97881 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean i Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Louise Cord, María Eugenia Genoni, and Carlos Rodríguez-Castelán, editors WORLD BANK GROUP © 2015 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 1 2 3 4 18 17 16 15 This work is a product of the staff of The World Bank with external contributions. The find- ings, 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 repre- sent. 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 IGO license (CC BY 3.0 IGO) http://creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribu- tion 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: Cord, Louise, María Eugenia Genoni, and Carlos Rodríguez-Castelán, editors. 2015. Shared Prosperity and Poverty Eradication in Latin America and the Caribbean. World Bank, Washington, DC. doi:10.1596/978-1-4648-0357-4. License: Creative Commons Attribu- tion CC BY 3.0 IGO 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. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the con- tent contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Divi- sion, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-0357-4 ISBN (electronic): 978-0-4648-0358-1 DOI: 10.1596/978-1-4648-0357-4 Cover image: “Caminos de Bojaca” (1984) © Gonzalo Ariza / World Bank Art Program. Further permission required for reuse. Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword xvii Acknowledgments xix About the Editors and Authors xxi Abbreviations xxv 1. Overview 1 Louise Cord, María Eugenia Genoni, and Carlos Rodríguez-Castelán Introduction 1 Transformational Change in Living Standards in the Region 3 The Asset-Based Approach to Gauging Household Income 16 The Income Generating Capacity of the Less Well Off 20 Final Remarks 43 Notes 44 References 47 2. Shared Prosperity and Poverty Reduction in Urban Argentina 53 Santiago Garriga, Emmanuel Skoufias, and Liliana D. Sousa Introduction 53 Background 54 Diagnostics 56 Policy Discussion 63 Notes 71 References 74 3. Poverty and Shared Prosperity in Brazil: Where to Next? 77 Javier E. Báez, Aude-Sophie Rodella, Ali Sharman, and Martha Viveros Introduction 77 v vi Contents The Impressive Pace of Poverty Reduction 79 A Positive Performance, but Challenges Remain 86 What Is Behind the Rapid Reduction in Poverty? 89 The Challenges Ahead and the Role of Policy in Poverty Reduction 93 Final Remarks 108 Annex 3A Labor Market Characteristics, Brazil 109 Notes 110 Bibliography 111 4. Toward Shared Prosperity in Colombia 115 Lea Giménez, Carlos Rodríguez-Castelán, and Daniel Valderrama Introduction 115 Background 116 Building the Foundations of Shared Prosperity: Recent Trends 118 The Drivers of the Observed Changes in Poverty and Inequality 125 Boosting Shared Prosperity 129 Final Remarks 141 Annex 4A Decomposing Poverty Reduction 142 Annex 4B Incomes and the MPI 143 Notes 146 References 149 5. Shared Prosperity and Opportunities in El Salvador 155 Megan Rounseville, Mateo Salazar, and Kinnon Scott Introduction 155 Poverty, Shared Prosperity, and Inequality: Levels and Trends 156 Who Has Moved Out of Poverty in the Past Decade? 162 What Has Driven Poverty and Inequality Reduction? 165 Bringing about Change in Welfare and Shared Prosperity 171 Final Remarks 184 Annex 5A Supplementary Data 186 Notes 190 References 192 6. Is Mexico on the Path to Shared Prosperity? 195 Kiyomi Cadena, Kinnon Scott, and Erwin R. Tiongson Introduction 195 The Macroeconomic Context 196 Welfare Trends over the Past 20 Years 199 Drivers of the Trends in Welfare 207 Policy Channels for Poverty Reduction 214 Final Remarks: The Policy Challenge 226 Annex 6A Data Sources 228 Annex 6B Migration and the Labor Force 232 Notes 233 References 238 Contents vii 7. Poverty and Shared Prosperity in Paraguay 245 Santiago Garriga, Luis F. López-Calva, María Ana Lugo, Alejandro Medina Giopp, Miriam Müller, and Liliana D. Sousa Introduction 245 Trends in Poverty and Shared Prosperity 246 Drivers behind the Trends 249 Key Challenges 256 Final Remarks 265 Notes 266 References 267 8. Steering toward Shared Prosperity in Peru 269 María Eugenia Genoni and Mateo Salazar Introduction 269 Outstanding Performance in Poverty Reduction 270 Inequality Has Narrowed, but Remains Significant 276 Economic Growth: The Main Driver of Improvement 278 Opportunities to Boost Shared Prosperity 282 Final Remarks 295 Annex 8A Poverty Rates and Gini Coefficients 296 Annex 8B Profile of the Poor and the Bottom 40 297 Annex 8C Macrodata 297 Annex 8D Decomposition of Changes in Extreme Poverty and Inequality by Income Components 298 Notes 299 References 300 9. Poverty and Shared Prosperity in Uruguay 303 Oscar Barriga Cabanillas, Marina Gindelsky, María Ana Lugo, Carlos Rodríguez-Castelán, and Liliana D. Sousa Introduction 303 Trends in Growth, Poverty, and Shared Prosperity 304 Drivers of the Reductions in Poverty and Inequality 309 Key Challenges 317 Final Remarks 322 Notes 323 References 324 Boxes 1.1 Poverty Trends in the Caribbean 10 1.2 Stagnation in the Contraction of Income Inequality in the Region 17 1.3 Explaining the Decline in Labor Force Participation among the Bottom 40 26 1.4 Connectivity Infrastructure in Latin America and the Caribbean 27 1.5 The Asset-Based Approach: Indigenous Populations 30 1.6 The Poverty Effects of High Food Prices, Paraguay 35 viii Contents 1.7 Shocks, Coping, and the Impact on Household Welfare, Haiti 38 3.1 Poverty Measurement in Brazil 80 3.2 The Bolsa Família Program, Brazil 91 3.3 The Brasil sem Misería Plan 96 3.4 The National Education Plan, Brazil 99 4.1 The Growth of the Middle Class in Colombia and the Region 120 5.1 Inequality and Shared Prosperity: From Statistics to Experiences in San Salvador 160 6A.1 Income Aggregates, Mexico 231 8.1 Comparing Mobility Out of Poverty in Peru and the Region 275 Figures 1.1 Socioeconomic Composition of the Population, Latin America and the Caribbean, 2003 and 2012 5 1.2 Shared Prosperity: Annualized Income Growth, Developing Regions, around 2006–11 6 1.3 Average GDP Growth Rates, Latin America and the Caribbean, 1990–2013 7 1.4 Trends in the Gini Coefficient, Latin America and the Caribbean, 2003–12 8 1.5 Extreme Poverty Rates, Latin America and the Caribbean, 2003–12 11 1.6 Composition of the Bottom 40, Latin America and the Caribbean, 2003 and 2012 13 1.7 Income Growth among the Bottom 40, Latin America and the Caribbean, around 2003–12 14 1.8 Income Growth, Bottom 40 and the Entire Population, Latin America and the Caribbean, around 2003–12 15 1.9 Contributions of Growth and Redistribution to Falls in Extreme Poverty, Latin America and the Caribbean, around 2003–12 16 B1.2.1 Gini Coefficient: Annualized Changes, Latin America and the Caribbean, 2003–10 and 2010–12 17 1.10 The Asset-Based Approach to the Generation of Household Market Income 19 1.11 Labor Income, Bottom 40 and Top 60, Latin America and the Caribbean, around 2012 21 1.12 The Reduction in Extreme Poverty, by Income Component, Latin America and the Caribbean, 2003–12 22 1.13 Educational Attainment, Bottom 40 and Top 60, Latin America and the Caribbean, around 2003–12 23 1.14 Completion of Sixth Grade on Time, Latin America and the Caribbean, 2000–12 24 1.15 Gaps in Labor Force Participation, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 25 B1.4.1 Electricity Coverage Rates, Latin America and the Caribbean, 2000–12 27 Contents ix B1.4.2 Cell Phone Coverage Rates, Latin America and the Caribbean, 2000–12 28 B1.4.3 Internet Coverage Rates, Latin America and the Caribbean, 2000–12 28 1.16 The Rise in Hourly Wages, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 29 B1.5.1 $2.50 and $4.00-a-Day Poverty Rates, Indigenous Populations, Latin America and the Caribbean, 2000–12 30 1.17 Transfers, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 32 1.18 Food Consumption in Total Consumption, Latin America and the Caribbean, around 2010 34 B1.6.1 Changes in the Extreme Poverty Rate, Paraguay, 2003–11 and 2011–13 35 1.19 Incidence and Poverty Effects of Natural Disasters, World Regions and Latin America and the Caribbean, 1970–2009 36 1.20 Shocks Reported by the Bottom 40 and Top 60, Peru, 2013 37 1.21 Policy Areas That Affect Household Income Generating Capacity 40 2.1 Poverty Rates and the Share of Income Held by the Bottom 40, Argentina, 1991–2012 55 2.2 Poverty Headcounts, Urban Areas, Argentina, 2004–12 56 2.3 Annualized Income Growth Ratio, the Bottom 40, Urban Argentina vs Region, 2003–12 57 2.4 Trends in Inequality, Urban Argentina and the Region, 2004–12 58 2.5 Employment Profile, Argentina, 2004–12 59 2.6 Average Monthly Earnings, by Gender and Educational Attainment, Argentina, 2004 and 2012 60 2.7 Improved Sanitation: Disparities in the HOI and Coverage, by Location and Region, Argentina, 2012 62 2.8 Sector of Employment, by Educational Attainment, Argentina, 2012 65 2.9 Government Spending as a Share of GDP, Argentina, 2003 and 2009 67 2.10 The Impact of the Pension Moratorium on Pension Coverage and Poverty, Argentina, 2004–12 68 2.11 Public Transfers, Households with Children and Low Educational Attainment, Argentina, 2004–12 69 3.1 Poverty Lines, Brazil, 1999–2013 81 3.2 The Reduction in Poverty, by State, Brazil, 2001–12 82 3.3 Matrix of Multidimensional and Income Poverty, Brazil, 2004 and 2012 83 3.4 Convergence in Poverty Reduction, Brazil, 2004–2013 84 3.5 The Poor, the Vulnerable, and the Middle Class, Brazil and the Region, 2004 and 2012 84 3.6 Shares of the Country’s Bottom 40, by State and Macroregion, Brazil, 2012 87 x Contents 3.7 Income Growth, Bottom 40 and Overall Population, by State, Brazil, 2002–12 88 3.8 Trends in Inequality, Brazil, 2001–12 88 3.9 Income Distribution, Brazil, 2012 89 3.10 Annualized GDP per Capita Growth Rate, Latin America and Caribbean, 1999–2012 90 3.11 Annualized Growth Incidence Curve, Brazil, 2001–12 90 3.12 Formal and Informal Jobs, Brazil, 2001–11 92 3.13 Income Components in the Decline in Poverty, Brazil, 2003–12 93 3.14 The Gini Coefficient before and after Government Transfers and Taxes, Brazil, 2009 97 3.15 Ratio of the Share of Taxes Paid to the Share of Total Market Income, Brazil, 2009 98 3.16 Trends in Selected Opportunities, Service Coverage, Brazil, 1981–2012 100 3.17 The HOI for Completing Grade 6 on Time, by State, Brazil, 2012 101 3.18 Attendance in Secondary School, 13–17 Age Group, by Income Decile, Brazil, 2012 102 3.19 Households Connected to Sewerage Networks, by Income Decile, Brazil, 2012 103 3.20 Share of Informality, by Location and Welfare-Consumption Group, Brazil, 2012 104 3.21 Labor Productivity per Person Employed, Selected Countries and Regions, 2012 105 3.22 Labor Productivity, the Real Average Wage, and the Minimum Wage, Brazil, 2004–14 107 3.23 Use of Financial Instruments, Region vs. Brazil, 2011 107 4.1 The Reductions in Moderate, Extreme, and Multidimensional Poverty, Colombia, 2002–13 119 B4.1.1 The Growth of the Middle Class, Latin America and the Caribbean, 2002–12 120 4.2 The Incidence of Poverty and Extreme Poverty, by Population and Urban or Rural Location, Colombia, 2002–13 121 4.3 Annualized Growth Rate in Income, the Bottom 40 and the Mean, Colombia, 2002–13 123 4.4 Improvements in the SPI, by Department, Colombia, 2002–13 124 4.5 The Gini Coefficient, Selected Countries, Latin America and the Caribbean, 2002–12 125 4.6 Inequality, by Income Source, Colombia, 2002–13 128 4.7 The Impact of Fiscal Policy on the Gini Coefficient, Four Countries in Latin America, 2009 130 4.8 The Distribution of Monetary Transfers, by Income Decile, Colombia, 2010 131 4.9 The Concentration Index of Public Spending, Colombia, 2010 131 4.10 The HOI, Colombia, 2012 133 4.11 Index of Relative Service Coverage, Colombia, 2008 and 2012 134 Contents xi 4.12 Indicators of Access to Financial Institutions, Selected Countries of Latin America 138 4B.1 Income, by Source and Income Quintile, Colombia, 2002–13 143 4B.2 Growth Incidence Curves of per Capita Income, Colombia, 2002–13 144 4B.3 Income Source Contributions to Moderate and Extreme Poverty Reduction, Colombia, 2002–13 145 5.1 Poverty Rates, El Salvador, 2000–12 156 5.2 Shared Prosperity, Central America and the Region, 2004–12 158 5.3 Trends in the Gini Coefficient, El Salvador, 2000–12 160 5.4 Intragenerational Mobility, El Salvador, 2004, 2007, and 2012 162 5.5 Income Growth Rate, by Decile, El Salvador, 2000–12 163 5.6 Poverty Mobility, El Salvador, 2004, 2007, and 2012 164 5.7 Household Income Growth, by Mobility Category, El Salvador, 2004, 2007, and 2012 165 5.8 The Human Opportunity Index and Basic Service Access, El Salvador and the Region, 2000–12 166 5.9 Annual per Capita GDP Growth Rate, Central America, 2000–12 167 5.10 The Decomposition of Poverty Reduction, El Salvador, 2000–12 168 5.11 Remittance Inflows, El Salvador, 1976–2012 169 5.12 Remittance Recipients and Nonrecipients, the Poor, and the Nonpoor, El Salvador, 2012 170 5.13 Ratio of Private Transfers to Total Income, by Decile, El Salvador, 2000–12 171 5.14 The Contribution of Income Components in Reducing the Gini Coefficient, El Salvador, 2000–12 172 5.15 Incidence of Spending on Social Programs and Universal Subsidies, El Salvador 173 5.16 Sectoral Employment and Productivity, El Salvador, 2000–12 174 5.17 Unemployment, El Salvador, 2000–12 175 5.18 Households with Access to Water and Electricity, by Poverty Status, El Salvador, 2012 176 5.19 Municipal Competitiveness Indicators, El Salvador, 2009–13 177 5.20 Trust in Government Institutions, El Salvador, 2004–12 178 5.21 Confidence in the Government to Achieve Poverty Reduction and Citizen Security, El Salvador, 2004–12 179 5.22 Perceptions of Political Agency, El Salvador, 2008–12 179 5.23 Homicide Rates, Central America and Mexico, 1995–2013 180 5.24 Changes in Behavior because of Crime, El Salvador, 2012 181 5.25 Effects of Crime and Violence on Businesses, El Salvador, the Region, and the World, 2010 182 5.26 Correlation: Homicide Rates in El Salvador and Unaccompanied Child Migrants from El Salvador at the U.S. Border 182 5.27 Crime Rates and the Poverty Rate, by Department, El Salvador, 2013 183 xii Contents 5.28 Emigration from El Salvador, 2004–12 185 5A.1 Rate of Growth of Private Transfers, by Decile, El Salvador, 2000–2012 188 5A.2 Trust in Government Institutions, El Salvador 189 6.1 Economic Indicators, Rank among Upper-Middle-Income Countries, the Region, and the World, Mexico, 1992–2013 196 6.2 Annual GDP Growth Rate, Mexico, 1991–2013 197 6.3 Postcrisis Economic Performance, Mexico, 1995–99 and 2009–14 198 6.4 Index of Real GDP and GDP per Capita, Mexico, 1990–2013 199 6.5 Trends in Monetary Poverty, Mexico, 1992–2012 200 6.6 Income Beta Convergence, Mexico, 1990–2010 201 6.7 Municipal Growth Incidence Curve, Mexico, 1990–2010 202 6.8 Trends in Income Inequality, Mexico, 1996–2012 204 6.9 Municipal Beta Convergence in Inequality, Mexico, 2000–10 205 6.10 Trends in Nonmonetary Well-Being, Mexico, 1990–2010 206 6.11 Trends in the Share of the Population Facing Social Deprivations, Mexico, 1990–2010 207 6.12 The Decomposition of Changes in Income Poverty (Extreme), Mexico, 2006–12 208 6.13 Unemployment and Underemployment Rates, by Gender, Mexico, 2005–14 210 6.14 Labor Earnings Relative to the Minimum Wage, Mexico, 2005–13 211 6.15 The Labor Income Poverty Index, Mexico, 2005–14 212 6.16 Changes in Employment and Productivity, by Sector, Mexico, 2000 and 2011 213 6.17 Homicides, Mexico, 1997–2011 213 6.18 Annual Growth in Remittances, Mexico, 2000–13 214 6.19 Federal Income, Tax Revenue, and Redistributive Spending, Mexico, 1990–2013 215 6.20 Informality Rate, by Gender, Mexico, 2005–13 218 6.21 Rigidity in the Employment Index, Regions and Selected Countries, 2012 219 6.22 Financial Service Use, Mexico, 2011 220 6.23 The HOI, by State, Mexico, 2000 and 2010 222 6.24 Quality of Education, 2012 PISA Results, OECD and the Region, 2012 224 6.25 Natural Disasters and Persons Affected, Mexico, 1940–2019 225 6B.1 Migration Patterns, Mexico, 1991–2009 233 7.1 GDP per Capita, Poverty Rates, and Inequality, Paraguay, 2003–13 246 7.2 Income Growth, the Bottom 40, Paraguay, 2003–13 247 7.3 Composition of the Population, by Socioeconomic Status, Paraguay, 2003–13 248 7.4 Multidimensional Poverty and Income Poverty Indicators, Paraguay, 2003–13 249 7.5 Per Capita Household Income Distribution, Paraguay, 2003, 2011, and 2013 250 Contents xiii 7.6 Changes in the Extreme Poverty Rate, Paraguay, 2003–11 and 2011–13 251 7.7 Growth Incidence Curves, Paraguay, 2003–11 and 2011–13 252 7.8 Decomposition of Changes in Extreme Poverty Rates, by Rural and Urban Location, Paraguay, 2003–13 253 7.9 Labor Income Growth and the Wage Employment Rate, Paraguay, 2003–13 253 7.10 Tekopora Transfers and Changes in Extreme Poverty without Family Transfers, Paraguay, 2003–13 254 7.11 Earnings among Workers with Incomplete Primary School and Employment, by Employer Type, Paraguay, 2003–13 255 7.12 Primary Sector Income and Rural Employment Sectors, by Poverty Status, Paraguay, 2013 257 7.13 Urban Unemployment and Employment Sectors, by Poverty Status, Paraguay, 2003–13 258 7.14 Monthly Earnings as a Share of the Minimum Wage, by Employment Type, Paraguay, 2013 259 7.15 Access to Sanitation and Piped Water in the Home, Children, Paraguay, 2013 260 7.16 Comparative Test Scores among Sixth Graders, Latin America and the Caribbean, 2006 261 7.17 Overall Inequality and the Inequality of Opportunity, Paraguay, 2003–13 262 7.18 Comparative Redistribution Effectiveness of Fiscal Systems, Latin America and the Caribbean, 2009 264 8.1 Total and Extreme Poverty Rates, Peru, 2004–13 271 8.2 Changes in the Extreme Poverty Rate, by Region, Peru, 2004–13 272 8.3 The Extreme Poor in Urban and Rural Areas, Peru, 2004–13 273 8.4 Households with Multiple Nonmonetary Deprivations, Peru, 2004 and 2013 274 8.5 Chronic Poverty, Peru, 2007 and 2010 274 B8.1.1 Share of the Poor, Vulnerable, and Middle Class, Peru and the Region, around 2004 and 2012 275 8.6 The Gini Coefficient, Urban and Rural Areas and Nationwide, Peru, 2004–13 276 8.7 Shared Prosperity: Mean Annual Growth in Average Income, by Region, Peru, 2004–13 277 8.8 Distribution of the Bottom 40, by Region, Peru, 2013 277 8.9 GDP Growth, Peru, 2000–13 279 8.10 Growth-Poverty Elasticity, by Geographical Region and Urban or Rural Area, Peru, 2004–13 279 8.11 Labor Market Performance, Peru, 2004 and 2012 280 8.12 Growth Rates Needed to Achieve Sen’s Welfare Index Benchmark in GDP per Capita and the Gini by 2030, Peru 281 8.13 Changes in per Capita Public Expenditures, by District Household Consumption, Peru, 2007–11 283 xiv Contents 8.14 Access to Water, Sanitation, Electricity, and Telephone, Peru, 2004 and 2013 285 8.15 Index of Utility Coverage Rates, by Region, Peru, 2004 and 2013 286 8.16 Access to Water, Sanitation, Electricity, and Telephone Services, Urban and Rural Areas, Peru, 2013 288 8.17 The Human Opportunity Index, Peru, 2012 289 8.18 Dependent Workers, by Firm Size and Income Decile, Peru, 2004 and 2013 291 8.19 Workers in the Pension System, by Income Decile, Peru, 2004–13 292 8.20 The Population with Financial Access Points in the District of Residence, Peru, 2013 292 8.21 Access to Financial Services, Bottom 40 and Top 60, Peru, 2011 293 8.22 Events Resulting in Household Income or Asset Loss during the Previous Year, Peru, 2013 294 8.23 Type of Events Reported, Bottom 40 and Top 60, Peru, 2013 295 8D.1 The Reduction in Extreme Poverty, by Income Components, Peru, 2004–12 298 8D.2 The Reduction in Inequality Measured by the Gini, Peru, 2004–12 298 9.1 Real Growth of Gross Domestic Product, Uruguay, 1990–2013 305 9.2 Trends in the Poverty Headcount, Uruguay, 2002–12 306 9.3 Growth Incidence Curves of per Capita Household Income, Urban Areas, Uruguay, 