iraq 1 3 1 7 1 4 296 127 2.0 Ireland 0 0 7 4 1 1 97 57 1.2 Israel 0 3 4 1 1 0 79 45 2.1 Italy 0 2 24 19 1 1 78 41 0.9 Japan 0 2 47 55 1 0 85 42 0.4 Jordan 1 0 5 0 0 0 141 98 .. Kazakhstan 0 0 4 7 3 0 361 145 8.8 Kenya 4 0 11 34 18 12 370 348 20.1 Korea, Rep. 1 0 25 21 2 1 84 39 2.6 Kyrgyz Republic 5 0 2 9 2 1 305 130 20.1 Lao PDR 0 0 11 5 4 1 203 162 4.6 Lebanon 0 0 1 1 0 0 148 100 2.2 Liberia 5 2 2 5 7 3 340 305 10.1 Libya 0 2 1 0 0 0 135 83 2.9 Lithuania 0 1 2 4 0 0 271 93 6.6 Madagascar 6 1 14 29 3 2 213 166 8.1 Malawi 3 2 12 23 8 3 396 400 36.0 Malaysia 1 0 14 22 8 2 144 73 .. Selected indicators 301 Volatility of GDP growth Volatility of household Risk preparation Poverty headcount ratio per capita consumption growth per capita index $2.50 a day, PPP $10 a day, PPP (% population) (% population) Standard deviation Standard deviation 0–100 scale 1990b 2010c 1990b 2010c 1990s 2000s 1990s 2000s 2013 Afghanistan .. .. .. .. .. 5.40 .. .. 11 Albania .. 11.0 .. 91.7 15.75 1.67 13.53 10.11 60 Algeria 36.6 .. 95.7 .. 2.73 1.50 3.89 8.83 59 Angola .. 77.1 .. 98.9 11.65 6.90 .. .. .. Argentina 5.1d 2.6d 56.7d 29.2d 5.56 6.75 .. 7.84 60 Armenia .. 36.1 .. 97.5 19.68 8.07 12.85 4.31 59 Australia .. .. .. .. 1.79 1.06 1.27 1.47 94 Austria .. .. .. .. 1.19 2.02 1.26 0.63 80 Azerbaijan 54.0 7.3 98.4 86.4 14.85 8.32 .. 15.36 49 Bangladesh 95.0 86.2 99.9 99.6 0.72 0.85 2.28 1.14 34 Belarus 0.3 0.2 57.1 19.8 8.89 2.94 10.39 5.21 78 Belgium .. .. .. .. 1.41 1.76 1.02 0.86 86 Benin .. 84.4 .. 99.3 1.02 0.92 5.71 .. 24 Bolivia 27.8 31.0 88.1 82.5 1.55 1.40 0.94 1.43 39 Bosnia and Herzegovina .. 0.5 .. 36.7 .. 2.66 .. .. 60 Brazil 37.6 15.1 81.0 65.4 2.97 2.28 3.12 2.34 58 Bulgaria 0.1 2.5 14.1 68.3 5.36 3.63 10.57 4.94 76 Burkina Faso 90.3 82.2 99.3 99.1 3.69 2.05 11.82 .. 38 Burundi 97.4 96.1 100.0 100.0 4.88 2.06 .. .. 22 Cambodia 84.4 64.5 99.3 98.3 .. 3.04 .. 3.95 30 Cameroon .. 42.9 .. 95.2 4.70 0.78 5.39 3.53 37 Canada .. .. .. .. 2.50 1.97 1.86 1.15 84 Central African Republic 93.5 86.2 99.6 99.3 4.57 3.13 .. 3.18 9 Chad .. 89.5 .. 99.8 7.99 9.06 11.74 .. 13 Chile 21.2 4.3 78.4 52.9 3.40 1.98 3.78 3.28 70 China 91.7 36.5 99.9 91.1 2.99 1.62 3.46 1.42 69 Hong Kong SAR, China .. .. .. .. 3.64 3.23 4.56 3.26 73 Colombia 21.1 22.0 79.5 75.5 3.00 1.72 4.29 2.06 53 Congo, Dem. Rep. .. 97.0 .. 100.0 5.26 4.26 .. 4.70 8 Congo, Rep. .. 81.8 .. 98.7 3.21 2.93 32.03 14.30 27 Costa Rica 20.7 8.1 80.1 59.4 2.64 2.81 2.69 1.98 58 Côte d’Ivoire 48.5 58.7 97.2 97.4 3.50 2.67 4.60 .. 19 Croatia .. 0.1 .. 8.1 .. 3.78 .. 4.05 72 Czech Republic .. .. .. .. 5.20 3.05 .. .. 88 Denmark .. .. .. .. 1.39 2.52 1.92 2.37 87 Dominican Republic 36.0 16.1 88.8 75.9 4.57 3.22 5.98 4.80 45 Ecuador 29.8 15.9 85.4 75.5 3.16 2.35 3.27 1.47 61 Egypt, Arab Rep. 44.1 32.0 97.8 97.7 1.45 1.71 0.93 2.95 63 El Salvador 37.1 23.1 87.9 83.0 1.60 1.87 6.41 4.34 53 Eritrea .. .. .. .. .. 8.36 .. .. .. Ethiopia 90.9 79.9 99.9 99.4 7.28 4.39 9.36 5.28 30 Finland .. .. .. .. 4.08 3.80 3.71 1.94 94 France .. .. .. .. 1.20 1.67 1.33 0.92 82 Georgia .. 46.6 .. 95.7 23.68 4.31 .. .. 61 Germany .. .. .. .. 1.71 2.51 1.18 0.80 87 Ghana 86.5 63.9 99.7 97.8 0.68 2.73 .. .. 30 Greece .. .. .. .. 1.81 4.29 1.27 4.40 78 Guatemala 63.5 33.9 94.7 85.2 0.69 1.41 0.95 0.91 43 Guinea 99.9 79.5 100.0 99.3 2.09 1.57 5.05 .. 5 Haiti .. 83.0 .. 97.8 .. 3.04 .. .. 16 Honduras 64.1 36.5 95.8 81.8 2.76 2.36 2.13 2.79 47 Hungary .. 0.6 .. 42.4 5.03 3.29 4.82 4.25 81 India 90.8 81.1 99.7 99.2 2.11 2.25 1.83 2.00 31 Indonesia 91.4 60.4 99.9 98.4 6.85 0.91 5.94 1.10 42 Iran, Islamic Rep. 20.3 14.8 83.8 85.1 4.47 1.98 2.69 .. 73 Iraq .. 36.4 .. 97.4 .. 20.52 .. .. 38 Ireland .. .. .. .. 3.02 4.12 2.43 3.98 85 Israel .. .. .. .. 1.19 2.79 .. 2.36 76 Italy .. .. .. .. 0.99 2.43 1.85 1.15 72 Japan .. .. .. .. 2.02 2.49 1.45 0.98 80 Jordan 24.6 5.4 89.3 81.9 5.25 2.10 11.89 5.08 64 Kazakhstan .. 4.2 .. 87.4 6.41 3.78 11.39 4.34 78 Kenya 68.6 76.6 97.3 98.4 1.88 2.15 4.32 3.02 29 Korea, Rep. .. .. .. .. 4.77 2.09 7.16 3.25 78 Kyrgyz Republic 37.2 34.4 86.5 96.5 10.66 3.19 13.36 9.74 55 Lao PDR .. 78.1 .. 99.1 1.34 1.09 .. 5.05 38 Lebanon .. .. .. .. 10.77 3.16 .. 3.84 40 Liberia .. 97.0 .. 99.9 38.04 16.65 .. 16.98 10 Libya .. .. .. .. .. 4.74 .. .. .. Lithuania .. 0.9 .. 41.7 11.52 6.88 .. 9.36 79 Madagascar 92.7 95.4 99.6 99.8 3.11 6.20 3.15 3.80 12 Malawi .. 88.5 .. 99.4 7.62 3.82 .. .. 39 Malaysia 18.3 6.2 80.8 55.1 5.18 2.71 6.28 3.21 67 302 WORLD DEVELOPMENT REPORT 2014 TABLE 3 Selected risk indicators (continued) Incidence of natural hazards Large recessions (droughts, earthquakes, floods, storms) Incidence of epidemics Adult mortality rate Homicide rate Per 1,000 Per 1,000 Per 100,000 Years in recession Total events Total epidemics male female people   1991–2000 2001–10 1993–2002 2003–12 1993–2002 2003–12 2007–11a 2007–11a 2010 Mali 2 0 7 15 4 6 356 293 8.0 Mauritania 5 2 9 11 2 1 286 217 14.7 Mexico 1 2 68 55 1 1 130 72 22.7 Moldova 8 1 7 5 1 0 300 145 7.5 Morocco 3 0 11 12 0 0 141 89 1.4 Mozambique 1 0 17 28 9 11 477 443 .. Myanmar 0 0 7 14 0 1 231 181 10.2 Nepal 0 0 10 18 11 3 182 155 2.8 Netherlands 0 0 7 4 1 0 75 56 1.1 New Zealand 1 0 11 11 0 1 87 58 1.1 Nicaragua 3 0 21 17 6 2 194 109 13.6 Niger 5 0 9 13 15 14 309 267 3.8 Nigeria 0 0 18 24 23 16 387 359 12.2 Norway 0 0 4 3 0 0 82 50 0.6 Pakistan 0 0 31 46 6 3 188 157 7.6 Panama 0 0 8 21 2 0 131 69 21.6 Papua New Guinea 5 3 15 13 4 3 310 233 13.0 Paraguay 3 3 8 10 1 7 166 119 11.5 Peru 0 0 23 21 4 3 156 96 10.3 Philippines 3 0 100 163 5 7 257 142 5.4 Poland 0 0 7 13 0 0 198 76 1.1 Portugal 0 0 7 6 0 0 122 53 1.2 Romania 5 1 18 29 2 1 179 73 2.0 Russian Federation 1 1 49 37 9 1 367 137 10.2 Rwanda 2 0 6 8 7 2 345 312 17.1 Saudi Arabia 3 2 1 10 3 0 122 94 .. Senegal 4 0 8 11 4 3 287 235 8.7 Serbia 1 0 6 4 2 0 147 81 1.2 Sierra Leone 6 0 2 6 10 4 459 438 14.9 Singapore 1 2 0 0 2 1 75 44 0.4 Slovak Republic 3 1 5 7 0 0 184 74 1.5 Somalia 0 0 12 25 14 9 365 309 1.5 South Africa 3 0 28 20 4 3 572 574 31.8 South Sudan 0 0 .. .. .. .. .. .. .. Spain 0 3 16 12 2 1 94 43 0.8 Sri Lanka 0 0 17 24 3 3 182 78 3.6 Sudan 0 0 17 17 11 17 262 208 24.2 Sweden 3 2 2 1 1 1 71 44 1.0 Switzerland 0 0 9 10 0 1 76 42 0.7 Syrian Arab Republic 1 0 2 3 0 0 108 70 2.3 Tajikistan 6 1 16 18 3 2 221 125 2.1 Tanzania 4 0 17 21 12 8 351 331 24.5 Thailand 2 0 44 38 1 5 201 99 5.3 Togo 6 2 5 6 6 3 335 292 10.9 Tunisia 0 0 .. .. .. .. 121 68 1.1 Turkey 1 3 29 32 0 2 133 74 3.3 Turkmenistan 7 0 2 0 0 0 302 158 .. Uganda 0 0 13 15 11 17 393 377 36.3 Ukraine 8 1 11 9 3 0 334 128 5.2 United Arab Emirates 5 0 .. .. .. .. 89 67 .. United Kingdom 0 2 27 19 1 1 95 58 1.2 United States 0 2 211 184 3 1 135 79 4.8 Uruguay 2 2 10 5 0 0 130 58 6.1 Uzbekistan 6 0 1 2 1 0 242 138 3.1 Venezuela, RB 4 4 12 17 1 1 169 87 45.1 Vietnam 0 0 48 67 3 5 129 87 1.6 West Bank and Gaza 0 0 .. .. .. .. 140 103 .. Yemen, Rep. 0 0 13 12 1 1 226 182 4.2 Zambia 6 0 4 12 7 7 486 489 38.0 Zimbabwe 4 7 6 10 10 10 517 571 14.3 World 1u 1u 2,561s 3,132s 560s 425s 207w 147w 5.7w Low income 3 1 387 640 275 214 291 254 14.5 Middle income 1 0 1,431 1,809 251 191 199 134 6.2 Lower middle income 2 0 659 888 179 133 240 171 4.8 Upper middle income 1 0 772 921 72 58 160 100 7.2 Low & middle income 2 1 1,818 2,449 526 405 210 150 5.8 High income 0 1 743 683 34 20 147 72 4.0 a. Data are for the most recent year available between 2007 and 2011. b. Country data are for the year closest to 1990 between 1985 and 1995. c. Country data are for the most recent year available between 2001 and 2011. d. Urban only. e. Based on the 1990 income classification. Selected indicators 303 Volatility of GDP growth Volatility of household Risk preparation Poverty headcount ratio per capita consumption growth per capita index $2.50 a day, PPP $10 a day, PPP (% population) (% population) Standard deviation Standard deviation 0–100 scale 1990b 2010c 1990b 2010c 1990s 2000s 1990s 2000s 2013 Mali 95.9 87.2 99.7 100.0 3.52 2.45 4.31 .. 18 Mauritania 74.9 60.9 99.0 97.8 4.57 4.97 .. 8.88 19 Mexico 21.8 8.8 77.9 67.5 3.61 3.32 7.08 3.74 56 Moldova .. 10.7 .. 87.8 14.13 3.92 .. 8.19 57 Morocco 26.3 24.5 90.6 91.6 6.22 1.84 7.05 2.14 51 Mozambique .. 88.4 .. 99.5 4.72 2.37 7.08 6.86 12 Myanmar .. .. .. .. 3.17 .. .. .. 30 Nepal .. 71.8 .. 99.3 1.44 1.58 .. .. 23 Netherlands .. .. .. .. 1.15 2.12 1.67 1.44 93 New Zealand .. .. .. .. 3.31 1.90 3.07 1.77 89 Nicaragua 46.0 43.8 93.1 95.1 3.02 1.93 13.59 1.94 39 Niger 94.8 85.0 99.8 99.5 4.18 3.26 6.22 .. 18 Nigeria 86.9 86.3 100.0 99.8 2.13 2.41 .. .. 27 Norway .. .. .. .. 1.24 1.72 1.52 1.72 94 Pakistan 93.3 76.4 99.9 99.3 1.82 1.95 3.79 4.90 33 Panama 34.0 18.8 79.7 72.1 2.53 3.55 8.77 5.82 56 Papua New Guinea .. .. .. .. 7.62 3.60 12.39 .. 22 Paraguay 7.7 18.4 68.5 73.2 1.91 5.36 3.82 6.12 50 Peru .. 18.3 .. 73.2 5.04 3.08 5.22 2.24 52 Philippines 66.4 53.3 97.4 95.5 2.35 1.73 0.92 1.02 45 Poland 0.6 0.5 68.1 48.7 4.43 1.71 1.91 1.29 76 Portugal .. .. .. .. 2.20 1.77 2.06 2.10 77 Romania .. 4.3 .. 85.3 6.42 4.97 8.12 7.56 65 Russian Federation 12.9 0.3 70.7 45.4 6.54 4.58 3.31 4.92 71 Rwanda 93.6 88.3 100.0 99.8 24.31 2.70 8.77 5.47 26 Saudi Arabia .. .. .. .. 2.89 2.50 .. .. 83 Senegal 86.9 67.5 99.0 98.3 2.54 1.68 3.31 4.49 26 Serbia .. 1.3 .. 61.3 17.42 2.98 .. 4.75 61 Sierra Leone 80.8 87.6 99.1 99.6 9.09 5.54 .. .. 14 Singapore .. .. .. .. 3.73 4.99 4.31 3.57 72 Slovak Republic .. 0.2 .. 44.1 7.16 3.71 .. 2.63 79 Somalia .. .. .. .. .. .. .. .. .. South Africa 48.9 39.5 85.5 79.5 2.01 1.91 2.16 2.58 45 South Sudan .. .. .. .. .. .. .. .. .. Spain .. .. .. .. 1.69 2.19 1.96 2.41 82 Sri Lanka 66.4 38.2 98.9 96.1 1.00 1.98 4.32 2.91 51 Sudan .. 58.5 .. 98.8 4.19 1.90 1.76 5.71 22 Sweden .. .. .. .. 2.65 3.03 2.64 1.73 96 Switzerland .. .. .. .. 1.58 1.85 1.03 0.70 87 Syrian Arab Republic .. 29.0 .. 94.2 4.07 1.90 6.97 .. 46 Tajikistan .. 42.2 .. 98.3 13.22 9.17 30.57 3.96 46 Tanzania 95.2 92.8 100.0 99.8 2.09 0.72 4.47 34.86 32 Thailand 50.1 9.6 93.9 81.2 6.62 2.79 6.88 2.02 62 Togo .. 64.9 .. 98.6 8.93 2.47 12.54 6.51 28 Tunisia 28.5 8.3 92.0 78.8 2.03 2.28 2.03 2.41 59 Turkey 14.7 8.3 84.9 67.9 4.87 5.12 5.17 5.09 70 Turkmenistan 67.5 .. 100.0 .. 15.40 4.36 .. .. .. Uganda 90.9 76.0 99.7 98.7 2.27 1.80 3.35 3.91 25 Ukraine 14.2 0.2 84.3 56.3 7.73 7.07 9.87 9.00 71 United Arab Emirates .. .. .. .. 4.79 6.08 .. 8.11 88 United Kingdom .. .. .. .. 1.77 2.41 2.11 2.61 82 United States .. .. .. .. 1.55 2.03 1.55 1.77 91 Uruguay 3.8 2.8 48.6 48.5 3.66 5.03 5.64 8.20 73 Uzbekistan .. .. .. .. 5.41 1.82 .. .. 54 Venezuela, RB 20.2 18.8 83.2 80.2 4.66 7.67 4.42 8.17 52 Vietnam 91.3 58.2 99.8 98.2 1.72 0.91 .. 2.39 56 West Bank and Gaza .. 1.3 .. 65.1 .. .. .. .. 46 Yemen, Rep. .. 61.7 .. 98.4 1.19 4.64 .. .. 30 Zambia 82.0 90.6 99.0 99.3 4.68 1.27 17.90 15.81 25 Zimbabwe .. .. .. .. 5.58 8.99 .. .. 25 World w w w w 3.23 w 2.42 w 3.45 w 2.70 w 55 u Low income 90.0 83.2 99.4 99.3 3.83 2.88 .. .. 23 Middle income 73.6 48.1 95.1 89.8 3.32 2.32 3.61 2.47 52 Lower middle income 82.1 68.5 98.4 97.2 2.98 2.14 3.11 2.56 43 Upper middle income 65.8 26.8 92.0 82.0 3.64 2.50 4.03 2.38 61 Low & middle income 71.7e 50.0e 94.2e 88.8e 3.37 2.38 3.79 2.83 44 High income .. .. .. .. 2.72 2.58 2.22 2.21 81 304 WORLD DEVELOPMENT REPORT 2014 TABLE 4 Selected indicators related to risk management at the household level Under-five Maternal Access to social Educational attainment Education quality mortality rate mortality ratio insurance Savings % of adults 25+ % of adults 25+ % of adults 25+ PISA mean PISA mean % of people with complete with complete with complete score, score, Per 1,000 Per 100,000 Survey % coverage, who saved in primary secondary tertiary math reading live births live births year adults 60+ the past year   2010 2010 2010 2009 2009 2011 2010     2011 Afghanistan 8.2 5.2 5.3 .. .. 101 460 2006 8.5 14.6 Albania 9.6 30.6 5.0 377 385 14 27 2009 103.7 22.7 Algeria 1.4 34.5 5.6 .. .. 30 97 2002 42.9 20.9 Angola .. .. .. .. .. 158 450 .. .. 36.6 Argentina 32.0 30.8 3.3 388 398 14 77 2010 103.8 24.4 Armenia 4.8 64.6 12.7 .. .. 18 30 2008 74.2 10.5 Australia 2.1 39.0 22.4 514 515 5 7 .. .. 68.3 Austria 17.2 45.3 9.5 496 470 4 4 2006 85.1 78.3 Azerbaijan .. .. .. 431 362 45 43 2007 104.6 10.8 Bangladesh 21.5 18.4 3.0 .. .. 46 240 2011 73.6 26.8 Belarus .. .. .. .. .. 6 4 2008 116.0 26.4 Belgium 11.1 33.3 20.4 515 506 4 8 2006 61.1 57.8 Benin 10.1 12.4 2.0 .. .. 106 350 2004 2.9 32.4 Bolivia 11.8 35.5 9.5 .. .. 51 190 2007 70.0 44.2 Bosnia and Herzegovina .. .. .. .. .. 8 8 2009 22.2 13.5 Brazil 26.2 25.1 5.2 386 412 16 56 2010 68.7 21.1 Bulgaria 24.4 27.1 12.1 428 429 12 11 2008 92.4 10.9 Burkina Faso .. .. .. .. .. 146 300 2005 2.6 38.0 Burundi 20.7 3.4 0.6 .. .. 139 800 2004 6.3 25.2 Cambodia 33.3 5.3 0.5 .. .. 43 250 2005 2.9 31.0 Cameroon 31.0 13.5 1.6 .. .. 127 690 2002 4.4 51.9 Canada 2.6 37.5 28.0 527 524 6 12 2007 54.5 65.5 Central African Republic 16.5 8.5 1.2 .. .. 164 890 2003 4.4 25.4 Chad .. .. .. .. .. 169 1,100 2001 0.6 28.7 Chile 13.6 29.7 11.7 421 449 9 25 2010 60.9 27.1 China 17.7 40.3 3.9 .. .. 15 37 2010 59.2 38.4 Hong Kong SAR, China 14.3 35.4 7.3 555 533 .. .. 2005 38.5 59.0 Colombia 28.9 31.8 8.5 381 413 18 92 2010 17.8 32.9 Congo, Dem. Rep. 8.4 7.9 1.1 .. .. 168 540 .. .. 24.1 Congo, Rep. 8.0 8.9 1.3 .. .. 99 560 2001 5.6 30.4 Costa Rica 29.7 19.4 13.2 .. .. 10 40 2009 42.9 41.0 Côte d’Ivoire 17.8 6.4 3.9 .. .. 115 400 2004 6.6 .. Croatia 17.1 29.2 5.3 460 476 5 17 2010 60.6 21.9 Czech Republic 10.0 65.2 5.6 493 478 4 5 2007 91.6 49.0 Denmark 28.3 26.8 12.5 503 495 4 12 2006 82.7 72.6 Dominican Republic 7.8 8.1 4.5 .. .. 25 150 2008 5.7 37.3 Ecuador 29.1 16.1 10.5 .. .. 23 110 2009 38.4 30.2 Egypt, Arab Rep. 6.4 22.3 5.7 .. .. 21 66 2004 31.1 8.1 El Salvador 15.9 16.9 8.5 .. .. 15 81 2010 14.6 25.9 Eritrea .. .. .. .. .. 68 240 .. .. .. Ethiopia .. .. .. .. .. 77 350 .. .. .. Finland 22.1 26.4 13.1 541 536 3 5 2006 81.2 68.7 France 6.5 40.1 10.6 497 496 4 8 2006 103.1 61.8 Georgia .. .. .. .. .. 21 67 .. .. 7.0 Germany 2.8 55.4 12.8 513 497 4 7 2006 86.1 67.3 Ghana 8.0 17.5 2.1 .. .. 78 350 2010 8.1 36.6 Greece 28.2 32.3 22.4 466 483 4 3 2006 75.1 27.8 Guatemala 16.4 9.7 2.2 .. .. 30 120 2008 6.9 24.8 Guinea .. .. .. .. .. 126 610 2001 2.9 27.1 Haiti 4.6 22.4 0.8 .. .. 70 350 .. .. 31.6 Honduras 35.3 13.3 4.0 .. .. 21 100 2009 2.7 21.8 Hungary 3.9 52.2 12.6 490 494 6 21 2008 95.3 26.7 India 16.6 0.8 4.1 .. .. 61 200 2010 23.7 22.4 Indonesia 32.0 11.1 1.7 371 402 32 220 2010 5.6 40.5 Iran, Islamic Rep. 12.5 24.4 12.9 .. .. 25 21 2001 16.3 32.3 Iraq 20.4 11.8 6.5 .. .. 38 63 2009 3.9 26.2 Ireland 10.9 32.4 20.2 487 496 4 6 2006 56.7 64.6 Israel 17.0 26.1 24.3 447 474 4 7 .. .. 44.9 Italy 19.0 32.5 6.6 483 486 4 4 2006 84.6 26.4 Japan 14.0 30.3 24.0 529 520 3 5 2003 83.5 63.3 Jordan 8.0 38.0 6.4 387 405 21 63 2006 19.2 18.2 Kazakhstan 3.2 39.5 11.8 405 390 28 51 2009 106.8 21.9 Kenya 40.7 0.6 2.0 .. .. 73 360 2006 6.9 40.1 Korea, Rep. 10.2 36.8 17.9 546 539 5 16 2005 28.1 64.5 Kyrgyz Republic 10.1 35.9 8.4 331 314 31 71 2008 103.9 36.4 Lao PDR 20.8 5.7 3.2 .. .. 42 470 2005 5.7 54.5 Lebanon .. .. .. .. .. 9 25 2003 9.1 30.8 Liberia 6.2 5.8 5.3 .. .. 78 770 .. .. 34.9 Libya 21.4 19.4 10.3 .. .. 16 58 .. .. .. Lithuania 4.0 55.5 16.4 477 468 6 8 2010 102.3 32.7 Selected indicators 305 TABLE 4 Selected indicators related to risk management at the household level (continued) Under-five Maternal Access to social Educational attainment Education quality mortality rate mortality ratio insurance Savings % of adults 25+ % of adults 25+ % of adults 25+ PISA mean PISA mean % of people with complete with complete with complete score, score, Per 1,000 Per 100,000 Survey % coverage, who saved in primary secondary tertiary math reading live births live births year adults 60+ the past year   2010 2010 2010 2009 2009 2011 2010     2011 Madagascar .. .. .. .. .. 62 240 .. .. 19.7 Malawi 12.6 8.1 0.3 .. .. 83 460 .. .. 33.0 Malaysia 12.9 33.8 5.0 .. .. 7 29 2007 54.7 51.0 Mali 6.3 2.5 1.2 .. .. 176 540 2010 20.0 37.4 Mauritania 23.7 6.3 1.5 .. .. 112 510 2002 6.8 22.9 Mexico 18.9 17.7 13.9 419 425 16 50 2010 18.6 27.1 Moldova 4.5 39.2 9.0 .. .. 16 41 2009 83.3 22.2 Morocco 17.5 9.9 5.9 .. .. 33 100 2007 13.2 30.5 Mozambique 12.5 1.6 0.3 .. .. 103 490 2004 8.8 42.6 Myanmar 24.3 9.0 4.0 .. .. 62 200 .. .. .. Nepal 9.8 6.3 2.0 .. .. 48 170 2006 39.2 18.4 Netherlands 7.6 40.5 16.5 526 508 4 6 2007 77.5 73.1 New Zealand 20.7 15.9 24.4 519 521 6 15 2007 68.5 72.6 Nicaragua 8.1 11.5 10.0 .. .. 26 95 2008 13.8 26.1 Niger 9.8 2.1 0.7 .. .. 125 590 2006 4.8 25.1 Nigeria .. .. .. .. .. 124 630 .. .. 64.4 Norway 0.5 45.7 14.6 498 503 3 7 2006 70.3 .. Pakistan 14.3 19.0 5.2 .. .. 72 260 2012 4.1 7.5 Panama 21.3 24.6 16.4 360 371 20 92 2009 32.0 34.9 Papua New Guinea 40.8 4.7 0.8 .. .. 58 230 2005 0.8 .. Paraguay 25.4 30.4 3.1 .. .. 22 99 2004 2.9 18.1 Peru 7.2 27.5 16.6 365 370 18 67 2008 20.0 29.1 Philippines 18.0 19.9 22.4 .. .. 25 99 2007 13.0 45.5 Poland 15.9 11.4 9.1 495 500 6 5 2009 71.7 30.8 Portugal 42.4 13.6 3.8 487 489 3 8 2006 86.7 33.7 Romania 1.4 40.2 6.5 427 424 13 27 2009 76.7 18.2 Russian Federation 2.4 30.2 23.3 468 459 12 34 2007 120.5 22.7 Rwanda 27.0 3.4 0.7 .. .. 54 340 2004 7.0 30.5 Saudi Arabia 15.3 24.0 8.7 .. .. 9 24 .. .. 33.4 Senegal 31.1 6.0 2.3 .. .. 65 370 2010 34.5 15.4 Serbia 25.5 28.4 7.6 442 442 7 12 2007 32.7 14.9 Sierra Leone 8.7 1.2 0.9 .. .. 185 890 .. .. 32.6 Singapore 16.6 15.8 12.2 562 526 3 3 2009 30.1 60.9 Slovak Republic 10.6 36.7 6.5 497 477 8 6 2008 115.1 49.3 Somalia .. .. .. .. .. 180 1,000 .. .. 21.8 South Africa 6.2 22.0 0.6 .. .. 47 300 2010 81.7 31.5 South Sudan .. .. .. .. .. 121 .. .. .. .. Spain 18.9 21.9 16.6 483 481 4 6 2006 67.9 46.3 Sri Lanka 8.4 47.1 10.5 .. .. 12 35 2010 20.9 36.3 Sudan 25.7 4.1 1.7 .. .. 86a 730 2003 4.0 22.7 Sweden 8.2 52.3 16.7 494 497 3 4 2006 81.4 82.8 Switzerland 15.6 41.4 12.4 534 501 4 8 .. .. .. Syrian Arab Republic 14.5 5.2 2.3 .. .. 15 70 .. .. 48.3 Tajikistan 4.9 43.7 4.6 .. .. 63 65 2004 98.5 13.8 Tanzania 49.0 1.2 0.6 .. .. 68 460 2005 0.1 40.1 Thailand 27.4 10.1 8.8 419 421 12 48 2010 73.5 60.0 Togo 21.4 12.2 1.8 .. .. 110 300 2003 3.7 19.6 Tunisia 18.0 15.1 6.2 371 404 16 56 2005 34.6 25.1 Turkey 45.8 17.7 5.9 445 464 15 20 2008 92.2 9.6 Turkmenistan .. .. .. .. .. 53 67 .. .. 44.5 Uganda 24.5 3.1 2.6 .. .. 90 310 2003 0.5 44.4 Ukraine 4.2 41.3 25.3 .. .. 10 32 2010 111.3 25.0 United Arab Emirates 11.7 32.1 10.1 .. .. 7 12 .. .. 30.1 United Kingdom 24.2 1.3 13.6 492 494 5 12 .. .. 56.7 United States 1.9 36.2 31.6 487 500 8 21 2008 58.1 66.8 Uruguay 35.3 18.6 6.4 427 426 10 29 2010 65.3 16.9 Uzbekistan .. .. .. .. .. 49 28 2005 136.1 31.4 Venezuela, RB 27.9 4.6 4.1 .. .. 15 92 2006 18.6 28.4 Vietnam 38.5 11.4 3.0 .. .. 22 59 2008 30.7 35.3 West Bank and Gaza .. .. .. .. .. 22 64 2009 5.5 16.2 Yemen, Rep. 8.6 6.1 1.9 .. .. 77 200 2006 8.5 11.9 Zambia 29.8 10.0 1.0 .. .. 83 440 2003 4.2 32.2 Zimbabwe 21.8 9.4 0.7 .. .. 67 570 2005 3.7 39.9 World 16.7 w 24.1 w 8.7w     51w 210w   46.0w 35.9w Low income 21.0 10.2 2.3 95 410 .. 29.9 Middle income 18.9 22.4 5.3 46 190 40.0 31.0 Lower middle income 18.7 8.1 5.2 62 260 22.3 27.5 Upper middle income 19.1 33.8 5.4 20 64 56.8 34.5 Low & middle income 19.1 21.3 5.1 56 240 39.1 30.9 High income 9.3 32.9 20.0 6 16 77.0 53.9 a. Excludes South Sudan. 306 WORLD DEVELOPMENT REPORT 2014 TABLE 5 Selected indicators related to risk management at the enterprise sector level Wage employment Goods market efficiency Labor market efficiency Pension contributors Formal production Wage and salaried workers, % of economy, % of total employed, annual average 1–7 scale 1–7 scale     annual average   1991–2000 2001–10 2006–07 2012–13 2006–07 2012–13 Survey year % of labor force 1999–2003 2004–07 Afghanistan .. .. .. .. .. .. 2006 3.7 .. .. Albania .. .. 3.46 4.33 4.05 4.40 2008 37.9 65.0 66.6 Algeria .. 46 3.66 2.99 3.52 2.79 2007 74.6 66.4 68.8 Angola .. .. .. .. .. .. .. .. 51.9 55.5 Argentina 71 75 3.66 3.18 3.44 3.29 2010 47.0 73.9 75.8 Armenia 55 52 3.74 4.22 4.59 4.72 2008 32.1 54.7 57.6 Australia 85 87 5.39 4.87 4.84 4.60 2005 90.7 85.8 86.4 Austria 86 87 5.33 4.91 4.47 4.69 2005 93.7 90.2 90.3 Azerbaijan .. 37 3.81 4.31 4.47 4.80 2007 35.4 39.8 44.7 Bangladesh 13 14 3.88 4.10 4.12 3.91 2004 2.5 64.3 65.2 Belarus .. .. .. .. .. .. 2008 93.5 52.2 55.3 Belgium 83 85 5.18 5.12 4.02 4.54 2005 91.4 77.8 78.4 Benin .. 10 3.60 3.66 3.76 4.40 2005 5.5 50.0 50.5 Bolivia 48 35 3.16 3.40 3.73 3.58 2009 12.2 32.6 35.6 Bosnia and Herzegovina .. 73 3.52 3.92 4.21 4.08 2009 24.5 66.0 66.9 Brazil 62 64 3.82 3.94 3.91 4.39 2010 59.3 60.0 62.2 Bulgaria 83 86 3.75 4.17 4.12 4.54 2008 78.7 63.5 66.2 Burkina Faso 4 6 3.70 3.80 4.18 4.42 2009 .. 58.9 60.2 Burundi 6 5 2.94 3.28 4.21 3.97 2006 3.5 60.6 60.4 Cambodia 15 21 3.97 4.42 4.76 4.78 2010 0.5 50.1 52.7 Cameroon 14 19 3.55 4.15 3.82 4.48 2006 16.2 67.5 68.5 Canada 84 86 5.34 5.12 5.21 5.45 2009 87.4 84.1 84.6 Central African Republic .. .. .. .. .. .. 2003 1.5 56.1 53.7 Chad 5 .. 2.67 3.08 3.73 4.12 2005 2.7 54.6 58.4 Chile 70 71 4.94 4.74 4.87 4.68 2010 57.7 80.3 81.2 China .. .. 4.17 4.31 4.27 4.60 2010 33.5 87.0 87.7 Hong Kong SAR, China 89 88 5.80 5.44 5.59 5.65 2009 78.9 83.4 84.7 Colombia 65 49 3.94 3.98 4.20 4.17 2010 27.8 61.2 64.6 Congo, Dem. Rep. .. .. .. .. .. .. 2009 .. 52.3 53.2 Congo, Rep. .. 22 .. .. .. .. 2008 9.7 52.3 55.3 Costa Rica 71 71 4.27 4.30 4.72 4.51 2010 58.6 73.8 74.9 Côte d’Ivoire .. 20 .. 3.78 .. 4.38 2004 12.8 55.9 53.5 Croatia 74 77 3.99 3.85 4.25 4.00 2010 76.0 67.0 69.0 Czech Republic 86 83 4.69 4.53 4.62 4.32 2007 95.4 81.0 82.4 Denmark 90 91 5.45 5.03 5.45 5.22 2007 92.9 81.9 82.7 Dominican Republic 56 52 3.67 3.97 3.99 4.00 2010 26.9 67.8 68.6 Ecuador 54 53 3.27 3.70 3.61 3.49 2007 26.4 66.3 69.2 Egypt, Arab Rep. 59 60 3.96 3.76 3.22 3.06 2009 55.1 64.6 65.8 El Salvador 55 55 4.39 4.21 4.53 3.86 2010 22.9 54.0 56.0 Eritrea .. .. .. .. .. .. .. .. 60.5 59.1 Ethiopia 7 8 3.44 3.79 4.13 4.18 .. .. 60.0 63.1 Finland 85 87 5.40 5.05 4.70 5.00 2005 89.7 82.0 82.7 France 87 89 5.10 4.47 4.06 4.41 2005 87.3 84.8 85.2 Georgia 41 35 3.75 4.18 4.25 4.67 .. .. 32.8 35.9 Germany 89 88 5.31 4.92 4.35 4.51 2005 86.9 83.9 84.3 Ghana .. 20 .. 4.20 .. 4.08 2012 8.7 58.3 60.7 Greece 55 63 4.28 3.92 3.63 3.56 2005 86.0 71.8 73.3 Guatemala 51 43 3.72 4.29 3.68 4.16 2008 20.3 48.7 50.6 Guinea .. .. .. 3.71 .. 4.49 2005 12.1 60.8 61.3 Haiti .. .. .. 3.03 .. 4.24 2010 8.1 44.2 42.9 Honduras 48 47 3.45 4.10 3.96 3.52 2009 17.3 50.4 53.3 Hungary 85 87 4.42 4.28 4.50 4.27 2008 92.0 75.2 76.1 India 15 17 4.60 4.21 3.90 4.24 2006 10.3 77.2 78.6 Indonesia 34 33 4.69 4.29 4.34 3.87 2010 11.0 80.6 81.6 Iran, Islamic Rep. 52 52 .. 4.00 .. 3.18 2010 40.5 81.2 82.3 Iraq .. .. .. .. .. .. 2009 43.1 .. .. Ireland 78 83 5.48 5.24 4.85 5.00 2005 88.9 84.0 84.4 Israel 85 87 5.08 4.51 4.93 4.61 2008 89.1 77.5 78.6 Italy 71 73 4.30 4.29 3.55 3.72 2005 90.1 72.9 73.1 Japan 81 85 5.21 4.98 5.20 4.89 2005 95.4 88.7 89.4 Jordan .. 83 4.42 4.50 4.04 4.02 2010 52.9 80.9 82.3 Kazakhstan .. 63 4.28 4.24 4.93 4.98 2009 62.5 57.5 60.6 Kenya 33 .. 4.00 4.10 4.19 4.62 2009 .. 65.7 68.3 Korea, Rep. 62 67 4.83 4.75 4.40 4.35 2011 79.9 72.6 73.9 Kyrgyz Republic .. 48 3.50 3.78 4.26 4.36 2008 40.4 58.9 60.4 Lao PDR 10 12 .. .. .. .. 2008 1.4 69.7 71.3 Lebanon .. 62 .. 4.57 .. 4.00 2003 34.5 66.3 67.6 Liberia .. 17 .. 4.54 .. 4.45 .. .. 56.3 55.3 Libya .. .. .. 3.45 .. 3.46 2003 68.5 65.4 67.5 Lithuania 80 84 4.38 4.36 4.43 4.41 2009 82.9 66.9 69.3 Selected indicators 307 TABLE 5 Selected indicators related to risk management at the enterprise sector level (continued) Wage employment Goods market efficiency Labor market efficiency Pension contributors Formal production Wage and salaried workers, % of economy, % of total employed, annual average 1–7 scale 1–7 scale     annual average   1991–2000 2001–10 2006–07 2012–13 2006–07 2012–13 Survey year % of labor force 1999–2003 2004–07 Madagascar 13 14 3.49 3.84 4.33 4.50 2009 5.3 58.7 59.8 Malawi .. .. .. 3.86 .. 4.58 .. .. 57.9 58.6 Malaysia 73 75 5.26 5.16 4.90 4.82 2010 53.5 68.5 69.8 Mali .. 11 3.58 3.87 4.00 3.89 2010 7.9 58.9 59.9 Mauritania .. .. 3.32 3.58 4.06 3.60 2000 13.1 64.2 66.3 Mexico 59 65 4.12 4.20 3.89 4.01 2010 27.8 69.6 70.5 Moldova 65 65 .. 3.98 .. 4.26 2011 71.0 55.2 56.1 Morocco 43 41 3.89 4.27 3.37 3.84 2011 29.1 64.2 66.2 Mozambique .. 9 3.31 3.77 3.98 3.72 2006 1.9 59.7 60.9 Myanmar .. .. .. .. .. .. .. .. 48.9 51.6 Nepal .. 25 3.75 3.78 3.64 3.75 2011 .. 63.1 63.6 Netherlands 88 87 5.34 5.29 4.63 4.99 2005 90.7 86.8 86.9 New Zealand 79 82 5.56 5.35 5.19 5.19 .. .. 87.4 88.0 Nicaragua 58 49 3.46 3.79 3.86 3.98 2008 21.7 54.7 56.4 Niger .. 5 .. .. .. .. 2006 1.9 59.1 60.3 Nigeria .. .. 4.13 4.16 4.11 4.50 2010 .. 42.5 46.0 Norway 91 92 5.04 4.79 4.97 4.98 2005 93.2 80.9 81.7 Pakistan 35 38 4.20 4.02 3.70 3.65 2009 .. 63.2 65.6 Panama 66 66 4.22 4.59 4.01 4.17 .. .. 35.4 38.3 Papua New Guinea .. .. .. .. .. .. 2009 4.4 63.5 63.1 Paraguay 58 47 3.33 4.19 3.47 3.92 2004 12.4 60.7 62.0 Peru 52 58 3.98 4.37 4.03 4.56 2009 21.7 40.4 43.9 Philippines 50 51 4.24 4.17 3.85 4.01 2011 26.3 57.1 60.1 Poland 71 75 4.26 4.39 4.44 4.48 2008 81.4 72.4 73.4 Portugal 72 75 4.49 4.31 4.12 3.80 2005 92.0 77.2 76.9 Romania 62 64 4.04 3.86 4.01 4.01 2008 67.9 66.3 68.9 Russian Federation 93 92 3.84 3.62 4.44 4.23 2011 65.1 54.7 58.1 Rwanda 6 .. .. 4.54 .. 5.10 2004 4.6 59.6 60.5 Saudi Arabia .. .. .. 5.12 .. 4.47 2010 .. 81.3 82.7 Senegal 11 22 .. 4.20 .. 4.27 2008 .. 55.2 57.6 Serbia .. 70 .. 3.57 .. 4.04 2007 45.0 .. .. Sierra Leone .. 8 .. 3.84 .. 3.92 2004 5.5 53.0 56.2 Singapore 85 85 5.79 5.60 5.65 5.80 2009 62.1 86.8 87.5 Slovak Republic 93 88 4.59 4.37 4.73 4.20 2003 78.9 81.3 82.6 Somalia .. .. .. .. .. .. .. .. .. .. South Africa .. 83 4.74 4.68 4.04 3.94 2010 6.7 71.8 73.8 South Sudan .. .. .. .. .. .. .. .. .. .. Spain 76 82 4.67 4.37 4.01 3.98 2005 69.4 77.4 77.6 Sri Lanka 59 57 4.13 4.33 3.28 3.66 2006 24.1 55.5 56.9 Sudan .. .. .. .. .. .. 2005 5.2 65.9 .. Sweden 89 89 5.22 5.14 4.47 4.81 2005 88.8 80.9 81.7 Switzerland 83 84 5.24 5.26 5.58 5.90 2005 95.4 91.3 91.6 Syrian Arab Republic .. 58 .. .. .. .. 2008 26.8 80.8 81.2 Tajikistan .. 53 3.50 4.04 4.12 4.55 .. .. 57.1 58.7 Tanzania 9 9 3.92 3.89 4.33 4.55 2007 .. 42.4 45.1 Thailand 36 43 4.72 4.56 5.02 4.32 2009 22.5 48.0 51.2 Togo .. 11 .. .. .. .. 2009 .. 65.1 65.1 Tunisia 69 67 .. .. .. .. 2011 .. 62.0 63.8 Turkey 43 55 4.47 4.55 3.53 3.79 2008 58.6 67.6 70.1 Turkmenistan .. .. .. .. .. .. .. .. .. .. Uganda .. 18 3.67 3.95 4.72 4.83 2004 10.3 57.0 58.5 Ukraine 88 84 3.75 3.82 4.21 4.44 2010 62.1 48.6 52.3 United Arab Emirates .. 96 4.85 5.31 4.74 5.24 .. .. 73.3 75.4 United Kingdom 87 87 5.48 5.09 5.41 5.42 2005 93.2 87.4 87.7 United States 92 93 5.55 4.88 5.80 5.37 2005 92.2 91.2 91.5 Uruguay 73 71 3.94 4.38 4.10 3.49 2009 78.5 47.8 51.3 Uzbekistan .. .. .. .. .. .. 2005 .. .. .. Venezuela, RB 61 62 3.42 2.78 3.52 2.88 2009 33.9 65.3 67.2 Vietnam 19 22 3.95 4.13 4.43 4.51 2010 20.7 84.5 85.3 West Bank and Gaza 63 61 .. .. .. .. 2009 14.0 .. .. Yemen, Rep. 42 .. .. 3.68 .. 3.44 2006 10.4 72.7 73.2 Zambia 20 18 3.23 4.53 4.02 3.97 2010 .. 51.6 54.5 Zimbabwe 38 38 3.29 3.63 3.50 3.40 2011 20.3 38.6 37.7 World ..w ..w 4.24u 4.25u 4.28u 4.29u 37.9w 83.8w 83.1w Low income .. .. 3.54 3.84 4.11 4.29 .. 58.3 59.9 Middle income .. .. 3.95 4.08 4.03 4.04 27.4 72.0 73.7 Lower middle income 26 26 3.84 4.06 3.92 4.03 15.3 69.6 71.1 Upper middle income .. .. 4.05 4.09 4.12 4.05 37.9 72.8 74.5 Low & middle income .. .. 3.84 4.00 4.05 4.12 25.5 71.7 73.4 High income 85 86 4.98 4.78 4.70 4.68 85.7 86.4 85.7 308 WORLD DEVELOPMENT REPORT 2014 TABLE 6 Selected indicators related to risk management at the financial sector level Financial systems structure, Use of formal financial risk-management tools % of GDP, 2005–10 average Loan Personally Purchased Population Population Use of Saved at a from a paid for agriculture using using electronic financial financial health insurance informal informal payments institution institution insurance % saving credit of agriculture Volume of Stock Mutual % % % workers % % transactions market Bank fund Insurance Pension age 15+ age 15+ age 15+ age 15+ age 15+ age 15+ million capitalization assets assets assets assets   2011 2011 2011 2011 2011 2011 2009           Afghanistan 2.8 7.4 0.1 10.8 11.7 36.8 .. .. 7.5 .. .. .. Albania 8.6 7.5 11.2 73.3 14.2 12.9 7 .. 50.8 0.0 1.4 0.0 Algeria 4.3 1.5 3.5 0.0 16.5 27.9 .. .. 33.4 .. 0.9 .. Angola 15.9 7.9 3.1 .. 20.7 24.9 44 .. 17.9 .. 1.2 .. Argentina 3.8 6.6 9.1 0.0 20.6 9.0 47 24.2 23.4 2.0 2.4 12.7 Armenia 0.8 18.9 0.6 5.9 9.7 37.6 7 1.1 16.4 0.4 0.4 .. Australia 61.9 17.0 .. .. 6.3 27.6 5,761 120.0 117.1 22.4 33.9 83.2 Austria 51.6 8.3 .. .. 26.8 9.3 2,129 35.3 129.5 48.6 34.0 4.8 Azerbaijan 1.6 17.7 1.1 19.5 9.2 37.1 52 .. 15.4 .. 0.7 .. Bangladesh 16.6 23.3 2.1 0.0 10.2 14.0 .. 7.5 53.3 .. 2.1 .. Belarus 6.8 16.1 3.1 4.7 19.6 36.6 .. .. 33.9 .. 0.8 .. Belgium 42.6 10.5 .. .. 15.2 7.4 2,211 68.1 109.7 30.2 65.4 3.5 Benin 7.0 4.2 0.7 1.4 25.3 32.4 .. .. 21.4 .. 1.9 .. Bolivia 17.1 16.6 3.7 4.3 27.2 9.1 4 17.6 36.0 3.0 4.0 23.1 Bosnia and Herzegovina 6.1 13.0 3.8 3.7 7.5 15.8 .. .. 54.6 .. .. .. Brazil 10.3 6.3 7.6 11.2 10.8 17.5 16,509 61.4 75.8 40.8 7.5 14.9 Bulgaria 4.8 7.8 3.9 2.0 6.1 24.0 66 23.0 50.2 0.7 3.6 3.8 Burkina Faso 7.9 3.1 0.8 0.5 30.1 32.9 .. .. 18.0 .. 0.7 .. Burundi 3.3 1.7 3.5 7.6 21.9 50.6 .. .. 23.1 .. .. .. Cambodia 0.8 19.5 2.6 14.3 30.2 40.1 .. .. 17.1 .. 0.5 .. Cameroon 9.9 4.5 1.2 3.4 42.0 45.6 .. .. 11.9 .. 1.7 .. Canada 53.2 20.3 .. .. 12.3 21.6 8,441 120.9 138.3 45.3 34.6 60.3 Central African Republic 2.5 0.9 0.4 1.5 22.9 24.1 .. .. 9.0 .. 1.0 .. Chad 6.8 6.2 1.3 17.0 21.9 34.9 .. .. 5.3 .. 0.1 .. Chile 12.4 7.8 5.7 0.0 14.6 13.9 224 108.4 70.8 10.2 19.3 61.4 China 32.1 7.3 47.2 7.2 6.3 22.1 43,094 81.7 118.0 8.1 9.7 0.7 Hong Kong SAR, China 42.8 7.9 .. .. 16.3 19.8 4,184 474.0 167.4 421.9 32.6 29.8 Colombia 9.2 11.9 5.6 8.3 23.7 22.0 287 40.9 36.4 2.6 4.9 15.1 Congo, Dem. Rep. 1.5 1.5 0.7 1.2 22.6 32.2 0 .. 3.4 .. 0.5 .. Congo, Rep. 5.5 2.8 0.1 2.8 25.0 29.2 .. .. 3.1 .. .. .. Costa Rica 19.9 10.0 4.1 0.0 21.1 11.4 115 6.4 45.2 4.3 1.8 5.7 Côte d’Ivoire .. .. .. .. .. .. .. 26.1 18.8 .. 3.3 .. Croatia 12.2 14.4 .. .. 9.7 23.7 269 50.6 73.6 5.6 7.6 7.4 Czech Republic 35.5 9.5 .. .. 13.5 19.9 175 28.9 59.5 3.6 10.0 5.1 Denmark 56.5 18.8 .. .. 16.1 12.0 1,461 67.3 204.2 56.5 75.4 57.9 Dominican Republic 16.0 13.9 8.4 0.9 21.3 23.5 188 .. 22.3 .. 1.4 2.9 Ecuador 14.5 10.6 3.1 8.4 15.7 16.7 2 8.4 24.2 1.1 0.7 .. Egypt, Arab Rep. 0.7 3.7 0.5 6.6 7.4 26.9 1,292 66.7 71.3 4.1 3.6 2.5 El Salvador 12.9 3.9 1.0 5.4 13.0 7.4 7 23.8 44.0 .. 2.3 20.8 Eritrea .. .. .. .. .. .. 0 .. .. .. .. .. Ethiopia .. .. .. .. .. .. .. .. 26.6 .. 0.8 .. Finland 56.1 23.9 .. .. 12.5 23.8 1,976 90.9 86.2 27.4 26.0 72.0 France 49.5 18.6 .. .. 12.3 5.9 12,970 82.4 119.6 67.9 89.3 1.0 Georgia 1.0 11.0 3.2 0.4 5.9 18.0 32 6.8 24.4 .. 1.6 .. Germany 55.9 12.5 .. .. 11.4 12.7 .. 45.7 131.4 43.6 59.5 12.0 Ghana 16.1 5.8 11.8 4.0 20.5 28.8 .. 12.4 21.7 .. 0.9 .. Greece 19.9 7.9 .. .. 7.9 22.2 129 50.3 107.9 7.2 5.7 .. Guatemala 10.2 13.7 1.7 1.7 14.6 10.5 0 .. 35.2 .. 1.4 .. Guinea 2.0 2.4 0.3 49.5 25.1 42.1 .. .. 7.7 .. .. .. Haiti 18.0 8.3 4.0 17.2 13.6 37.4 .. .. 14.8 .. .. .. Honduras 8.5 7.1 1.3 2.8 13.3 12.9 .. .. 47.3 .. 3.0 .. Hungary 17.3 9.4 .. .. 9.4 9.3 355 26.4 74.0 10.4 8.6 11.0 India 11.6 7.7 6.8 6.6 10.8 22.9 4,102 80.0 59.1 6.8 15.3 5.3 Indonesia 15.3 8.5 0.9 0.0 25.2 40.6 1,787 31.9 31.6 .. 1.8 2.2 Iran, Islamic Rep. 19.7 30.7 19.3 24.3 12.6 40.4 1,658 16.9 33.1 .. 1.0 .. Iraq 5.4 8.0 0.2 7.8 20.8 46.7 16 .. 10.1 .. .. .. Ireland 51.3 15.7 .. .. 13.3 13.6 .. 42.0 200.6 363.4 83.4 45.5 Israel 24.8 16.7 .. .. 20.1 23.3 940 97.0 99.3 .. .. 38.7 Italy 15.5 4.6 .. .. 10.9 6.6 3,252 36.0 121.6 15.9 33.8 2.2 Japan 51.3 6.1 .. .. 12.0 11.6 1,428 87.7 164.1 13.2 74.4 19.9 Jordan 8.3 4.5 1.1 .. 9.9 26.8 385 181.6 98.5 0.1 4.8 42.8 Kazakhstan 6.7 13.1 1.9 1.5 15.2 28.8 164 31.8 43.6 .. 1.3 9.6 Kenya 23.3 9.7 5.4 3.3 16.9 57.7 .. 38.4 38.6 .. 7.6 12.9 Korea, Rep. 46.9 16.6 .. .. 17.6 16.0 9,696 84.2 101.2 27.2 40.5 3.9 Kyrgyz Republic 0.9 11.3 0.0 2.7 35.5 24.2 6 2.0 10.1 .. 0.4 .. Lao PDR 19.4 18.1 4.5 4.6 35.1 14.4 .. .. 10.3 .. .. .. Lebanon 17.1 11.3 7.9 0.0 13.7 15.9 10 32.1 142.3 1.0 7.1 .. Liberia 13.9 6.5 5.9 18.8 20.9 44.7 .. .. 11.2 .. .. .. Libya .. .. .. .. .. .. 0 .. 10.7 .. .. .. Lithuania 20.5 5.6 14.6 31.1 12.2 30.2 217 21.1 58.0 0.7 3.2 2.3 Madagascar 1.4 2.3 0.3 1.2 18.3 61.4 3 .. 12.2 .. 1.8 .. Malawi 8.2 9.2 0.5 1.2 24.8 42.1 .. 19.6 12.7 .. 5.7 .. Malaysia 35.4 11.2 16.4 7.2 15.6 21.3 332 133.5 114.6 23.6 18.5 49.3 Mali 4.5 3.7 1.0 9.2 32.9 25.4 .. .. 18.9 .. 0.5 .. Mauritania 6.4 7.9 2.0 17.0 16.4 35.8 .. .. 27.1 .. .. .. Selected indicators 309 Insurance premium Foreign currency mismatches on Bank savings Credit (life + non-life) bank balance sheets Loan-to-deposit ratio Loan Deposit dollarization dollarization % of GDP % of GDP % of GDP % % % 2000 2011 2000 2011 2000 2011 2000 2010 2000 2010 2000 2011 Afghanistan .. .. .. .. .. .. .. .. .. .. .. .. Albania 43.8 66.7 4.7 39.0 0.4 0.6 53.2 72.6 29.1 43.1 10.7 58.4 Algeria 26.1 45.7 5.9 14.5 0.5 0.6 .. .. .. .. 22.8 31.8 Angola 25.6 52.7 2.0 21.6 0.9 1.0 74.8 50.2 83.0 61.0 7.7 41.0 Argentina 27.4 20.6 23.2 16.0 2.4 2.4 66.1 12.5 44.1 .. 84.5 77.9 Armenia 8.9 20.4 9.9 34.3 0.1 0.5 .. 65.2 .. 67.6 111.3 168.1 Australia 61.3 99.2 84.6 124.1 8.8 5.3 .. .. .. .. 138.0 125.1 Austria 80.4 95.0 102.2 118.9 4.7 4.6 .. .. .. .. 127.1 125.2 Azerbaijan 10.6 13.4 5.9 17.9 0.3 0.3 .. .. .. .. 55.9 133.9 Bangladesh 29.8 54.2 24.3 48.6 0.5 1.1 .. .. .. .. 81.7 89.6 Belarus 14.9 35.9 8.8 41.7 0.6 0.8 54.9 30.7 72.2 53.8 58.8 116.2 Belgium 82.3 106.1 77.8 92.6 7.4 7.6 .. .. .. .. 94.6 87.2 Benin 15.6 29.7 11.6 25.3 0.6 0.8 .. .. .. .. 74.3 85.3 Bolivia 46.1 44.7 58.7 36.8 0.8 1.0 96.3 69.2 92.6 52.0 127.2 82.3 Bosnia and Herzegovina 16.6 41.6 40.8 48.2 1.9 1.9 .. 73.7 53.7 50.2 73.2 115.7 Brazil 44.3 66.5 31.7 58.0 1.4 2.4 18.0 .. 6.1 .. 71.5 87.2 Bulgaria 22.7 63.9 12.3 72.0 1.3 2.0 35.9 58.3 59.2 53.6 54.0 112.7 Burkina Faso 13.1 26.8 11.7 20.8 0.6 0.4 .. .. .. .. 89.4 77.6 Burundi 14.5 19.0 20.8 19.2 0.4 0.2 .. .. .. .. 143.9 100.7 Cambodia 18.7 32.3 6.4 28.8 .. 0.1 97.0 98.1 93.2 96.6 34.2 89.2 Cameroon 11.1 18.2 8.2 12.7 0.7 0.9 .. .. .. .. 73.8 70.1 Canada 70.1 .. 76.1 .. 5.4 1.8 .. .. .. .. 108.5 .. Central African Republic 3.4 9.3 4.7 9.4 0.3 0.3 .. .. .. .. 138.6 101.7 Chad 3.9 7.2 3.4 6.0 0.3 0.2 .. .. .. .. 86.7 82.5 Chile 50.9 43.4 63.8 70.5 3.7 3.6 15.7 10.8 10.4 13.8 125.4 162.2 China 110.4 164.4 112.2 127.4 1.5 3.0 .. .. .. .. 101.7 77.5 Hong Kong SAR, China 220.8 321.8 152.6 206.2 4.3 11.4 .. .. .. .. 69.1 64.1 Colombia 22.4 20.9 20.8 35.1 1.6 2.1 .. 4.2 .. .. 92.7 168.0 Congo, Dem. Rep. 9.2 12.4 3.2 6.3 0.1 0.4 .. .. 43.5 .. 35.2 50.8 Congo, Rep. 8.6 21.5 4.8 7.3 0.1 0.4 .. .. .. .. 55.8 33.8 Costa Rica 14.1 23.2 24.0 47.4 2.0 1.8 41.6 .. 41.3 .. 170.9 198.6 Côte d’Ivoire 13.3 .. 14.9 .. 1.2 1.4 .. .. .. .. .. 68.0 Croatia 37.0 67.4 31.9 73.8 2.2 2.5 84.6 73.0 71.1 65.5 86.1 109.5 Czech Republic 58.1 64.1 47.0 55.8 3.0 3.7 21.6 13.4 15.3 8.9 80.9 87.0 Denmark 46.8 52.1 135.2 .. 6.6 9.5 .. 25.4 .. .. 66.3 .. Dominican Republic 25.5 20.8 29.0 22.3 1.4 1.2 .. 21.8 .. .. 113.7 107.0 Ecuador 21.1 32.9 29.3 30.9 1.2 2.2 .. .. .. .. 138.9 93.8 Egypt, Arab Rep. 64.7 63.1 52.0 31.3 0.6 0.7 23.1 31.2 .. .. 80.3 49.5 El Salvador 42.5 38.6 45.1 39.0 1.4 1.8 .. .. .. .. 106.2 101.0 Eritrea 131.0 89.7 29.9 13.6 1.2 0.4 .. .. .. .. 22.8 15.2 Ethiopia 27.0 .. 17.8 .. 0.6 0.5 .. .. .. .. 65.7 .. Finland 46.0 63.6 53.0 96.4 4.9 3.2 .. .. .. .. 115.2 151.5 France 62.6 88.3 85.0 116.2 8.4 8.6 .. .. .. .. 135.9 131.6 Georgia 5.2 23.4 7.4 32.8 0.2 0.5 81.4 74.3 77.9 71.0 143.1 140.5 Germany 91.5 117.2 119.4 105.5 5.1 5.3 .. .. .. .. 130.5 90.0 Ghana 16.4 24.2 13.8 14.5 0.9 0.6 .. .. .. 30.7 84.3 60.0 Greece 50.5 83.1 46.9 118.0 0.0 2.1 .. .. .. .. 92.9 142.0 Guatemala 18.2 38.5 19.8 23.4 0.9 1.0 18.4 28.5 .. 24.1 108.5 60.8 Guinea 4.8 23.7 3.4 9.1 0.1 0.0 .. .. .. .. 70.3 38.5 Haiti 29.9 39.2 15.1 14.2 .. 0.4 .. .. .. .. 50.4 36.2 Honduras 35.2 46.4 34.1 48.0 1.5 1.6 23.2 26.1 29.0 30.1 96.8 103.6 Hungary 38.8 45.7 32.5 65.2 2.9 2.8 33.1 65.5 20.4 18.4 83.5 142.4 India 45.8 66.7 28.8 50.6 1.8 3.5 .. .. .. .. 63.0 75.8 Indonesia 48.2 34.4 19.4 28.3 1.0 1.5 41.4 12.9 26.6 14.6 40.3 82.4 Iran, Islamic Rep. 31.6 14.5 19.5 13.7 0.6 1.1 2.6 14.9 1.1 8.3 61.7 94.9 Iraq .. .. .. .. .. .. .. .. .. .. .. .. Ireland 77.6 103.1 104.6 204.3 9.7 8.0 .. .. .. .. 134.8 198.2 Israel 79.0 94.9 77.2 94.8 4.5 4.3 34.9 14.8 31.6 27.2 97.7 99.9 Italy 49.9 86.2 74.8 122.1 5.3 6.6 .. .. .. .. 149.9 141.7 Japan 228.4 222.1 190.8 105.0 6.9 1.6 .. .. .. .. 83.5 47.3 Jordan 90.5 100.3 71.9 73.4 1.5 1.6 .. .. 95.5 70.5 79.5 73.3 Kazakhstan 11.1 29.1 11.2 37.0 0.3 0.6 68.5 45.0 50.9 43.7 100.9 127.0 Kenya 29.8 45.9 25.8 38.0 2.0 2.6 .. .. .. .. 86.5 82.6 Korea, Rep. 65.5 74.7 75.8 99.8 10.8 11.9 1.2 2.7 .. .. 115.7 133.6 Kyrgyz Republic 5.0 .. 4.1 .. 0.1 0.2 68.6 55.8 57.9 52.2 81.3 .. Lao PDR 16.0 32.7 7.9 23.3 0.2 0.4 72.3 .. 75.4 61.5 49.2 71.1 Lebanon 182.9 223.9 85.5 81.1 2.2 1.8 87.0 81.3 62.3 60.0 46.7 36.2 Liberia 1.6 11.1 0.8 5.7 .. .. .. .. .. .. 51.4 51.7 Libya 36.9 83.1 22.7 17.6 0.5 0.3 .. .. .. .. 61.4 21.2 Lithuania 16.9 38.0 13.1 53.7 0.8 1.5 67.7 72.4 45.3 32.4 77.3 141.2 Madagascar 14.0 17.4 8.8 10.9 0.6 0.6 .. .. .. .. 62.8 62.7 Malawi 13.8 28.6 5.6 18.4 1.7 1.0 .. .. .. .. 40.5 64.1 Malaysia 112.9 128.9 126.7 112.1 4.0 4.3 .. .. 2.0 4.7 112.3 87.0 Mali 13.7 21.9 15.0 21.4 0.5 0.5 .. .. .. .. 109.2 98.1 Mauritania .. 23.9 .. 25.6 .. 0.4 .. .. .. .. .. 107.4 310 WORLD DEVELOPMENT REPORT 2014 TABLE 6 Selected indicators related to risk management at the financial sector level (continued) Financial systems structure, Use of formal financial risk-management tools % of GDP, 2005–10 average Loan Personally Purchased Population Population Use of Saved at a from a paid for agriculture using using electronic financial financial health insurance informal informal payments institution institution insurance % saving credit of agriculture Volume of Stock Mutual % % % workers % % transactions market Bank fund Insurance Pension age 15+ age 15+ age 15+ age 15+ age 15+ age 15+ million capitalization assets assets assets assets   2011 2011 2011 2011 2011 2011 2009           Mexico 6.7 7.6 8.5 4.9 20.4 22.1 1,640 31.5 29.8 7.3 3.2 8.1 Moldova 3.5 6.4 1.6 2.6 18.7 41.5 27 .. 33.4 .. 2.2 .. Morocco 12.2 4.3 4.5 18.7 18.2 41.6 28 68.6 75.7 22.0 17.1 20.0 Mozambique 17.5 5.9 3.7 5.3 25.2 35.4 4 .. 23.8 .. 2.7 .. Myanmar .. .. .. .. .. .. .. .. 3.5 .. .. .. Nepal 9.9 10.8 1.8 2.2 8.5 43.4 0 29.2 45.6 .. 0.6 .. Netherlands 57.8 12.6 .. .. 15.2 7.2 4,807 87.4 194.7 12.9 61.5 119.9 New Zealand 60.4 26.6 .. .. 12.2 23.7 .. 36.8 138.2 10.9 .. 11.6 Nicaragua 6.5 7.6 0.8 5.9 19.5 6.7 .. .. 37.5 .. 0.5 .. Niger 1.2 1.3 0.2 5.8 24.0 46.8 .. .. 10.0 .. 0.8 .. Nigeria 23.6 2.1 0.4 2.3 40.8 46.2 .. 24.7 30.2 .. 1.7 2.9 Norway .. .. .. .. .. .. 1,835 61.3 79.8 16.2 36.6 7.2 Pakistan 1.4 1.6 0.5 2.6 6.0 27.3 .. 29.3 37.6 1.8 .. .. Panama 12.5 9.8 4.9 0.0 22.5 17.4 .. 30.5 81.0 .. 5.2 2.7 Papua New Guinea .. .. .. .. .. .. .. 117.6 32.3 .. .. .. Paraguay 9.7 12.9 5.9 1.7 8.4 16.6 .. 3.5 23.2 .. 1.2 .. Peru 8.6 12.7 3.5 11.1 20.6 16.4 103 56.5 22.1 3.2 3.5 15.5 Philippines 14.7 10.5 5.5 0.0 30.8 47.6 .. 47.5 40.1 1.1 6.5 3.5 Poland 18.0 9.6 .. .. 12.8 16.7 2,027 33.2 43.6 7.6 10.3 11.9 Portugal 25.6 8.3 .. .. 8.2 6.9 1,457 41.2 169.1 13.9 33.0 12.4 Romania 8.7 8.4 5.9 6.5 9.5 18.3 278 18.8 33.9 0.4 2.8 0.4 Russian Federation 10.9 7.7 6.7 3.7 11.8 24.2 2,833 68.7 37.6 0.3 2.2 1.3 Rwanda 17.8 8.4 5.3 4.3 12.6 30.1 1 .. 11.7 .. .. .. Saudi Arabia 17.2 2.1 .. .. 16.1 34.8 159 104.6 52.5 6.9 1.7 .. Senegal 3.7 3.5 0.9 3.4 11.6 27.5 .. .. 25.8 .. 2.2 .. Serbia 3.2 12.3 3.7 2.2 11.7 31.4 398 30.6 37.4 .. 3.2 0.2 Sierra Leone 14.5 6.1 0.5 17.3 18.1 43.0 0 .. 11.4 .. .. .. Singapore 58.4 10.0 .. .. 2.5 22.7 289 186.0 114.6 48.8 48.2 57.2 Slovak Republic 36.8 11.4 .. .. 12.4 23.7 415 6.5 55.3 4.8 5.3 3.5 Somalia 13.6 1.6 0.5 20.1 8.2 34.7 .. .. .. .. .. .. South Africa 22.1 8.9 7.4 43.3 9.4 35.3 970 233.8 80.4 31.9 36.8 104.6 South Sudan .. .. .. .. .. .. .. .. .. .. .. .. Spain 35.0 11.4 .. .. 11.2 15.3 3,564 90.0 192.0 24.1 23.8 8.0 Sri Lanka 28.1 17.7 7.5 8.1 8.2 16.4 29 20.6 36.5 0.2 3.9 0.6 Sudan 3.4 1.8 9.5 20.7 19.3 63.3 10 .. 13.3 .. 0.5 .. Sweden 63.6 23.4 .. .. 19.3 18.9 2,846 107.4 127.0 38.2 82.7 3.4 Switzerland .. .. .. .. .. .. 1,268 241.3 178.6 37.3 86.1 99.7 Syrian Arab Republic 5.1 13.1 9.5 2.6 43.2 49.3 .. .. 36.1 .. .. .. Tajikistan 0.3 4.8 0.8 9.0 13.5 26.9 .. .. 15.0 .. .. .. Tanzania 11.9 6.6 2.6 7.4 28.2 45.1 4 4.0 19.3 .. 1.2 .. Thailand 42.8 19.4 24.1 7.4 17.2 7.8 558 62.8 106.1 17.9 11.4 5.3 Togo 3.6 3.8 0.7 2.2 16.0 20.4 .. .. 23.4 .. 1.9 .. Tunisia 5.0 3.2 6.1 0.0 20.1 23.2 .. 14.0 59.7 7.0 0.8 .. Turkey 4.2 4.6 4.4 0.0 5.4 61.3 1,910 30.7 50.4 3.2 1.8 0.6 Turkmenistan 0.1 0.8 0.0 2.1 44.4 38.5 .. .. .. .. .. .. Uganda 16.3 8.9 0.7 9.1 28.1 43.9 29 9.6 16.4 .. 1.1 .. Ukraine 5.4 8.1 1.7 0.0 19.6 36.3 .. 29.8 55.0 .. 2.8 0.1 United Arab Emirates 19.2 10.8 .. .. 11.0 24.6 94 .. .. 0.3 4.2 .. United Kingdom 43.8 11.8 .. .. 12.9 17.0 13,486 123.9 183.3 36.3 95.9 76.2 United States 50.4 20.1 .. .. 16.4 24.5 79,011 122.7 65.3 76.6 44.2 70.6 Uruguay 5.7 14.8 9.3 22.1 11.2 9.5 51 0.5 28.2 0.0 3.8 13.3 Uzbekistan 0.8 1.5 0.8 4.3 30.6 12.3 .. .. .. .. 0.5 .. Venezuela, RB 13.6 1.7 6.0 41.2 14.8 11.7 27 2.9 19.5 .. .. .. Vietnam 7.7 16.2 17.5 3.1 27.5 27.8 .. 13.2 90.1 0.2 4.2 .. West Bank and Gaza 5.5 4.1 5.1 9.8 10.7 49.6 2 68.4 9.5 .. .. .. Yemen, Rep. 1.1 0.9 0.0 0.0 10.8 56.2 8 .. 13.1 .. .. .. Zambia 11.8 6.1 1.2 11.8 20.4 41.8 2 13.5 16.0 .. 1.4 3.8 Zimbabwe 17.3 4.9 14.6 6.2 22.6 57.7 1 .. .. .. .. .. World 22.4w 9.1w 17.1w 6.5w 13.5w 24.7w 249,527s 40.8m 42.1m 10.7m 4.3m 7.8m Low income 11.5 11.4 2.2 5.1 18.5 32.7 49 17.6 15.2 .. 1.1 12.9 Middle income 18.1 7.6 19.1 6.7 12.9 25.9 77,239 28.6 35.5 3.6 3.1 5.2 Lower middle income 11.1 7.3 5.2 5.1 16.4 29.3 7,349 22.5 29.2 2.6 2.8 3.5 Upper middle income 24.9 7.9 32.6 8.3 9.6 22.7 69,890 32.5 47.6 3.9 3.4 6.4 Low & middle income 17.5 8.0 17.4 6.6 13.5 26.6 77,288 27.9 29.8 3.6 2.8 5.5 High income 40.4 13.0 .. .. 13.5 18.1 172,239 69.9 99.1 17.9 25.7 11.6 Selected indicators 311 Insurance premium Foreign currency mismatches on Bank savings Credit (life + non-life) bank balance sheets Loan-to-deposit ratio Loan Deposit dollarization dollarization % of GDP % of GDP % of GDP % % % 2000 2011 2000 2011 2000 2011 2000 2010 2000 2010 2000 2011 Mexico 23.5 26.7 17.2 19.8 1.8 1.6 .. 9.7 6.4 8.6 72.9 74.4 Moldova 13.1 36.6 12.6 33.6 0.5 1.1 40.8 46.3 42.4 48.0 96.4 91.7 Morocco 59.2 89.5 50.7 71.6 2.3 2.6 1.0 .. .. .. 85.7 80.0 Mozambique 23.3 33.7 16.7 23.6 0.5 1.1 40.2 32.4 46.7 35.8 71.7 70.2 Myanmar 18.0 17.9 9.5 7.9 0.1 .. .. .. .. .. 53.0 44.0 Nepal 38.8 63.9 30.3 52.4 0.5 1.5 .. .. .. .. 78.1 82.0 Netherlands 92.3 131.9 134.1 197.9 7.2 4.9 .. .. .. .. 145.3 150.1 New Zealand 78.7 94.3 110.0 147.7 3.8 2.7 .. .. .. .. 139.8 156.7 Nicaragua 34.4 38.1 30.5 32.1 1.2 1.5 83.0 88.7 70.3 .. 88.6 84.1 Niger 5.9 11.8 5.2 14.0 0.5 0.6 .. .. .. .. 87.2 119.4 Nigeria 14.9 30.6 11.7 21.3 0.5 0.7 .. .. 5.4 .. 78.1 69.5 Norway 45.9 .. 65.5 .. 4.3 4.8 12.1 12.6 16.2 25.4 142.6 .. Pakistan 27.7 29.1 22.3 18.2 0.5 0.6 .. .. .. .. 80.7 62.6 Panama 76.7 82.7 95.5 84.6 2.7 2.9 .. .. .. .. 124.4 102.3 Papua New Guinea 29.0 41.8 17.1 22.4 1.5 0.1 .. .. .. .. 59.1 53.5 Paraguay 22.2 27.3 27.1 41.0 1.1 1.1 49.0 42.0 59.7 39.6 122.4 150.4 Peru 25.8 29.5 25.7 26.3 0.9 1.3 81.4 52.3 77.2 52.6 99.4 89.1 Philippines 50.7 52.2 36.8 31.8 1.2 1.4 27.1 18.4 .. 