2000–07 307 9.4 Growth Incidence Curves of per Capita Household Income, Uruguay, 2007–12 308 9.5 Inequality, Uruguay, the Region, and the OECD, 2006–13 309 9.6 Socioeconomic Groups, by Poverty Status, Uruguay, 2002–11 310 9.7 Decomposition of Shifts in Moderate Poverty, Urban Areas, Uruguay, 2003–12 311 9.8 Unemployment and Formal Employment, Uruguay, 2006–12 312 9.9 Employment and Participation Rates, by Skill Level, Uruguay, 2007–12 313 9.10 Sector of Low-Skilled Employment, Uruguay, 2007–12 314 9.11 Real Wage Index, Uruguay, 2000–13 315 9.12 Growth in Real Mean Monthly Labor Earnings, by Skill Level, Uruguay, 2007–12 316 9.13 The Gini Coefficient and Pre- and Postfiscal Incomes, Uruguay, 2009 318 9.14 The Impact of a Crisis on Poverty and Inequality, Uruguay, 2011–14 319 9.15 Households in the Crisis Scenario in Year 2 (2014), Uruguay 320 9.16 15- to 18-Year-Olds Not in School and Not Working, by Gender, Uruguay and the Region, 2000–12 321 Maps 1.1 Heterogeneity in Living Standards, Bolivia and Peru, 2007 and 2011 12 Contents xv 5.1 Extreme Poverty, by Department, El Salvador, 2012 157 5.2 Bottom 40, by Department and Mean Income Growth, El Salvador, 2000–12 159 6.1 Extreme Poverty Headcount, Mexico, 2012 203 Tables 1.1 Extreme Poverty Rates, Developing Regions, 2002 and 2011 4 1.2 Bottom 40 and Top 60: Household Characteristics, Latin America and the Caribbean, 2003 and 2012 6 3.1 Profile of the Extreme Poor, the Poor, and the Nonpoor, Brazil, 2012 85 3A.1 Labor Market Characteristics, Brazil, 2004 and 2012 109 4.1 Indicators of Inequality, Colombia, 2002–13 124 4A.1 Participation in Poverty Reduction, Intrasectoral Effect and Intersectoral Effect, Selected Household Characteristics, Colombia, 2002–13 142 5A.1 Profile of the Poor, El Salvador, 2012 186 5A.2 Remittance Recipients and Nonrecipients, El Salvador, 2000 and 2012 187 5A.3 Change in the Employment Mix among the Nonpoor, El Salvador, 2004–12 188 5A.4 Change in the Employment Mix among the Poor, El Salvador, 2004–12 189 6.1 Food and Asset Poverty, by Area and Indigenous Status, Mexico, 2012 203 6.2 Multidimensional Poverty Measurement and Social Deprivation Indicators, Mexico, 2010–12 205 6A.1 Measures of Poverty, by Data Source, Level of Disaggregation, and Availability, Mexico 230 B6A.1.1 Differences between the Traditional ENIGH and the MCS-ENIGH, Mexico 231 6B.1 Labor Force Growth Rates, Mexico, 2005–10 232 8A.1 Official Poverty Rates and Gini Coefficients, Nationwide and Urban and Rural Areas, Peru, 2004–13 296 8B.1 Average Characteristics of the Extreme Poor, the Poor, and Others, Peru, 2013 297 8C.1 GDP and Fiscal Data, Peru, 2004–13 297 9.1 Sectoral Output and Employment Growth, Uruguay, 2003–13 313 Foreword T he Latin America and the Caribbean Region has seen marked and criti- cal progress for its people over the last decade. Extreme poverty has been halved; inequality has declined; and the growth rate among the bot- tom 40 percent of the population in the region eclipses the performance of that group in every other region in the world. These are all great strides that have helped transform the socioeconomic makeup of the region and grow the middle class to unprecedented levels. Continuing with the status quo, however, will not be enough, and the last decade’s progress is at risk in the face of the global economic slowdown and declining incomes across the region. Moreover, with 75 million people still living in extreme poverty and nearly two-thirds of the population either poor or vulnerable to falling into poverty, the region has not yet enabled and harnessed the full potential of all of its people. A persistent lack of opportunities, quality basic services, and good jobs has kept many of the poor in poverty, and made it harder to break the cycle of poverty and vul- nerability between generations. The region’s overall advances mask significant differences between coun- tries, with strong performers canceling out some of the losses of those that were perhaps less successful in reducing poverty and boosting the welfare of the least well off. And, even in countries where progress has been sub- stantial, poverty is often persistent and geographically concentrated. Take Peru, for example, one of the countries that has done quite well in reducing poverty over the last 10 years. Just one-third of the country’s population lives in rural areas; however, those same areas account for half of the poor and 80 percent of the extreme poor. It is important to keep in mind that Latin America and the Caribbean includes countries with varying levels of development, and thus the compo- sition of the bottom 40 percent and the impact of growth on this group may xvii xviii Foreword look markedly different from country to country. Some of the strongest performers, Argentina, Bolivia, Brazil, and Panama, saw income growth rates among the bottom 40 at well over 7 percent. Compare this to some of the weakest performers, Guatemala and Mexico, which saw growth rates among the bottom 40 of –1.0 and 1.3 percent, respectively. Shared Prosperity and Poverty Eradication in Latin America and the Caribbean takes a closer look at the region, presenting eight country case studies to better understand where poverty persists and how best to design policies and programs that will reach the least well off both today and in the years to come. This country-specific approach helps offer tailored analysis for countries, taking into account their socioeconomic structure, progress on the World Bank Group’s twin goals, and level of development, rather than applying the region’s overall good performance to each country uniformly. As the World Bank Group continues to work with its partners to end poverty by 2030 and boost shared prosperity around the world, knowing who remains poor and vulnerable and how to increase the welfare of the bottom 40 percent in each country will be crucial. Policies and programs, to be effective, cannot be designed with no evidence to support them, or targeted solely on the basis of what we think might work. This study will help policy makers do a better job of building on the last decade’s progress, promoting growth and incomes regardless of the global slowdown, and moving forward into an even more successful decade to come for the people of Latin America and the Caribbean. Jorge Familiar Ana Revenga Vice President, Latin America Senior Director, and the Caribbean Poverty Global Practice World Bank Group World Bank Group Acknowledgments T his set of country case studies has been produced by World Bank experts working in the Poverty Global Practice (GPVDR), Latin America and the Caribbean Region, the World Bank. The coordination of the country studies has been led by Louise Cord, María Eugenia Genoni, and Carlos Rodríguez-Castelán. The core team members are Giselle Del Carmen, Stephanie Majerowicz, and Daniel Valderrama. The team benefited from valuable inputs provided by Alan Fuchs, Santiago Garriga, Lea Giménez, María Ana Lugo, and Martha Viveros. Administrative support was supplied by Karem Edwards. Robert Zimmermann conducted editorial reviews. Publishing and distribution sup- port was assured by Mark Ingebretsen, Patricia Katayama, and Marcela Sánchez-Bender. The work was carried out under the direction of Louise Cord, Augusto de la Torre, Humberto López, and Ana Revenga. The peer reviewers were Francisco Galrão Carneiro, Wendy Cunning- ham, Samuel Freije-Rodríguez, Michele Gragnolati, Magnus Lindelow, Kathy A. Lindert, Gladys López-Acevedo, Luis F. López-Calva, Julian Mes- sina, Zafer Mustafaoglu, Jamele Rigolini, Peter Siegenthaler, Emily Sinnott, Miguel Székely, and Renos Vakis. The team also received useful observa- tions from Javier E. Báez and Daniel Lederman, and it benefited from inter- nal discussions with members of Poverty Global Practice, Latin America and the Caribbean Region. The authors of the chapters would like to express their appreciation of the following: Javier E. Báez, Augusto de la Torre, Daniel Lederman, Julian Messina, Ana Revenga, and Miguel Székely (Overview); Pablo Acosta, Louise Cord, Zafer Mustafaoglu, Rafael Rofman, Emily Sinnott, and Trang Van Nguyen (Argentina); Rita Almeida, Oscar Barriga Cabanillas, Louise Cord, Michael Drabble, Cornelius Fleischhaker, Magnus Lindelow, Luis F. López-Calva, Miriam Müller, Elizaveta Perova, Carlos Rodríguez-Castelán, xix xx Acknowledgments Philip Schellekens, Joana Silva, Emmanuel Skoufias, and Anna Wellenstein (Brazil); Tania Díaz-Bazán, Patricia Caraballo, Giselle Del Carmen, Lou- ise Cord, Mauricio Cuellar, Eric Dickson, María Eugenia Genoni, Samuel Freije-Rodríguez, Stephanie Majerowicz, and Mary Alexander Sharman (Colombia); Pablo Acosta, Oscar Calvo González, Kathy Lindert, Mateo Salazar, Liliana D. Sousa, and Miguel Székely (El Salvador); Louise Cord, Wendy Cunningham, Samuel Freije-Rodríguez, Gladys López-Azevedo, Luis F. López-Calva, Carlos Rodríguez-Castelán, Isidro Soloaga, and Miguel Székely (Mexico); Louise Cord, María Eugenia Dávalos, Caro- lina Díaz-Bonilla, Michele Gragnolati, Jesko Hentschel, Rafael de Hoyos, Dante Mossi, Zafer Mustafaoglu, and Rossana Polastri (Paraguay); Amparo Ballivian, Malva Baskovich, Livia Benavides, Oscar Barriga Caba- nillas, Louise Cord, María Eugenia Dávalos, Tania Díaz-Bazán, Santiago Garriga, Stephanie Majerowicz, Karina Olivas, Gustavo Perochena, Adam Kahn Ratzlaff, Megan Rounseville, Kinnon Scott, Peter Siegenthaler, Renos Vakis, and Nobuo Yoshida (Peru); and Louise Cord, Carolina Díaz-Bonilla, Luis F. López-Calva, Zafer Mustafaoglu, Cristina Savescu, and Emily Sinnot (Uruguay). About the Editors and Authors About the Editors Louise Cord is practice manager in the Poverty Global Practice for the World Bank’s Latin America and the Caribbean Region, leading a diverse program on poverty, equity, and gender equality. Prior to her experience in Latin America, Louise was the poverty sector manager in the central unit of the Poverty Reduction and Economic Management (PREM) Network, where previously she had been a lead economist working on pro-poor growth, aid effectiveness, rural poverty, and poverty reduction strategies in Africa. Before coming to the PREM Network, she worked for seven years in the World Bank’s rural development group of the Latin America and Caribbean Region on rural poverty, agricultural trade and price policy, and rural finance. She has published several articles and reports on poverty and agricultural policy in Mexico, Eastern Europe, and Central Asia, and more recently on pro-poor growth, inequality, and political economy. She holds a PhD in development and economic policy from the Fletcher School of Law and Diplomacy, Tufts University. María Eugenia Genoni is an economist in the Poverty Global Practice at the World Bank, working in the Latin America and Caribbean Region. Cur- rently, she leads the poverty and equity program in Peru and co-leads the program in Bolivia. She also has contributed to the World Bank’s poverty and equity agenda in Central America and to the Regional Gender Impact Evaluation Initiative. Her research has focused on development economics and applied microeconomics, particularly on survey design, poverty and inequality, migration, and risk management. She received her PhD in eco- nomics from Duke University. xxi xxii About the Editors and Authors Carlos Rodríguez-Castelán is senior economist in the Poverty Global Prac- tice at the World Bank, working in Latin America and the Caribbean Region. Currently, he leads the poverty and equity program in Colombia and co-leads the Global Solutions Area on Markets and Institutions for Poverty Reduction and Shared Prosperity. He also has contributed to the World Bank’s poverty and equity agenda in Chile, Mexico, Paraguay, and Uruguay and to the corporate activity of Data for Goals. Prior to joining the World Bank, he was a postdoctoral fellow in the Foreign Policy and Global Economy and Development programs of the Brookings Institution. His research has focused on development economics and applied microeconom- ics, particularly on poverty and inequality analysis, noncompetitive market structures, social protection, education, and risk management. He received his PhD in economics from Cornell University. About the Authors Javier E. Báez is senior economist in the Poverty Global Practice of the World Bank, where he specializes in poverty analysis and impact evalua- tion. He holds a PhD in economics from Syracuse University and an MA in development economics from Harvard University. Oscar Barriga Cabanillas is a PhD candidate in agricultural and resource economics at the University of California, Davis. Previously, he worked at the Poverty Global Practice of the World Bank as a junior professional associate. Kiyomi Cadena is a consultant in the Poverty Global Practice of the World Bank, where she has conducted poverty and social impact analyses for proj- ects in Central America, the Latin America and Caribbean Region, and Mexico. Santiago Garriga is a junior professional associate in the Poverty Global Practice of the World Bank. He is part of the team for statistical develop- ment in the World Bank’s Latin America and Caribbean Region, where he has conducted data analysis on poverty and inequality. Lea Giménez is an economist in the Poverty Global Practice of the World Bank, focusing on poverty measurement, institutional and statistical capac- ity building, social impact, and activities supporting evidence-based policy making. Marina Gindelsky is a consultant in the Poverty Global Practice of the World Bank, where she has conducted research on poverty and inequality for projects in Iraq and Uruguay. About the Editors and Authors xxiii Luis F. López-Calva is lead economist and regional poverty adviser in the World Bank’s Europe and Central Asia Region and was previously lead economist in the Poverty Unit in Latin America and the Caribbean Region. He holds a PhD in economics from Cornell University. María Ana Lugo is an economist in the Poverty Global Practice of the World Bank, specializing in equality of opportunities, poverty, and multidi- mensional poverty analyses. She holds a PhD in economics from the Univer- sity of Oxford. Alejandro Medina Giopp is a senior monitoring and evaluation specialist at the Poverty Global Practice of the World Bank. He holds a PhD in manage- ment sciences from the Escuela Superior de Administración y Dirección de Empresas, Barcelona. Miriam Müller is a research analyst in the Poverty Global Practice of the World Bank, where she specializes in gender. She is currently a PhD candi- date in the sociology program at Humboldt University of Berlin. Aude-Sophie Rodella is an economist in the Poverty Global Practice of the World Bank, focusing on urban poverty analysis and inclusive growth. She holds a PhD in microeconomics from the Centre for Studies and Research on International Development, Université d’Auvergne. Megan Rounseville is a consultant in the Poverty Global Practice of the World Bank, where she specializes in econometric impact evaluation and poverty analyses. She has worked with projects in several countries in Latin America and the Caribbean. Mateo Salazar is a junior professional associate in the Poverty Global Prac- tice of the World Bank. He has coauthored various analytical reports on the measurement, monitoring, and diagnosis of the causes and nature of pov- erty in several countries in Latin America and the Caribbean. Kinnon Scott is a senior economist in the Poverty Global Practice of the World Bank, specializing in social mobility and poverty analyses. She holds a PhD in economic and social development from the University of Pittsburgh. Ali Sharman is a junior professional associate in the Poverty Global Practice of the World Bank, where she has worked on conducting data analysis and research for projects in Brazil, Colombia, the Dominican Republic, and Ecuador. Emmanuel Skoufias is lead economist in the Poverty Global Practice of the World Bank, with experience working on targeting and the impacts of xxiv About the Editors and Authors social transfers, child malnutrition, risk management among the poor, and the impacts of climate change on welfare. He holds a PhD in economics from the University of Minnesota. Liliana D. Sousa is an economist in the Poverty Global Practice of the World Bank, where she is the co–task team leader of the team for statistical development in the Latin America and Caribbean Region. She holds a PhD in economics from Cornell University. Erwin R. Tiongson is a professor in the Walsh School of Foreign Service at Georgetown University. Previously, he was senior economist, World Bank, focusing on poverty analysis. He holds a PhD in economics from the George Washington University. Daniel Valderrama is a junior professional associate in the Poverty Global Practice of the World Bank, where he has conducted data analysis on pov- erty, inequality, and shared prosperity for projects in several Latin Ameri- can countries. Martha Viveros is a consultant in the Poverty Global Practice of the World Bank, where she has conducted data analyses on poverty, inequality, vul- nerability, and shared prosperity in Brazil and in Latin America and the Caribbean. Abbreviations BF Bolsa Família (family allowance program) (Brazil) BRIC Brazil, Russian Federation, India, and China BSM Brasil sem Misería (Brazil without Misery plan) (Brazil) CONEVAL Consejo Nacional de Evaluación de la Política de Desarrollo Social (National Council for the Evaluation of Social Development Policy) (Mexico) DANE Departamento Administrativo Nacional de Estadística (National Administrative Department of Statistics) (Colombia) ENIGH Encuesta Nacional de Ingresos y Gastos de los Hogares (Household Income and Expenditure Survey) (Mexico) FONDEN Fondo de Desastres Naturales (National Disaster Fund) (Mexico) GDP gross domestic product HOI human opportunity index IDP internally displaced person INDEC Instituto Nacional de Estadística y Censos (National Institute of Statistics and Censuses) (Argentina) INEGI Instituto Nacional de Estadística y Geografía (National Institute of Statistics and Geography) (Mexico) IPEA Institute for Applied Economic Research (Brazil) IPS-8 Índice de Privación Social (recortado) (adjusted social deprivation index) (Mexico) MCS-ENIGH Modulo de Condiciones Socioeconomicas (Socioeco- nomic Conditions Module of ENIGH) (Mexico) xxv xxvi Abbreviations MESEP Misión de Empalme de las Cifras de Pobreza y Mercado Laboral (Colombia) MPI multidimensional poverty index OECD Organisation for Economic Co-operation and Development PANES Plan de Atención Nacional a la Emergencia Social (Uruguay) PISA Program for International Student Assessment (OECD) PPP purchasing power parity SEDLAC Socio-Economic Database for Latin America and the Caribbean SPI shared prosperity indicator Note: All dollar amounts are U.S. dollars ($) unless otherwise indicated. CHAPTER 1 Overview Louise Cord, María Eugenia Genoni, and Carlos Rodríguez-Castelán Introduction I n 2013, the World Bank adopted two overarching goals to guide its work: (1) to end extreme poverty or to reduce the share of people living in extreme poverty to 3 percent of the global population by 2030 and (2) to promote shared prosperity in every country through a sustainable increase in the well-being of the poorer segments of society, roughly defined as the lowest 40 percent of the income distribution (the bottom 40).1 The adop- tion of these complementary objectives is helping to renew the focus of the global development community on the welfare of those at the bottom of the income distribution. Moreover, these goals provide a line of sight that development agencies and countries may use to prioritize actions and funds. Over the last decade, the Latin America