22.6 72.5 60.9 Poland 36.0 50.5 26.6 54.9 2.5 3.3 21.9 30.2 15.7 8.8 73.8 108.7 Portugal 89.3 133.5 126.1 192.1 4.7 5.9 .. .. .. .. 141.2 143.9 Romania 19.7 31.4 7.1 38.0 0.8 1.3 59.5 60.1 47.0 34.8 36.2 121.1 Russian Federation 15.5 40.8 13.3 45.0 1.7 0.9 37.8 24.1 46.2 29.8 85.6 110.2 Rwanda 13.0 .. 10.2 .. 0.4 0.5 .. .. .. .. 78.6 .. Saudi Arabia 37.4 51.7 24.4 39.7 0.2 0.5 25.4 13.6 18.6 11.9 65.2 76.8 Senegal 18.4 32.0 18.6 29.5 1.0 1.2 .. .. .. .. 101.1 92.3 Serbia 14.1 43.2 49.1 51.0 2.1 1.5 .. 84.1 76.7 73.1 .. 118.1 Sierra Leone 9.1 18.9 2.0 8.9 .. 0.5 .. .. .. .. 22.3 47.0 Singapore 98.2 128.1 97.8 112.6 6.2 6.1 .. 42.9 .. .. 99.7 87.9 Slovak Republic 66.9 54.8 51.1 49.7 2.8 3.0 19.0 18.1 17.2 17.1 76.3 90.7 Somalia .. .. .. .. .. .. .. .. .. .. .. .. South Africa 51.5 62.6 69.1 68.9 17.5 11.2 49.8 .. 4.2 .. 134.2 110.1 South Sudan .. .. .. .. .. .. .. .. .. .. .. .. Spain 78.9 150.9 97.8 205.4 6.0 4.9 .. .. .. .. 123.9 136.1 Sri Lanka 33.5 34.4 28.8 30.6 1.2 1.2 .. .. .. .. 86.1 89.2 Sudan 5.7 15.9 2.1 11.2 0.3 0.4 .. .. .. .. 37.5 70.6 Sweden 36.2 59.8 42.3 .. 6.9 7.5 14.1 23.0 26.7 26.1 116.8 .. Switzerland 113.8 151.6 154.7 168.4 9.7 7.7 15.7 15.1 17.1 33.4 135.9 111.1 Syrian Arab Republic 38.3 55.0 8.5 23.2 0.4 0.6 .. .. .. .. 22.1 42.1 Tajikistan 3.2 .. 13.6 .. 0.1 0.4 48.4 29.8 .. 72.2 .. .. Tanzania 13.8 28.8 4.6 17.7 0.6 0.7 .. .. .. .. 33.2 61.5 Thailand 106.6 103.8 108.3 108.6 2.5 4.2 9.2 .. .. 0.9 101.6 104.5 Togo 16.0 38.3 16.0 31.1 1.0 1.5 .. .. .. .. 99.9 81.0 Tunisia 46.8 55.1 60.5 72.2 1.4 1.6 .. .. .. .. 129.3 131.1 Turkey 32.5 50.3 17.8 50.1 0.9 1.1 62.3 31.7 48.2 33.5 54.6 99.6 Turkmenistan .. .. .. .. .. 0.3 90.3 .. 30.1 .. 21.9 .. Uganda 12.1 18.1 5.6 15.2 0.4 0.6 .. .. .. .. 46.3 83.8 Ukraine 11.0 37.1 11.1 55.8 1.1 2.0 46.0 60.3 38.4 .. 100.4 150.7 United Arab Emirates 45.1 59.3 46.2 61.9 1.2 1.4 .. .. .. 18.9 102.5 104.5 United Kingdom 103.9 162.6 128.6 186.7 17.0 11.7 36.5 40.1 55.9 52.9 123.7 114.9 United States 65.1 79.2 50.4 55.1 7.1 6.9 .. .. .. .. 77.5 69.6 Uruguay 40.6 40.3 44.9 23.7 1.7 1.5 86.3 75.8 90.7 .. 110.6 58.7 Uzbekistan .. .. .. .. 0.2 0.3 .. .. 8.1 .. .. .. Venezuela, RB 15.2 32.1 10.5 20.4 1.2 2.0 .. .. 0.2 .. 69.0 63.6 Vietnam 32.8 94.8 35.3 111.6 0.6 1.3 20.7 17.8 .. .. 107.6 117.8 West Bank and Gaza .. .. .. .. .. .. .. .. .. .. .. .. Yemen, Rep. 13.9 15.3 4.3 4.1 0.2 0.2 41.9 45.0 52.7 39.3 30.9 27.0 Zambia 21.0 20.4 8.2 12.3 1.4 1.1 48.3 34.7 53.7 .. 38.9 60.1 Zimbabwe .. .. .. .. .. .. .. .. .. .. .. .. World 34.5m 46.2m 28.8m 39.0m 1.3m 1.9m         80.7m 87.2m Low income 13.8 26.3 9.5 15.2 0.5 0.6 71.0 70.2 Middle income 31.6 41.2 22.7 34.7 1.0 1.3 77.4 85.6 Lower middle income 28.3 36.8 17.9 30.8 0.7 0.9 75.3 81.2 Upper middle income 32.5 51.5 25.7 47.8 1.5 1.6 78.7 89.4 Low & middle income 24.5 37.2 17.5 30.8 0.8 1.2 74.4 82.4 High income 62.6 79.2 63.8 90.0 4.3 3.9 97.7 106.6 312 WORLD DEVELOPMENT REPORT 2014 TABLE 7 Selected indicators related to risk management at the macroeconomy level Worldwide Governance CPI inflation rate Government primary surplus Gross public debt International reserves Flexible exchange rate regimes Indicators average Average annual % of GDP, % of GDP, % of GDP, Coarse classification, (–2.5 to 2.5) growth, % annual average annual average annual average 1–6 scale (least to most)   2010–12 2005–07 2010–12 2005–07 2010–12 2005–07 2010–12 1996–2000 2001–05 2006–10 2011 Afghanistan 6.2 –1.3 0.0 .. .. .. 33.8 .. 6 3 –1.75 Albania 2.7 –0.5 –0.4 56.0 59.0 19.1 19.9 4 3 3 –0.20 Algeria 6.7 11.1 –2.0 22.5 10.7 70.7 97.9 2 2 2 –0.93 Angola 11.9 .. .. 29.3 32.8 16.8 27.0 .. 1 2 –1.06 Argentina 9.8 3.7 0.1 77.0 46.3 16.0 11.0 1 3 2 –0.22 Armenia 5.1 –1.8 –2.2 17.0 36.3 16.2 19.1 2 2 3 –0.28 Australia 2.6 1.6 –3.6 10.2 23.9 5.6 3.4 4 4 4 1.63 Austria 2.9 0.7 –1.1 62.2 72.7 4.2 5.0 1 1 1 1.49 Azerbaijan 4.4 .. .. 10.7 11.0 11.3 15.0 1 1 1 –0.84 Bangladesh 9.7 –1.2 –1.4 .. .. 6.2 9.9 2 2 1 –0.87 Belarus 56.2 0.8 1.6 12.6 40.8 5.8 10.2 5 1 3 –1.01 Belgium 3.2 3.1 –0.6 88.0 97.6 3.4 5.1 1 1 1 1.37 Benin 4.7 0.0 –0.4 24.8 31.3 18.8 13.3 1 1 1 –0.29 Bolivia 7.2 .. .. 58.7 35.4 29.1 47.6 2 2 1 –0.54 Bosnia and Herzegovina 2.9 .. .. 21.8 41.3 28.6 24.9 1 1 1 –0.43 Brazil 6.0 3.5 2.6 67.0 66.2 9.1 14.7 3 3 3 0.13 Bulgaria 3.6 3.9 –1.8 23.8 16.3 35.7 34.8 1 1 1 0.18 Burkina Faso 3.3 1.9 –2.9 29.6 28.0 10.9 10.2 1 1 1 –0.38 Burundi 13.8 –0.2 –2.3 131.9 35.9 10.8 13.8 3 2 2 –1.19 Cambodia 4.2 –0.2 –3.1 33.1 28.7 20.9 32.0 .. 2 2 –0.78 Cameroon 2.9 .. .. 26.5 13.6 9.9 14.1 1 1 1 –0.89 Canada 2.2 2.4 –3.6 69.5 84.0 2.9 3.7 2 3 3 1.62 Central African Republic 1.3 3.0 –0.5 94.2 31.8 8.4 7.9 1 1 1 –1.30 Chad 2.4 2.3 –0.5 29.1 34.5 9.5 9.0 1 1 1 –1.30 Chile 3.2 6.8 0.7 5.3 10.3 12.0 15.0 3 3 3 1.21 China 4.0 .. .. 17.8 27.3 40.3 44.7 1 1 1 –0.58 Hong Kong SAR, China 4.7 4.3 2.7 32.5 33.1 69.8 117.6 1 1 1 1.40 Colombia 3.3 1.8 0.0 36.0 35.0 9.9 9.7 3 3 3 –0.23 Congo, Dem. Rep. .. 1.4 2.7 141.0 35.8 1.8 9.1 4 4 .. –1.64 Congo, Rep. 2.6 17.5 13.2 101.7 22.5 20.7 38.9 1 1 1 –1.01 Costa Rica 4.7 2.8 –2.4 32.8 31.6 13.7 13.2 2 2 1 0.58 Côte d’Ivoire 3.1 0.4 –1.8 82.1 70.1 10.5 16.6 .. .. .. –1.16 Croatia 2.8 –0.7 –2.6 35.5 48.7 21.9 24.6 2 2 2 0.38 Czech Republic 2.6 –1.4 –3.1 28.2 40.6 21.1 20.9 1 3 3 0.95 Denmark 2.6 5.7 –2.5 32.5 46.4 11.9 25.8 1 1 1 1.86 Dominican Republic 6.1 0.8 –1.9 21.0 30.9 6.5 7.0 2 3 2 –0.36 Ecuador 4.8 .. .. 30.3 19.8 5.5 3.0 1 1 1 –0.76 Egypt, Arab Rep. 8.6 –3.7 –4.6 91.3 76.7 24.4 9.8 1 2 2 –0.74 El Salvador 3.4 .. .. 38.6 50.6 11.0 12.1 1 1 1 –0.07 Eritrea .. .. .. 154.8 134.2 2.4 4.9 .. 1 1 –1.40 Ethiopia 28.2 .. .. 50.6 25.0 6.9 .. 2 2 .. –0.96 Finland 3.1 3.6 –2.0 38.8 50.3 4.3 3.8 1 1 1 1.85 France 2.0 –0.2 –3.2 65.0 86.2 4.1 4.9 1 1 1 1.21 Georgia 3.7 .. .. 27.6 35.2 11.0 19.0 3 2 2 0.01 Germany 2.0 0.9 0.5 67.3 81.7 3.9 5.0 1 1 1 1.42 Ghana 8.9 .. .. 35.1 48.7 12.6 14.8 5 2 2 0.14 Greece 2.4 –1.4 –2.8 105.3 159.0 1.1 1.7 1 1 1 0.36 Guatemala 5.0 –0.3 –1.3 21.7 24.6 13.3 13.3 2 2 2 –0.57 Guinea 18.2 1.5 –4.3 126.6 78.2 1.4 2.5 3 2 2 –1.19 Haiti 7.3 .. .. 40.3 15.1 5.3 17.6 4 3 3 –1.16 Honduras 6.0 –2.2 –3.5 40.6 32.2 23.0 15.8 2 1 1 –0.55 Hungary 4.8 –3.6 3.0 64.9 80.7 17.9 35.2 2 3 3 0.74 India 9.1 –1.4 –4.2 78.4 67.2 19.2 16.1 2 2 2 –0.30 Indonesia 4.8 .. .. 40.1 25.1 12.3 13.0 3 3 3 –0.46 Iran, Islamic Rep. 23.9 4.2 1.7 18.7 13.7 .. .. 3 1 1 –1.16 Iraq .. 7.3 2.0 162.8 42.0 33.1 34.6 .. 1 2 –1.34 Ireland 2.1 2.3 –14.5 25.6 105.3 0.4 0.8 1 1 1 1.45 Israel 2.6 2.9 –0.8 85.8 74.9 19.4 31.7 2 3 3 0.59 Italy 2.9 2.1 1.1 105.1 122.4 4.1 6.0 1 1 1 0.52 Japan –0.2 –3.5 –9.0 185.1 228.1 20.5 20.8 4 4 4 1.17 Jordan 4.6 –1.9 –4.6 78.1 72.5 45.4 39.8 1 1 1 –0.12 Kazakhstan 6.7 5.6 3.9 6.9 11.2 17.6 15.2 2 2 2 –0.59 Kenya 11.7 –0.2 –3.0 47.9 48.9 10.9 13.8 2 2 3 –0.69 Korea, Rep. 3.1 1.5 1.0 30.1 33.8 25.0 28.3 3 3 3 0.76 Kyrgyz Republic 9.4 .. .. 71.7 53.1 28.2 31.6 2 2 2 –0.83 Lao PDR 5.9 –2.3 –2.9 73.7 56.2 13.8 14.9 6 .. .. –0.91 Lebanon 5.8 .. .. 176.6 139.6 81.3 108.6 1 1 1 –0.64 Liberia 7.7 2.9 –3.6 634.6 29.3 10.9 32.5 4 4 4 –0.74 Libya 10.7 29.5 4.9 0.4 0.0 107.1 .. 6 6 6 –1.34 Lithuania 3.6 –0.1 –3.4 17.7 38.7 17.9 18.9 1 2 2 0.69 Selected indicators 313 TABLE 7 Selected indicators related to risk management at the macroeconomy level (continued) Worldwide Governance CPI inflation rate Government primary surplus Gross public debt International reserves Flexible exchange rate regimes Indicators average Average annual % of GDP, % of GDP, % of GDP, Coarse classification, (–2.5 to 2.5) growth, % annual average annual average annual average 1–6 scale (least to most)   2010–12 2005–07 2010–12 2005–07 2010–12 2005–07 2010–12 1996–2000 2001–05 2006–10 2011 Madagascar 7.9 .. .. 50.9 37.3 10.6 12.7 3 3 3 –0.71 Malawi 14.2 .. .. 71.0 44.8 5.6 5.0 3 2 2 –0.33 Malaysia 2.4 –1.7 –2.8 41.8 54.5 50.9 45.0 1 1 3 0.32 Mali 4.1 8.8 –1.9 31.5 31.2 15.7 13.4 1 1 1 –0.49 Mauritania 5.3 0.2 1.3 121.9 89.9 5.3 14.1 2 2 3 –0.88 Mexico 3.8 .. .. 38.7 43.4 8.4 12.7 3 3 3 –0.13 Moldova 6.2 .. .. 30.0 24.5 24.4 30.7 2 2 2 –0.30 Morocco 1.1 0.4 –3.9 59.5 55.1 30.8 21.2 2 1 1 –0.33 Mozambique 10.4 .. .. 58.8 47.0 17.6 21.3 1 3 3 –0.30 Myanmar 5.0 –3.1 –4.4 84.3 51.0 .. .. 6 6 6 –1.65 Nepal 9.5 0.6 0.0 48.1 34.0 20.1 18.7 2 1 1 –0.89 Netherlands 2.4 1.8 –3.3 48.2 66.8 3.4 5.0 1 1 1 1.71 New Zealand 2.6 3.8 –4.2 19.4 35.8 11.2 12.0 3 3 3 1.83 Nicaragua 7.6 1.5 0.9 85.9 57.0 13.3 19.5 2 2 2 –0.61 Niger 1.7 .. .. 39.6 27.6 10.5 13.6 1 1 1 –0.58 Nigeria 11.5 9.5 –0.3 17.7 16.8 28.7 16.1 3 2 2 –1.15 Norway 1.0 14.5 10.6 54.3 39.2 15.9 11.0 3 3 3 1.70 Pakistan 10.8 –0.5 –2.9 59.4 61.4 10.4 7.5 2 2 2 –1.14 Panama 5.8 4.0 –0.1 54.2 39.4 8.5 8.1 1 1 1 0.08 Papua New Guinea 8.4 8.1 2.4 .. .. 24.8 31.1 2 2 2 –0.69 Paraguay 5.9 .. .. 25.3 12.3 16.2 19.3 3 3 1 –0.60 Peru 3.5 3.3 2.2 33.7 22.1 20.9 29.4 2 2 2 –0.18 Philippines 3.9 3.9 1.4 51.8 42.4 19.8 31.3 2 2 3 –0.49 Poland 4.0 –0.6 –2.7 46.6 55.5 14.6 20.0 3 3 3 0.83 Portugal 3.2 –2.0 –2.8 64.8 108.1 5.1 6.3 1 1 1 0.93 Romania 4.6 –1.0 –2.9 14.3 34.1 23.4 26.3 5 3 3 0.15 Russian Federation 6.7 8.2 –0.2 10.6 11.2 30.5 27.3 2 2 3 –0.74 Rwanda 6.0 –1.2 –0.7 41.4 25.0 14.9 15.5 .. 2 2 –0.21 Saudi Arabia 4.7 20.9 10.4 26.8 5.8 64.8 99.2 1 1 1 –0.47 Senegal 2.4 –3.2 –4.4 30.3 40.2 14.2 14.7 1 1 1 –0.39 Serbia 9.2 .. .. 44.4 53.4 33.4 36.1 .. 3 3 –0.12 Sierra Leone 14.5 8.2 –2.4 93.4 49.2 10.1 14.5 .. 4 2 –0.64 Singapore 4.9 7.6 5.2 88.5 105.2 97.9 100.1 3 3 3 1.47 Slovak Republic 3.8 –1.3 –4.5 31.3 45.5 22.4 2.0 2 2 1 0.79 Somalia .. .. .. .. .. .. .. .. 6 6 –2.30 South Africa 5.4 3.8 –1.8 31.9 39.2 9.9 11.9 4 4 4 0.25 South Sudan 47.3 .. .. .. .. .. .. .. .. .. –1.48 Spain 2.8 3.1 –7.9 39.7 71.5 1.5 2.7 1 1 1 0.94 Sri Lanka 6.8 –1.9 –1.5 .. .. 10.7 12.2 2 2 2 –0.29 Sudan .. .. .. 80.2 80.6 4.9 0.7 2 2 2 –1.60 Sweden 1.9 2.2 –1.0 45.3 38.6 6.8 9.5 3 3 3 1.80 Switzerland –0.2 1.5 0.8 62.7 49.0 15.9 58.1 3 3 3 1.71 Syrian Arab Republic 19.7 .. .. 46.8 29.4 52.1 34.9 3 .. .. –1.10 Tajikistan 9.1 .. .. 37.2 34.8 5.9 6.5 5 2 2 –1.10 Tanzania 14.3 .. .. 42.4 39.7 15.8 15.7 2 2 2 –0.36 Thailand 3.4 2.4 –0.2 42.6 42.9 32.4 50.7 3 3 3 –0.29 Togo 3.1 –1.0 –2.9 93.8 47.5 14.5 18.4 1 1 1 –0.89 Tunisia 4.6 0.0 –1.3 49.1 43.0 18.3 19.0 2 2 2 –0.18 Turkey 7.7 4.5 1.7 46.4 39.3 11.5 11.9 3 4 3 –0.01 Turkmenistan .. .. .. 3.7 11.2 .. .. 6 6 6 –1.41 Uganda 16.3 .. –3.3 57.8 31.2 18.2 15.7 3 2 2 –0.59 Ukraine 4.2 –1.2 –2.6 14.9 38.2 22.0 19.2 1 1 1 –0.58 United Arab Emirates 0.9 16.5 8.5 7.1 19.2 18.0 10.7 .. 1 1 0.48 United Kingdom 3.7 –1.5 –6.2 42.8 85.1 2.0 3.7 3 3 3 1.34 United States 2.6 –0.7 –7.9 66.6 102.4 1.7 2.6 4 4 4 1.23 Uruguay 8.1 3.8 1.3 71.1 56.5 17.0 23.2 2 3 3 0.84 Uzbekistan .. 4.1 6.2 21.8 9.2 .. .. .. 3 3 –1.29 Venezuela, RB 23.6 2.0 –17.2 36.4 40.8 18.4 6.3 2 1 1 –1.28 Vietnam 13.8 –0.2 –2.4 42.5 52.3 24.1 11.3 .. 2 2 –0.54 West Bank and Gaza .. .. .. .. .. .. .. 1 1 1 –0.76 Yemen, Rep. 16.8 –0.4 –0.6 41.7 43.6 37.5 16.6 .. 2 2 –1.33 Zambia 6.4 .. .. 25.3 25.9 8.0 13.3 5 3 4 –0.30 Zimbabwe .. .. .. 58.0 67.8 2.4 7.3 1 5 1 –1.47 World 1.5m –1.8m 42.6m 41.3m Low income 0.3 –1.9 58.0 35.9 Middle income 0.8 –1.7 41.8 40.8 Lower middle income –0.3 –2.1 42.5 38.9 Upper middle income 2.6 –0.9 38.7 41.3 Low & middle income 0.8 –1.8 46.4 39.7 High income 2.1 –1.6 34.8 48.6 314 WORLD DEVELOPMENT REPORT 2014 TABLE 8 Natural disasters and climate change indicators Consequences of natural hazards (droughts, earthquakes, floods, storms) Average annual deaths Total damages Average annual damages CO2 emissions Total deaths per million population $ millions % of GDP Per capita, metric tons 1993–2002 2003–12 1993–2002 2003–12 1993–2002 2003–12 1993–2002 2003–12 2009 Afghanistan 8,787 1,908 45.1 7.1 20 167 .. 0.16 0.2 Albania 11 5 0.3 0.2 18 0 0.06 0.00 1.0 Algeria 1,292 2,694 4.2 7.5 362 6,179 0.08 0.36 3.3 Angola 169 372 1.3 2.0 10 0 0.02 0.00 1.4 Argentina 86 109 0.2 0.3 3,307 1,198 0.11 0.04 4.4 Armenia 4 1 0.1 0.0 141 0 0.75 0.00 1.5 Australia 82 80 0.4 0.4 7,974 18,907 0.20 0.18 18.4 Austria 20 9 0.3 0.1 2,640 2,300 0.12 0.06 7.4 Azerbaijan 49 3 0.6 0.0 156 55 0.35 0.01 5.5 Bangladesh 5,891 7,772 4.6 5.3 6,342 5,384 1.44 0.68 0.3 Belarus 7 0 0.1 0.0 137 10 0.09 0.00 6.3 Belgium 16 11 0.2 0.1 183 947 0.01 0.02 9.6 Benin 34 85 0.5 0.9 3 0 0.01 0.00 0.5 Bolivia 323 305 4.0 3.1 291 746 0.34 0.45 1.4 Bosnia and Herzegovina 0 7 0.0 0.2 158 227 0.38 0.12 7.8 Brazil 401 2,142 0.2 1.1 262 7,619 0.00 0.05 1.9 Bulgaria 3 56 0.0 0.7 1 461 0.00 0.09 5.6 Burkina Faso 28 101 0.3 0.7 0 150 0.00 0.18 0.1 Burundi 6 180 0.1 2.1 0 0 0.00 0.00 0.0 Cambodia 1,029 311 8.8 2.2 315 592 1.01 0.57 0.3 Cameroon 63 57 0.4 0.3 0 0 0.00 0.00 0.3 Canada 77 26 0.3 0.1 2,549 3,186 0.04 0.02 15.2 Central African Republic 10 13 0.3 0.3 0 0 0.00 0.00 0.1 Chad 102 131 1.3 1.2 1 10 0.01 0.01 0.0 Chile 221 666 1.5 4.0 505 30,231 0.06 1.68 3.9 China 22,136 100,853 1.8 7.6 120,564 205,511 1.18 0.45 5.8 Hong Kong SAR, China 91 7 1.4 0.1 249 0 0.01 0.00 5.3 Colombia 1,935 1,635 5.0 3.6 1,863 3,442 0.19 0.14 1.6 Congo, Dem. Rep. 331 110 0.7 0.2 9 0 0.01 0.00 0.0 Congo, Rep. 2 32 0.1 0.8 0 0 0.00 0.00 0.5 Costa Rica 101 119 2.7 2.6 577 370 0.41 0.12 1.8 Côte d’Ivoire 28 24 0.2 0.1 0 0 0.00 0.00 0.4 Croatia 0 2 0.0 0.0 0 410 0.00 0.06 4.9 Czech Republic 56 39 0.5 0.4 4,438 698 0.69 0.03 10.3 Denmark 8 4 0.2 0.1 2,605 1,400 0.15 0.04 8.3 Dominican Republic 421 954 5.0 9.8 1,982 587 0.94 0.13 2.1 Ecuador 401 108 3.3 0.7 305 1,155 0.11 0.19 2.0 Egypt, Arab Rep. 669 71 1.0 0.1 142 0 0.02 0.00 2.8 El Salvador 1,669 479 28.3 7.8 2,432 2,342 2.02 1.09 1.0 Eritrea 3 0 0.1 0.0 5 0 0.07 0.00 0.1 Ethiopia 543 1,367 0.9 1.7 22 9 0.03 0.00 0.1 Finland .. .. .. .. .. .. .. .. 10.0 France 247 137 0.4 0.2 15,939 10,760 0.11 0.04 5.6 Georgia 14 13 0.3 0.3 582 96 1.61 0.08 1.3 Germany 112 70 0.1 0.1 19,619 10,202 0.09 0.03 9.0 Ghana 213 179 1.2 0.8 34 0 0.04 0.00 0.3 Greece 236 10 2.2 0.1 5,670 946 0.42 0.03 8.4 Guatemala 532 1,856 5.0 13.6 774 1,910 0.40 0.49 1.1 Guinea 25 10 0.3 0.1 0 0 0.00 0.00 0.1 Haiti 1,499 229,306 18.1 2,379.1 231 8,356 0.62 13.04 0.2 Honduras 15,270 333 255.3 4.5 4,061 340 7.81 0.24 1.0 Hungary 49 17 0.5 0.2 393 598 0.08 0.04 4.9 India 60,760 30,870 6.0 2.6 20,325 22,273 0.47 0.18 1.7 Indonesia 3,048 178,093 1.5 76.0 1,674 11,981 0.18 0.23 1.9 Iran, Islamic Rep. 4,786 28,093 7.5 38.7 5,726 1,183 0.56 0.03 8.2 Iraq 0 36 0.0 0.1 0 1 0.00 0.00 3.6 Ireland 16 2 0.4 0.0 239 325 0.03 0.01 9.3 Israel 18 2 0.3 0.0 118 0 0.01 0.00 9.0 Italy 214 395 0.4 0.7 25,552 22,586 0.21 0.10 6.7 Japan 5,885 20,781 4.7 16.3 123,726 280,055 0.32 0.58 8.6 Jordan 16 0 0.3 0.0 1 0 0.00 0.00 3.8 Kazakhstan 122 50 0.8 0.3 41 239 0.02 0.02 14.0 Kenya 300 874 1.0 2.3 12 201 0.01 0.07 0.3 Korea, Rep. 1,163 358 2.5 0.7 7,614 6,414 0.22 0.07 10.4 Kyrgyz Republic 1 81 0.0 1.5 4 3 0.03 0.01 1.2 Lao PDR 81 80 1.6 1.3 304 100 2.38 0.18 0.3 Lebanon 0 0 0.0 0.0 0 0 0.00 0.00 4.9 Liberia 10 4 0.4 0.1 0 0 0.00 0.00 0.1 Libya 0 0 0.0 0.0 42 0 0.02 0.00 10.5 Lithuania 8 4 0.2 0.1 5 256 0.00 0.05 3.8 Madagascar 571 1,101 3.9 5.5 69 805 0.19 0.86 0.1 Malawi 573 46 5.4 0.3 8 0 0.04 0.00 0.1 Malaysia 328 172 1.5 0.6 54 1,500 0.01 0.06 7.1 Mali 24 45 0.2 0.3 0 0 0.00 0.00 0.0 Mauritania 26 20 1.0 0.6 0 0 0.00 0.00 0.6 Selected indicators 315 TABLE 8 Natural disasters and climate change indicators (continued) Consequences of natural hazards (droughts, earthquakes, floods, storms) Average annual deaths Total damages Average annual damages CO2 emissions Total deaths per million population $ millions % of GDP Per capita, metric tons 1993–2002 2003–12 1993–2002 2003–12 1993–2002 2003–12 1993–2002 2003–12 2009 Mexico 2,261 542 2.2 0.5 8,960 20,186 0.21 0.18 3.8 Moldova 62 4 1.7 0.1 386 414 2.36 0.68 1.3 Morocco 991 791 3.5 2.6 1,166 429 0.29 0.05 1.6 Mozambique 1,232 237 7.1 1.0 483 194 1.12 0.20 0.1 Myanmar 230 139,047 0.5 271.7 10 4,564 .. .. 0.2 Nepal 2,620 1,254 11.8 4.8 237 63 0.49 0.05 0.1 Netherlands 8 7 0.1 0.0 2,065 578 0.05 0.01 10.3 New Zealand 4 191 0.1 4.5 177 24,842 0.03 1.91 7.4 Nicaragua 3,489 332 70.7 5.9 1,020 0 2.20 0.00 0.8 Niger 84 140 0.8 0.9 0 3 0.00 0.01 0.1 Nigeria 438 935 0.4 0.6 76 538 0.02 0.03 0.5 Norway 1 4 0.0 0.1 303 130 0.02 0.00 9.7 Pakistan 4,173 79,205 3.0 47.4 596 22,117 0.10 1.35 0.9 Panama 11 83 0.4 2.3 9 17 0.01 0.01 2.2 Papua New Guinea 2,359 180 46.2 2.7 162 27 0.43 0.03 0.5 Paraguay 100 9 1.9 0.1 7 7 0.01 0.00 0.7 Peru 1,055 958 4.2 3.3 362 600 0.06 0.05 1.6 Philippines 5,142 10,834 6.9 12.0 2,089 4,847 0.29 0.28 0.7 Poland 104 52 0.3 0.1 4,211 3,336 0.24 0.06 7.8 Portugal 50 48 0.5 0.5 48 2,958 0.00 0.12 5.4 Romania 143 212 0.6 1.0 1,062 1,313 0.25 0.06 3.7 Russian Federation 2,706 270 1.8 0.2 2,503 3,977 0.09 0.02 11.1 Rwanda 126 85 1.8 0.8 0 0 0.00 0.00 0.1 Saudi Arabia 19 295 0.1 1.1 0 1,200 0.00 0.03 16.2 Senegal 215 46 2.3 0.4 41 10 0.08 0.01 0.4 Serbia 12 2 0.2 0.0 0 0 0.00 0.00 6.3 Sierra Leone 25 154 0.6 2.8 0 0 0.00 0.00 0.3 Singapore 0 0 0.0 0.0 0 0 0.00 0.00 6.4 Slovak Republic 57 9 1.1 0.2 227 408 0.08 0.04 6.3 Somalia 2,485 482 36.0 5.3 0 100 .. .. 0.1 South Africa 500 163 1.2 0.3 401 673 0.03 0.02 10.1 South Sudan .. .. .. .. .. .. .. .. 0.0 Spain 109 46 0.3 0.1 8,661 3,249 0.14 0.02 6.3 Sri Lanka 42 35,931 0.2 177.7 4 2,012 0.00 0.49 0.6 Sudan 369 330 1.4 1.0 42 484 0.04 0.09 0.3 Sweden 4 7 0.0 0.1 160 2,800 0.01 0.06 4.7 Switzerland 18 8 0.3 0.1 2,348 3,569 0.08 0.07 5.4 Syrian Arab Republic 27 11 0.2 0.1 0 0 0.00 0.00 3.1 Tajikistan 113 148 1.9 2.0 132 311 1.00 0.60 0.4 Tanzania 289 171 0.9 0.4 4 0 0.00 0.00 0.2 Thailand 1,244 10,008 2.0 15.1 2,929 42,573 0.26 1.56 4.1 Togo 3 69 0.1 1.2 0 0 0.00 0.00 0.2 Tunisia .. .. .. .. .. .. .. .. 2.4 Turkey 18,515 1,111 30.2 1.6 23,019 2,746 0.85 0.04 3.9 Turkmenistan 11 0 0.3 0.0 100 0 0.38 0.00 9.7 Uganda 319 160 1.4 0.5 73 0 0.11 0.00 0.1 Ukraine 57 46 0.1 0.1 426 2,816 0.10 0.16 5.9 United Arab Emirates .. .. .. .. .. .. .. .. 20.3 United Kingdom 108 58 0.2 0.1 11,439 13,938 0.08 0.05 7.7 United States 2,317 4,242 0.8 1.4 128,109 474,999 0.15 0.33 17.3 Uruguay 5 21 0.2 0.6 280 45 0.11 0.01 2.3 Uzbekistan 0 13 0.0 0.0 50 0 0.03 0.00 4.2 Venezuela, RB 30,233 212 128.7 0.8 3,249 330 0.36 0.01 6.5 Vietnam 7,852 2,692 10.4 3.2 3,250 5,605 1.19 0.62 1.7 West Bank and Gaza .. .. .. .. .. .. .. .. 0.6 Yemen, Rep. 562 278 3.4 1.3 1,212 0 1.92 0.00 1.1 Zambia 5 55 0.1 0.4 21 0 0.06 0.00 0.2 Zimbabwe 119 37 1.0 0.3 127 0 0.20 0.00 0.7 World 240,551s 911,500s 4.1w 13.7w 656,227s 1,340,427s 0.21w 0.22w 4.7wa Low income 27,921 386,806 4.6 50.5 31,429 21,249 0.56 0.51 0.3 Middle income 195,846 495,936 4.8 10.6 222,706 389,427 0.44 0.26 3.3 Lower middle income 109,266 344,583 5.5 14.6 41,440 80,026 0.39 0.24 1.6 Upper middle income 86,580 151,353 4.1 6.6 181,266 309,401 0.46 0.26 5.1 Low & middle income 223,767 882,742 4.7 16.3 254,135 410,675 0.45 0.26 2.9 High income 16,784 28,758 1.4 2.3 402,093 929,752 0.16 0.21 11.2 Selected island states 7,237 241,643 7.4 205.9 5,859 17,287 0.48 0.61 1.3 a. Includes emissions not allocated to specific countries. 316 WORLD DEVELOPMENT REPORT 2014 TABLE 9 Global temperature anomalies: Difference relative to 1951–80 Temperature means (degrees Celsius) Dec–Feb Mar–May Jun–Aug Sep–Nov Jan–Dec 5 year average 1951 –0.32 –0.09 0.01 0.04 –0.06 –0.05 1952 0.12 –0.01 0.03 –0.04 0.02 –0.05 1953 0.10 0.14 0.05 0.03 0.09 –0.04 1954 –0.08 –0.16 –0.15 0.00 –0.12 –0.06 1955 –0.05 –0.23 –0.03 –0.13 –0.12 –0.06 1956 –0.23 –0.23 –0.17 –0.17 –0.18 –0.07 1957 –0.09 –0.01 0.12 0.06 0.04 –0.04 1958 0.25 0.05 –0.07 –0.01 0.04 –0.02 1959 0.04 0.11 0.02 –0.08 0.03 0.02 1960 0.03 –0.21 –0.03 –0.00 –0.04 0.02 1961 0.13 0.11 0.04 0.02 0.05 0.03 1962 0.02 0.04 0.01 0.02 0.04 –0.02 1963 0.06 –0.09 0.14 0.18 0.07 –0.03 1964 –0.06 –0.26 –0.12 –0.25 –0.20 –0.05 1965 –0.17 –0.13 –0.10 –0.09 –0.10 –0.06 1966 –0.07 –0.05 –0.00 –0.04 –0.04 –0.08 1967 –0.09 0.06 –0.02 –0.01 –0.01 –0.03 1968 –0.11 0.05 –0.07 –0.03 –0.05 0.00 1969 –0.13 0.10 0.02 0.12 0.06 –0.00 1970 0.22 0.05 –0.02 0.06 0.04 0.00 1971 –0.10 –0.09 –0.08 –0.01 –0.07 0.04 1972 –0.15 0.01 0.10 0.05 0.02 0.02 1973 0.26 0.26 0.11 0.09 0.16 0.01 1974 –0.15 –0.05 0.01 –0.08 –0.07 –0.00 1975 0.03 0.11 –0.06 –0.10 –0.01 0.02 1976 –0.10 –0.18 –0.14 –0.15 –0.12 –0.00 1977 0.10 0.24 0.21 0.05 0.15 0.04 1978 0.07 0.11 –0.05 0.07 0.05 0.09 1979 0.02 0.08 0.05 0.21 0.12 0.17 1980 0.35 0.28 0.20 0.18 0.23 0.16 1981 0.35 0.31 0.29 0.13 0.29 0.20 1982 0.18 0.03 0.06 0.09 0.09 0.20 1983 0.41 0.33 0.19 0.25 0.27 0.17 1984 0.17 0.21 0.09 0.09 0.12 0.14 1985 0.02 0.10 0.07 0.08 0.08 0.18 1986 0.23 0.23 0.09 0.04 0.15 0.20 1987 0.27 0.18 0.32 0.27 0.29 0.22 1988 0.44 0.41 0.36 0.26 0.36 0.29 1989 0.23 0.24 0.24 0.25 0.24 0.33 1990 0.34 0.54 0.33 0.34 0.40 0.31 1991 0.41 0.38 0.45 0.32 0.38 0.28 1992 0.37 0.31 0.13 0.01 0.19 0.29 1993 0.29 0.26 0.18 0.10 0.21 0.30 1994 0.14 0.30 0.28 0.38 0.29 0.29 1995 0.54 0.38 0.45 0.39 0.43 0.34 1996 0.33 0.31 0.36 0.28 0.33 0.43 1997 0.36 0.42 0.41 0.59 0.46 0.45 1998 0.68 0.64 0.70 0.45 0.61 0.44 1999 0.56 0.32 0.37 0.39 0.40 0.48 2000 0.41 0.51 0.42 0.34 0.41 0.51 2001 0.38 0.55 0.53 0.56 0.53 0.51 2002 0.66 0.70 0.56 0.59 0.62 0.54 2003 0.57 0.56 0.55 0.63 0.60 0.59 2004 0.65 0.54 0.36 0.60 0.52 0.60 2005 0.58 0.65 0.62 0.74 0.66 0.60 2006 0.62 0.49 0.59 0.65 0.59 0.58 2007 0.78 0.68 0.57 0.58 0.63 0.59 2008 0.34 0.55 0.46 0.61 0.49 0.59 2009 0.52 0.55 0.63 0.64 0.59 0.59 2010 0.66 0.80 0.59 0.65 0.67 0.57 2011 0.45 0.56 0.64 0.54 0.55 2012 0.42 0.60 0.56 0.70 0.57 Dec–Feb Mar–May Jun–Aug Sep–Nov 10 year average 1951–1960 –0.02 –0.06 –0.02 –0.03 –0.03 1961–1970 –0.02 –0.01 –0.01 –0.00 –0.01 1971–1980 0.03 0.08 0.03 0.03 0.05 1981–1990 0.26 0.26 0.20 0.18 0.23 1991–2000 0.41 0.38 0.37 0.33 0.37 2001–2010 0.58 0.61 0.55 0.63 0.59 Selected indicators 317 TABLE 10 Aid commitments Emergency response Reconstruction relief Prevention and preparedness Total $ millions % of total $ millions % of total $ millions % of total $ millions 1981 417.3 80.3 101.3 19.5 1.1 0.2 519.6 1982 698.4 71.3 242.0 24.7 39.1 4.0 979.5 1983 615.0 54.4 466.2 41.2 49.9 4.4 1,131.1 1984 1,226.2 65.3 596.6 31.8 53.6 2.9 1,876.4 1985 813.7 44.0 863.3 46.7 172.8 9.3 1,849.9 1986 838.1 62.7 497.5 37.3 0.0 0.0 1,335.6 1987 1,134.7 66.6 562.0 33.0 7.0 0.4 1,703.7 1988 535.2 52.4 461.2 45.1 25.9 2.5 1,022.3 1989 638.0 43.6 817.4 55.9 6.8 0.5 1,462.2 1990 930.0 58.1 579.5 36.2 91.6 5.7 1,601.1 1991 1,316.0 89.6 142.6 9.7 9.4 0.6 1,468.0 1992 1,252.4 76.5 377.1 23.0 8.3 0.5 1,637.7 1993 1,031.3 68.7 456.8 30.4 12.5 0.8 1,500.7 1994 1,253.9 71.0 511.3 29.0 0.5 0.0 1,765.7 1995 1,804.5 83.6 339.5 15.7 13.4 0.6 2,157.4 1996 2,818.5 93.7 186.5 6.2 3.5 0.1 3,008.6 1997 2,329.6 83.9 220.4 7.9 227.8 8.2 2,777.7 1998 3,107.5 72.4 1,143.5 26.7 38.1 0.9 4,289.2 1999 5,254.9 78.4 1,393.0 20.8 52.1 0.8 6,699.9 2000 2,403.7 70.0 975.4 28.4 56.5 1.6 3,435.6 2001 2,819.9 52.8 2,473.8 46.3 44.1 0.8 5,337.8 2002 3,871.9 69.6 1,609.1 28.9 84.5 1.5 5,565.5 2003 3,244.5 74.6 878.3 20.2 228.4 5.2 4,351.2 2004 4,097.1 83.1 789.7 16.0 43.0 0.9 4,929.8 2005 4,489.6 55.9 2,964.2 36.9 579.0 7.2 8,032.7 2006 3,039.4 65.5 1,374.0 29.6 227.8 4.9 4,641.3 2007 3,472.9 58.5 2,173.4 36.6 290.8 4.9 5,937.1 2008 3,227.9 74.9 573.7 13.3 509.5 11.8 4,311.1 2009 3,698.7 81.2 484.3 10.6 374.4 8.2 4,557.5 2010 225.9 37.2 276.4 45.5 105.6 17.4 607.9 Total 62,606.6 69.2 24,530.0 27.1 3,357.0 3.7 90,493.6 1981–1990 7,846.5 58.2 5,187.1 38.5 447.7 3.3 13,481.3 1991–2000 22,572.2 78.5 5,746.1 20.0 422.1 1.5 28,740.4 2001–2010 32,187.8 66.7 13,596.9 28.2 2,487.2 5.2 48,271.9 318 WORLD DEVELOPMENT REPORT 2014 Technical notes Population density: Population density is the total population (as defined above) divided by land Data consistency and reliability area, in square kilometers. Land area is a country’s Considerable effort has been made to standardize the total area, excluding area under inland water bodies, data, but full comparability cannot be assured, and national claims to continental shelves, and exclusive care must be taken in interpreting the indicators. economic zones. In most cases the definition of in- Many factors affect data availability, comparability, land water bodies includes major rivers and lakes. and reliability: statistical systems in many developing Data sources: Food and Agriculture Organization, economies are still weak; statistical methods, cover- at http://faostat.fao.org; population data from World age, practices, and definitions differ widely; and cross Development Indicators, at http://data.worldbank. country and intertemporal comparisons involve org/indicator/SP.POP.TOTL. complex technical and conceptual problems that Population age composition, ages 0–14: Popula- cannot be resolved unequivocally. Data coverage may tion age composition is the population between the not be complete because of special circumstances af- ages 0 to 14 as a percentage of the total population fecting the collection and reporting of data, such as (as defined above). Data source: United Nations problems stemming from conflicts. For these reasons, Population Division, World Population Prospects, at although data are drawn from sources thought to be http://esa.un.org/wpp. the most authoritative, they should be construed Gross national income: Gross national income only as indicating trends and characterizing major (GNI; formerly gross national product or GNP) differences among economies, rather than as offering is the sum of value added by all resident produc- precise quantitative measures of those differences. ers, plus any product taxes (less subsidies) not in- cluded in the valuation of output, plus net receipts of primary income (compensation of employees and Table 1: Key indicators of development property income) from abroad. Data sources: World Population: Total population is based on the de facto Bank national accounts data, at http://data.world definition of population, which counts all residents bank.org; Organisation for Economic Co-operation regardless of legal status or citizenship—except for and Development (OECD) National Accounts data, refugees not permanently settled in the country of at http://stats.oecd.org/. asylum, who are generally considered part of the GNI per capita: GNI per capita is gross national population of their country of origin. The values income (as defined above) divided by the total shown are midyear estimates. Data sources: United population. Data source: World Bank, World De- Nations Population Division, World Population velopment Indicators, at http://data.worldbank.org/ Prospects; United Nations Statistical Division, Pop- data-catalog/world-development-indicators. ulation and Vital Statistics Report (various years); Gross national income, PPP: PPP GNI (formerly census reports and other statistical publications from PPP GNP) is gross national income (as defined national statistical offices; Eurostat: Demographic above) converted to current international dollars Statistics; Secretariat of the Pacific Community: Sta- using purchasing power parity (PPP) rates. An in- tistics and Demography Programme; U.S. Census ternational dollar has the same purchasing power Bureau: International Database. over gross national income as a U.S. dollar has in the Average annual population growth rate: Annual United States. Data sources: World Bank national population growth rate for year t is the exponen- accounts data, at http://data.worldbank.org; OECD tial growth rate of the total population (as defined National Accounts data, at http://stats.oecd.org. above) from year t - 1 to t, expressed as a percent- PPP GNI per capita: PPP GNI per capita is GNI age. Data sources: United Nations Population Divi- per capita (as defined above), based on purchasing sion, World Population Prospects; United Nations power parity (PPP). Data source: World Bank, In- Statistical Division, Population and Vital Statistics ternational Comparison Program database, at http:// Report (various years); census reports and other www.worldbank.org/data/icp. statistical publications from national statistical of- Gross domestic product per capita growth: fices; Eurostat: Demographic Statistics; Secretariat Gross domestic product (GDP) per capita growth of the Pacific Community: Statistics and Demogra- is the annual percentage growth rate of GDP per phy Programme; U.S. Census Bureau: International capita, based on constant local currency. (Aggregates Database. are based on constant 2005 U.S. dollars.) GDP per Selected indicators 319 capita is gross domestic product divided by the to- Incidence of natural hazards (droughts, earth- tal population. GDP (at purchaser’s prices) is the quakes, floods, storms): The incidence of natural sum of gross value added by all resident producers hazards is the number of droughts, earthquakes, in the economy, plus any product taxes and minus floods, and storms (as defined by the EM-DAT data- any subsidies not included in the value of the prod- base) that occurred in the specified time period. Di- ucts. It is calculated without making deductions for sasters that affect multiple countries are considered depreciation of fabricated assets, or for depletion to be separate events for the purpose of calculat- and degradation of natural resources. Data sources: ing summary groups. Data source: EM-DAT: The World Bank national accounts data, at http://data OFDA/CRED International Disaster Database, at .worldbank.org; OECD National Accounts data, at http://www.emdat.be. http://stats.oecd.org/. Incidence of epidemics: The incidence of epi- Life expectancy at birth: Life expectancy at birth demics is the number of epidemics that were recorded indicates the number of years a newborn infant would in the specified time period. An epidemic is either an be expected to live if prevailing patterns of mortality unusual increase in the number of cases of an infec- at the time of its birth were to stay the same through- tious disease that already exists in a region or popula- out its life. Data sources: United Nations Population tion, or the appearance of an infection previously Division, World Population Prospects; United Na- absent from a region. Disasters that affect multiple tions Statistical Division, Population and Vital Statis- countries are considered to be separate events for tics Report (various years); census reports and other the purpose of calculating summary groups. Data statistical publications from national statistical offices; source: EM-DAT: The OFDA/CRED International Eurostat: Demographic Statistics; Secretariat of the Disaster Database, at http://www.emdat.be. Pacific Community: Statistics and Demography Pro- Adult mortality rate (per 1,000 people): The gramme; U.S. Census Bureau: International Database. adult mortality rate is the probability of dying be- Adult literacy rate: The adult literacy rate is the tween the ages 15 and 60, if subject to the age specific percentage of people aged 15 and above who can— mortality rates of that year between those ages. Data with understanding—read and write a short, simple sources: United Nations Population Division, World statement on their everyday life. Data source: United Population Prospects, at http://esa.un.org/wpp; Uni- Nations Educational, Scientific, and Cultural Or- versity of California, Berkeley and Max Planck Insti- ganization Institute for Statistics, at http://www.uis tute for Demographic Research, Human Mortality .unesco.org. Database, at http://www.mortality.org. Homicide rate (per 100,000 people): Intentional homicides are estimates of unlawful homicides pur- Table 2: Key indicators of development for posely inflicted as a result of domestic disputes, in- other economies terpersonal violence, violent conflicts over land re- See indicator descriptions for table 1. sources, intergang violence over turf or control, and predatory violence and killing by armed groups. In- tentional homicide does not include all intentional Table 3: Selected risk indicators killing; the difference is usually in the organization Large recessions, years in recession: Using real GDP of the killing. Individuals or small groups usually per capita from the World Development Indicators, commit homicide, whereas killing in armed con- large recessions are identified by following Barro and flict is usually committed by fairly cohesive groups Ursúa 2012 and using as a threshold a 5% decline in of up to several hundred members and is thus usu- GDP per capita growth from peak to trough. The ally excluded. Data source: United Nations Office on constructed variable is a dummy variable (1 if a coun- Drugs and Crime’s International Homicide Statis- try was in a large recession, and 0 otherwise). Data tics, at http://www.unodc.org/unodc/en/data-and- sources: Barro, Robert J., and José F. Ursúa, 2012, analysis/homicide.html. “Rare Macroeconomic Disasters,” Annual Review Poverty headcount ratio ($2.50 a day and $10 a of Economics 4 (1): 83–109), with data available at day, PPP; % population): The poverty headcount http://scholar.harvard.edu/barro/publications/barro- ratio is the percentage of the population living on ursua-macroeconomic-data; World Bank national less than $2.50 a day and $10 a day, respectively, at accounts data, at http://data.worldbank.org; OECD 2005 international prices. Data source: World Bank, National Accounts data, at http://stats.oecd.org/. PovcalNet (an online tool for poverty measurement), 320 WORLD DEVELOPMENT REPORT 2014 at http://iresearch.worldbank.org/PovcalNet/index Assessment (PISA) is an internationally comparable .htm. student assessment, coordinated by the OECD. The Volatility of household consumption growth assessment has evaluated the knowledge and skills per capita: Historical volatility is calculated by look- of 15 year olds by testing reading, mathematical, and ing at past changes in household consumption per scientific literacy every three years since 1997. Mean capita. Household consumption growth per capita performance by subject refers to the average score is approximated using natural logarithm differences. of all students’ scores in that country. Data source: The standard deviation of percentage changes in OECD, at http://www.oecd.org/pisa. household consumption per capita (observed vola- Under-five mortality rate (per 1,000 live births): tility) is calculated within decades with at least nine The under-five mortality rate is the probability that observations. Data sources: World Bank national a newborn baby will die before reaching age five, if accounts data, at http://data.worldbank.org; OECD subject to the age specific mortality rates of that year. National Accounts data, at http://stats.oecd.org. Data source: Level and Trends in Child Mortality, Volatility of GDP growth per capita: Historical UN Inter-agency Group for Child Mortality Estima- volatility is calculated by looking at past changes in tion (United Nations Children’s Fund (UNICEF), GDP per capita. GDP per capita growth is approxi- World Health Organization (WHO), World Bank, mated using natural logarithm differences. The stan- and United Nations Population Division), at http:// dard deviation of percentage changes in GDP per www.childmortality.org. capita (observed volatility) is calculated within de- Maternal mortality ratio (per 100,000 live cades with at least nine observations. Data sources: births): The maternal mortality ratio is number World Bank national accounts data, at http://data of women who die from pregnancy related causes .worldbank.org; OECD National Accounts data, at while pregnant or within 42 days of pregnancy ter- http://stats.oecd.org. mination. Data source: Trends in Maternal Mor- Risk preparation index: The risk preparation tality: 1990–2010, WHO, UNICEF, United Nations index is a composite index that estimates prepara- Population Fund, and the World Bank, at http:// tion for risk across countries. The components of www.who.int/reproductivehealth/publications/ the index are average years of schooling; immuniza- monitoring/9789241503631/en. tion rate (measles); proportion of households with Access to social insurance: Access to social insur- less than $1,000 in net assets; access to finance index; ance is the percentage of the population aged 60 and contributors to a pension scheme (% of labor force); over who are beneficiaries of social insurance. Data proportion of respondents stating that “in general, sources: World Bank, Pensions (database), at http:// people can be trusted”; access to improved sanita- go.worldbank.org/8KO0DUVDS0; United Nations tion facilities (% of population); and gross public Population Division, World Population Prospects at debt (% of revenues). The index follows the meth- http://esa.un.org/wpp. odology suggested in Foa, R, 2013, “Household Risk Savings, % of people who saved in the past year: Preparation Indices: Construction and Diagnostics,” The percentage of people who saved in the past year Background Paper for the World Development Report denotes the percentage of respondents (aged 15+) 2014. Data source: calculations by WDR staff. who report saving or setting aside any money in the past 12 months. Data source: World Bank, Global Financial Inclusion Database, at http://data.world Table 4: Selected indicators related to risk bank.org/data-catalog/financial_inclusion. management at the household level Educational attainment (primary, secondary, and Table 5: Selected indicators related to risk tertiary): Educational attainment is the percentage management at the enterprise sector level of the population aged over 25 who have completed each level of education (primary, secondary, and Wage employment: Wage and salaried workers (em- tertiary). Observations are estimates based on pop- ployees) are those workers who hold the type of jobs ulation censuses. Data source: Robert J. Barro and defined as “paid employment jobs,” where the in- Jong-Wha Lee, Barro-Lee Educational Attainment cumbents hold explicit (written or oral) or implicit Dataset, at http://www.barrolee.com. employment contracts that give them a basic remu- Education quality, PISA mean score (math and neration that is not directly dependent upon the rev- reading): The Programme for International Student enue of the unit for which they work. Data source: Selected indicators 321 International Labour Organization, Key Indicators saving or setting aside any money by using an ac- of the Labour Market (database), at http://www.ilo count at a formal financial institution such as a bank, .org/kilm. credit union, microfinance institution, or coopera- Goods market efficiency: Goods market ef- tive in the past 12 months. Data source: World Bank, ficiency is a composite indicator of economies’ ca- Global Financial Inclusion Database, at http://data pacities to produce the right mix of products and .worldbank.org/data-catalog/financial_inclusion. services given their particular supply and demand Loan from a financial institution: The percent- conditions, as well as to ensure that these goods can age of people who have obtained a loan from a fi- be most effectively traded. The indicator uses a 1–7 nancial institution is the percentage of respondents scale, where 1 indicates the lowest efficiency and 7 (aged 15+) who report borrowing any money from a the highest efficiency. Data source: World Economic bank, credit union, microfinance institution, or other Forum, Global Competitiveness Report, 2006-07 financial institution such as a cooperative in the past and 2012-13 editions, at http://www.weforum.org/ 12 months. Data source: World Bank, Global Finan- issues/competitiveness-0/gci2012-data-platform. cial Inclusion Database, at http://data.worldbank Labor market efficiency: Labor market efficiency .org/data-catalog/financial_inclusion. is a composite indicator of economies’ efficiency in Personally paid for health insurance: The per- allocating workers to their most effective uses and centage of people who have personally paid for providing incentives for them to give their best ef- health insurance is the percentage of respondents forts in their jobs. The indicator uses a 1–7 scale, (aged 15+) who have personally purchased health or where 1 indicates the lowest efficiency and 7 the medical insurance (in addition to any nationally pro- highest efficiency. Data source: World Economic Fo- vided health insurance). Data source: World Bank, rum, Global Competitiveness Report, 2006-07 and Global Financial Inclusion Database, at http://data 2012-13 editions, at http://www.weforum.org/issues/ .worldbank.org/data-catalog/financial_inclusion. competitiveness-0/gci2012-data-platform. Purchased agriculture insurance: The percentage Pension contributors: Pension contributors of people who have purchased agriculture insurance are the total number of people who actively con- is the percentage of respondents (aged 15+) who are tribute to the pension system in a given year, as a farming, fishing, or forestry workers and have per- percentage of the labor force. Data sources: World sonally paid for crop, rainfall, or livestock insurance Bank, Pensions (database), at http://go.worldbank in the past 12 months. Data source: World Bank, .org/8KO0DUVDS0; Pallares-Miralles, Montserrat, Global Financial Inclusion Database, at http://data Carolina Romero, and Edward Whitehouse, 2012, .worldbank.org/data-catalog/financial_inclusion. “International Patterns of Pension Provision II: A Population using informal savings: The popula- Worldwide Overview of Facts and Figures,” Social tion using informal savings is calculated as the dif- Protection and Labor Discussion Paper SP 1211, ference between the percentage of respondents (aged World Bank, Washington D.C. 15+) who have “saved any money in the past year” Formal production: Formal production is pro- and those who “saved at a financial institution in the duction by the formal sector, as a share of the econ- past year”. Data source: World Bank, Global Finan- omy, computed as 100 minus the estimated size cial Inclusion Database, at http://data.worldbank of the shadow economy. Data source: Schneider, .org/data-catalog/financial_inclusion. Friedrich, Andreas Buehn, and Claudio E. Montene- Population using informal credit: The popula- gro, 2010, “Shadow Economies All over the World: tion using informal credit is calculated as the differ- New Estimates for 162 Countries from 1999 to ence between the percentage of respondents (aged 2007,” at http://documents.worldbank.org/curated/ 15+) who have obtained a “loan in the past year” and en/2010/06/12864844/shadow-economies-all-over- those who obtained a “loan from a financial institu- world-new-estimates-162-countries-1999-2007. tion in the past year”. The percentage of people who have obtained a loan in the past year is the percentage of respondents who borrowed money in the past 12 Table 6: Selected indicators related to risk months from a formal financial institution, a store (by management at the financial sector level using installment credit), family or friends, an em- Saved at a financial institution: The percentage of ployer, or another private lender. Data source: World people who have saved at a financial institution is the Bank, Global Financial Inclusion Database, at http:// percentage of respondents (aged 15+) who report data.worldbank.org/data-catalog/financial_inclusion. 