and Caribbean region achieved important progress toward the twin goals by cutting extreme poverty in half and realizing the highest income growth rate among the bottom 40 across all regions of the world in absolute terms and relative to total popula- tion. These gains have transformed the configuration of the socioeconomic groups in the region. In 2012, more than one-third of the bottom 40 in the region was comprised of vulnerable households (those that have moved out of poverty, but do not have enough income to be considered part of the middle class); this compares with 2003, when the bottom 40 was exclu- sively comprised of households living in poverty. The inclusive nature of the growth process in the region has also been evident in the decline in the region’s notoriously high levels of inequality, which dropped from a Gini coefficient of 0.56 in 2003 to 0.52 in 2012. Some projections estimate the share of households that will be living in extreme poverty ($1.25 a day) in the region in 2030 at 3.1 percent, down from 4.6 percent in 2011, and thus reaching the World Bank’s goal of 3 percent by 2030 (World Bank 2015b).2 1 2 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Despite this impressive performance, social progress has not been uni- form over this period, and certain countries, subregions, and even groups have participated less in the growth process, thereby constraining oppor- tunities for poverty reduction and shared prosperity in countries and the region. More than 75 million people are still living in extreme poverty in the region, half of them in Brazil and Mexico, and extreme poverty rates (using the $2.50-a-day per capita line) are above 40 percent in Guatemala and reach nearly 60 percent in Haiti.3 This means that extreme poverty is still an important issue in both low- and middle-income countries in the region. The recent slowdown in economic activity and the decline in the pace of inequality reduction pose additional barriers to rapid progress toward the institutional goals (Cord et al. 2014; de la Torre et al. 2014).4 According to a recent study by Narayan, Saavedra-Chanduvi, and Tiwari (2013), the shared prosperity indicator (SPI) is highly correlated with growth in aver- age incomes, but, if inequality is high, mean income growth will not accrue proportionally to the bottom segment of the distribution. The purpose of this overview is to assess the performance of the region in reducing poverty and boosting shared prosperity during the last decade, while using a simple asset-based framework to highlight some of the key ele- ments affecting the capacity of less well-off households to generate income. The descriptions presented in this chapter set the stage for the eight country studies that follow and that assess the heterogeneous advances toward the goals and identify some of the key policy variables that have affected the outcomes within the countries. The first part of this chapter provides a baseline analysis of the region in terms of the institutional goals, while emphasizing the diversity of outcomes. This analysis takes advantage of comprehensive harmonized household survey data from the Socio-Economic Database for Latin America and the Caribbean (SEDLAC) database; such data are key for cross-country compa- rability.5 These data cover 17 countries in Latin America and the Caribbean and account for about 90 percent of the population in the region.6 The second part of the chapter illustrates an asset-based framework. The framework identifies the main elements that contribute to household income generation and that can be intuitively related to poverty reduction and shared prosperity. The simple framework depicts the realization of household market income as a function of four major components: (1) the capacity of households to generate income based on the productive assets they own, (2) the private transfers—the monetary value of domestic and international private contributions—they receive and the public transfers that are incorporated as a policy variable, (3) the set of prices of the basket of goods and services that the households consume, and (4) the external shocks that generate variability in the incomes. The capacity of households to generate income based on the productive assets they own can be further disaggregated into the interaction between the role of assets (human capi- tal, housing, and capital and land), the intensity of asset use (participation in labor and financial markets, agency), and the returns to assets (labor demand factors, including uneven returns by race, gender, and location). Chapter 1: Overview 3 This asset-based approach integrates macroeconomic and microeco- nomic dimensions so that growth and the incidence of growth can be understood as mutually determined processes. The framework considers the distribution of assets as a given in the short run; thus, changes in the income generation capacity of households depend mostly on macroeco- nomic variables that affect the demand for labor across sectors, relative prices (returns and consumer prices), and the intensity of the use of assets over the economic cycle. In the long run, the main drivers of income growth will be the level and distribution of assets—human, physical, financial, social, and natural capital—that people own and accumulate, as well as the intensity with which they are used and the associated returns, which will reflect asset productivity. The third part of the chapter relies on the asset-based framework to characterize the bottom 40 in terms of their capacity to generate income relative to the top 60 percent of the distribution (the top 60). The analysis focuses mainly on describing the capacity of households to generate labor income given the importance of this source of income in total income and as a driver of trends in poverty and shared prosperity in the past decade. Exploring the asset composition of households can provide information important to understanding the factors that contribute to boosting the capacity of individuals to generate income, climb out of poverty, and avoid the risk of downward mobility. Finally, the chapter links the twin goals to four fundamental policy areas that have a direct impact on the capacity of households to gener- ate income, but with a particular focus on those households that are poor and that belong to the bottom 40. These four broad policy areas, which have also been defined in previous studies (World Bank 2013a, 2014a), are (1) equitable, efficient, and sustainable fiscal policy and macroeconomic sta- bility (direct and indirect taxes and transfers, inflation targets); (2) fair and transparent institutions capable of delivering universal, good-quality basic services (a greater and better supply of public goods, protection of property rights); (3) well-functioning markets (improved connectivity to markets, competition policy); and (4) adequate risk management at the macro and household levels (macroprudence, safety nets). The country study cases pre- sented in the rest of this volume organize the discussion around these four policy areas in a way that is relevant for poverty reduction and the promo- tion of shared prosperity. Transformational Change in Living Standards in the Region Recent trends in poverty reduction and shared prosperity Poverty reduction Over the past decade, the Latin America and Caribbean region experienced remarkable reductions in extreme poverty.7 According to extreme poverty 4 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Table 1.1 Extreme Poverty Rates, Developing Regions, 2002 and 2011 Extreme poverty rate, Extreme poverty rate, $1.25 a day $2.50 a day Region 2002 2011 Change, % 2002 2011 Change, % Sub-Saharan Africa 57.1 46.8 18.0 84.2 78.0 7.4 South Asia 44.1 24.5 44.4 86.7 74.5 14.0 East Asia and the Pacific 27.3 7.9 71.0 62.4 31.9 48.8 Latin America and the Caribbean 10.2 4.6 54.7 27.1 13.3 51.0 Middle East and North Africa 3.8 1.7 55.9 31.9 22.1 30.7 Europe and Central Asia 2.1 0.5 77.0 11.6 3.8 67.2 Source: World Bank calculations using PovcalNet (online analysis tool), World Bank, Washington, DC, http:// iresearch.worldbank.org/PovcalNet/. Note: The poverty data on Latin America and the Caribbean differ slightly from the data in the SEDLAC database because of variations in the methodology used to calculate poverty rates. measures using an income-based aggregate and an international poverty line of $1.25 a day in 2005 prices, the extreme poverty rate fell from 10.2 to 4.6 percent between 2002 and 2011. Based on a higher international poverty line of $2.50 a day calculated from an average of national poverty lines in the region to identify the extreme poor, the headcount fell by half, from 27.1 to 13.3 percent over the same period (table 1.1). Compared with other developing regions, Latin America and the Carib- bean also performed well in reducing extreme poverty over the last decade. Based on a $1.25-a-day poverty line, the region’s extreme poverty reduc- tion of about 55 percent surpassed South Asia and Sub-Saharan Africa, but lagged Europe and Central Asia and East Asia and the Pacific. Based on the $2.50-a-day poverty line, the region’s extreme poverty reduction of 51 percent exceeded the declines observed in all other regions except Europe and Central Asia, which cut this rate by 67 percent.8 The improvements in living conditions in Latin America and the Carib- bean dramatically shifted the socioeconomic composition of the popula- tion. In 2012, more Latin Americans were living in the middle class than in total poverty, 34.4 versus 21.2 percent in 2003 (figure 1.1, panel a). Moreover, whereas in 2003, 6 in 10 people in the bottom 40 were among the extreme poor, by 2012, only 3 in 10 were in this condition. In 2012, the vulnerable (people earning between $4 and $10 a day) made up a third of the bottom 40 in the region (figure 1.1, panel b).9 Shared prosperity The reduction in poverty rates and the significant expansion in the middle class observed in Latin America and the Caribbean has been accompanied by strong growth in the incomes of the bottom 40. Between 2003 and 2012, the average income of the bottom 40 in the region increased by 5 percent a year, from $2.10 a day per capita in 2005 prices to $3.30 a day. This growth rate was greater than the corresponding rate observed for the whole population, which was 3.3 percent a year (from $8.80 a day per capita to Chapter 1: Overview 5 Figure 1.1 Socioeconomic Composition of the Population, Latin America and the Caribbean, 2003 and 2012 a. Total population b. Bottom 40 100 100 Share of population (%) Share of bottom 40 (%) 21.2 34.4 39.6 37.2 80 80 35.6 60 60 37.3 32.5 40 40 17.3 60.4 20 13.0 20 24.1 30.2 12.1 0 0 2003 2012 2003 2012 Year Year Middle class Poor but not extreme poor Vulnerable Extreme poor Source: Calculations based on data in the SEDLAC database. Note: The estimates of poverty, vulnerability, and the middle class are population-weighted averages of country estimates. The extreme poor are people living on less than $2.50 a day; the poor but not extreme poor are those living on $2.50 to $4.00 a day; the vulnerable are those living on $4.00 to $10.00 a day; and the middle class are those living on $10.00 to $50.00 a day (all in 2005 purchasing power parity [PPP] international U.S. dollars). $11.70). The region’s performance in shared prosperity was also positive compared with that of other regions. Between 2006 and 2011, the average growth rate per year in the mean income of the bottom 40 across countries in the region was approximately 5.2 percent. This was the highest rate in all regions (figure 1.2, panel a). Moreover, the region’s bottom 40 enjoyed the most rapid income growth relative to the total population; thus, based on these indicators, Latin America and the Caribbean has been the most inclusive region in the world over the last decade (figure 1.2, panel b). Demographic changes and the composition of the bottom 40 Over the last decade, the observed progress in poverty reduction and shared prosperity has been accompanied by a transformational change in the basic demographic characteristics of households in the region (table 1.2). House- holds in Latin America have become smaller and more likely to be headed by older, more well educated, and women household members. These trends are similar among households in the bottom 40 and households in the top 60. Despite the similar trends, households in the bottom 40 are significantly different from those in the top 60, and the gaps have not changed substan- tially. Households in the bottom 40 are younger, larger, and more likely to be headed by women and less well-educated individuals. For instance, the education gap of household heads was approximately three years between the two groups in 2012. Moreover, 2 in 3 households in the bottom 40 resided in urban areas, compared with 9 in 10 among the top 60. 6 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.2 Shared Prosperity: Annualized Income Growth, Developing Regions, around 2006–11 a. The bottom 40 b. Ratio: bottom 40 to entire population 6 2.0 1.9 Simple averages, bottom 40 Ratio of bottom 40 growth 5.2 5.0 5 to average growth 1.6 1.5 4.1 income growth 1.3 1.2 4 3.5 1.2 1.1 1.0 3 2.2 2.2 0.8 2 0.4 1 0 0 be ica c ia a Sa Afr nd a be ica a fic a a fr d So cifi si ic si si ric A an s i an ha ica ou an a h ta rib er A lA fr rib er A lA c f ic a Pa A A th st Ca Am Ca Am h th Su ort Eas P ra ra or a ut n n e e N le E nt nt ra ra th th th atin th atin N le Ce S Ce ha d d d d an an id Sa id L d L d e e an an M M b- b- a a si si Su pe pe d d A A an an ro ro st st Eu Eu Ea Ea Region Region Source: GDSP (Global Database of Shared Prosperity), World Bank, Washington, DC, http://www.worldbank.org /en/topic/poverty/brief/global-database-of-shared-prosperity. Note: The data are simple averages across countries in the regions calculated using household surveys. They may not be strictly comparable because some regions use expenditure survey data, while Latin America and the Caribbean uses income data. Table 1.2 Bottom 40 and Top 60: Household Characteristics, Latin America and the Caribbean, 2003 and 2012 Bottom 40 Top 60 Indicator 2003 2012 2003 2012 Average age, household head, years 43.3 45.3 48.2 50.0 Woman-headed households, % 28.1 36.3 27.4 34.7 Average education, household head, years 4.7 5.8 8.0 8.9 Average household size, number 4.4 4.1 3.4 3.0 Urban households, % of total 66.6 66.2 86.3 87.5 Source: Calculations based on data in SEDLAC. Note: The data represent population-weighted averages across countries in the region. Transformational change reflects strong growth and significant redistribution Strong growth and a significant narrowing in the region’s high level of income inequality drove the gains in poverty reduction and shared pros- perity between 2003 and 2012. The combination of prudent macrofiscal economic policies, global liquidity, and positive terms of trade because of Chapter 1: Overview 7 Figure 1.3 Average GDP Growth Rates, Latin America and the Caribbean, 1990–2013 7 5.9 5.9 6 5.6 5.6 5.4 Average GDP growth (%) 5 4.7 3.7 4.3 4 4.0 4.5 3.7 3.8 3.6 2.8 3 2.8 2.3 2 1.8 2.4 1 0.3 0.3 0.5 0.4 0.1 0 –1 –1.6 –2 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Source: WDI (World Development Indicators) (database), World Bank, Washington, DC, http://data.worldbank .org/data-catalog/world-development-indicators. Note: The regional average is the regional aggregate of the countries in the region, excluding high-income countries. the commodity boom helped foster a decade of strong growth in the region, which was largely able to weather well the financial crisis. In particular, during the past decade, real incomes rose by more than 25 percent across the region; annual gross domestic product (GDP) increased at an average of 3.2 percent. Moreover, growth proved resilient across the region: many countries maintained positive growth rates throughout the global financial crisis of 2008.10 However, while GDP growth was an important driver of poverty reduction and shared prosperity, it did not seem to be the only force behind the progress. In fact, while the region’s GDP growth during the 2000s was high, the region did not grow much more quickly relative to the previous decade (figure 1.3). GDP growth was 3.1 percent during the 1990s, compared with 3.2 percent during the 2000s.11 Despite similar growth rates, the region’s performance in poverty reduction was different in the 1990s and 2000s. While poverty fell less than 1 percent a year dur- ing the 1990s, poverty rates decreased at a much higher rate in the 2000s, approximately 6 percent a year.12 The different poverty gains across two decades with similar levels of growth highlight the importance of the nature of growth and the redistributive policies applied. An important difference between the 1990s and the 2000s was the region’s progress in narrowing household income inequality. While the Gini coefficient barely changed during the 1990s, it fell from 0.56 to 0.52 between 2003 and 2012 (figure 1.4). This trend was widespread: income inequality declined in all 17 countries for which frequent household survey 8 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.4 Trends in the Gini Coefficient, Latin America and the Caribbean, 2003–12 0.56 0.55 0.56 0.55 0.54 Gini coefficient 0.53 0.52 0.52 0.51 0.50 0.50 0.49 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Year Region, pooled Regional weighted average Source: Cord et al. 2014. Note: Because the Gini coefficient does not satisfy group decomposability, the regional Gini coefficient is computed based on pooled country-specific data for 17 countries. To test the robustness of the results, the unweighted average is also presented. data are available.13 Even though this decline likely reflects a combination of pro-poor social policies and growth, there is still debate about the spe- cific drivers behind it. Recent evidence highlights the change in the distribu- tion of labor income as the main factor behind the progress, followed by the expansion of government transfers and, for the countries in the Southern Cone, the broadening of pension coverage (Cord et al. 2014; López-Calva and Lustig 2010; Lustig, López-Calva, and Ortiz-Juárez 2013). The decline in labor income inequality is largely explained by a fall in the skill pre- mium, that is, a reduction in the wage differential between more highly educated workers relative to less highly educated workers. This reduction seems to reflect a combination of lower excess demand for skilled labor and improved access to education that increased the supply of skilled workers (Gasparini et al. 2011). In particular, the expansion of education coverage over the period implied a rise in the share of new students at lower socio- economic status, which may have reduced the average quality of education. A deterioration at the margin of the quality of educational institutions may have also accompanied this trend (de la Torre et al. 2014). One potential demand-side explanation of the observed narrowing in wage inequality is the effect of the commodity boom, which promoted growth in the nontrad- able sectors and, in this way, raised the demand for unskilled workers rela- tive to skilled workers. Chapter 1: Overview 9 In sum, during the past decade, both growth and redistribution contrib- uted toward the progress achieved in eradicating extreme poverty and pro- moting shared prosperity. Two-thirds of the observed decline in extreme poverty in the region between 2003 and 2012 can be explained by eco- nomic growth, while the rest is explained by changes in income distribution (World Bank 2014a). Progress was heterogeneous across countries While the region’s progress on the twin