322 WORLD DEVELOPMENT REPORT 2014 Use of electronic payments: Use of electronic Credit (% of GDP): Credit to GDP is the ra- payments is the total transaction volume of direct tio of domestic private credit to the real sector credits and credit transfers, direct debits, payments by deposit money banks to GDP. Data source: by debit card, and payments by credit cards. Data World Bank, Global Financial Development Data- source: World Bank, Global Payment Systems Sur- base, at http://data.worldbank.org/data-catalog/ vey, at http://go.worldbank.org/5MYOUCYBR0. global-financial-development. Stock market capitalization: Stock market capi- Insurance premiums (% of GDP): Insurance talization is the total value of all listed shares in a premiums to GDP is the ratio of insurance pre- stock market, as a percentage of GDP. Data sources: miums (life and nonlife) to GDP. Data source: Standard & Poor’s, Global Stock Markets Factbook and World Bank, Global Financial Development Data- supplemental S&P data; World Bank, Global Finan- base, at http://data.worldbank.org/data-catalog/ cial Development Database, at http://data.worldbank global-financial-development. .org/data-catalog/global-financial-development. Loan dollarization: Loan dollarizarion is the ra- Bank assets (% of GDP): Banks assets to GDP tio of foreign currency denominated loans to total is the ratio of total assets held by deposit money loans. Data source: Chitu, Livia, 2013, “Was Un- banks to GDP. Assets include claims on the domes- official Dollarisation/Euroisation an Amplifier of the tic real nonfinancial sector (which includes cen- ‘Great Recession’ of 2007–09 in Emerging Econo- tral, state, and local governments); nonfinancial mies,” Comparative Economic Studies 55: 233–65, at public enterprises; and the private sector. Deposit http://www.palgrave-journals.com/ces/journal/v55/ money banks comprise commercial banks and n2/full/ces20131a.html. other financial institutions that accept transferable Deposit dollarization: Deposit dollarization is deposits, such as demand deposits. Data source: the ratio of foreign currency denominated deposits to total deposits. Data source: Chitu, Livia, 2013, World Bank, Global Financial Development Data- “Was Unofficial Dollarisation/Euroisation an Am- base, at http://data.worldbank.org/data-catalog/ plifier of the ‘Great Recession’ of 2007–09 in Emerg- global-financial-development. ing Economies,” Comparative Economic Studies 55: Mutual fund assets (% of GDP): Mutual fund as- 233–65, at http://www.palgrave-journals.com/ces/ sets to GDP is the ratio of the total assets held by mu- journal/v55/n2/full/ces20131a.html. tual funds to GDP. A mutual fund is a type of man- Loan-to-deposit ratio: The loan-to-deposit ratio aged collective investment scheme that pools money is the ratio of private credit by deposit money banks from many investors to purchase securities. Data to the sum of domestic demand, time, and saving de- source: World Bank, Global Financial Development posits in deposit money banks. Data source: World Database, at http://data.worldbank.org/data-catalog/ Bank, FinStats (internal database). global-financial-development. Insurance assets (% of GDP): Insurance as- sets to GDP is the ratio of the total assets held Table 7: Selected indicators related to risk by insurance companies to GDP. Data source: management at the macroeconomy level World Bank, Global Financial Development Data- CPI inflation rate: The consumer price index (CPI) base, at http://data.worldbank.org/data-catalog/ inflation rate is the annual percentage change in the global-financial-development. cost to the average consumer of acquiring a bas- Pension assets (% of GDP): Pension assets to ket of goods and services (which may be fixed or GDP is the ratio of the total assets held by pension changed at specified intervals). The Laspeyres for- funds to GDP. A pension fund is any plan, fund, mula is generally used. Data source: International or scheme that provides retirement income. Data Monetary Fund (IMF) International Financial source: World Bank, Global Financial Development Statistics, at http://elibrary-data.imf.org/FindData Database, at http://data.worldbank.org/data-catalog/ Reports.aspx?d=33061&e=169393. global-financial-development. Government primary surplus: The government Bank savings (% of GDP): Banks savings to GDP primary surplus is the gross fiscal balance plus net is the sum of domestic demand, time, and savings de- interest payments. Data source: IMF World Eco- posits in deposit money banks, expressed as a share nomic Outlook Database April 2013, at http://www of GDP. Data source: World Bank, Global Financial .imf.org/external/ns/cs.aspx?id=28. Development Database, at http://data.worldbank Gross public debt: Gross public debt is all liabili- .org/data-catalog/global-financial-development. ties that require payment or payments of interest Selected indicators 323 and/or principal by the debtor to the creditor at a flood, or storm (as defined by the EM-DAT data- date or dates in the future. Gross public debt includes base) during the specified time period. Deaths per debt liabilities in the form of special drawing rights, million population is the number of deaths divided currency and deposits, debt securities, loans, insur- by the population for the same time period. Data ance, pensions and standardized guarantee schemes, sources: Guha-Sapir, D., and P. Heudtlass, forthcom- and other accounts payable. Data sources: Abbas, S. ing, “Standardized Indicators of Human and Eco- Ali, Nazim Belhocine, Asmaa El Ganainy and Mark nomic Loss from Natural Disasters,” CRED working Horton, 2010, “A Historical Public Debt Database”, paper, Université catholique de Louvain, Brussels; IMF Working Paper WP/10/245, at http://www.imf. EM-DAT: The OFDA/CRED International Disaster org/external/pubs/cat/longres.cfm?sk=24332.0; IMF Database, at http://www.emdat.be. World Economic Outlook Database April 2013, at Damages from natural disasters (total, or % http://www.imf.org/external/ns/cs.aspx?id=28. GDP): Damages from natural disasters are the re- International reserves: International reserves ported economic impacts as a consequence of are holdings of monetary gold, special drawing droughts, earthquakes, floods, and storms (as de- rights, reserves of IMF members held by the IMF, fined by the EM-DAT database) during the specified and holdings of foreign exchange under the con- time period. The economic impact usually consists trol of monetary authorities. The gold component of direct (such as damage to infrastructure, crops, of these reserves is valued at year end (December housing) and indirect (such as loss of revenues, un- 31) London prices. Data are in current U.S. dollars. employment, market destabilization) consequences Data source: IMF International Financial Statistics, to the local economy. Total damages are in current at http://elibrary-data.imf.org/FindDataReports. US dollars, according to when the events occurred. aspx?d=33061&e=169393. Damages as a percentage of GDP is the ratio of Flexible exchange rate regimes: For each coun- total reported damages to midyear GDP for the same try, five year averages of observed exchange rate flex- time period. Data sources: Guha-Sapir, D., and P. ibility (1996–2000, 2001–05, 2006–10) are computed Heudtlass, forthcoming, “Standardized Indicators of based on a taxonomy of de facto exchange rate ar- Human and Economic Loss from Natural Disasters,” rangements. The coarse classification of exchange CRED working paper, Université catholique de Lou- rate regimes takes values between 1 and 6, ranging vain, Brussels; EM-DAT: The OFDA/CRED Interna- from the least to most flexible exchange rate regimes. tional Disaster Database, at http://www.emdat.be. Data source: Calculations based on “Annual Coarse Classification 1946–2010” data in Ilzetzki, Ethan, CO2 emissions per capita: CO2 (carbon diox- Carmen M. Reinhart, and Kenneth S. Rogoff, 2010, ide) emissions per capita are the emissions from the “Exchange Rate Arrangements Entering the 21st burning of fossil fuels and the manufacture of ce- Century: Which Anchor Will Hold?”, at http://per ment—including carbon dioxide produced during sonal.lse.ac.uk/ilzetzki/IRRBack.htm. consumption of solid, liquid, and gas fuels and gas Worldwide Governance Indicators average: flaring—divided by the midyear population. Data Worldwide Governance Indicators average is the sources: Carbon Dioxide Information Analysis Cen- average of six indicators reflecting broad dimen- ter, Oak Ridge National Laboratory, at http://cdiac sions of governance (voice and accountability; po- .ornl.gov; population data from World Development litical stability and absence of violence; government Indicators, at http://data.worldbank.org/indicator/ effectiveness; regulatory quality; rule of law; control SP .POP.TOTL. of corruption) as defined by the Worldwide Gover- nance Indicators project. Data source: World Bank, Table 9: Global temperature anomalies: Worldwide Governance Indicators, at http://info Difference relative to 1951–80 .worldbank.org/governance/wgi/index.asp. Temperature anomalies are global average tem- peratures relative to the same time of year in the Table 8: Natural disasters and climate change base period 1951–80. Temperature anomalies are indicators shown in degrees Celsius. Data source: Combined Deaths from natural disasters (total, or per mil- Land-Surface Air and Sea-Surface Water Tempera- lion): Deaths from natural disasters are the number ture Anomalies dataset (Land-Ocean Temperature of persons reported as dead or missing or presumed Index, LOTI), NASA, at http://data.giss.nasa.gov/ dead as a consequence of a drought, earthquake, gistemp. 324 WORLD DEVELOPMENT REPORT 2014 Table 10: Aid commitments Aid commitments are commitments for emergency including the OECD’s Creditor Reporting System. response, reconstruction relief, and prevention and Data sources: Disaster Aid Tracking Global Facility preparedness. Commitment amounts are in constant for Disaster Reduction and Recovery, at http://gfdrr 2009 US dollars and are five year running averages. .aiddata.org/dashboard; AidData Center for Devel- These amounts are aggregated from AidData proj- opment Policy, at http://www.aiddata.org/content/ ect level results and are collected from many sources index/Research/research-datasets. Index The following letters denote special treatment in the material: “b” for boxes; “ f ” for cartoons, diagrams, figures, and illustrations; “m” for maps; “n” for notes; and “t” for tables and glossaries. A preparation for risk, 15b, 69b Aadhaar identification project (India), 123 stimulus package as response to financial crisis, 235b access to finance Asia. See also individual countries and regions demand-side factors influencing, 200–202, 201f informal vs. formal employment rates, 178b enterprise sector approach to risk, 177, 184 regional economic cooperation, 256 finance system approach to risk, 28, 194, 197–202, 198f Asian Bond Markets and Asian Bond Fund Initiatives, 216 household approach to risk, 22, 125–26 asset management, 236–38, 237t informal credit, 204 asymmetric information, 81t, 115, 265–66 preparation for risk, 69b Australia access to labor markets disclosure of balance sheet fiscal risks, 240 household approach to risk, 22, 124–25 government guaranteeing loans, 240 women’s empowerment, 22–23 long-term fiscal sustainability reports, 240 accountability universal pension and health care programs, 282 community approach to risk, 159 avian flu (H5N1) pandemic, 36, 249b, 254, 260 empowerment of communities, 154–55, 155f Azito Energie SA, 181b financial system approach to risk, 207 government regulation of business, 183 B independent fiscal councils providing, 280 bailouts in financial crises, 15, 18, 33, 40, 174, 210–11b, 240–41, macroeconomic approach to risk, 226, 242 245n59, 267 adversarial system, 88b Baltic Stock Exchange, 216 adverse selection, 67, 81t, 145 Bangladesh Afghanistan community approach to risk management, 153 households facing shocks, 55t, 113 empowerment with voice and accountability, 154 informal credit and assistance, 143, 143f conditional cash transfer programs, 123b National Solidarity Program, 25, 140, 154 cyclone casualty reduction, 5, 8, 57 Africa. See also individual countries and regions domestic violence, 146 drought-caused mortality, 57 early warning system for storms, 99, 274n13 rainfall predictions, 94, 94m educational attainment, benefits of, 118 regional economic cooperation, 256 fertility rates and reliance of women on children for support, agriculture 120 climate change and, 146 flood protection, costliness of, 16, 89 frequency of agricultural shocks, 140, 141f garment factory collapse (2013), 27, 185 husbands and wives engaging in separate agricultural labor, 112 Grameen Pension Scheme, 200b risk aversion of farmers, 8, 59 household approach to risk, 21, 111 Weather Based Crop Insurance Scheme (India), 126b diversification of income sources, 120 AIDS. See HIV/AIDS organizational membership, 149 Allais, Maurice, 10 Rural Advancement Committee (BRAC), 154 altruism, 111b, 144 safety net programs, 41, 129 American Recovery and Reinvestment Act (2009), 235b seasonal migration to urban areas, 59, 60f, 121 analytical framework for risk management, 11, 17b Bank for International Settlements, 256, 262 Arab Spring, 140 banking crises. See financial and macroeconomic crises arbitrage, 204 bank runs, 206 Argentina bankruptcy laws, 177, 182 banking crisis (1980), 209 banks as distributors of social security benefits, 123b, 125 job loss, 4 Barro, Robert, 31 325 326 WORLD DEVELOPMENT REPORT 2014 Basel Committee on Bank Supervision, 215 food security and, 104 Basel III framework, 269 policy design for, 122–23 Becker, Gary, 12, 21 as safety net, 129–30 behavioral failures, 10, 16, 17b, 81–85, 83b, 98f Catastrophe Deferred Drawdown Option (CAT DDO), 76, 77n1, Benin, clean water in, 85f 242, 261, 274nn19–20 Bernoulli, Daniel, 10 cattle plague (rinderpest), 249 best-practice policies for managing financial crises, 207–11 Central Asia. See Eastern Europe and Central Asia birth weight and prenatal care, 114 central banks, role of Black, Duncan, 10 best-practice policies, 207 Bolsa Familia Program (Brazil), 68, 85, 93 emergency liquidity assistance, 219n56 bonds, 216, 238 inflation, 229, 231 Brazil integrated financial management, 239, 281 banks as distributors to social program beneficiaries, 123b, 125 Chicago Board Options Exchange Market Volatility Index (VIX), conditional cash transfer programs, 68, 85, 93 238 cost of doing business, 182 Chicago heat wave (1995) and community response, 139 crime and violence prevention, 164–65 child abuse, 114, 114f disclosure of balance sheet fiscal risks, 240 child care, 120–21, 124 electronic payments vs. use of cash, 196b child labor, 172 global systemically important financial institutions (G-SIFIs), child mortality, 57, 115, 156, 254 206 children health insurance programs, 127t effect of large shocks on, 113 long-term bonds issued by, 238 effect of maternal employment on, 178b private nonprofit hospitals, 123b as support for women, 120 public procurement, 196b Chile safety net programs, 41, 129 budget planning, 41, 280 sovereign spread and sound fiscal regulation, 233 commodity revenues, management of, 239b vulnerable population, government spending on, 132 conditional cash transfer programs, 123b Buchanan, James, 10 correction for cyclical influences of business cycle, 239b budgets and budgeting. See also independent fiscal councils earthquake (2010), 57 coping with budgetary surprises, 236, 237t Economic and Social Stabilization Fund, 62 for disaster aid and relief, 76–77 financial system approach to risk, 211 medium-term expenditure frameworks (MTEFs), 236 disclosure of balance sheet fiscal risks, 240 procyclical fiscal bias, 34b fiscal expansions and spending programs, 235 structural budget balance, rules targeting, 280 Fiscal Responsibility Law (2006), 239b building codes, enforcement of, 57, 62, 77, 89–90 independent fiscal council, 34b, 243 Bulgaria Medium-Term Fiscal Framework (MTFF), 241b enterprise flexibility, 175 natural resources, discovery of, 62 job loss, 4 pension system, 62, 240, 281 Busan Partnership. See New Deal for Engagement in Fragile States preparation for risk, 15b, 69b business sector. See enterprise sector approach to risk recovery from depression (1980s), 175b Responsibility and Transparency Law (Law 819, 2003), 241b C sovereign wealth funds (SWFs), 238 Cameroon China cash transfer program and worker training, 130 climate change measures, 273, 283 manufacturing firms sharing risk with workers, 170 community approach to risk, 140, 141f, 144 Canada earthquake (2008), 61, 62 government guarantee of loans, 240 elderly living with other family members, 112f independent fiscal council, 281 enterprise sector approach to risk, 171 integrated financial management by government, 239 famine (1959–61), 114 capacity building, 71b, 76, 99, 184, 271 fiscal expansions and spending programs, 235 cap-and-trade mechanisms, 273 grandparents caring for grandchildren, 111b, 112 CAPRA (Central American Probabilistic Risk Assessment), 263b households facing shocks, 113 carbon emissions. See climate change insured losses, 73n16 carbon taxes, 273 sovereign wealth funds (SWFs), 238 CARD (Philippines private pension insurance), 200b stimulus package as response to financial crisis, 235b Caribbean Catastrophe Risk Insurance Facility (CCRIF), 36, 262b total factor productivity, 25 Caribbean regional economic cooperation, 256 universal pension and health care programs, 282 cash transfer programs. See also conditional cash transfer programs choice under uncertainty, 10–11 fairness within households and, 22 cholera, 272b Index 327 civil society, approach to risk, 19f, 21, 92, 154, 260, 283 informal insurance, 144–45 Clean Air Coalition, 284 vulnerable groups, inclusion of, 145 climate change knowledge, 159t “coalitions of the willing,” 272, 284–85 legal environment improvements, 152 community approach to risk, 145–48 location-based communities, 140–42 conflict and violence linked to, 147–48 long-run perspective, 159–60 free riders, 18, 285 prosperity, 23 global sea level rise, 54b, 250f protection, 145–48, 159t impact of, 4, 70, 146, 268, 273f, 274n9 recommendations and policies to improve, 38b, 152–60 international community approach to risk, 37b, 88–89, 251, 254, priorities, 159t 255b, 263, 266, 268, 275n53 research priorities, 160 incremental approach, 273, 283 resilience, 23, 159–60 national measures to limit carbon emissions, 283–84, 284t risks faced by, 23, 139–40 public opinion on, 266, 267f systemic risk, 20b, 140 tragedy of the commons, 18, 254 trust of government, police, and religious leaders, 154, 155f “coalitions of the willing,” 272, 284–85 uncertainty, avoidance of, 159 Coase, Ronald, 25, 169 vulnerable groups, inclusion of, 160 cohesiveness Compagnie Ivoirienne de Production de l’Electricité (CIPREL), community approach to risk, 24, 148–52 181b international community approach to risk, 36 comparative advantage, principle of, 20b in risk management, 3, 4b competition in banking sector, 213–14, 217, 218n8 Coleman, James, 23 Comunidades Solidarias Rurales (CSR, El Salvador), 105 collective action. See also systemic risk conditional cash transfer programs community approach to risk, 23, 25, 87–89, 148, 149 distributional inequities and, 93 incentive constraints, 10–11 education demand and, 118 international community approach to risk, 35, 256, 268, 269 food security and, 104–5 Colombia health risk management via, 85 banking crisis (1982), 29, 204 as incentive for self-reliance, 41 budget planning, 41 policy design for, 122–23 CAT DDO (Catastrophe Deferred Drawdown Option), 242 preparation and resilience improvements via, 68 conditional cash transfer program, 123b public-private partnerships and, 123b disclosure of balance sheet fiscal risks, 240, 241b women’s empowerment and, 23 government guaranteeing loans, 240 confidence. See trust and confidence health insurance programs, 127t conflict labor market flexibility, 190 community-based conflict, 146–48, 147f long-term bonds issued by, 238 communication technologies, use of, 153b pension system, 281 between neighboring countries, 256 preparation for disaster management, 77 conflict-affected countries. See fragile and conflict-affected Prevention of and Attention to Disasters System, 241b countries school voucher program, 123b conflict filters, 160 common-pool problem, 81t connectedness communication campaigns to change social norms, 86–87 community approach to risk, 24, 148–52 community approach to risk, 19, 19f, 23–25, 139–63 in financial crises, 205–6 characteristics that improve contribution to risk management, international economic, financial, and social interconnectedness, 24, 140–43 35, 35f cohesiveness, 24, 148–52, 149f consumer protection, 25–26, 27, 171–72, 172b, 185 confronting common risks and opportunities, 23, 145–46 contagion of financial crises, 206, 218n41 connectedness, 24, 148–52 contingent liabilities, management of, 239–42, 241b coping, 23, 24f, 143, 143f, 159t coordination failures, 81t, 89, 90–92, 226, 268 crime, violence, and conflict, 146–48, 147f coping cultural, ethnic, and kinship communities, 142, 145, 147–48 in analytical framework for risk management, 17b definition of community, 23, 139, 161n6 community approach to risk, 23, 24f, 143, 143f, 159t empowerment of communities, 152–55 defined, 12 flexibility, 159t, 160 enterprise sector approach to risk, 187t incentives, 160 financial system approach to risk, 217t informal approach to risk, 143, 143f household approach to risk, 111–14, 113t, 130t infrastructure and services, 155–59 international community approach to risk, 257f, 260–61, 270t infrastructure to complement community capacity, 155–59 macroeconomic approach to risk, 228, 236–39, 244t insurance provided by community, 14, 143–45, 159t with market and government failures, 85–87 328 WORLD DEVELOPMENT REPORT 2014 as part of risk management, 13f, 14–15, 63, 65f, 68–70 deep uncertainty. See uncertainty relationship with preparation, 13f, 14, 65f, 70, 267, 268f Deepwater Horizon oil spill (2010), 178 Corporación Nacional Del Cobre de Chile, 239b default situation, 82, 83b corporatist approach, 88b demand-side factors influencing access to finance, 200–202, 201f corruption, 18, 89, 254, 266, 274n27 Denmark Costa Rica flexicurity model for labor market, 41, 183, 183b CAT DDO (Catastrophe Deferred Drawdown Option), 242 independent fiscal council, 281 flexible and formal