goals was substantial during the period, the averages mask significant heterogeneity across and within coun- tries. While certain countries took advantage of a decade of high growth rates to drive steep declines in poverty and boost shared prosperity, such as Bolivia, Brazil, and Peru, others grappled with lackluster growth, such as Guatemala and Mexico. Other countries achieved substantial growth, but struggled to convert the gains into better livelihoods among the poor- est. One clear example is the Dominican Republic, where GDP per capita grew by 53 percent from 2000 to 2012, while extreme poverty remained stagnant (box 1.1). The region still presented wide disparities in extreme poverty rates. In 2012, about 4 in 10 people in Guatemala and Honduras were living in extreme poverty. In contrast, 3 in 100 people were among the extreme poor in Chile and Uruguay (figure 1.5). Nonetheless, there is evidence of a regional convergence in poverty rates: countries with high poverty rates at the beginning of the decade experienced large reductions thereafter. Some of the top performers were the Andean countries and Brazil. Notable excep- tions were Guatemala and Honduras, which both had high initial extreme poverty rates; Guatemala even saw a subsequent rise in extreme poverty. In addition, even among the strong performers, there were significant geographical disparities, including pockets of high and persistent poverty. For instance, Peru, one of the best performers on the twin goals in the region, presented strong disparities in poverty across its 1,800 districts. In 2007, almost half the extreme poor were concentrated in approximately 11 percent of the districts (map 1.1, panel a), while these same 11 percent of districts accounted for a third of the total population. In addition, in 2013, the rural areas of Peru contained 33 percent of the country’s population, but accounted for half of the poor and 80 percent of the extreme poor. Mean- while, in Bolivia between 2001 and 2011, approximately half the munici- palities reduced extreme poverty substantially. However, some areas were still lagging in 2011, particularly small rural municipalities where the pov- erty rates had been higher at the beginning of the decade. In 2011, nearly a third of Bolivia’s municipalities still showed an incidence of extreme poverty greater than 50 percent (map 1.1, panel b). In the case of Colombia, histori- cally large disparities between urban and rural areas persist, and the rate of income convergence across the country’s departamentos has been limited over the past decade. According to official data, the difference between the departamento with the highest poverty rate and the departamento with the 10 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.1 Poverty Trends in the Caribbean Even though the improvement in economic conditions was significant throughout Latin America, progress was sluggish and limited in the Caribbean. Extreme poverty rates in the Dominican Republic have remained stagnant despite the strong economic growth over the past decade (World Bank 2014b). Between 2000 and 2012, the extreme poverty headcount ($2.50 a day) fell less than 1 percent- age point (from 15.7 to 14.6 percent) below the regional average. In Jamaica, poverty rates based on official figures reached 17.6 percent in 2010, compared with 12.3 percent in 2008. The country was negatively affected by the global crisis, as well as rising food and energy prices, and this hindered poverty reduction (World Bank 2014c). Similarly, while extreme poverty in Haiti—based on a consumption aggregate and a national pov- erty line of $1.23 a day—dropped from 31 to 24 percent between 2001 and 2012, the gains appear to have been linked to the greater aid flows, particularly into urban areas, and higher remittances, which soared after the earthquake (World Bank and ONPES 2014). In addition, the moderate poverty rate remains high (58.5 percent in 2012). The lack of official poverty and inequality data in the eastern Caribbean makes it challenging to evaluate trends in poverty there. Nonetheless, the patterns of asset ownership and the high rates of unemployment and underemployment suggest that social disparities have been exacerbated by the 2008 financial crisis (World Bank, forthcoming). The evidence from household survey data suggests that the financial crisis had significant negative and long-lasting impacts on household welfare in St. Lucia. While the unemployment rate was around 16.9 percent among all welfare quintiles from early 2008 through late 2009 (according to an asset-based welfare measure), the unemployment rate among the bottom 40 (29 percent) was nearly double the rate among the two highest quintiles (15.7 percent) from 2011 to 2013. Prior to the crisis, the characteristics of the bottom 40 and the top 60 were relatively similar in St Lucia, while, since the crisis, there has been a widening gap between the two groups. For example, in 2008, although they were more likely to be self-employed and less likely to be working in the professional services sector, the bottom 40 were virtually indistinguishable from the top 60. By 2013, however, the bottom 40 were significantly more likely to be unemployed (by 11 percentage points), significantly less likely to be an employee or an employer, had significantly less educational attainment, showed a higher probability of residing in urban areas, typically had smaller households, and were more likely to be living in woman-headed households. By 2013, relative to the top 60, they were twice as likely to be working in the agricultural sector, were more likely to be working in construc- tion or manufacturing, and were significantly less likely to be working in education, health care, or social or professional services. These outcomes are not surprising given that the economies in the Caribbean greatly depend on industries such as tourism, agriculture, and financial services that rely heavily on the external demand of the developed economies where the crisis originated. In addition, most Caribbean countries suffer from substantial national debt and lack a stable financial sector to channel financial resources effi- ciently. These challenges make especially difficult the establishment of the social protection mecha- nisms necessary to shield the vulnerable from the relatively large shocks faced by the region. lowest rate was 38 percentage points in 2002, whereas, in 2014, the differ- ence was 53 percentage points. (See the country chapters.) Levels of development differ across Latin America, which implies that levels of income and other characteristics of the bottom 40 in each country may also differ, especially because participation in this population segment is measured in relative terms. In some countries, there is a large overlap Chapter 1: Overview 11 Figure 1.5 Extreme Poverty Rates, Latin America and the Caribbean, 2003–12 a. Rates, circa 2012 45 42 40 37 Extreme poverty rate (%) 35 29 30 25 20 18 16 15 15 14 15 13 12 12 11 10 10 8 5 5 3 3 0 a ar s lo ua El Bo ia ic al ia pu r ra lic Ec uay na r a M ru o tin a R l a, ica gu C n , u le an en st zi Re do Pa do ic ra on al m ic ba b in S liv ay hi Pe rg o ra Pa b Co ag rb m ex N u m a ua g ur B d an v te ua H C G ru U om A D Country b. Convergence in rates, 2003–12 0 extreme poverty rate (%) Annual average change, –0.5 Uruguay Mexico Dominican Republic –1.0 El Salvador Costa Rica Paraguay Panama Colombia Honduras –1.5 Argentina Peru Brazil Ecuador –2.0 Bolivia Nicaragua 0 5 10 15 20 25 30 35 40 45 50 Extreme poverty, 2003 (%) Source: Calculations based on data in SEDLAC. Note: The extreme poverty rate is calculated using a $2.50-a-day poverty line. Panel b excludes Guatemala, which is the only country in the region in which extreme poverty grew over the period. between the bottom 40 and the extreme poor (for example, Guatemala, Honduras, and Nicaragua), while, in other countries, the bottom 40 is mainly comprised of people living above the poverty line (such as Chile and Uruguay). The heterogeneous progress over the past decade in shared prosperity can also be illustrated through changes in the composition of the bottom 40. For instance, while 8 in 10 people in the bottom 40 in 12 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Map 1.1 Heterogeneity in Living Standards, Bolivia and Peru, 2007 and 2011 a. The extreme poor, Peru, 2007 b. Extreme poor, municipalities, Bolivia, 2011 Share of extreme poor (%) 47.2–90.1 39.7–47.2 1,628 districts (89%): circa 65% of population, 50% of extreme poor 26.1–39.7 204 districts (11%): circa 35% of population, 50% of extreme poor 3.0–26.1 Source: Calculations using monetary poverty maps of Bolivia and Peru. Note: District poverty maps in Peru are based on consumption using data from the 2007 National Household Survey and the 6th National Housing Census and 11th Population Census (both 2007). The municipal poverty map of Bolivia is estimated based on income using data from the 2011 Household Survey and the 2012 National Census of Housing and Population. The computation of poverty rates follows the official poverty methodologies of the countries. Both maps have been estimated using the Elbers, Lanjouw, and Lanjouw (2003) small area methodology. Ecuador were among the extreme poor in 2003, only 3 in 10 were in this condition in 2012. In contrast, in several Central American countries, such as Guatemala, Honduras, and Nicaragua, an overwhelming proportion of the bottom 40 continued to be composed of the extreme poor, with little change (figure 1.6). While the average income of the bottom 40 grew approximately 5 per- cent a year across the region between 2003 and 2012, the heterogeneity was significant in shared prosperity by country. The strongest performers, Argentina, Bolivia, Brazil, and Panama, with income growth rates among the bottom 40 well over 7 percent, far outpaced the weakest performers, Guatemala and Mexico, with growth rates among the bottom 40 of −1.0 and 1.3 percent, respectively. Guatemala was the only country in the region in which the incomes of the bottom 40 declined over the decade (figure 1.7). For most countries in the region, income growth among the bottom 40 outpaced the average growth among the population over the decade (fig- ure 1.8). However, the size of the gap also varied. In some countries, such as Argentina, Bolivia, and Nicaragua, the growth rate was significantly higher Chapter 1: Overview 13 Figure 1.6 Composition of the Bottom 40, Latin America and the Caribbean, 2003 and 2012 a. Circa 2003 100 19 19 Share of bottom 40 (%) 37 36 80 46 42 60 56 55 63 62 61 75 38 32 78 77 60 86 93 100 33 40 40 35 35 40 49 39 44 45 43 43 20 38 23 25 37 31 22 23 23 14 0 s ua ua Ec la r ia a ru Re l Pa on ay Re or Pa ic a a o ca ay ile i do ra bi az m tin ic bl a liv Pe d Ri gu gu Ch ag gi m ex du m na Br ua va pu en Bo a ar te ra ru lo M on al st rg ic Co U S Co H A N G an om El ic in D Country b. Circa 2012 100 12 26 20 29 29 29 14 Share of bottom 40 (%) 37 36 36 32 32 18 80 44 40 15 29 73 28 60 23 29 100 93 37 33 41 33 32 47 78 40 39 50 73 75 35 48 46 51 42 31 20 27 31 31 18 27 13 17 0 on a s Co ua a om El S ivia r ic ay r on a ru co il ca a ile ay do do ra al bi m az tin bl Pe Ri gu i Ch gu ag gi m ex du m na Br va ua l pu en Bo Re a te ar ra ru lo M Pa al Ec st Re rg ua ic Pa U Co H A N G an ic in Country D Extreme poor Poor (not extreme poor) Vulnerable Middle class Source: Calculations based on data in SEDLAC. Note: Estimates of poverty, vulnerability, and the middle class in the region are population-weighted averages of country estimates. The poor are defined as people living on less than $4 a day; the vulnerable are those living on $4–$10 a day; and the middle class are those living on $10–$50 a day (all in 2005 PPP international U.S. dollars). among the bottom 40, while, in Costa Rica, Guatemala, and Mexico, the rates were almost the same. Colombia was the only country in the set that was analyzed in which average income growth among the bottom 40 did not surpass the income growth of the total population. 14 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.7 Income Growth among the Bottom 40, Latin America and the Caribbean, around 2003–12 Annualized growth rate (%) 10 8.8 7.6 7.3 7.3 8 6.9 6.5 6 5.7 5 4.8 4.4 4.4 4.3 4.3 4.3 4 3 2.5 2 1.3 0 –1 –1 –2 12 12 2 11 , 2 –12 12 12 20 n 20 09 9 2 11 1 1 ex 200 2 2 2 1 –1 –0 –1 –1 –1 1 o, 3–1 –1 –1 o 4– 4– 4– an 04– 3– – 3– 4– gi 07 3 05 05 3 3 3 04 00 Re Br 200 na 200 0 Pe 200 00 00 00 00 00 bl 00 Bo , 20 20 0 20 20 ,2 on ia, 2 ,2 ,2 ,2 pu , 2 n, Pa il, Ec ia, r, a, a, , a, a ru as ay ile Re or ic do az m Co gu c al ba ic liv b d Ri r gu Ch rb m m du ua an lva ur a ,u a ar ra te lo M st in Sa a, ua ic Co Pa ay tin H N G gu l om E en ic ru rg U A D Country Source: Calculations based on data in SEDLAC. Note: Annualized growth rate of the income of the bottom 40. The numbers for the region are calculated using pooled data of countries. To analyze the same set of countries every year, interpolation has been applied if country data were not available for a given year. Even though there was a positive correlation between total income growth and income growth among the bottom 40 during the last decade, the relationship was not perfect. Some countries, such as Chile, Colom- bia, Costa Rica, Honduras, and Paraguay, had similar growth rates in the average income among the bottom 40, but different overall income growth rates. Other countries, such as Argentina, Brazil, and Colombia, experi- enced similar total income growth rates, but performed differently in the mean income of the bottom 40. This heterogeneity indicates that the out- comes in shared prosperity were dependent not only on growth, but also on the sources of growth and specific policies and redistribution efforts. Similarly, the responsiveness of poverty to growth was heterogeneous in the region. For instance, Mexico showed low GDP growth over the period (about 0.7 percent a year), but poverty levels were responsive to this growth (about 2 percent of poverty reduction for each 1 percent in GDP growth). In contrast, the Dominican Republic experienced high GDP growth, but this did not translate into a commensurate reduction in poverty (about 0.2 percent of poverty reduction for each 1 percent in GDP growth). There was also significant variation across countries in the relative importance of redistribution and growth for poverty reduction. Thus, in Colombia, poverty reduction was only driven by growth, while in other Chapter 1: Overview 15 Figure 1.8 Income Growth, Bottom 40 and the Entire Population, Latin America and the Caribbean, around 2003–12 Argentina, urban 8 Annualized growth rate, bottom 40 (%) Brazil Bolivia Panama 6 Ecuador Peru Region, bottom 40 Uruguay, urban Honduras 4 Nicaragua Chile Colombia Paraguay Costa Rica El Salvador 2 Dominican Republic Mexico 0 Guatemala Region, entire population –2 0 2 4 6 8 Annualized growth rate, entire population (%) Source: Calculations based on data in SEDLAC. Note: Blue line = the 45º line. The data on the region are calculated using the pooled data on the countries. countries, such as the Dominican Republic, El Salvador, and Nicaragua, redistribution was almost exclusively responsible for the reductions in extreme poverty. Most countries fell somewhere in between: important components of poverty reduction were attributable to growth, but others were associated with redistributive policies such as the expansion of social safety nets (figure 1.9). The sustainability of the social gains achieved by most countries in the region may be jeopardized by less positive prospects for economic growth and by stagnation in the pace of the reduction in income inequality. Accord- ing to de la Torre et al. (2014), growth in Latin America and the Caribbean has been decelerating since 2012 relative to the significant growth rates that characterized the region during the golden precrisis years. According to the latest projections, GDP growth in the region will reach only 1.7 percent in 2015 and 2.9 percent in 2016 (World Bank 2015c). Moreover, Cord et al. (2014) find evidence of stagnation in the pace of the reduction in income inequality in Latin America since 2010 (box 1.2). To identify opportunities to maintain the progress toward achieving the twin goals of ending extreme poverty and boosting shared prosperity, the next section presents a conceptual framework that is useful for understand- ing the factors that may contribute to boosting the capacity of individuals 16 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.9 Contributions of Growth and Redistribution to Falls in Extreme Poverty, Latin America and the Caribbean, around 2003–12 10 Reduction in extreme poverty (%) 5 0 –5 –10 –15 –20 r Bo l ia lo u rg ia Re a on n Pa ras ay in Sa ua Re dor ru c Pa ay a ca M le ua ico a i do i az tin m al r o bl liv b i Pe Ri gu gu Ch om E rag gi ex m m du na Br ua an lva pu en a ra te a Ec st Co ic U Co H A N G l ic D Country Redistribution Growth Source: Calculations based on data in SEDLAC. Note: The figure shows a Datt-Ravallion decomposition. Changes in extreme poverty are decomposed into changes associated with economic growth (or mean income) in the absence of changes in inequality (or income distribution) and changes in inequality in the absence of growth. For more information about the method, see Datt and Ravallion (1992). to generate income, climb out of poverty, and avoid the risk of downward mobility. The framework takes account of the concept of sustainability and the interaction of macro- and microeconomic variables in achieving and sustaining the goals socially, economically, and environmentally. The Asset-Based Approach to Gauging Household Income The World Bank goals of reducing extreme poverty and boosting shared prosperity have three important characteristics in common. First, both are measured using a monetary welfare indicator, such as income or consump- tion, as a proxy for the capability of individuals to achieve a certain stan- dard of living.14 The extreme poverty rate measures the share of individuals currently living below the $1.25-a-day threshold, while the shared prosper- ity goal aims to capture a relevant sustainable increase in income among the poorer segments of society, roughly defined as the bottom 40. Second, both Chapter 1: Overview 17 Box 1.2 Stagnation in the Contraction of Income Inequality in the Region The within-country trends in income inequality are significantly different in Latin America and the Caribbean if one views the last decade as two periods, 2003–10 and 2010–12 (figure B1.2.1). Such a split is useful because it showcases the stagnation in the pace of the contraction in income inequality in the region after the global financial crisis of 2008 (see Cord et al. 2014). Of the 17 countries on which data are available for 2003–10, 15 exhibited a decline in the Gini coefficient; Colombia and Costa Rica were the only exceptions. Since 2010, 4 of the 15 countries on which data are available experienced a rise in the Gini coefficient (Costa Rica, Honduras, Mexico, and Peru), while Panama showed no change. The rise of the Gini coefficient in Honduras was substantial, from 0.53 to 0.57 in 2010–12. Meanwhile, the increase in the Gini from 0.48 to 0.49 in Mexico in 2010–12 explains a good part of the recent regional slowdown in the decline of income inequality.a At the same time, while inequality reduction continued in 10 countries after 2010, the pace of the decline weakened in Brazil, the most populous country in the region.b Figure B1.2.1 Gini Coefficient: Annualized Changes, Latin America and the Caribbean, 2003–10 and 2010–12 5 4.0 4 3 2 0.9 1 0.2 0.3 ࡗ ࡗ 0 ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ ࡗ –0.0 ࡗ ࡗ ࡗ –1 ࡗ –0.8 –0.5 –0.4 –1.4 –1.3 –1.0 –1.0 –1.0 –2 –2.0 –1.9 –3 ru ca la ,u a ra n Sa uay Ec dor en Bo r ia in Co ban pu a ic ile il a st ru M ca on ico s do ra ay gu Re bi az m Pa ba bl a tin liv Co Pe Ri Ch m ex an lom du na Br a ua g gu ra r ur lv a te Pa a, ua i H N G El ic rg U om A D Country 2010–12 ࡗ 2003–10 Source: Cord et al. 2014, based on data in SEDLAC. Note: The figure shows changes in the Gini coefficient between 2003–10 and 2010–12, or the nearest years, in case data for these years are not available. Data on Guatemala and Nicaragua are available only for the first period. Cord et al. (2014) find that the declines in inequality before 2010 were driven by labor markets in the Andean and Southern Cone subregions, including Brazil, while in parts of Central America and Mexico, the decline was mainly determined by equalizing nonlabor income sources and the impact of (continued) 18 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.2 Stagnation in the Contraction of Income Inequality in the Region (Continued) the financial crisis, which especially affected the incomes of the top end of the distribution. They also find that the stagnation experienced since 2010 reflects, to a large extent, the subsequent recovery in Central America and Mexico. Moreover, even in countries in which income inequality continued to fall, this was mostly driven by zero or negative growth among the top of the income distribution, rather than greater growth among the poorest. a. The Gini coefficients in this study are calculated using the SEDLAC database, a regional harmonization effort. The effort generates income aggregates that are comparable across countries and, as a result, often differ from official income aggregates. The trends in Mexico’s Gini coefficient are comparable with the trends in the Gini calculated by Mexico’s National Institute of Statistics and Geography (INEGI) using the traditional Household Income and Expenditure Survey (ENIGH). The latter Gini increased from 0.435 to 0.440 between 2010 and 2012, while the Gini calculated by Mexico’s National Council for the Evaluation of Social Development Policy (relying on the socioeconomic conditions module of the household survey) fell from 0.509 to 0.498. b. Brazil is home to 37 percent of the total population of the 17 countries under analysis. focus on the welfare of those at the bottom of the income (or consumption) distribution; the poverty rate is an absolute measure, while shared prosper- ity is a relative concept. Third, both track economic progress by focusing on trends in household welfare. Based on these three shared characteristics of the twin goals, this section presents a simple asset-based approach as a macro-micro framework to guide the discussion in the following section, which describes key aspects of the capacity of households in the bottom 40 in Latin America and the Caribbean to generate income compared with the top 60 in the region. The framework is an extension of a model presented by Attanasio and Székely (2001) and Bussolo and López-Calva (2014) and that aims to unpack the elements of the market incomes of households to shed light on the potential determinants of outcomes in poverty and shared prosperity.15 In the framework, the realization of household market income is a func- tion of four main components: (1) the capacity of households to generate income based on the assets they own; (2) the private transfers households receive, which may include domestic and international remittances and in- kind transfers from other households; (3) the set of prices of the basket of goods and services that the households consume; and (4) a positive prob- ability of being affected by the realization of (negative or positive) shocks (health, natural disasters, crime, and loss of employment) (figure 1.10).16 The capacity of households to generate income based on the assets they own can be disaggregated into three additional elements: (1) the stock of income-earning assets owned by each household member, which may include human capital (such as educational attainment and years of experi- ence in the labor market), financial and physical assets (such as ownership of machinery or financial assets such as stocks and bonds), social capital (such as the set of norms and social networks that facilitate collective action; see Putnam 1993), and natural capital (such as land, soil, forests, and water); (2) the rate at which these assets are utilized by each household member to produce income (this may include labor market participation, the use of Chapter 1: Overview 19 Figure 1.10 The Asset-Based Approach to the Generation of Household Market Income Growth Returns to Private assets transfers Household market = Assets x Intensity of use x + x External shocks income Prices Prices Distribution machinery, and the exploitation of land through agricultural production); and (3) the returns to assets (such as the price of factors of production, including wages and interest rates). For ease of illustration, the elements of the asset-based framework are sometimes presented somewhat independently of each other. However, the elements do interact with each other as part of the dynamics of household income generation. For instance, nominal wages and the number of hours of work are important in decisions to participate in the labor market, and con- sumer prices may impact income earnings through the returns to the assets of producer households (Bussolo and López-Calva 2014; López-Calva and Rodríguez-Castelán 2014). Moreover, in the framework, both the observed accumulation of income-earning assets and the observed rate at which these assets are used by individuals are assumed to incorporate the desire of indi- viduals to realize their aspirations, one of the manifestations of agency. Some examples of how a lack of aspirations may prevent households from accumulating assets and participating in productive activities include sub- optimal investment in human capital and production technologies or the abandonment of the search for employment in formal sector firms.17 Furthermore, actual household market income may differ from poten- tial household market income because of shocks that may affect private transfers and the income from the use of assets. There are multiple exter- nal risks, including macroeconomic crisis, extreme climate-related events, health-related shocks, and crime and violence, that individuals and societies face and that can have pernicious consequences for the income generating capacity of households (World Bank 2013b). Risks turned into negative shocks could potentially lead to asset loss, disinvestment, unemployment, malnutrition, and child labor if people lack the means to manage and cope with them. A large body of empirical evidence shows that the poor are often more vulnerable to the negative consequences of shocks. Thus, in the 20 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean framework, the probability of being affected by external shocks is expected to be greater among low-income households. The asset-based approach integrates both the macroeconomic and the microeconomic dimensions so that growth and the incidence of growth can be understood as jointly determined processes. The framework facilitates an explanation not only of the ways macrofactors affect income growth among different population groups, but also of the ways the distribution of assets across such groups may determine the capacity of these groups to contribute to overall growth. According to Bussolo and López-Calva (2014), the framework considers the distribution of assets as a given in the short run, and changes in the income generating capacity of households thus depend mostly on the macroeconomic variables that influence the demand for labor across sectors, relative prices (returns), and the intensity of the use of assets over the economic cycle. In the long run, the main driv- ers of income growth will be the level and distribution of the assets—the human, physical, financial, social, and natural capital—that people own and accumulate, as well as the intensity with which the assets are used and the volume of the associated returns, which will reflect the productivity of the assets. Finally, the asset framework allows for a cohesive description of intra- and intergenerational economic mobility, chronic and transient poverty, and between-group inequities (the poor and the nonpoor, the bottom 40 and the top 60, minorities, and so on) that potentially thwart the possibility of certain vulnerable populations to participate in and benefit fully from the development process. In the next section, the asset-based framework is used to describe trends in selected central components of the income generating capacity of house- holds in the bottom 40 relative to the top 60 to shed more light on the significant progress achieved in poverty reduction and underscore the sub- stantial heterogeneity of the countries of Latin America and the Caribbean. The Income Generating Capacity of the Less Well Off Data from household surveys across the region show that labor makes up a significant majority of income across all countries among the bottom 40 and the top 60 (figure 1.11). Labor income accounts for 60 to 80 percent of total income among households in the bottom 40, while the corresponding share is even higher among households in the top 60. It has been the main driver of poverty and inequality declines over the past decade. The majority (60 percent) of the decline in extreme poverty in the region is explained by higher labor incomes. Higher earnings among women were responsible for 22 percent of the decline, while the earnings of men accounted for 38 per- cent (figure 1.12). Similarly, labor incomes explained approximately two- thirds of the total poverty reduction and about 45 percent of the inequality reduction between 2003 and 2012. Chapter 1: Overview 21 Figure 1.11 Labor Income, Bottom 40 and Top 60, Latin America and the Caribbean, around 2012 100 24 20 32 20 24 35 33 26 28 34 26 27 24 40 29 31 35 Share of total income (%) Bottom 40 50 76 80 68 70 76 65 67 74 72 66 74 73 76 60 71 69 68 0 100 29 17 29 21 24 27 29 20 25 19 24 22 22 28 21 30 38 71 83 71 79 76 73 71 80 75 81 76 78 78 72 79 70 62 Top 60 50 0 n ia il Co hile in os bia Re ica Ec lic Sa or ua dor on la M s ic ico Pa ua Pa ma ay ,u u an ra az ba ay er liv a d b gu ag rb an a R m ex C m du na Br P ua a pu C ur Bo lv te ar ra lo t a, tin H N G gu El en ic ru rg U om A D Country Labor income Nonlabor income Source: Calculations based on data in SEDLAC. Given the importance of labor income as a share of total income among the less well off, a description of how the capacity to generate labor income has evolved over the past decade across the region can promote a better understanding of the progress and divergence of countries with respect to the twin goals. In particular, this section focuses on the ability of the bottom 40 to generate labor income and explores the asset stock, intensity of use, and returns that determine labor income. It then illustrates the importance of private transfers, prices, and exposure to external shocks in determining the market income of households. It concludes with a brief discussion of how policies can be linked to the capacity of households to generate income through the asset-based approach. The stock of assets: human capital Human capital is generally defined as the stock of knowledge, competen- cies, and personal attributes that determine a person’s capacity to perform in a labor market. It can be built up through education or training, but also includes intrinsic talents and skills, such as creativity and discipline, that are more difficult to measure. Human capital is the main asset that allows individuals to generate labor income. Hanushek and Woessmann (2012) find that differences in human capital can account for half to two-thirds of the variations in income between Latin America and the rest of the world. In large part, this is driven by differences in educational attainment and in the 22 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.12 The Reduction in Extreme Poverty, by Income Component, Latin America and the Caribbean, 2003–12 Income component and employment level Men, labor Women, labor Nonlabor income Share of the change in the 0 –10 –10 –9 –10 headcount (%) –13 –12 –20 –22 –18 –30 –28 –38 –40 –40 le ent Em me r le ent m r ns s m r co o co o co he er in Lab in Lab io l l e e sf ve ve m m in Ot ns an oy oy Pe Tr pl pl Em Employment level Labor level Nonlabor income Source: Calculations based on data in SEDLAC. Note: Estimates of poverty at the regional level are population-weighted averages of countries. The figure shows the Shapley Decomposition of poverty changes between 2003 and 2012 by components of the income aggregate. See Azevedo, Sanfelice, and Nguyen (2012) for details about the decomposition technique. quality of schooling. Educational attainment is an imperfect, but important measure of human capital. In the past decade, there have been substantial improvements in educational attainment among the bottom 40 across the region, but the group continues to lag the top 60 (figure 1.13). Most countries in the region have achieved nearly universal coverage in primary education. With a few exceptions in Central America, the gaps in access to primary education between the bottom 40 and the top 60 have practically closed. While progress has also been made in access to secondary education (above 80 percent in most countries), access to tertiary educa- tion remains a privilege of the wealthier top 60, with more persistent gaps between the bottom 40 and top 60, and achieving universality among either group is a distant goal. For instance, in Uruguay, while access to secondary education was at 86 percent among the bottom 40 and 95 percent among the top 60 in 2012, access to tertiary education among these two groups was 21 and 55 percent, respectively. Despite the improvements in access and educational attainment, the quality of education remains an important challenge across the entire income distribution in Latin America and the Caribbean. There is signifi- cant variation in the quality of education within the region, which is heavily Chapter 1: Overview 23 Figure 1.13 Educational Attainment, Bottom 40 and Top 60, Latin America and the Caribbean, around 2003–12 Average educational attainment, 12 15+ age group (years) 10 8 6 4 2 0 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 60 Bo 60 Bo 60 Bo 60 60 60 m m m m m m m m m m m m m m p p p p p p p p p p p p p p tto tto tto tto tto tto tto tto tto tto tto tto tto tto Bo Bo Bo ARG BRA BOL CHL COL DOM ECU GTM HON MEX PER PRY SLV URY Country 2003 2012 Source: Calculations based on data in SEDLAC. Country codes are ISO 3166 standard. correlated with top 60 or bottom 40 status. While the rate of completion of the sixth grade on time has improved, especially among the bottom 40, there is still evidence of gaps across socioeconomic groups (figure 1.14). As of 2012, the gap in the completion of sixth grade on time between children in households in the bottom 40 and children in households in the top 60 was widest—more than 20 percentage points—in Colombia, the Domini- can Republic, and Nicaragua. Internationally comparable measures of educational quality such as the scores of the Program for International Student Assessment (PISA) of the Organisation for Economic Co-operation and Development (OECD) dem- onstrate that the region lags all other regions except Sub-Saharan Africa in learning outcomes. The assessment scores have improved among some countries in the region that apply the test, most notably Brazil and Peru and, to a lesser extent, Chile and Uruguay. However, overall performance is significantly behind the performance of the OECD countries. Thus, the average student in the region scores 100 points lower than the average OECD student in mathematics, which is equivalent to two full years of education in mathematics (Bruns and Luque 2015). Intensity of use: labor force participation To turn human capital into labor income, the poor and bottom 40 need access to the labor market. This includes not only the ability to participate in the labor market, but also sufficient labor demand so that the bottom 40 are able to work an adequate amount of time. The labor force participation 24 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.14 Completion of Sixth Grade on Time, Latin America and the Caribbean, 2000–12 95 On-time completion rate (%) 85 75 65 55 45 35 25 15 5 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 p 40 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 60 To m To m To m To m To m To m To m To m To m To m To m To m To m To m To m To m To m tto tto tto tto tto tto tto tto tto tto tto tto tto tto tto tto tto Bo ARG BOL BRA CHL COL CRI DOM ECU SLV GTM HON MEX NIC PAN PRY PER URY Country 2000 2012 Source: Calculations based on data in SEDLAC. Note: The figure reflects a simulation for 12- to 16-year-olds. For Brazil, Guatemala, and Nicaragua, the simulation represents 13- to 17-year-olds because primary education starts one year later in these three countries. Country codes are ISO 3166 standard. rate in the region was slightly above 65 percent between 2003 and 2012. However, regional trends in labor force participation diverged among indi- viduals in the bottom 40 and individuals in the top 60: the rate increased from 66.7 to 68.6 percent among the latter, but fell from 62.8 to 59.4 per- cent among the former. This phenomenon, which was related to a decline in the use of produc- tive assets among the less well off between 2003 and 2012, was the norm in many countries in Latin America (figure 1.15). With the exception of a few countries in Central America, the Dominican Republic, Mexico, and Para- guay, the share of the bottom 40 participating in the labor force dropped during these years. The trends were similar among men and women except in Chile and Uruguay, where labor force participation narrowed among men and widened among women. Moreover, in the countries in which the labor force participation of the bottom 40 increased, female labor force participation drove the change. Labor earnings among women can thus make a key contribution to poverty reduction and greater shared prosper- ity. Indeed, female labor market participation grew by 15 percent in Latin America from 2000 and 2010, which contributed to the substantial drop in poverty rates observed across the region (World Bank 2012a). Among the Chapter 1: Overview 25 Figure 1.15 Gaps in Labor Force Participation, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 90 Average labor force participation, 85 80 15+ age-group (%) 75 70 65 60 55 50 45 40 40 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 60 60 Bo 60 Bo 60 Bo 60 Bo 60 60 60 Bo 60 Bo 60 60 Bo 60 Bo 60 60 m m m m m m m m m m m m m m p p p p p p p p p p p p p p tto tto tto tto tto tto tto tto tto tto tto tto tto tto To To Bo Bo Bo Bo Bo Bo ARG BRA BOL CHL COL DOM ECU GTM HON MEX PER PRY SLV URY Country 2003 2012 Source: Calculations based on data in SEDLAC. Country codes are ISO 3166 standard. top 60, labor force participation rose in most countries, mainly also driven by the higher participation of women in the labor market. The higher labor force participation rates of the top 60 relative to the bottom 40 is somewhat endogenous, but is nonetheless indicative that the bottom 40 may face higher barriers or enjoy fewer opportunities or incen- tives to access labor markets. The decline in the share of the bottom 40 participating in the labor force suggests that the reduction in poverty and in the promotion of shared prosperity observed in the region would have