economy, 179 unemployment, 26, 167 Intel investment in workers, 178 depressions. See financial and macroeconomic crises cost-effectiveness of risk management De Soto, Hernando, 179 bias toward less cost-effective ex post actions, 92, 267, 274n30 developed countries early warning systems for storms, 274n13 climate change initiatives, 273 financial and information constraints and, 81–82, 84 countercyclical fiscal policy, 232 precautionary principle and, 97b drought no longer cause of death, 57 preparation vs. ex post actions, 8–10, 8f, 43n12, 70–72, 71f, preparation for risk, 15b 73n34, 73n52 universal pension and health care programs, 282 transport safety in France, 84 developing countries Côte d’Ivoire countercyclical fiscal policy, 32, 33m, 230–31, 231f, 232b, 233, household approach to risk 234m diversification of income sources, 120 educational attainment, 116–18, 117–18f husbands and wives engaging in separate agricultural labor, financial stability committees, 215, 215t 112 health insurance, 22 power sector’s resilience during civil unrest, 181, 181b informal mechanisms for risk management, 20 safety net programs, 130 mortality rates, 57, 253m Council of Risk Analysts, 278 pension programs, 281–82, 282f countercyclical and sustainable fiscal policy, 32–33, 33m, 41, preparation for risk, 15b, 69b 230–35, 231f, 232b, 234m reserve requirement policy, 231, 232b creative destruction, 25, 171 self-employment, 25 credit, 194. See also access to finance shadow banking sector, 206 crime shocks, impact of, 4–5, 5f, 57, 58b community approach to risk, 24, 139, 146–48, 147f susceptibility of households to risk, 55–56, 56f integrated urban crime and violence prevention, 164–65 development example of Peruvian family resilience in face of risk, 7b community approach to risk, 160 homicide trends across regions, 9b, 54b negative effect on GDP, 4, 53, 56 cost-effectiveness of risk management and, 70–72, 71f crises. See climate change; conflict; crime; domestic violence; goal of risk management, 62–63 financial and macroeconomic crises; food insecurity; jobs as driver of, 174 illness; natural disasters life-saving aspect of risk management, 5, 8, 57 crisis volatility, 70, 73n48 opportunity, risk as, 4–5, 6f, 53–55, 54b crowdsourcing, 153b preparation for risk accelerating, 70 cultural communities, 142 resilience and successful risk management for, 63 currency unions, 230, 231b risk management as powerful instrument for, 3, 5–10, 53–75 Czech Republic setbacks averted through risk management, 57–58 central bank role, 30, 207, 281 development assistance, 20, 152, 267, 268f, 269 financial system approach to risk, 193 Diamond, Douglas, 28 labor market flexibility, 190–91 disaster preparedness and response, 68, 70, 71b, 261b. See also resilience to financial crisis, 8, 222–23 warning of natural disasters Disaster Risk Finance Insurance (DRFI) program, 242, 262b D disclosure of public sector fiscal decision making, 240 data collection. See also information constraints discretionary fiscal policy, 233–34 financial system approach to risk, 203, 217t disease prevention. See also flu outbreak; immunization; specific technology leveraging for, 82, 123 diseases debt and equity investments, 194 avoiding future shocks through, 62 decision making international community costs of, 267–68, 274n36 disclosure of public sector fiscal decision making, 240 lack of resources information available for, 65 in developing countries, 57 irrationality in, 82–84, 83b as obstacle to, 116, 117f robust decision making, 95, 95f, 96b noncommunicable disease, 115, 116f women’s voice in, 124 pandemic risk, management of, 248–49, 254, 260, 263, 267 Index 329 social norms and, 86 electronic payment systems, 196b technology improvements and, 13, 66b El Salvador displaced persons, 151b family sacrificing child’s education in time of disaster, 114 distortionary policies, 18, 182, 206 food security in face of natural disaster and economic shocks, distributional problems, 93 104–5 diversified income portfolio, benefits of, 120 public works programs, 130 domestic violence, 23, 86, 114, 114f, 124, 124f, 146, 147f emerging markets. See developing countries Dominican Republic employment. See enterprise sector approach to risk banking crisis, 57 empowerment housing and safety, 145 of communities, 152–55, 155f returns to education, 124 of women, 124, 138f donors. See development assistance enterprise sector approach to risk, 19, 19f, 25–27, 166f, 167–89 dowry, 120 access to finance, 177, 184 droughts, 57, 58, 104, 146 bankruptcy laws, 177, 182 dueling certitudes, 94 better business environment, 27, 185–86 Dybvig, Phillip, 28 business registration, 182 characteristics that improve contribution to risk management, E 10, 26–27, 168–72 early warning systems. See warning of natural disasters coping, 187t ease of doing business, 182 employment services, 125 East Asia. See also individual countries flexibility, 26, 28f, 172–76, 173f, 178–79 domestic violence, 124 benefits of, 59, 167 macroeconomic and sound monetary policies, 229 countries classified by levels of, 180f Eastern Europe and Central Asia. See also individual countries government policy enhancing, 179–85 banking crises, effect of, 204 in India, 190–91 community approach to risk, 143, 144f labor market flexibility, 173, 173f, 183, 183b, 190–91 education budgets, 32 product market flexibility, 173, 173f pension reform, 128, 240, 281 formality, 26–27, 28f, 176–79, 176f economic crises. See financial and macroeconomic crises benefits of, 167, 168 economic terminology, glossary of, 81t countries classified by levels of, 180f Ecuador globalization, relationship with, 178b banks as distributors to social program beneficiaries, 125 government policy enhancing, 179–85 community approach to risk, 140–41, 148 informal sector’s role in weak economy, 173 conditional cash transfer programs, 123b gender differences in employment, 174–75 economic crisis (1998–2000), 57, 114 gender equality, 22–23 education community approach to risk, 25 household approach to risk dropping out of school to work, 178b access to labor markets, 124–25 educational attainment, 22f, 116–18, 117f increasing labor supply, 120–21 example of Peruvian family resilience in face of risk, 7b incentives, 186 family sacrificing child’s education in time of disaster, 114 innovation and resource reallocation, 25, 167–68, 171, 175, 184 financial and macroeconomic crises causing cuts, 32 insurance, 187t financial education programs, 202 job creation, 167–68, 174 household benefits from, 21, 116 job loss, effect of, 4, 57 loans, 196 knowledge, 187t positive contribution to risk management, 59, 116–18 limited liability, 170 worker training. See skills training for workers long-run perspective, 184–85, 186 effective risk management, 10–15. See also cost-effectiveness of risk minimum wage, 182 management property rights, 180–81. See also property rights Egypt prosperity, 25, 167–68, 172, 185 flood protection, 89 protection, 187t food insecurity, 53 recommendations and policies to improve, 38b, 185–86 Harassmap technology to report women’s harassment, 153b priorities, 186t Ehrlich, Isaac, 12 regulations, enforcement of, 27, 174, 174f, 177, 181–83 elderly people. See also old age income support resilience, 25–26, 181b abuse of, 114 risk sharing, 19, 25, 170 countries with fast growing old-age dependency ratio, 129, 281, skills training for workers, 130, 183–84, 185 283f social protection and insurance, 183–84, 183b living with other family members, 112f technology adoption and global collaboration, 184–85 330 WORLD DEVELOPMENT REPORT 2014 transition to wage employment, 168, 169f, 170 financial and macroeconomic crises government policy encouraging, 179–85 bailouts, 15, 18, 33, 40, 174, 210–11b, 240–41, 245n59, 267 types of enterprises, 168–69 bank runs, 206 uncertainty, 181, 186 Basel III framework, 269 urban areas, employment shocks in, 55 best-practice policies for managing, 207–11 vulnerable groups, inclusion of, 186 contagion of, 206, 218n41 worker, consumer, and environmental protection, 25–26, coping responses, 145 171–72, 185 crisis preparedness for future, 208–9, 209b work status tied to social insurance, 29b, 184, 281–82 developing countries Envía CentroAmérica (remittances program), 125 with banking crises, 57 environmental protection. See also climate change GDP growth after 2008–09 crisis, 58, 209 enterprise sector and, 25–26, 27, 171–72, 172b development impact of, 4 forest management, 150 divergent national interests, effect of, 269 mining regulation, 43–44n70 diverting resources from growth and development, 267 ozone layer protection, 36, 37b, 269, 273, 284 failure of financial firms and past public policies, 15, 16, 28–29, precautionary actions in environmental regulation, 97b 206–7 regulatory framework for, 43–44n70, 185 fair and effective solutions, 204–5, 209–11 environmental shocks, 60–62. See also natural disasters financial indicators, choice of, 207 Ethiopia fiscal rules’ failure leading to, 236 community approach to risk, 145 GDP loss, 174 droughts global, 252, 254 effect on household worth, 57 harmful effects of, 204–6 program to build resilience to, 104 interconnectedness leading to systemic risk, 205–6 health insurance programs, 127t lessons learned, 175b household approach to risk, 21, 111 macroeconomic policy women’s household value, 119 effect of, 204 illness, effect on household consumption, 4, 53 uncertainty, 227b preparation for risk, 15b, 69b volatility and deep recessions, 228, 228f Productive Safety Net Program, 23, 41, 104, 129 macroprudential tools matched to countries, 207–8, 208t reservoir’s effect on mosquito population and health map (1970–2011), 31m consequences, 94 monetary policy, 32 risk aversion of farmers, 8 moral hazard in bailouts, 18, 210–11b ethnic communities, 142, 147–48, 150 poverty and, 53, 232 European Banking Authority, 215 prevention of, 8, 14, 204–11 European Insurance and Occupational Pensions Authority, 215 private sector solutions, 209 European Monetary Union, 41 public-private cooperation, solutions requiring, 90 European Securities and Markets Authority, 215 resilience to economic shocks, 222–23 European Systemic Risk Board, 215 stimulus packages, 232, 234–35, 235b, 280 European Union “too big to fail,” 210–11b Emissions Trading System, 284 trends across regions, 9b, 14, 54b Fiscal Compact Treaty, 280 Financial Stability Board, 215, 262 formation of, 272 financial system approach to risk, 19, 19f, 27–31, 192f, 193–221, 203f interconnectedness of banks leading to systemic risk, 205 access to finance, 197–202. See also access to finance labor market regulation, 182 bailouts. See bailouts in financial crises monetary policy and Euro Area crisis, 231b, 260 bank runs, 206 precautionary principle, application of, 97b characteristics that improve contribution to risk management, Stability and Growth Pact rules, 280 27–29, 193–94 exchange rate regimes, 32, 194, 229–30 competition, beneficial effect of, 213–14, 217, 218n8 flexible (floating) regimes, 230, 231b coping, 217t exclusion for gender, ethnicity, or culture, 86, 145, 150 crises. See financial and macroeconomic crises expected utility theory, 83b data collection frameworks, 203, 217t debt and equity investments, 194 F definition of financial system, 194 Factories Act (1948, India), 190 demand-side factors influencing access, 200–202, 201f FCSs. See fragile and conflict-affected countries diversified financial strategy, 196, 198f, 216 fear of risk. See risk aversion electronic payments, 195, 196b Federal Emergency Management Agency (U.S.), 82 financial education programs, 202 fertility rates and reliance of women on children for support, 120 financial infrastructure, 29–30, 202 Index 331 financial stability committees in developing countries, 215, 215t information constraints on preparation for, 82 globalization and cross-border banking, 206 psychological effects of flooding, 86 governance of financial institutions, 201, 206–7, 215–16 public-private coordination, 90 incentives and improving confidence, 202–3 in Singapore, 156b inclusion and depth, 27–29, 30f, 203, 211–16 in urban areas, 24 trade-offs and synergies, 30–31, 212–13f, 212–16, 214t flu outbreak innovative financial instruments, 199b, 202 avian flu (H5N1) pandemic, 36, 249b, 254, 260 knowledge, 203, 217t Mexico (2009), 66b long-term financing, need for, 205 pandemic (1919), 114 market insurance, 67, 194, 196, 217t, 218n15, 238 Food and Agriculture Organization (FAO), 249b overcoming obstacles, 202 Food and Drug Administration (U.S.), 176 poverty, financial tools aimed to reduce, 203 food insecurity, 23, 53, 54b, 104–5, 256 prosperity, 27 forecasting weather threats. See warning of natural disasters protection, 217t. See also financial and macroeconomic crises foreign exchange services, 194 public deposit insurance, 218n15 forest management, 150 public trading of assets, 194–95 formality reasons people reject financial tools, 201–2, 201f defined, 26 recommendations and policies to improve, 38b, 216–17 enterprise sector approach to risk, 26–27, 176–78, 176f, 178b priorities, 217t benefits of globalization for, 178b regulatory reforms, 29–31, 203 country variance in flexibility and formality, 28f remittances. See migration and remittances formal labor, 176–77, 176f resilience, 27, 222–23 formal production, 176, 176f risk pricing information, 195 government policy enhancing, 179–85 risk-taking capital, 194 relationship to flexibility, 178–79 saving instruments, 119, 194, 196, 201 fragile and conflict-affected countries shadow banking sector, 206 border states, effect on, 269 stability and, 28–29, 211–12 development opportunities when risks are well managed, trade-offs and synergies, 30–31, 212–16, 214t 264–65b supply-side factors influencing access, 197–200 donors’ concern about effective use of aid, 269, 271 systemically important financial institutions (SIFIs), 206, 209–11 informal mechanisms for risk management, 20 systemic risk, 30, 205, 205f restoring confidence and institutional capacity, 160 types of financial tools, 194–203, 195f risks exceeding national capacity, 20, 34, 252–54 uncertainty, reduction of, 211 France and cost-benefit analysis of transport safety, 84 fiscal councils. See independent fiscal councils free riders, 18, 149, 285 fiscal dominance, 245n12 Fukushima nuclear accident (Japan 2011), 71b, 178 fiscal policies. See macroeconomic approach to risk Fiscal Responsibility Law (2006, Chile), 239b G flexibility garment factory collapse (Bangladesh 2013), 27, 185 community approach to risk, 159t, 160 gender differences defined, 26, 172 enterprise sector approach to risk, 174–75 enterprise sector approach to risk, 26, 172–76, 173f, 186 household approach to risk, 22–23 benefits of, 59 education enrollment and attainment, 118 country variance in flexibility and formality, 28f fertility rates and reliance of women on children for support, government policy enhancing, 179–85 120 labor market flexibility, 173, 173f, 183, 183b, 190–91 ownership rights, 119 product market flexibility, 173, 173f shocks hurting girls disproportionately, 114 relationship to formality, 178–79 women participating in labor force, 120, 121f, 123–24 financial system approach to risk, 203 Germany, unemployment and labor market flexibility in, 167, 183, household approach to risk, 131 183b macroeconomic approach to risk, 242 Ghana overcoming obstacles to risk management, 99–100, 280 informal workers, 168 promotion of, 41 manufacturing firms sharing risk with workers, 170 floating regimes, 230 risk taking of farmers, 8, 59 flooding and flood protection technology use by farmers, 13, 66b climate change and, 70 Global Earthquake Model, 263b community approach to risk, 145, 154 Global Emerging Markets Local Currency Bond program, 216 financial constraints on preparation for, 16, 18 Global Facility for Disaster Reduction and Recovery (GFDRR), flood insurance, 84, 85 261b 332 WORLD DEVELOPMENT REPORT 2014 globalization coping mechanisms, 111–14, 113t enterprise sector approach to risk, 184–85 definition of household, 109 formality’s relationship with globalization, 178b diversifying income sources and increasing labor supply, 120–21 financial system approach to risk, 206 example of Bangladesh family, 109 health collaboration, 249b, 254 fairness within household, 22–23, 113–14, 123–24 incremental approaches to risk management. See incremental flexibility, 131 approaches to global solutions government policies to assist, 110, 121–30, 122t international community approach to risk, 37b, 251–52, 269 health care and disease prevention, 115–16, 115–16f, 126–28, 127t technology adoption and global collaboration, 5, 184–85, 273 household employment, 179 global systemically important financial institutions (G-SIFIs), 206 incentives, 131 governance income support for old age, 128–29 community involvement in, 25, 154–55 insurance and financial preparation, 119 financial corporate governance, 201, 206–7, 215–16 informal insurance through family formation, fertility, and government failures, 18, 85–87, 89 marriage, 119–20 Graham, John D., 278 investing for future, 21–22, 115–18 Grameen Bank (Bangladesh), 200b knowledge, acquiring and sharing, 115 Great Depressions of the Twentieth Century (Kehoe & Prescott), 175b long-run perspective, 131 greenhouse gases. See climate change migration and remittances, 125. See also migration and grievances, ethnic, 147 remittances Group of 20, 269 obstacles to, 115–21 Guatemala preparation to manage risk, 115–21 CAT DDO (Catastrophe Deferred Drawdown Option), 242 prosperity, 21–22 diversification of household income sources, 120 protection and risk pooling, 21, 109–13 resource constraints on disease prevention, 116 recommendations and policies to improve, 38b, 110, 122–23, 130–32 H priorities, 130t Haiti earthquake (2010), 35, 57, 66b, 272b resilience, 7b, 21–22 health insurance. See insurance; social insurance risk pooling, 21 health issues. See also disease prevention; immunization; insurance risks and shocks faced by, 55, 55t, 73n7, 110–14, 110f benefits of health preventive programs, 8, 10, 14, 59 safety nets for bad times, 129–30 community approach to risk, 25 susceptibility of households to risk, 55–56, 56f, 73n7 frequency of health shocks, 140, 141f uncertainty, 132 household approach to risk, 126–28, 127t vulnerable, protection of, 131–32 international approach to risk, 254 housing loans, 196, 197b knowledge improving, 65 housing prices and financial crises, 204 lack of indicators for risk, 92 human capital, 21, 67, 69b, 113, 115–18, 120. See also education public health and social norms, 87–88 humanitarian crises, 252–54 reservoir’s effect on mosquito population and health Hungary consequences, 94 financial crises vulnerable groups, protection for, 126–28, 127t currency devaluation due to, 204 heuristics, 84 education cuts due to, 32 HFA (Hyogo Framework for Action), 259t, 261b fiscal council abolished, 280 high-income countries. See developed countries integrated financial management by central bank, 239 Hindu Succession Act (India), 124 Hurricane Andrew (1992), 90 HIV/AIDS, 156, 248, 254, 259t Hurricane Katrina (2005), 84, 85 Ho Chi Minh City, risk of flooding in, 96b Hurricane Sandy (2012), 66b holistic approach to risk management, 18–21, 19f, 231 Hurwicz, Leonid, 10, 36 homicide. See crime; violence Hyogo Framework for Action (HFA), 259t, 261b Honduras housing affected by criminal activity, 197b I hurricane (1998), 57 Iceland, financial bailouts in, 33, 241 horizontal coordination, 90 idiosyncratic risks household approach to risk, 19, 19f, 21–23, 108f, 109–35 communities sharing, 19, 23, 140, 141f, 143–45 access to financial products, 22, 119, 125–26 costs of, 57 access to labor markets, 22, 124–25 defined, 4 access to social programs, 22, 123, 123b in developing countries, 57 accumulating financial and physical assets, 119 informal insurance and, 21, 144 characteristics that improve contribution to risk management, market insurance and, 196 22–23, 111b monetary policy and, 231b Index 333 size of risk, 20b Integrated Social Security Programme, 200 vulnerable households and, 53, 111 National Rural Employment Guarantee Act (NREGA), 30, 155 illness. See disease prevention; health issues G2P program, 202, 218n20 immunization, 21, 22f, 68, 85, 115, 115f National Rural Livelihoods Mission, 154 InaSAFE (software to assess future disaster events), 263b psychological effects of flooding, 86 incentives rainfall insurance for farmers, 59, 60f assessment of, 17b, 98f risk taking of farmers, 8, 59 community approach to risk, 160 safety net programs, 41, 129 enterprise sector approach to risk, 177, 186 sanitary conditions leading to stunting, 85 financial system approach to risk, 202–3 Self Employed Women’s Association (SEWA), 148 household approach to risk, 131 total factor productivity, 25 international community approach to risk universal pension and health care programs, 282 aligned incentives, 36, 37b, 270–71 Weather Based Crop Insurance Scheme, 126b detrimental effect of incentives, 36, 267–69 whistleblower laws, 92 financial incentives, 273 individual management of risk, 80–87, 82f. See also behavioral incremental approaches for nonaligned incentives, 37b, failures; self-insurance; self-protection against risks 271–73, 283–85 Indonesia technological incentives, 273 Aceh tsunami and earthquake (2004), 35, 272b macroeconomic approach to risk, 242 banking crisis (1997), 57, 209 for people and institutions to plan and prepare for risk, 40–41 community approach to risk, 142, 144, 154 inclusion construction costs for flood- and earthquake-proofing, 93 community approach to risk, 159 disclosure of balance sheet fiscal risks, 240 financial system approach to risk, 27–29, 30f, 203, 211–16 diversification of household income sources, 120 trade-offs and synergies, 30–31, 212–13f, 212–16, 214t domestic abuse, 114 income support. See also old age income support early warning system for tsunamis, 42 programs as safety nets, 157–58, 160. See also safety nets for bad financial bailouts, 33, 267 times health insurance, 23, 127t incremental approaches to global solutions, 37b, 271–73, 283–85 household approach to risk, 111 independent fiscal councils, 32, 34b, 242–43, 280–81 National Program for Community Empowerment (PNPM), 152 index-based insurance, 125–26, 126b poverty programs, community involvement in, 157 India tsunami and earthquake (2010), 53 Aadhaar identification project, 123 women employees during East Asian crisis (1990s), 178b Bhopal gas leak (1984), 178 infant mortality, 115, 115f, 144 Brimstowad Report, 79 inflation, 32, 41, 59, 229–31 Chitale Committee, 79 inflation targeting, 230, 280 community approach to risk management, 154 informal insurance, 23, 143–45, 143f community-sponsored public meetings on health and education informal sector. See enterprise sector approach to risk; formality entitlements, 25 information and communication technologies, 66b, 82, 115, 153b, cost of doing business, 182 263 decentralized electoral systems, 154 information constraints, 16, 81–82. See also asymmetric disease prevention, resource constraints on, 116, 117f information; knowledge enterprise sector approach to risk infrastructure global supply chain’s effect on worker conditions, 172b community capacity, infrastructure to complement, 155–59 labor market flexibility, 190–91, 191t financial infrastructure, 29–30, 202 resource allocation, 171 “lumpy” nature of, 87 ethnic divisions, 150 provision of basic infrastructure needed for success of risk Factories Act (1948), 190 management, 85, 85f flooding and drainage system in Mumbai, 16, 79, 80f, 99 public investment in, 233–36 health insurance programs, 127, 127t, 282 innovation and resource reallocation. See also technology household approach to risk enterprise sector approach to risk, 25, 167–68, 171, 175, 184 consumption, protection of, 111 financial instruments, 202 diversification of income sources, 120 social norms working against, 86 extended family assistance, 21 instability, reduction of. See stability low-risk, low-return crops grown by poor rural households, institutional reforms for risk management, 36–40, 99, 278–85 120 independent fiscal councils, 32, 34b, 242–43, 280–81 marriage as method of creating informal insurance, 120 macroprudential capacity, 243, 281 shocks faced by households, 55, 55t mechanisms that transcend political cycles to take long-run women in workforce and delayed decision to marry, 124–25 perspective, 41 Industrial Disputes Act (1947), 190 national risk boards, 37–40, 39f, 92–93, 99, 278–79 334 WORLD DEVELOPMENT REPORT 2014 insurance. See also financial system approach to risk; self-insurance divergent national interests, effect of, 268 in analytical framework for risk management, 17b effectiveness in resolving global risks, 262–69, 263b automobile insurance, 86, 98 enforcement of agreements, 36, 268, 272 community approach to risk, 14, 143–45, 159t examples of global actions, 258–59t deductibles, 86 financial assistance to less developed countries, 20, 267 Disaster Risk Finance Insurance (DRFI) program, 242 financial crises, 252–54 enterprise sector approach to risk, 187t humanitarian crises, 252–54 financial system approach to risk, 217t incentives flood insurance, 84, 85 aligned incentives, 36, 37b, 270–71 health insurance as detrimental, 36, 267–69 catastrophic expenses avoided with, 22 financial incentives, 273 demand for, 119 technological incentives, 273 reforms to reach most vulnerable, 126–28, 127t incremental approaches to global solutions, 37b, 271–73, 283–85 regulation of, 85 insurance, 257f, 261–62, 270t self-insurance vs. market insurance, 67 international risk boards, creation of, 272, 285 universal health insurance, 23, 41, 126, 136–37, 232 knowledge sharing, 100, 256–57, 257f, 262, 263b, 270t, 271, household approach to risk, 21, 130t 272–73 building assets for, 119 mobilization of resources, 36, 260–62 building informally through family formation, fertility, and obstacles to risk management, 263–67 marriage, 119–20 overcoming, 256–62, 257f index-based, 125–26, 126b prosperity, 33–34 informal insurance, 143–45, 143f protection of the vulnerable, 42 as safety net for low-income households, 23 protection through rules, capacity building, and coordination, shortcomings of, 145 257–60, 257f, 270t innovative insurance mechanisms, 199b recommendations and policies to improve, 38b, 269–73 international community approach to risk, 257f, 261–62, 270t priorities, 270t livestock insurance, 199b resilience, 33–35 market insurance, 67, 194, 196, 238 risks crossing national borders, 35, 44n101, 254–56 Mexican Catastrophe Climate Contingency Insurance Program, risks exceeding national capacity, 34–35, 252–56, 253f 126b systemic risk, 20b as part of preparation, 12, 13f, 14, 65f, 67–68, 68b uncertainty, 266–67, 271 public deposit insurance, 218n15 International Labour Organization, 172 public-private scheme, 90 International Monetary Fund (IMF), 225, 256, 260, 261, 262, rainfall insurance for farmers, 59, 60f 274n20 ratio of insured losses in U.S. to China, 73n16 International Organization of Securities Commissions, 215 regional insurance pools, 36, 261–62, 262b International Partnership on Avian and Pandemic Influenzas, 249b social insurance, 27, 29b, 128–29, 128f, 183–84, 281–82, 282f, international risk boards, creation of, 272, 285 285n28 international trade policies, 43n36, 256 unemployment insurance, 119, 232 investing vulnerable groups and, 145 debt and equity investments, 194 Weather Based Crop Insurance Scheme (India), 126b household approach to risk, 21–22, 115–18 Integrated Social Security Programme (India), 200 Intel investment in workers, 178 Intel investment in workers, 178 Investor Motivation Surveys, 182 intellectual property laws, 177 public investment in infrastructure, 233–36 interest group capture, 18, 89 IPCC (Intergovernmental Panel on Climate Change), 272, 285n44 Intergovernmental Panel on Climate Change (IPCC), 272, 285n44 Iraq Intergovernmental Platform on Biodiversity and Ecosystem informal credit and assistance, 143, 143f Services, 285n44 risk pooling, 24f international agreements, enforcement of, 36, 268, 272 Ireland International Association of Insurance Supervisors, 215 devaluation during recession, 231b International Civil Aviation Organization, 99 financial bailouts, 33, 241 international community approach to risk, 19, 20, 33–36, 250f, Islam 251–77 banking products, 195–96, 206, 216 agents of international community, 251, 252f zakat (charitable giving), 142b characteristics that improve risk management, 35–36 climate change, 37b, 88–89, 251, 254, 255b, 263, 266, 268, 273, J 275n53 Jamaica cohesiveness through shared preferences and objectives, 36 banking crisis, 57 coping, 257f, 260–61, 267, 268f, 270t domestic abuse, 114 Index 335 integrated risk management function in government, 39 lack of resources as obstacle to risk management, 16, 17b, 80, 81–82, long-term bonds issued by, 238 89–90, 98f, 116, 117f, 271 whistleblower laws, 92 land conflicts. See property rights Japan landslides, reducing risk of, 62 bank bailouts, 174 Lao DPR devaluation during recession, 231b community approach to risk management, 153–54 earthquake and tsunami (2011), 86, 100 households facing shocks, 55t, 113 Fukushima nuclear accident (2011), 71b, 178 La Porta, Rafael, 179 Johnson & Johnson painkiller Tylenol, 176 Latin America. See also individual countries justice contributory and noncontributory pensions, 128, 281 local dispute resolution, 150 crime’s negative effect, 4, 53, 56, 146, 146f moral hazard, 86 financial and macroeconomic crises causing public services regional or international courts, 256 reduction, 32 macroeconomic and sound monetary policies, 229 K organizational membership, 149 Kahneman, Daniel, 10 reporting crime and violence via technology, 153b Kehoe, Timothy, 175b Latvia Kenya devaluation during recession, 231b communication technology to map locations of violence, 153b integration into Baltic Stock Exchange, 216 community approach to risk, 156 workfare program, 158 disease prevention, resource constraints on, 116, 117f legal environment, 152, 177. See also justice domestic abuse, 114 Lehman Brothers, 206 independent fiscal council, 34b life-saving benefits of risk management, 5, 8, 57 manufacturing firms sharing risk with workers, 170 limited liability businesses, 170 M-PESA and M-KESHO financial projects, 199 Limited Test Ban Treaty, 37b, 284 old age pension scheme, 200b livestock insurance, 199b resilience, 8, 222–23 long-run perspective Keynesian multiplier, 233 community approach to risk, 159–60 Khan, Akhtar Hameed, 152 enterprise sector approach to risk, 184–85, 186 kinship, communities based on, 142, 145 household approach to risk, 131 kinship norms, 111b institutional mechanisms that transcend political cycles, 41 Knight, Frank, 25, 169 macroeconomic approach to risk, 229, 243–44 Knightian uncertainty, 18 in risk management design, 17b, 99–100 knowledge loss aversion, 61b in analytical framework for risk management, 17b low-income countries. See developing countries community approach to risk, 159t enterprise sector approach to risk, 187t. See also skills training M for workers financial system approach to risk, 203, 217t Maastricht treaty (1992), 97b households acquiring and sharing, 115, 130t macroeconomic approach to risk, 31–33, 224f, 225–47 international community acquiring and sharing, 256–57, 257f, adequacy of institutional capacity, 243, 281 262, 263b, 270t, 271, 272–73 automatic stabilizers, strengthening of, 235 new technologies and, 65, 66b building foundation for long-term risk management, 242–43 as part of preparation, 12, 13, 13f, 63–65, 65f contingent liabilities, management of, 239–42, 241b sharing lessons learned from mistakes of others, 99–100 coping in bad times and managing in good times, 228, 228f, translating into action, 82–85, 265–66 236–39 Korea, Republic of cyclically adjusted balance (CAB), 239b levy on bank noncore financial liabilities, 30 countercyclical and sustainable fiscal policy, 32–33, 33m, 230–35, macroeconomic and sound monetary policies, 229 234m. See also countercyclical and sustainable fiscal policy pension system and universal health coverage, 232 generating sustainable fiscal resources to finance, 236–42 unemployment insurance, 232 Disaster Risk Finance Insurance (DRFI) program, 242 Kyoto Protocol (1997), 258t financial shocks. See financial and macroeconomic crises Kyrgyz Republic fiscal rules, use of, 208, 236, 239b business environment, 182 flexibility, 242 universal health insurance, 126, 127t, 136–37, 137f flexible exchange rate regimes, 32 incentives, 242 L long-run perspective, 229, 243–44 labor and employment. See enterprise sector approach to risk market insurance, 238 lack of information and cognitive failures, 16, 80, 81–82, 98f, 271. prosperity, 31 See also data collection prudent asset management, 236–38 336 WORLD DEVELOPMENT REPORT 2014 public debt management strategies, 238 flu outbreak (2009), 66b recommendations and policies to improve, 32–33, 38b, 225, Fund for Natural Disasters (FONDEN), 199b 237t, 242–44 global systemically important financial institutions (G-SIFIs), 206 priorities, 244t health insurance programs, 127t resilience, 8, 31–32, 225–26, 236–39 household approach to risk, 21, 111 social security expenditures, 239–40 diversification of income sources, 120 sound fiscal policy expansions, 235–36 innovative insurance mechanisms, 199b stability, 31–32, 58, 226–36 integrated risk management function in government, 39 transparent and credible monetary policy, 32, 229–31, 240 long-term bonds issued by, 238 uncertainty, 226–29, 227b, 242 MultiCat Program of World Bank in collaboration with, 274n18 vulnerable, protection of, 243 National Households Survey of Financial Services Usage, 203 macroprudential regulation, 30, 62, 207–8, 231, 281 pension system, 281 macroprudential policy committees, 207, 218n47 safety net programs, 41, 129, 232 taxonomy of terms for macroprudential, 208t microeconomic policy, 73n38, 204 mad cow disease, 84, 88 microfinance institutions, 119, 154, 199–200, 203 malaria, 2f, 57, 62, 65–66, 67, 70, 94, 197b microprudential regulation and bank supervision, 230 Malawi Middle East and North Africa conditional cash transfer programs, 123 domestic violence, 124 households facing shocks, 55t financial and macroeconomic crises causing security cuts, 32 World Bank disaster protection program, 274n18 women’s labor participation rate, 23, 120, 121f Malaysia middle-income countries banking products, 206 banking systems’ resilience, 211 central bank role, 31, 41, 215 CAT DDO (Catastrophe Deferred Drawdown Option), 76, 77n1 fiscal expansions and spending programs, 235 educational attainment, lag in, 117f, 118 labor market flexibility, 179, 190 shocks affecting households, 55 macroeconomic crises, 204 universal pension and health care programs, 282 Mali migration and remittances household approach to risk, 21, 111 community treatment of migrants, 24 integrated risk management function in government, 39 desire of people in developing countries to emigrate, 59, 73n24 malnutrition, 144 electronic transfer of remittances, 195, 196b Manski, Charles, 94 foreign exchange services for sending remittances, 194 market failures, 85–87, 184 household view of, 21 market insurance, 67, 194, 196, 238 household view of migration, 121, 125 marriage as method of creating informal insurance, 120 informal insurance, remittances as, 121, 144 Maskin, Eric, 10, 36 international transmittal of remittances, 260 maternal mortality, 9b, 54b, 115 lower costs for remittances, 125 maximization of expected utility, 82, 83b resettlement of at-risk population in safe areas, 77 McDonald’s Corp., 172b seasonal migration to urban areas, 5, 59, 60f, 121 measles immunization, 21, 22f, 115, 115f Millennium Development Goals, 116, 254, 257, 263, 264b medium-term expenditure frameworks (MTEFs), 236 Miller, Merton, 27 Medium-Term Fiscal Framework (MTFF, Chile), 241b minimum wage, 182 men. See gender differences mining industry crime and, 146 hazardous worker conditions, 172 meteorological data, costs of, 82 informal enterprises, 168 Mexico labor agreements, 168 Catastrophe Climate Contingency Insurance Program, 126b regulation, 43–44n70 CAT DDO (Catastrophe Deferred Drawdown Option) bonds, missing markets and public goods, 16, 18, 79–81, 85–86 242 mobile banking, 29–30, 115, 203 conditional cash transfer programs, 68, 85 mobile phones, 115, 125, 153b, 155–56, 157f consumer protection, 29 mobilization. See also empowerment educational attainment, benefits of, 118 community approach to risk, 148 enterprise sector approach to risk international community approach to risk, 36, 260–62 dropping out of school to work, 178b monetary policy, 14, 32, 41, 226, 229–31. See also macroeconomic informal employment decreased after NAFTA, 178b approach to risk manufacturing jobs and female formal employment, 178b countercyclical policy applied to, 231–35, 232b recovery from depression (1980s), 175b fragmented approach of advanced countries, 231 social insurance tied to employment, 184 Mongolia financial system approach to risk, 201–2 domestic abuse, 114 data collection, 203 livestock insurance, 199b Index 337 Montreal Protocol (1987), 36, 37b, 258t, 272, 284 Nicaragua moral hazard declining to pursue weather-indexing program, 274n30 bailouts, 18, 210–11b, 267 diversification of household income sources, 120 bankruptcies, 182 housing and safety, 145 community approach to risk, 145 labor market regulation, 182 defined, 81t safety net programs, 129 enterprise sector approach to risk, 184 women’s educational attainment, 118 financial solution of unlimited liquidity and reduced interest Nigeria rates resulting in, 231 community risk sharing, 140 insurance, 67, 86 informal credit and assistance, 23, 24f, 143, 143f international community approach to risk, 271 noncommunicable diseases, 115, 116f public moral hazard, 90 noncontributory pensions, 128–29, 281–82, 282f Morgenstern, Oskar, 10 North Africa. See Middle East and North Africa Morocco Norway garment industry workers, 178 budget planning, 41, 280 independent fiscal council, 34b correction for cyclical influences of business cycle, 239b integrated risk management function in government, 39 sovereign wealth funds (SWFs), 238 migration programs, 125 Nuclear Non-Proliferation Treaty, 36, 37b, 269, 272, 284 multirisk approach to national risk strategy, 91b national risk assessments, 37 O mortality rates, low-income vs. high-income countries, 4, 57, 253m obstacles to risk management, 16–18, 78f, 79–103, 82f mortgages, 196, 197b behavioral failures, 10, 16, 81–85, 83b Multilateral Investment Guarantee Agency, 261 coordination failures, 81t, 89, 90–92 Mumbai flood control. See India deep uncertainty. See uncertainty Myerson, Roger, 10, 36 government failures, 18, 85–87, 89 myopia, 81t, 85, 268 households facing, 115–21 lack of information and cognitive failures, 16, 80, 81–82 N lack of resources, 16, 17b, 80, 81–82 nationalization of private sector, 181 missing markets and public goods, 16, 18, 79–81, 85–86 National Jua Kali Pension Scheme (Kenya), 200b overcoming, 98–100 national risk assessments (NRAs), 37, 91b, 95, 278–79 flexible solutions, 99–100, 280 national risk boards, creation of, 37–40, 39f, 92–93, 99, 278–79, 279f household solutions, 115–21 National Rural Employment Guarantee Act (India), 30, 202, institutional arrangements, time to create, 99. See also 218n20 institutional reforms for risk management natural disasters. See also droughts; flooding and flood protection low-cost interventions, 99 collective nature of risk management for, 87 prioritizing soft measures, 99 developing countries’ exposure to, 57 simplicity in approach, 98–99 international aid, effective use of, 272b technology tailored to local circumstances, 99 international support for disaster risk management, 261b win-win solutions, 99. See also win-win solutions loss of assets from, 4, 14–15 policy design to address, 17b macroeconomic strategy to manage liabilities associated with, political economy issues, 89, 92–93 241–42 repetition of crises when no preventive planning, 79, 80m national coordinating body to deal with, 241b sequence for assessing, 97–98, 98f preparedness. See disaster preparedness and response social and economic externalities, 18, 86 trends across regions, 9b, 54b OIE (World Organisation for Animal Health), 249, 249b, 262 warning of. See warning of natural disasters old age income support, 29b, 33, 128–29, 128f natural resources, discovery and mining of, 62. See also mining contributory and noncontributory pensions, 128, 281–82, 282f industry countries with fast growing old-age dependency ratio, 129, 282, Nestlé, 172b 283f Netherlands migration associated with, 158–59 Bureau for Economic Policy Analysis (CPB), 34b, 243, 280, 281 private pension insurance, 200b national risk assessments (NRAs), 37, 91b, 278, 279 Olson, Mancur, 10 New Deal for Engagement in Fragile States (Busan Partnership), 36, Open Data for Resilience Initiative, 263b 249b, 264b opportunity New Zealand defined, 11, 62 disclosure of balance sheet fiscal risks, 240 risk as, 4–5, 6f, 53–55, 54b KiwiSaver scheme, 129, 282 risk management unleashing, 59, 60f long-term fiscal sustainability reports, 240 Organisation for Economic Co-operation and Development universal pension and health care programs, 282 (OECD), 256 338 WORLD DEVELOPMENT REPORT 2014 Ostrom, Elinor, 150 risk aversion and, 61b ozone layer protection. See environmental protection saving deposits of the poor, 201 voice and accountability, 154 P power sector’s resilience during civil unrest in Côte d’Ivoire, 181, Pacific Catastrophe Risk Assessment and Financing Initiative, 262b, 181b 263b precautionary principle, 97b Pakistan preparation disclosure of balance sheet fiscal risks, 240 Colombia as example, 77 garment factory fire (2012), 178 components of, 12–13, 17b, 68 Orangi settlement projects, 23, 152, 154 cost-effectiveness of. See cost-effectiveness of risk management pandemic prevention. See disease prevention financial crisis preparedness for future, 208–9, 209b paradox of protection, 16, 84 household approach to risk, 115–21 Partnership of Market Readiness and the Climate, 284 incentives for people and institutions to plan and prepare for patronage system, 88b risk, 40–41, 271 pension programs. See old age income support; social insurance index across countries, 15b, 69b Peru international community approach to risk, 267, 268f, 271 budgeting, 154 as part of risk management, 3, 4b, 12–15, 63, 68, 69b community approach to risk management, 154 Philippines as example, 76–77 disaster risk management, coordination across government relationship with coping, 13f, 14, 65f, 70, 267, 268f ministries, 90 spending on, 68, 73n53 domestic violence, 146 Prescott, Edward, 175b example of family resilience in face of risk, 7b principal-agent problem, 81t financial crises, macroeconomic approach to, 225 private sector. See also public-private partnerships global systemically important financial institutions (G-SIFIs), financial crisis solutions, 209 206 systemic risk and, 20b households facing shocks, 55t working together with civil society and state, 21 language and judicial system, 86 probabilistic risk-modeling techniques, 263b long-term bonds issued by, 238 Procredit (microfinance bank), 200 macroeconomic policies procyclical bias of fiscal policy, 233, 280 resilience, 8, 222–23 productivity, 171–72, 175b sound monetary policies, 229 Programa de Educación, Salud y Alimentación (PROGRESA, mining industry Mexico), 232 hazardous worker conditions, 172 Programme for International Student Assessment (PISA), 117f, 118 informal enterprises, 168 “Project Medusa,” 172b labor agreements, 168 property rights, 89, 122, 124, 