been even more dynamic had the labor participation among the bottom 40 risen in more countries. Achieving a better understanding of the constraints faced by the bottom 40 in participating in labor markets is thus critical to efforts to enhance the inclusiveness of growth and the ability of the bottom 40 to contribute to growth. Box 1.3 discusses several hypotheses that may explain the decline in labor force participation among the bottom 40 in many countries in Latin America. Over the past decade, there have been important gains in access to hous- ing and communications infrastructure that, all else being equal, may have enhanced the access to markets and allowed for greater use of productive assets by households. Recent studies have found that greater access to elec- tricity services among informal women entrepreneurs and wider access to financial markets through mobile phone services can have positive effects 26 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.3 Explaining the Decline in Labor Force Participation among the Bottom 40 There may be several reasons for the drops in labor force participation among the bottom 40 in Latin America and the Caribbean. According to one hypothesis, younger segments of the population are delaying their participation in the labor market to invest in education. This would represent a potential trade-off involving a sacrifice of short-term gains in poverty reduction and shared prosperity for greater long-term human capital improvements. This hypothesis is consistent with the falloff in labor force participation among 15- to 20-year-olds in many countries in 2012 and the rise in enrollments in secondary and tertiary education among the poor in the region. This was evident in, for example, Bolivia, Brazil, Colombia, and Ecuador. According to a second hypothesis, the high unemployment rates observed among younger age- groups discourage labor force participation. This is consistent with data indicating that persistent shares of youth are out of school and out of work (Cárdenas, de Hoyos, and Székely 2014). Recent demographic trends have pushed youth above the threshold for working age, while the workforce, especially potential workers with less education or poorer-quality education, may not be able to take advantage of employment opportunities. A third hypothesis is related to the potential effects on labor force participation at the margin, particularly among the 25–65 age-group, caused by newly expanded social protection systems across the region, including conditional cash transfer programs, universal health insurance schemes, and unemployment insurance initiatives. This hypothesis is consistent with the findings of recent studies on the negative labor market outcomes generated by social protection schemes instituted in parallel to established social security programs for the formally employed (for instance, see Levy 2008 on the case of Mexico). Argentina, Brazil, and Ecuador may offer examples of this phenomenon. A fourth hypothesis focuses on the decline in labor force participation among the 65+ age-group. Because of the aging of the population, smaller, younger cohorts are unable to replace the older cohorts that are retiring, thereby cutting into overall participation rates. Moreover, the expansion of noncontributory pension programs and skills obsolescence among older workers, especially in the context of the demands of new information technologies, may also be contributing to a reduction in the labor force participation rates among the 65+ age-group. on the productive use of assets by households (Demombynes and Thegeya 2012; Dinkelman 2011). Box 1.4 presents evidence on access to services in Latin America that can be associated with the greater use of the productive assets of households, particularly among the poor and the bottom 40. Returns: wages Despite the drop in labor force participation among the bottom 40, there has been improvement in hourly wages among the bottom 40 in most of the countries of the region over the past decade.18 The rise in hourly wages has been especially strong in Argentina, Bolivia, and Brazil, while the rise has been more moderate in Chile, Colombia, Nicaragua, Paraguay, Peru, and Uruguay. The rest of the region has seen smaller increases in hourly wages among the bottom 40. In contrast, except for Honduras, the top 60 has enjoyed a smaller expansion in hourly wages (figure 1.16). This indicates that an important force behind the rise in the incomes of the bottom 40 has been higher returns in the labor market rather than greater labor market Chapter 1: Overview 27 Box 1.4 Connectivity Infrastructure in Latin America and the Caribbean While not a perfect indicator of connectivity to markets, access to electricity and new information technologies are a good proxy for the transaction costs and barriers associated with accessing mar- kets. Access to electricity, cell phones, and the Internet allows individuals to connect to markets to employ their assets and obtain returns. Access to electricity has improved across Latin America and the Caribbean over the past decade, and regional disparities have shrunk substantially (figure B1.4.1). Bolivia and Peru have made the big- gest advances in expanding electricity coverage among the bottom 40. However, substantial dispari- ties still exist within and across countries. While less than 70 percent of the population in Nicaragua has access to electricity, Brazil, Chile, and Uruguay have achieved almost universal coverage. Many countries have closed the electricity gap between the bottom 40 and the top 60, but the gap is sill large in Bolivia, many Central American countries, and Peru. Figure B1.4.1 Electricity Coverage Rates, Latin America and the Caribbean, 2000–12 100 Electricity coverage rate (%) 90 80 70 60 50 40 30 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 60 m m m m m m m m m m m m m m m p p p p p p p p p p p p p p p tto tto tto tto tto tto tto tto tto tto tto tto tto tto tto Bo BOL BRA CHL COL CRI DOM ECU SLV GTM HON MEX NIC PRY PER URY Country 2000 2012 Source: Calculations based on data in SEDLAC. As of 2012, access rates to cell phones were high in the region among both the bottom 40 and the top 60 (figure B1.4.2). The large gaps between the bottom 40 and top 60 observed at the beginning of the decade had been nearly erased 12 years later in countries such as Brazil and Chile. However, cover- age gaps of over 20 percentage points between households in the top 60 and the bottom 40 persist in Mexico, Nicaragua, and Peru, and this limits access to markets and information among the poorest. Internet access rates are much lower across the region, and there is significant heterogeneity (fig- ure B1.4.3). Available data suggest that countries have made enormous leaps in Internet connectivity over the past decade. Coverage rates in Brazil and Chile rose from low levels to 21 and 25 percent of the bottom 40, respectively. However, unlike electricity and cell phone coverage, which is now almost universal across the region, even the wealthiest Latin American countries barely reach 50 percent in Internet coverage, while coverage does not exceed 10 percent in Bolivia and in Central America. (continued) 28 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.4 Connectivity Infrastructure in Latin America and the Caribbean (Continued) Figure B1.4.2 Cell Phone Coverage Rates, Latin America and the Caribbean, 2000–12 100 90 Cell phone coverage rate (%) 80 70 60 50 40 30 20 10 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 p 0 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 To 4 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 Bo 60 60 m m m m m m m m m m m m m m m tto tto tto tto tto tto tto tto tto tto tto tto tto tto tto Bo BOL BRA CHL COL CRI DOM ECU SLV GTM HON MEX NIC PRY PER URY Country 2000 2012 Source: Calculations based on data in SEDLAC. Figure B1.4.3 Internet Coverage Rates, Latin America and the Caribbean, 2000–12 80 70 Internet coverage rate (%) 60 50 40 30 20 10 0 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 To 40 60 60 60 60 60 60 60 60 60 60 60 60 60 m m m m m m m m m m m m m p p p p p p p p p p p p p tto tto tto tto tto tto tto tto tto tto tto tto tto Bo Bo Bo Bo Bo Bo Bo Bo Bo Bo Bo Bo Bo BOL BRA CHL COL CRI ECU SLV GTM HON MEX PRY PER URY Country 2000 2012 Source: Calculations based on data in SEDLAC. Chapter 1: Overview 29 Figure 1.16 The Rise in Hourly Wages, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 Annualized growth rate 8 of hourly wages (%) 6 4 2 0 –2 n ia il ile a Ec ic Sa or or on a s o Pa ua ay ru an ra az bi al ba ic bl liv d ad Pe Ch gu ag rb m ex m du Br ua pu ur Bo lv ,u te ar ra lo M Re a, ua Co ic ay tin H N an G gu El en ic ru rg in U om A D Country Bottom 40 Top 60 Source: Calculations based on data in SEDLAC. participation, which is consistent with the falling skill premiums noted in many studies during the first decade of the 21st century. Despite the gains among the bottom 40, some population groups are lag- ging in wage compensation. Thus, for example, according to a recent report of the World Bank (2012a), women and men may not be compensated on par. After controlling for education, age, and the share of workers in each occupation between 2000 and 2010, the report finds evidence of a large and persistent wage gap affecting women in Brazil, Chile, Mexico, and Peru that is especially marked among the top-paid professions. One of the advantages of the simple asset-based framework is the frame- work’s suitability for the analysis of the capacity of various socioeconomic and demographic groups to generate income. Box 1.5 describes poverty rates and the capacity to generate income among indigenous populations based on a subset of countries on which household survey data on ethnicity are available. Private transfers In some countries and among some households, private transfers, such as remittances and in-kind transfers from other households, can be a major source of income and a determinant of household well-being. In the region, total transfers represent about 10 percent of total household income. Moreover, the share of private transfers in total household income tends to be larger among the bottom 40 than among the top 60. However, the 30 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.5 The Asset-Based Approach: Indigenous Populations Poverty reduction The poverty reduction in Latin America and the Caribbean between 2000 and 2012 was also evident among most indigenous groups. For instance, the share of indigenous people living on less than $2.50 a day in Bolivia and Ecuador (extreme poverty) fell 19 and 17 percentage points, respectively. In both cases, the decline was higher than the decline among the total population. In contrast, the share of indigenous people living on less than $2.50 a day in Guatemala rose from 45.7 to 54.9 percent over the period.a Figure B1.5.1 $2.50- and $4.00-a-Day Poverty Rates, Indigenous Populations, Latin America and the Caribbean, 2000–12 a. Extreme poverty rate ($2.50 a day) 60 50 40 Percentage 30 20 10 0 2000 2012 2000 2012 2000 2012 2000 2012 2000 2012 2000 2012 2012 Guatemala Ecuador Bolivia Brazil Peru Chile Mexico b. Moderate poverty rate ($4.00 a day) 80 70 60 Percentage 50 40 30 20 0 2000 2012 2000 2012 2000 2012 2000 2012 2000 2012 2000 2012 2012 Guatemala Ecuador Bolivia Brazil Peru Chile Mexico Indigenous Total population Source: Calculations based on data in SEDLAC. Note: The nearest year to 2000 or 2012 is used for countries on which data are not available in that year. Ethnic identity is based on self-reported data. Because the data presented here are based on SEDLAC, a regional data harmonization effort that increases cross-country comparability, they may differ from official statistics reported by governments and national statistical offices. All monetary values are reported in 2005 PPP international U.S. dollars. (continued) Chapter 1: Overview 31 Box 1.5 The Asset-Based Approach: Indigenous Populations (Continued) Level of assets: human capital The positive changes in poverty reduction in the past decade have been accompanied by improve- ments across various education measures on indigenous populations in the region. Among relevant groups in Brazil and Ecuador, school enrollments among 6- to 15-year-olds rose 9 and 10 percentage points, respectively. In 2000–12, the groups in Brazil also showed the greatest increase in average years of schooling—1.5 additional years—among people aged 18+ years. Guatemala experienced the greatest gains in the literacy rate (12 percentage points) and school enrollments among 6- to 15-year- olds (18 percentage points) during the period. Nonetheless, indigenous groups continue to lag the total population in human capital accumulation. As of 2012, Bolivia, Ecuador, and Guatemala had the widest gaps in educational attainment. Indigenous groups in Ecuador exhibited an average of four years less schooling than the total population. Similarly, in Bolivia, the literacy rate among indigenous groups was 13.7 percentage points lower. Intensity of asset use: labor force participation Trends in labor force participation rates among indigenous groups was heterogeneous in 2000–12. In Bolivia, participation rates among indigenous groups expanded by 3.6 percentage points, greater than the 1 percentage point increase among the total population and the largest rise among the countries in the analysis. Enhancements in human capital accumulation and employment have translated into greater poverty reductions in Bolivia. In contrast, labor force participation among indigenous groups in Ecuador declined by 10.7 percentage points, deeper even than the 7.2 percentage point fall among the overall population. The drop occurred mainly because of female labor force participation in both groups, which narrowed by nearly 16 and 10 percentage points, respectively, during the period. Despite the progress, indigenous groups still lag in the region, and this is hindering advances in shared prosperity and poverty reduction. a. The share of the indigenous population living on less than $1.25 a day in Guatemala increased from 17.3 to 18.5 percent over the period. However, this was smaller than the rise among the total population (11.8 to 13.7 percent). significance of private transfers as a share of total transfers varies widely across countries and between the bottom 40 and the top 60 (figure 1.17, panel a). Private transfers are especially important in countries in Central America, such as El Salvador and Guatemala, where they account for more than 80 percent of total households transfers. Evidence indicates that the positive effects of remittance flows include greater macroeconomic stability, higher savings, better access to health care and education, more entrepreneurship, and reductions in poverty and social inequality. The money migrant workers send back to their home countries is linked to lower poverty rates and enhancements in education and health indicators (Fajnzylber and López 2008). Between 2002 and 2008, remit- tance flows rose substantially each year, at an average rate of 17 percent. However, in 2006, the growth rate, though high, began slowing, and, because of the economic crisis in 2008, remittances fell more than 15 per- cent in the final two quarters of 2009. Given the importance of these flows for the recipient households, migrants adjusted their spending habits to continue to send money home despite the economic uncertainty. The year 32 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.17 Transfers, Bottom 40 and Top 60, Latin America and the Caribbean, 2003–12 a. Transfers in household income, circa 2012 household income (%) 30 Share of total 20 10 0 40 tto 60 To 0 tto 0 40 tto 0 To 0 tto 0 40 tto 0 To 0 tto 0 To 0 tto 0 To 0 tto 0 40 tto 0 40 tto 0 To 0 60 4 6 Bo p 6 4 Bo p 6 Bo p 6 4 Bo p 6 4 Bo p 6 4 Bo p 6 Bo p 6 Bo p 6 4 m p m p m m m m m m m m m p To To To To To tto Bo Bo Bo , ia bl an r a as o a y n , n a ba ay do al m ua ic ba in liv r pu nic or m ex ur gu du na ua El ic ur ent ag Bo ad te M Re mi ru on Pa Ec r rg ua Pa lv U o H Sa A D G Country Public transfers Private transfers b. Growth: private transfers, circa 2003–12 10 Annualized growth in private transfers (%) 5 0 –5 –10 –15 ba a, ia ic r or a s o a ay ba y, do ra al m ic bl ur tin liv ur ua ad gu n n m ex du na ua pu Bo g en lv te ra M ru on Pa Ec Re Sa rg ua Pa U H A an G El ic Country in om D Bottom 40 Top 60 Source: Calculations based on data in SEDLAC. Note: The figure covers only countries where data on private transfers are available and comparable. 2010 marked the start of an upward trend lasting throughout that year and reaching an annual positive growth with respect to the previous year. The flows in 2011 exceeded the amounts sent the previous year by 6 percent, the largest positive growth rate of the previous four years (Maldonado, Bajuk, and Hayem 2012). Over the past decade, the trends in the growth of private transfers, which includes remittances and other in-kind transfers, varied by country and among the top 60 and the bottom 40 (see figure 1.17, panel b). However, in most countries, private transfers grew more quickly or fell more slowly Chapter 1: Overview 33 among the bottom 40. The only two countries in which private transfers grew more slowly among the bottom 40 were Mexico and Uruguay. Among the top 60 in most countries on which data are available, private transfers showed negative growth rates. In El Salvador, one of the largest remittance-receiving countries in the region, private remittances played a major role in poverty reduction. In 2012, private remittances accounted for over 16 percent of GDP, a more than 10-fold increase since 1990. Remittances expanded in both size and scope. In 2000, 4 percent of households received remittances; by 2012, one in five households was receiving remittances, while the amount per migrant rose by almost a third between 2000 and 2010. Remittances do not necessarily go to the poorest households in El Salvador; the average per capita income in households receiving remittances is $8.90 (2005 U.S. dollars), compared with $3.10 among poor households. Reliance on remit- tances exposes countries to the business cycles of the countries in which the migrants reside. In El Salvador, this means a strong reliance on the U.S. economy because 88 percent of Salvadoran migrants reside in the United States.19 The sharp decline in remittances that occurred because of the 2008 financial crisis highlights the vulnerability associated with this dependence. In Paraguay, family transfers may not be an important driver behind the change in the incidence of poverty, but still play an important role in alleviating poverty and as a household mechanism for coping with adverse shocks. Without these transfers, the extreme poverty rate in rural areas would be 4 percentage points higher. The elderly and woman-headed households receive substantially larger family transfers, suggesting that migration is a household income diversification and coping mechanism. Prices of goods and services The market income of households is also directly affected by the prices of the goods and services they consume. During the past decade, macrosta- bility has translated into lower inflation rates, which has helped maintain the purchasing power of households relative to the situation in the 1990s. However, fluctuations in food prices have been an important source of vulnerability among some households in the bottom 40. Evidence shows that households in the lower deciles of the income distribution consume a higher share of food with respect to their total basket of goods, and these households are thus more exposed to changes in food prices (figure 1.18).20 Estimates based on the latest recorded worldwide increase in food prices, in 2011, show that high, volatile food prices pushed 44 million people further into poverty primarily in low- and middle-income countries (World Bank 2011a). Box 1.6 presents an interesting case of the potential negative effects of high food prices on poverty reduction. Risk and external shocks Uninsured risks often have permanent effects on the welfare of households by aggravating poverty traps because low-income people—the poor or the bot- tom 40—are often more vulnerable to the negative consequences of shocks 34 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.18 Food Consumption in Total Consumption, Latin America and the Caribbean, around 2010 80 Food share in total consumption (%) 60 40 20 0 il ile a r or a o ua a ru do az bi al m ic ad Pe Ch ag m ex m na Br ua lv te ar lo M Pa Ec Sa ua Co ic N G El Poorest quintile Quintile 2 Quintile 3 Quintile 4 Richest quintile Source: Calculations based on data in SEDLAC. (Barro 2006; Becker 1968; Carter et al. 2007; Dercon and Christiaensen 2011). Specifically, negative shocks can directly affect all components of the income generating capacity of households. For instance, the assets of any individual or household could be destroyed by a natural disaster, and such a disaster could also affect household decisions on the accumulation of cer- tain assets. Uncertainty in the realization of shocks may likewise affect the intensity of the use of assets; for example, an expected drought (or flood) could reduce the utilization of land for agricultural activities. Risk is also captured by relative prices similar to the interest rate, which certainly cap- tures the sovereign risk of an economy as a whole. Finally, macroeconomic contagion can cause a fiscal crisis that may reduce a government’s capacity to provide social assistance to the poor by reducing the coverage or the size of cash transfers. One increasingly important source of risk is climate change, which is expected to raise the frequency and severity of extreme weather events. The Latin America and Caribbean region has already experienced the greater variability, frequency, and strength of natural disasters in recent years. In particular, there appears to be a positive correlation between natural disas- ters in the region and a worsening trend in welfare indicators (figure 1.19). Poor and vulnerable populations tend to be more prone to episodes that result in the loss of income or assets. Poor households may be exposed not only to large, unusual shocks, but also to smaller high-frequency events that may prevent the households from escaping poverty. Chapter 1: Overview 35 Box 1.6 The Poverty Effects of High Food Prices, Paraguay Between 2003 and 2013, economic growth and improvements in income distribution combined to contribute to a large reduction in moderate poverty in Paraguay, from 44.0 to 23.7 percent. However, because the extreme poverty line is determined solely based on the price of a selected food basket, the reduction in the extreme poverty rate became less dynamic when food prices began rising at a higher rate than general prices. This was particularly evident in 2003–11, when extreme poverty fell by only 3.2 percentage points. In contrast, a slowdown in food price inflation in 2011–13 was an important contributing factor in the 7.9 percentage point decline in the extreme poverty rate during those years (figure B1.6.1, panel a). Figure B1.6.1 Changes in the Extreme Poverty Rate, Paraguay, 2003–11 and 2011–13 a. Extreme poverty rate b. Decomposition of changes in extreme poverty 25 8 Change in extreme poverty 6 and poverty change (%) 20 (in percentage points) Extreme poverty rate 21.2 4 6.3 15 18.0 18.0 2 10 0 –2.4 –2 5 10.2 –7.6 –4 –5.1 0 –6 2003 2011 –3.2 2011 2013 –7.9 –8 –1.8 –0.4 –5 Change Change –10 –10 –12 2003–11 2011–13 2003–11 2011–13 Years Years Poverty line Redistribution Growth Source: World Bank calculations based on data from the Permanent Household Survey for 2003, 2011, and 2013. A quantification of the effects of economic growth, redistribution, and an extreme poverty line based solely in food prices helps unpack the changes in extreme poverty over the last decade in Para- guay. Together, high economic growth rates and improved income distribution accounted for a decline by 9.5 percentage points in extreme poverty in 2003–11, while rapidly rising prices for the food items in the basket relative to general prices slowed the reduction in the extreme poverty rate by 6.3 percentage points (figure B1.6.1, panel b), leading to a net decline of only 3.2 percentage points in the rate. Thus, the food price rise relative to general prices cut into the positive effects on poverty reduc- tion of significant economic growth and gains in redistribution. In contrast, since 2011, all three forces have been trending in the same direction. The deceleration of the increase in food prices between 2011 and 2013 meant that, in real terms, the extreme poverty line—updated using food price data of the Banco Central del Paraguay—was marginally lower in 2013 than in 2011. As a consequence, prices played a limited, but positive role in the drop-off in the extreme poverty rate, whereas the enhanced income distribution reflected in the widening of the distribution was responsible for 65 percent of the total change in the headcount (5 percentage points out of close to 8), and average income growth (the shift to the right in the distribution) explains the remaining 35 percent of the fall. An additional contributing factor behind the sensitivity of the extreme poverty line to food prices is the fact that a large share of the population lives in households with incomes near the extreme poverty line. Because of this clustering, even slight shifts in the poverty line can have noticeable impacts on poverty rates. 36 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 1.19 Incidence and Poverty Effects of Natural Disasters, World Regions and Latin America and the Caribbean, 1970–2009 a. Incidence, by region, 1970–2008a 30 28.5 Events by region (number) 25 21 20 16.5 15.6 15 15 13.5 13.3 10.9 11.6 10.5 10 8.7 8.3 6.4 5.4 5 5.5 4.7 5 3.9 2.1 1.6 0 1970s 1980s 1990s 2000s Decade Africa Asia-Pacific Central and Eastern Europe Western Europe Latin America and the Caribbean b. Correlation with poverty, Latin America, 2009 45 ࡗ Extreme poverty headcount (%) 40 35 ࡗ 30 ࡗ 25 ࡗ ࡗ 20 ࡗ ࡗ ࡗ ࡗ ࡗ 15 ࡗ ࡗࡗ 10 ࡗ ࡗ 5 ࡗ ࡗ 0 0 0.5 1.0 1.5 2.0 Population affected by droughts, floods, and extreme temperatures (%) Sources: Cavallo and Noy 2011 based on data in, for panel a, EM-DAT (International Disaster Database), Centre for Research on the Epidemiology of Disasters, Université Catholique de Louvain, Brussels, http://www.emdat.be/database; and, for panel b, SEDLAC for the poverty headcount; and WDI (World Development Indicators) (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-develop- ment-indicators, for population impacts. a. The years for Latin America and the Caribbean in panel a are 1970–2012. Báez, de la Fuente, and Santos (2010) show that disasters produce del- eterious impacts on education, health, and many income generating pro- cesses. They also highlight that, in most disaster events, the poorest carry the heaviest burden of the effects. For instance, in Peru, while 30 percent of Chapter 1: Overview 37 Figure 1.20 Shocks Reported by the Bottom 40 and Top 60, Peru, 2013 15 14 Respondents (%) 10 10 9 5 5 5 5 5 4 1 1 1 1 0 Employment Health Household Crime Natural Other or business event breakup related disaster Event Bottom 40 Top 60 Source: Calculations based on data from the National Household Survey. Note: The events resulting in the loss of income or assets were self-reported within the previous 12 months. Employment or business = an episode involving job loss or the loss of a family business by a household member. Health event = a household member was sick. Natural disaster = drought, flood, storm, infectious disease or epidemic, and so on. Crime related = a household member was robbed or assaulted. Household breakup = household head left the household. The data on natural disasters are statistically different. households in the poorest decile reported experiencing a shock that trans- lated into a loss of income or assets, only 14 percent of households in the richest decile did so. Poor households are especially vulnerable to weather- related events in Peru. While many events that cause shocks affect the bot- tom 40 and top 60 similarly, households in the bottom 40 are substantially more likely to report they are affected by natural disasters and weather- related crises (figure 1.20). Box 1.7 presents the relevant case of Haiti. Another source of risk among many households in the region is crime and violence. Ongoing crime and violence across Central America and Mexico affect all aspects of development and intensifies inequities. They influence investment in human capital, raise the security costs of businesses, divert funds to combating crime, and discourage domestic and international investment because they impact the general investment climate (Cárdenas and Rozo 2008; Dell 2014; Powell, Manish, and Nair 2010; World Bank 2014d). For instance, the costs of crime and violence in El Salvador are high. Acevedo (2008) estimates that the costs of crime and violence represented almost 11 percent of GDP there in 2008.21 There is ample evidence of the effects of crime and violence among individuals and firms. Over 45 percent of men and 40 percent of women in El Salvador alter their shopping habits because of fear of crime and violence; 15 percent have moved; and over 38 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 1.7 Shocks, Coping, and the Impact on Household Welfare, Haiti The recent history of Haiti is characterized by a combination of shocks and slow economic growth. In 2004, political and extreme weather events led to a 5 percent contraction in GDP. In May 2004, Hurri- cane Jeanne killed some 3,000 people, left a quarter of a million people homeless, and generated economic losses estimated at nearly $300 million (Zapata Martí 2005). In 2008, four hurricanes (Fay, Gustav, Hanna, and Ike) led to a combined economic contraction of 1 percent of GDP per capita. The associated floods destroyed more than two-thirds of the country’s crops, resulting in child malnutri- tion and death. In 2010, a severe earthquake brought about the largest per capita GDP contraction in Haiti’s history, at 5.5 percent, and a death toll of over 300,000. In 2012, hurricanes Isaac and Sandy had a significant economic impact: the former destroyed nearly $250 million in crops, and the latter dev- astated 90,000 hectares of cropland leading to a fall in per capita GDP of 1 percent. A recent study (World Bank and ONPES 2014) includes the results of an analysis of the relationship between poverty incidence and shocks produced by natural disasters in Haiti using a survey of living conditions—the Enquête sur les Conditions de Vie des Ménages après le Séisme (postearthquake household living conditions survey)—collected through a partnership between the World Bank and the government of Haiti. The study also considers household risk coping mechanisms, such as using savings, receiving transfers from friends, changing nutritional inputs, or taking children out of school. The study finds that a typical Haitian household faces multiple shocks annually and that nearly 75 percent of households are economically impacted by at least one idiosyncratic shock each year. Households in poverty are more likely to experience shocks: 95 percent of households in extreme poverty experience at least one economically damaging shock annually. Although households impacted by climatic shocks are more likely to be affected by agricultural setbacks or covariate eco- nomic shocks, there are no clear patterns indicating that certain types of shocks occur together. The study also finds differences in the use of coping mechanisms by both the type of shock expe- rienced and the poverty status of the household. Most households are able to cope with idiosyncratic shocks without resorting to changes in nutritional inputs. However, nutritional inputs are less well protected if a household experiences a covariate economic or weather shock. If there is a covariate economic shock in the community, a staggering 56 percent of households in extreme poverty change their nutritional profile, compared with 37 percent among resilient households (that is, the nonpoor and nonvulnerable). The study also shows that shocks are more likely to impede the future economic activities of households because households are forced to sell assets or take on debt to cope; this also affects households in extreme poverty more frequently than resilient households. Relative to house- holds in extreme poverty, resilient households are two times more likely to rely on nonloan monetary help supplied by outsiders. 5 percent have changed jobs out of concern of being victimized. In 2010, over 85 percent of firms paid for security, which is 25 percentage points above the regional average, and slightly more than half of all firms identi- fied crime, theft, and disorder as the major constraint to doing business, which is also substantially higher than the regional average. In 2011, the homicide rate reached 90 per 100,000 deaths in Honduras, three times the level in Mexico and higher than the rate in El Salvador, which had the second highest rate. If crime were reduced by 10 percent in Honduras, then GDP could increase by 0.7 percent (World Bank 2011b). In 2012, the majority of Hondurans and Salvadorans reported crime and violence as the number one problem in their countries (Lagos and Dammert 2012). Chapter 1: Overview 39 Mexico has experienced an increase in the number of drug-related homi- cides, from 28 to 73 percent of total homicides from 2007 to 2011, respec- tively (SNSP 2012). Enamorado, López-Calva, and Rodríguez-Castelán (2014) find a negative impact of drug-related crime on income growth in municipalities in Mexico from 2005 to 2010 and no significant effect of non–drug-related crime on economic growth. Moreover, Enamorado et al. (2014) contend that a 1 percentage point rise in the Gini coefficient trans- lated into an increase of more than 10 drug-related homicides per 100,000 inhabitants between 2006 and 2010. Despite a lack of data on households, an analysis of municipal data sug- gests the relationship between poverty and crime in Mexico was convex in 2010: homicide rates were higher in both the poorest and richest municipal- ities (World Bank 2012b). This may arise because criminal organizations were diversifying their activities into richer municipalities through kidnap- ping and extortion or because of an effective security strategy in areas with high concentrations of crime and poverty. Using the World Bank (2012b) methodology, the United Nations Development Programme (UNDP 2013) finds parallel results in Brazil in 2011. In Colombia, the results suggest a contrast: the higher homicide rates occur in municipalities with the highest rates of multidimensional poverty. Drug-related violence is also associated with higher unemployment and poorer school performance and can have long-run detrimental conse- quences in human capital accumulation (Arias and Esquivel 2012; Caudi- llo and Torche 2014; Michaelsen and Salardi 2013). Similarly, Velásquez (2014) finds that the violence associated with the Mexican drug war may also have long-term consequences on the wealth and welfare of Mexican households. Not only does the evidence suggest drug-related crime hinders economic growth, but the costs of combating drug trafficking are estimated at $9 billion a year, nearly as much as the Mexican government spends on social development (Keefer and Loayza 2010). Links to policies The asset-based framework represents a valuable way to approximate the heterogeneity in shared prosperity in the region. The capacity to accumulate assets, use them intensively, and obtain returns from the assets are system- atically different among households in the bottom 40 and households in the top 60, and there are large variations across countries. The framework helps highlight how some macro and external variables that are not under the control of households may affect poor and less well-off households dif- ferently, such as food prices, climate change, or crime. The specifics largely depend on the context in each country and are examined in the country chapters. In particular, a meaningful discussion of effective policy interventions to further shared prosperity requires a more detailed analysis within coun- tries to understand the potential determinants of the diversity. The policy framework described below represents a systematic, concrete method to analyze the links between policies and the income growth of the bottom 40 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean 40. Interventions in specific policy areas can be weighed for their potential impact on the accumulation of assets, the intensity of asset use, and the returns to assets and on final market incomes. This can help gauge how the policies may eventually allow the less well off to contribute to growth. Thus, this subsection elaborates on the connection between the policies and household market income to provide a road map for profiting from the country chapters. The asset-based framework assumes that all agents are rational, that markets function perfectly, and, thus, that all individuals can take advan- tage of the full potential of their assets. However, in reality, the main fac- tors that affect the income generating capacity of households include, for instance, inequality in opportunities, risk, and market failures that explain why some individuals are able to accumulate more productive assets, while others are prevented from doing so. Based on an examination of interven- tions that address institutional and market imperfections and that are gener- ally used in microeconomic theory, the asset-based approach can be linked to four fundamental policy areas that have a direct impact on the capacity of households in an economy to generate income, but with a special focus on households in the bottom 40. The policies have also been identified in previous studies (World Bank 2013a, 2014a). They are (1) equitable, effi- cient, and sustainable fiscal policy and macroeconomic stability; (2) fair and transparent institutions capable of delivering good-quality basic ser- vices; (3) well-functioning markets; and (4) adequate risk management at the macro and household levels (figure 1.21). The policies can influence the realization of the total income of households by directly affecting the pri- vate income generating capacity of households through asset accumulation, Figure 1.21 Policy Areas That Affect Household Income Generating Capacity Equitable fiscal policy and macro Fair and transparent stability institutions Returns to Private assets transfers Household market = Assets x Intensity of use x + x External shocks income Prices Prices Well-functioning Risk markets management Chapter 1: Overview 41 asset use, returns to assets, and increases in the size of private transfers or adding public transfers, while mitigating the negative effects of external shocks.22 First, equitable and sustainable fiscal policy has an impact on income generating capacity through direct taxation; it also affects the decisions of individuals about the intensity of the use of assets by influencing returns through direct taxes and public transfers. Indirect taxes, such as the value added tax, can have an immediate effect on consumer prices and, thus, have an impact on the relative returns