152, 180–81 regulation, 43–44n70 prosperity Philippines community approach to risk, 23 CAT DDO (Catastrophe Deferred Drawdown Option), 242 enterprise sector approach to risk, 25, 167–68, 172, 185 Civilian Emergency Administration, 76 financial system approach to risk, 27 community approach to risk management, 154 household approach to risk, 21–22 Disaster Risk Reduction and Management Act (2010), 76–77 international community approach to risk, 33–34 financial sector approach to risk, 201 macroeconomic approach to risk, 31 migration programs, 125 protection, 12, 13–14, 13f, 65–67, 65f. See also insurance mobile banking, 29–30, 203 in analytical framework for risk management, 17b preparation for disaster risk management, 76–77 community approach to risk, 145–48, 159t private retirement insurance product, 199, 200b enterprise sector approach to risk, 187t remittances to, 121, 144, 196b financial system approach to risk, 217t seasonal forecasting for agriculture, 58 household approach to risk, 130t physical assets international community approach to risk, 257–60, 257f, 270t accumulating to manage risk, 119 paradox of protection, 16 natural disasters causing loss of, 14–15 prudent asset management, 236–38, 237t political capture, 34, 202, 278, 279, 281 public action, 40–42 political economy issues, 89, 92–93, 265 public choice approach, 10 poverty. See also vulnerable people public debt management strategies, 238–39 financial tools aimed to reduce, 203 public goods, 19f, 89–90 fragile and conflict-affected countries, 254 community provision of, 23, 24, 150 income-support programs, 157–58 continuous provision of, 32 prevalence by region, 58b global, 251, 268 Index 339 missing markets and public goods, 16, 18, 79–81, 85–86 remittances. See migration and remittances “public goods game,” 149 Republic of Korea. See Korea, Republic of state provision of, 69b research priorities, refocusing of, 160 public-private partnerships reserve requirement policy, 231, 232b, 237 coordination failures, impact on risk management, 90 reservoir’s effect on mosquito population and health consequences, to deliver services and identify needy beneficiaries, 123, 123b 94 infrastructure projects, 241b resilience Vienna Initiative, 206, 216 community approach to risk, 23, 159–60 public service delivery, 123, 123b defined, 12, 63 public trading of assets, 194 in developing countries, 64b public works programs, 129–30 economic shocks, 222–23 Putnam, Robert, 23 enterprise sector approach to risk, 25–26, 168–72, 181b example of, 7, 7b R financial system approach to risk, 222–23 rape, 86 household approach to risk, 7b, 21–22 realistic approach to risk management, 17b, 98–99 international community approach to risk, 33–35, 261b reasonable worst-case scenarios, 95 macroeconomic approach to risk, 8, 31–32, 225–26, 236–39 recessions. See financial and macroeconomic crises risk management, 3 reciprocal exchange of support, 111b rinderpest (cattle plague), 249 recommendations, 36–40, 38b risk community approach to risk, 38b, 152–60 acceptable level of risk, difficulty in defining, 87–88, 88b enterprise sector approach to risk, 38b, 185–86 avoidance. See risk aversion financial system approach to risk, 38b, 216–17 as burden vs. opportunity, 4–5, 6f, 53–55, 54b flexibility, 41. See also flexibility collective nature of, 87–88 household approach to risk, 38b, 110, 122–23, 130–32, 130t defined, 11, 62 incentives for people and institutions to plan and prepare for development, relevance to, 55–59 risk, 40–41. See also incentives exposure to, defined, 11, 62 international community approach to risk, 38b, 269–73 financial crises developing out of, 204–6 long-run perspective. See long-run perspective prevalence in everyday life across developing countries, 55, 55t macroeconomic approach to risk, 32–33, 38b, 225, 237t, 242–44 trends across regions, 9b, 54b national risk board, creation of, 37–40, 39f risk aversion, 8, 59, 61b, 79, 83b, 87, 264b overcoming obstacles to risk management, 98–100 risk chain, 11–12, 12f, 62, 63f flexible solutions, 99–100 risk indicators in political economy, 92–93 institutional arrangements, time to create, 99 risk management low-cost interventions, 99 analytical framework for, 11, 17b prioritizing soft measures, 99 community. See community approach to risk simplicity in approach, 98–99 components of, 12–15, 13f, 59–70 technology tailored to local circumstances, 99 coping, 14–15, 63, 68–70. See also coping win-win solutions, 99 insurance, 14, 67–68, 68b. See also insurance protecting the vulnerable, 41–42. See also vulnerable people knowledge, 13, 63–65. See also knowledge public action, 40–42 protection, 13–14, 65–67. See also protection uncertainty and unnecessary risks, avoidance of, 40. See also cost effectiveness. See cost-effectiveness of risk management uncertainty defined, 12 redistribution of wealth in banking crises, 204 developmental benefits of, 55–59 refugees and internally displaced people, 151b effective risk management, 10–15. See also cost-effectiveness of regional cooperation, 256, 262b risk management regional insurance pools, 36, 261–62 enterprise sector approach. See enterprise sector approach to regulations, implementation and enforcement of risk finance, 29, 203 financial system approach. See financial system approach to risk labor, environment, and consumer, 20, 27, 177, 181–83 goal of, 62–63 political reluctance to impose, 92 holistic approach to, 18–21 reduction of uncertainty through, 65 household. See household approach to risk weak enforcement and compliance, 90 institutional. See institutional reforms for risk management religion. See also Islam interaction between experts and policy makers to improve, 88b conflicts related to, 147 international community approach. See international mandate to share wealth, 142, 142b community approach to risk social exclusion based on, 150 key insights on process of, 4b trust of religious leaders, 154, 155f life-saving effect of, 5, 8, 57 340 WORLD DEVELOPMENT REPORT 2014 long-run perspective. See long-run perspective Samaritan’s dilemma, 267 macroeconomic approach. See macroeconomic approach to risk sanitary conditions, 157f obstacles to. See obstacles to risk management leading to stunting, 85 opportunity unleashed by, 59, 60f social marketing to increase access in Tanzania, 158b policy design to address, 17b Saudi Arabia powerful instrument for development, 3, 5–10, 6f, 53–75. See banking products, 206 also development electronic payments system, 196b preparation as part of. See preparation sovereign wealth funds (SWFs), 238 realistic approach to, 17b, 98–99 saving instruments, 119, 194, 196, 201 resilience and. See resilience school voucher programs, 123b state approach. See state approach to risk Schumpeter, Joseph, 25, 171 terminology of, 61t Self Employed Women’s Association (SEWA), 148 risk pooling self-employment, 25, 26f, 27, 168, 169f, 170, 179 by community, 23, 24f self-insurance, 14, 19, 67, 195 by household, 21, 109–13 self-protection against risks, 66–67, 81 insurance and, 67 self-reliance, promotion of, 41–42 by multilateral organizations, 265f severe acute respiratory syndrome (SARS), 248–49 risk sharing shadow banking sector, 206 civil society, private sector, and state working together to Shleifer, Andrei, 179 manage risk, 21, 34 shocks community approach to risk, 14, 19 beneficial vs. negative, 11 enterprise sector approach to risk, 19, 25, 170 coping with, 14–15, 24f. See also coping financial sector approach to risk, 19 economic shocks. See financial and macroeconomic crises holistic approach to risk, 18–20, 19f environmental shocks, 60–62. See also climate change; natural household approach to risk, 19, 67 disasters risk-taking capital, 194 households facing, 55, 55t, 73n7, 110–14, 110f road conditions, 85 macroeconomic shocks, 228, 228f robust decision making, 95, 95f, 96b vulnerable groups, impact on, 4, 5f, 53, 57 Rogers, Michael D., 97b Singapore Rousseau, Jean-Jacques, 109 national risk assessments (NRAs), 279 rule of law, 20, 24, 85 sovereign wealth funds (SWFs), 238 rural areas water and flood management, 156, 156b Comunidades Solidarias Rurales (CSR, El Salvador), 105 Whole-of-Government Integrated Risk Management crime and violence in, 146, 147 framework, 39, 279 electronic payments, use of, 196b Sistema de Pensiones Solidarias, 240 household approach to risk, 120 skills training for workers, 130, 183–84, 185, 187t informal insurance, 160 smallpox eradication, 249, 258–59t, 269, 273, 285 National Rural Employment Guarantee Act (India), 30, 202, social and economic externalities, 18, 86 218n20 social capital, 23, 24, 148 Rural Advancement Committee (BRAC, Bangladesh), 154 social cohesion, 148. See also cohesiveness safety net programs. See safety nets for bad times The Social Contract (Rousseau), 109 sanitation improvements, 158b social exclusion based on ethnicity, 150–51, 151f shocks more adversely affecting, 55, 55t, 111, 140, 141f social insurance, 27, 29b, 128–29, 128f, 184, 281–82, 282f, 285n28 women’s lack of employment opportunities, 120 social norms Russian macroeconomic crises, 204, 212, 224f changes in, 86–87, 87f Rwanda decision making and, 84 earthquake (2008), 66b domestic violence and, 23, 86, 124f integrated risk management function in government, 39 environmental practices and, 172 safety net programs, 130 health issues and, 87–88 household support of family members, 111b S women participating in labor force, 124 safety nets for bad times social protection and insurance, 19f, 183–84, 232–33. See also social food consumption needs, 104–5 insurance household approach to risk, 129–30 expansion during downturns, 235 income-support programs, 157–58 household access to, 22, 123, 123b international community approach to risk, 261 preparation for risk, 69b macroeconomic approach to risk, 236 vulnerable groups, protection for, 126–28 state approach to risk, 23, 67 social security benefits, 123b, 232 vulnerable, protection of, 41 Somalia famine relief, 269 Index 341 South Africa public works programs, 129 automobile insurance, 98 self-employment, 25, 26f, 170 banks as distributors to social program beneficiaries, 125 susceptibility of households to risk, 56 central bank role, 207, 281 women’s ownership rights, restrictions on, 119 conditional cash transfer programs, 22, 123b subsidiarity, principle of, 20b consumer protection, 29 supply-side factors influencing access to finance, 197–200 crime and violence prevention, 165 Sweden Extended Public Works program, 129 bailouts of financial institutions, 267 government guarantee of loans, 240 independent fiscal council, 34b, 243, 281 health insurance programs, 127t systemically important financial institutions (SIFIs), 206, 209–11 noncontributory pension system, 282 systemic risk, 87–89, 226f old-age social pensions, 158 community risk sharing, 140, 141f public works programs, 130 contagion possibility, 206, 218n41 violent crime, 146–47 financial crises as result of, 205–6 South Asia. See also individual countries financial indicators, choice of, 207 child mortality, 156 financial system approach to risk, 205, 205f informal employment, 179 best practices, 207–11 measles immunization, 115, 115f possibility frontier of systemic bank resolution, 210f, 212–13 organizational membership, 149 trade-offs and synergies in finance, 214–16, 214t self-employment, 25, 26f, 170 independent fiscal council’s management of, 281 women’s labor participation rate, 23 informal insurance insufficient in face of, 145 women’s ownership rights, restrictions on, 119 macroprudential regulation for, 30 Southeast Europe and the Caucasus Catastrophe Risk Insurance matched to countries in financial crises, 207–8, 208t Facility (SEEC-CRIF), 262b size of risk, 20b sovereign wealth funds (SWFs), 237, 238, 239b, 245n45 state action protecting against, 20b, 67, 122, 122t Spain, unemployment in, 26, 167–68 stability T financial system approach to risk, 28–29, 211–12 Tanzania trade-offs and synergies, 30–31, 212–16, 214t community approach to risk, 139, 140, 142 macroeconomic approach to risk, 31–32, 226–36 ethnic divisions, 151 state approach to risk, 4b, 19, 19f, 23, 24–25 punishment of free riders, 149–50 coordination failure, 90–92 sanitation improvements, 158b essential public goods, provision of, 24, 69b domestic violence, 146 infrastructure and, 24 formal employment and wages, 177 obstacles to risk management, 89–95 household approach to risk coordination failures, 81t, 89, 90–92 elderly people, abuse of, 114 deep uncertainty. See uncertainty family sacrificing child’s education in time of disaster, 114 insufficient resources. See lack of resources as obstacle to risk low-risk, low-return crops grown by poor rural households, management 120 political economy issues, 89, 92–93 insurance schemes, 23 promoting inclusion and respect for diversity, 24–25 remittances to, 121 rule of law and, 24 safety net programs, 130 social services provision, 23 taxes systemic risk and, 20b, 67, 122, 122t carbon taxes, 273, 284t women’s power in household, 23 to finance social insurance programs, 98, 137, 184 working together with civil society and private sector, 21 during financial and macroeconomic crises, 175b, 237t, 271 Stern Review estimate of cost of climate change, 268 fuel tax, 98 stigma of failure, 86 macroeconomic approach to risk, 227b stimulus packages, 232, 234–35, 235b, 280 as obstacles to enterprises, 181, 182 Stone-Geary utility functions, 61b research and development write-offs, 86 stunting, 85 technical assistance, 262b Sub-Saharan Africa. See also individual countries technology child mortality, 156 enterprise adoption, 25 crime’s negative effect, 146, 146f global collaboration and, 5, 184–85, 273 domestic violence, 124 insurance and, 67 grandparents caring for grandchildren, 111b, 112 knowledge improving, 65, 66b informal employment, 179 leveraging use of, 66b, 79, 123 measles immunization, 115, 115f tailored to local circumstances, 99, 153b organizational membership, 149 Telenovelas (televised soap operas), 87, 87f 342 WORLD DEVELOPMENT REPORT 2014 Thailand financial system approach to risk, 211 banking crisis (1997), 29, 57, 204 household approach to risk, 132 central bank role, 207, 281 international community approach to risk, 271 domestic abuse, 114 knowledge as way of overcoming, 65 flooding affecting U.S. auto manufacture, 86 macroeconomic approach to risk, 226–29, 227b, 242 political instability’s effect on business environment, 181 risk management and, 10–11, 40, 60 remittances to, 121 robust decision making in face of, 95, 95f, 96b shadow banking sector, 206 undocumented migrants, 125 universal pension and health insurance, 23, 41, 127, 127t, 282 unemployment insurance, 119, 232 time inconsistency in government policy, 81t, 93, 128 unemployment rates, 26, 183, 245n1 Timor Leste Unilever, 172b diversification of household income sources, 120 United Kingdom Petroleum Fund, 238 consensual system, 88b “too big to fail” vs. moral hazard in bailouts, 210–11b independent fiscal council (Financial Policy Committee), 243, total factor productivity, 25 281 Toward a Safer World Initiative (UN), 249b long-term fiscal sustainability reports, 240, 280 trade-offs macroeconomic policy uncertainty, 227b financial inclusion vs. stability, 30–31, 212–13f, 212–16, 214t national risk assessments (NRAs), 37, 91b, 95, 278, 279 flexibility vs. formality, 27, 168 universal pension and health care programs, 282 insurance and protection strategies, 14 United Nations lack of resources and, 16 climate change coordination, 285 national risk assessments (NRAs), 91b, 278, 279f Conference on Environment and Development, 97b risk management, 10, 30–31, 43n14 Environment Programme, 284 trade retaliation, 43n36 Framework Convention on Climate Change, 88–89, 262 tragedy of the commons, 18, 254 Office for Disaster Risk Reduction, 261b training for workers, 130, 183–84, 185 pandemic response plans, 249b, 257 transparency United States financial system approach to risk, 206 adversarial system, 88b government regulation of business, 183 climate change measures, 273, 283 institutional, 32, 160, 229–31, 240, 241b creative destruction, 171 A Treatise on the Family (Becker), 21 flood-prone areas trust and confidence, 119, 154, 155f, 193, 201–2 awareness of risk, 16, 82 truth and reconciliation committees, 151 flood insurance policies, 84 Turkey repetition of crises in absence of effective prevention, 80m banking crisis (2000–01), 40, 57, 210–11b Hurricane Andrew (1992), 90 fiscal expansions and spending programs, 235 Hurricane Katrina (2005), 84, 85 long-term bonds issued by, 238 Hurricane Sandy (2012), 66b management of central government debt and cash reserves, independent fiscal council (Congressional Budget Office), 281 239 individual world views study, 87 transition to wage employment, 170 insured losses, 73n16 universal pension and health insurance programs, 23, 41, 126, interconnectedness of banks leading to systemic risk, 205 127t, 136, 282 long-term fiscal sustainability reports, 240 World Bank disaster advisory services, 274n18 macroeconomic policy uncertainty, 227b Tversky, Amos, 10 national risk assessments, 37, 279 Tylenol, recovery from sales drop, 176 precautionary actions in environmental regulation, 97b stimulus package as response to financial crisis, 235b U subprime crisis, 212 Uganda Thai flooding affecting U.S. auto manufacture, 86 community approach to risk management, 25, 154 vulnerable population, government spending on, 132 conflict affecting crops, 147 whistleblower laws, 92 disease prevention, resource constraints on, 116 universal health insurance. See insurance households facing shocks, 55t, 113 urban areas independent fiscal council, 34b banking crises, 204 informal credit and assistance, 24f, 143, 143f crime in, 146, 164–65 Ukraine, banking crisis (2008), 29, 204 flood prevention, 24, 154 uncertainty migration to. See migration and remittances community approach to risk, 159 research needs, 160 deep uncertainty, 16, 18, 65, 70, 81t, 89, 93–95, 266–67 shocks, 55, 55t, 62, 111, 140 enterprise sector approach to risk, 181, 186 urban poor, 105 Index 343 Uruguay water insecurity and water management, 85f, 156b, 157f, 254, 256 global systemically important financial institutions (G-SIFIs), weak states, violent conflicts in, 147 206 wellness programs, 26, 171 long-term bonds issued by, 238 West Africa. See also individual countries utility theory, 10, 61b, 83b informal employment, 177 whistleblowers, 92 V Whole-of-Government Integrated Risk Management framework vertical coordination, 90 (Singapore), 39 Vienna Initiative, 206, 216 Wiener, Jonathan B., 97b, 278 Vietnam win-win solutions, 10, 59, 99, 154, 158 flooding women flood protection, costliness of, 16, 89 empowerment of, 124, 138f robust decision making to deal with, 96b equality with men. See gender differences pension and health insurance programs, 127t, 282 manufacturing jobs and female formal employment, 178b violence. See also crime; domestic violence microfinance self-help organizations, 154 community approach to risk, 146–48 worker protection, 25–26, 27, 171–72, 172b, 177, 178b, 185 communication technologies, use of, 153b workfare programs, 41, 158, 235 epidemics, incidence of, 54b World Bank types of and spillovers, 147f avian flu (H5N1) emergency response, 249b voice, 149, 154–55 Catastrophe Deferred Drawdown Option (CAT DDO), 76, 77n1, Voices of the Poor, 154 242, 261, 274nn19–20 voluntary organizations, membership in, 148–49, 149f Doing Business reports, 182, 202 von Neumann, John, 10 financial development and major disasters, 260, 274n18 vulnerability Global Emerging Markets Local Currency Bond program, 216 cost of coping with, 14, 70, 112 Immediate Response Mechanism, 274n19 definition of, 11, 61t, 62, 73n38, 132 knowledge-sharing programs, 256 household limitations leading to, 121, 130 Water and Sanitation Program, 158b macroprudential tools, effect on, 208t World Development Report 2012: Gender Equality and Development, obstacles to risk management, effect on, 110 175 policy design and, 95, 95f World Development Report 2013: Jobs, 174 to poverty, 58b World Economic Forum, 39, 278 research priority to investigate underlying causes of, 160 World Health Organization (WHO), 249b uncertainty and, 18 World Organisation for Animal Health (OIE), 249, 249b, 262 vulnerable people World Trade Organization, 272 community approach to risk, 24, 160 World Values Survey, 55, 154, 167 exclusion of most vulnerable, 145 world views, 87–88, 88b research priorities, refocusing of, 160 enterprise sector approach to risk, 186 Z household approach to risk, 110, 112, 131–32 Zambia insurance provided by community and, 145 employee safety and formal employment, 177 international community approach to risk, 42, 285 formal employment and wages, 177 macroeconomic approach to risk, 243 resource constraints on disease prevention, 116, 117f poverty traps and, 57, 58b, 81 Zimbabwe risk management to encourage protection of, 41–42 drought, 114 shocks, impact on, 4, 5f, 53, 57 manufacturing firms sharing risk with workers, 170 W supplier contracts, 177 warning of natural disasters, 8, 52f, 57–58, 76, 79, 99, 266, 274n13 community approach to risk, 146 household approach to risk, 115 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. The • 136 trees Publishing and Knowledge Division has cho- • 61 million Btu of total sen to print the World Development Report energy 2014: Risk and Opportunity—Managing • 11,704 pounds of net Risk for Development on recycled paper with greenhouse gases 50 percent postconsumer fiber in accordance • 63,481 gallons of with the recommended standards for paper waste water usage set by the Green Press Initiative, a non- • 4,249 pounds of solid profit program supporting publishers in using waste fiber that is not sourced from endangered for- ests. For more information, visit www.green- pressinitiative.org,