of households. Although there is evidence that fiscal policy has a limited impact on inequality in Latin America and the Caribbean, the expansion of cash transfers and noncontributory pen- sion programs in the region in the last decade has provided a safety net that has pulled people out of poverty by boosting their incomes directly and helping to protect them from falling back into extreme poverty if they are hit by external shocks (World Bank 2014a). While direct cash transfers complement household income directly, these programs assist in incentiviz- ing the accumulation of human capital by making the transfers conditional on school attendance and health care checkups. This also forces govern- ments to supply the schools and clinics necessary to meet the increased demand for these services, thereby boosting human capital. Moreover, the parameters of monetary policy related to macroeconomic stability, such as inflation targets linked to interest rates, directly affect rela- tive prices in an economy and, thus, the income generating capacity and productive choices of households. For instance, prudent macropolicies have allowed countries in the region to control inflation rates and achieve lower, steadier inflation rates for more than a decade. This regional improvement in the ability to control inflation impacts the real return on household assets. High inflation erodes the purchasing power of household wages, which effectively lowers the real returns on human capital and other types of assets. Prudent fiscal and monetary policies that are conducive to sustain- able and acceptable trends in fiscal deficits and inflation are also important in mitigating potential external shocks such as fiscal and financial crises. Overall, fiscal policies have efficiency and equity implications in both the short and the long run that can differentially affect the bottom 40 and the top 60. In the short run, the net system of fiscal incentives can reinforce or offset market income gaps. In the long run, they can impact decisions related to asset accumulation and use—as in the case of labor force partici- pation or hiring decisions by firms—and may induce factor misallocations or affect the size distribution of firms. Second, fair and transparent institutions capable of delivering good- quality basic services may directly affect the decisions of individuals to accu- mulate assets. In particular, strong institutional capacity linked to the deliv- ery of good-quality services in education and health care can enhance the ability of poorer households to improve their accumulation of net assets. More and better health care services and employment systems are funda- mental for mitigating the risks that households face from health-related 42 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean shocks and the consequences of employment loss. Basic services such as running water, electricity, and sewerage are important contributors to human capital accumulation, particularly among the poor. Over the past decade, there has been a large expansion in access to education, water, improved sanitation, and health care services across the region. However, coverage remains uneven across and within countries and is positively cor- related with income levels. The services are often inadequate in quality and only weakly coordinated with other key services, which undermines the overall impact, especially among the bottom 40. Governance failures can also act as a barrier to progress in achiev- ing the twin goals by imposing constraints on economic growth and job creation. Institutions can promote the protection of property rights and thus improve the investment climate in an economy by boosting the avail- ability of well-paid employment opportunities and affecting the returns to factors of production. Strong regulatory entities are also crucial in oversee- ing private market behaviors, thereby minimizing the risk of financial and sectoral macroeconomic crises. On a global scale, weaknesses in institutions influence the competitiveness of an economy. Robust competition policy that reduces entry barriers for new firms to certain markets directly affects the relative prices faced by all households by reducing consumer prices. While weak institutions are not considered a binding constraint on growth, there is evidence that they play a negative role. Overall, since the bottom 40 has more limited options, such as a lack of practical access to high-quality services in the private market, governance failures can fundamentally con- strain the capacity of the bottom 40 to build their human capital assets or to take advantage of economic opportunities, which undermines shared prosperity. Third, directly linked to better connectivity and competition, well- functioning markets are central to any effort to reduce the barriers to a more efficient utilization of household productive assets and can help grow the relative returns to assets. Enhanced transportation infrastructure that allows disadvantaged groups to connect to markets is an example of an opportunity to raise the utilization of assets that can create additional income. Poor-quality infrastructure adds to the negative effects of distance between regions and limits the connection of local markets with national and global markets. Infrastructure deficits can also have a negative impact on the investment climate and can compromise the ability of an economy to expand to its full potential. An extensive and well-functioning transporta- tion and communications infrastructure network is a necessary condition for access from poorer areas to major markets and services. Inequalities in coverage across regions limit the returns to other development initia- tives such as investments in education, health care, and social programs. Noncompetitive business environments and poor-quality infrastructure limit productivity growth, the labor demand that creates good jobs, and, therefore, the ability of the labor market to translate economic growth into higher incomes among the bottom 40. Chapter 1: Overview 43 Access to financial markets is also important for the income generating capacity of the poor in at least three ways. First, access to savings accounts and investment opportunities allows individuals to employ financial assets (such as savings) to obtain returns (interest rates) and thus complement labor incomes. Second, by encouraging and facilitating savings, access helps mitigate the impact of shocks and therefore protect against risks. If they have access to savings, the poor no longer need to sell assets or underinvest in human capital (by pulling children out of school) if an unexpected crisis strikes. Third, access to financial institutions that include access to credit allows individuals to finance small businesses or invest in fertilizer, physical assets, or human capital and thereby improve the level and intensity of their use of human capital and physical assets. Finally, adequate risk management can reduce the exposure to and impact of shocks among all households in an economy, but particularly the poor and vulnerable, who usually have a higher probability of risk and are thus forced to engage in negative coping mechanisms. Public safety nets such as public cash transfer schemes that are flexible so they may be scaled up during crisis and scaled down during recovery may be important instru- ments for supplying temporary income support to households affected by external shocks. Final Remarks Latin America and the Caribbean has experienced remarkable absolute and relative gains in achieving the twin goals. Moderate growth, combined with falling inequality, has propelled reductions in poverty and income growth among the bottom 40. Between 2002 and 2011, extreme poverty ($2.50 a day per capita) was cut in half, and higher incomes changed the demo- graphic composition of the bottom 40. In 2003, everyone in the bottom 40 was poor, and almost two-thirds of the bottom 40 were among the extreme poor, but, by 2012, only two-thirds of the bottom 40 were poor, and only 30 percent were among the extreme poor, while the largest group were the vulnerable (at 37.2 percent). These trends are reflected in higher house- hold incomes, mainly from higher wages. Greater human capital accumula- tion, economic growth, and falling inflation rates have been major factors behind the higher real-wage levels. Private and public transfers contributed almost 20 percent to the reduction in poverty. Some projections, drawing on the promising trends of the last decade in the region, estimate the share of households that will be living in extreme poverty ($1.25 a day) at 3.1 percent in 2030, down from 4.6 percent in 2011 (World Bank 2015b). Despite this impressive performance, extreme poverty is still a salient issue in middle- and low-income countries in the region. More than 75 mil- lion people are still living in extreme poverty in the region, half of them in Brazil and Mexico, and extreme poverty rates (based on the $2.50-a-day per capita line) are above 40 percent in Guatemala and reach nearly 60 44 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean percent in Haiti. Moreover, combining the share of the poor and vulnerable in the region suggests that nearly two-thirds of the region’s population is either poor or vulnerable to the risk of falling back into poverty. As growth wanes and progress in reducing the region’s high levels of inequality slows, it will be more important than ever for governments to focus policies on inclusive growth. For example, understanding the drivers behind the fall- ing labor force participation rates among the bottom 40 will be critical to ensuring the inclusiveness of growth, especially in a lower growth context that could limit labor market returns. Focusing on expanding the assets and market participation of indigenous households will also be crucial in clos- ing the gaps between the bottom 40 and the top 60. In addition, the specter of climate change suggests that severe weather events may become more frequent, which, the evidence indicates, are likely to affect the poor and the vulnerable more than the middle class. The goal of this chapter is to provide a baseline description on the stand- ing of the region in the effort to achieve the twin goals and information on a framework that can contribute to a better understanding of the compo- nents of the income of households that are directly linked to the monetary elements of the twin goals. The country studies presented in the rest of this book provide a more detailed discussion of recent trends, policy areas, and challenges related to the income generating capacity of the less well off. The presentation in each chapter is organized around four important pillars that are linked directly to the asset-based framework: (1) equitable, efficient, and sustainable fiscal policy and macroeconomic stability (direct and indirect taxes and transfers, inflation targets); (2) fair and transparent institutions capable of delivering universal, good-quality basic services (a greater and better supply of public goods, protection of property rights); (3) well-functioning markets (improved connectivity to markets, competi- tion policy); and (4) adequate risk management at the macro and household levels (macroprudence, safety nets). This comprehensive framework can be useful in approximating the diversity of results in poverty and shared pros- perity observed over the past decade and in helping to identify the chal- lenges ahead in the effort to reduce poverty and boost shared prosperity. Notes 1. The extreme poverty rate is measured by the number of people whose income or consumption falls below an international poverty line of $1.25 a day in 2005 purchasing power parity (PPP) international U.S. dollars, a poverty line that corresponds to an average of the national poverty lines of the 15 poorest developing countries. Because the aim is to end chronic poverty and because frictional poverty—poverty stemming from unexpected economic fluctuations in poor countries, political conflict, and war—cannot be brought to an end yet, the first goal is formalized as a target of bringing the number of people living below the $1.25-a-day poverty line to less than 3 percent of the world’s popula- tion (Basu 2013). Chapter 1: Overview 45 The second goal, boosting shared prosperity, places explicit attention on the least well off in a society by focusing on fostering the well-being of the bottom 40 in every country. Specifically, progress toward reaching the goal is assessed by measuring income or consumption growth among the bottom 40 in each country. According to the World Bank (2015a, 10): One way to think about the  .  .  . shared prosperity goal is as an alterna- tive to average income as the benchmark of development progress. Instead of assessing and measuring economic development in terms of the overall average growth in a country, the shared prosperity goal places emphasis on the bottom 40 percent of the population. In other words, good progress is judged to occur not merely when an economy is growing, but, more specifi- cally, when that growth is reaching the least well off in society. Although the shared prosperity indicator (SPI) focuses attention on the poor- est segments of a country’s population, it does not completely ignore the other segments. People above the bottom 40 may fall back into poverty if growth occurs only among the bottom 40 (Basu 2013). 2. In the region, most countries measure poverty using an income-based aggre- gate; this implies that it will always be reasonable to expect a positive extreme poverty rate because of frictional factors such as unemployment. For more details about the projections, see Ravallion (2003) and World Bank (2015a). 3. The poverty rate in Haiti is calculated using consumption instead of income as in the case of all other countries in the region for which data are available and harmonized. In the Latin America region, given the level of economic develop- ment, analysts use poverty lines that are higher than the global $1.25-a-day line. A $2.50-a-day extreme poverty line (an average of national extreme pov- erty lines) is considered more appropriate for the region. 4. According to recent World Bank studies (2013a, 2014a), the growth of gross domestic product (GDP) in the region declined from about 6.0 percent in 2010 to an estimated 2.5 percent in 2013, while the Gini coefficient was stagnant between 2010 and 2012. 5. See SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad Nacional de La Plata, La Plata, Argentina and World Bank, Washington, DC, http://sedlac .econo.unlp.edu.ar/eng/statistics.php. 6. These countries are Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Paraguay, Peru, and Uruguay. Regional poverty rates are population-weighted averages of country-specific poverty rates using interna- tional poverty lines. Whenever possible, annual household surveys from 2003 to 2012 have been used to estimate annual poverty rates. However, many countries do not conduct such surveys. To overcome this limitation, regional poverty rates have been estimated by generating artificial surveys using mac- roeconomic information on private consumption growth rates from the WDI (World Development Indicators) (database), World Bank, Washington, DC, http://data.worldbank.org/data-catalog/world-development-indicators. 7. The World Bank measures poverty rates according to the number of people whose income or consumption falls below a given threshold. To estimate the number of people living in extreme poverty, it currently uses an international poverty line of $1.25 a day in 2005 PPP international U.S. dollars, a poverty line that corresponds to an average of the national poverty lines of the 15 46 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean poorest developing countries. In Latin America and the Caribbean, given the level of economic development, analysts use poverty lines that are higher than the global $1.25 a day. A $2.50-a-day extreme poverty line (an average of national extreme poverty lines) and a $4.00-a-day total poverty line are more appropriate for the region. 8. In Latin America and the Caribbean, poverty is measured using income, while, in other regions, the World Bank uses consumption as the welfare aggregate. Consumption is typically assumed to be a better measure of current living standards given that it does not fluctuate as much as income. Consumption is usually more readily measured than income in countries with more informal labor markets. Relative to consumption measures, income measures usually imply that a larger share of households report zeroes and are thus classified as extreme poor. 9. The recent World Bank regional flagship report, Economic Mobility and the Rise of the Latin American Middle Class (Ferreira et al. 2013), characterizes the middle class based on the concept of economic security. A defining feature of membership in the group is household economic stability, which implies a low probability of falling back into poverty. The study defines a household as vulnerable if it faces more than a 10 percent likelihood of falling back into poverty over a five-year interval, which, surveys show, is approximately the average probability in countries such as Argentina, Colombia, and Costa Rica. This yields an income threshold of about $10 a day per capita (2005 PPP inter- national U.S. dollars) for the middle class. The report defines three economic classes: (a) the poor (people who have a per capita income below $4 a day), (b) the vulnerable ($4–$10 a day); and (c) the middle class ($10–$50 a day), all in 2005 PPP international U.S. dollars. The remainder, people with more than $50 a day in income, makes up less than 3 percent of the region’s population. 10. Nonetheless, the crisis had a significant negative effect on economic growth and income inequality in the Caribbean, Central America, and Mexico. In Cen- tral America and Mexico, labor market incomes and remittances dropped as a direct consequence of the recession in the United States, whereas the Caribbean countries suffered losses in incomes because of a decline in tourism and the higher prices of food imports. 11. The precise growth rates for the decades are sensitive to how the decades are defined. If the year 1990 (2000) is picked as the starting point rather than 1991 (2001), the respective growth rates for the two decades are 2.75 and 2.99 percent. 12. Calculations based on data in SEDLAC. 13. The decline has been documented in several studies using alternative sources of data, time periods, and income and inequality measures (see de la Torre et al. 2014; Gasparini et al. 2008; López-Calva and Lustig 2010; Lustig, López- Calva, and Ortiz-Juárez 2013). 14. Ravallion (2012) constructs a poverty measurement framework that is con- sistent with the utility theory and can capture the multidimensional aspect of poverty. 15. The proposed asset-based conceptual framework has been supported by aca- demic research and has also been extensively applied in other studies that have analyzed the determinants of progress in poverty reduction and shared prosper- ity around the world (for example, see Attanasio and Székely 2001; Carter and Chapter 1: Overview 47 Barrett 2006; Székely and Montes 2006; World Bank 2014a). For a more for- mal presentation of the framework, see López-Calva and Rodríguez-Castelán (2014). 16. The framework represents private transfers as independent of household income-earning assets, but these, particularly international remittances, may be correlated with access to markets and the probability that households will migrate. 17. Studies that discuss the role of aspirations in household decision making include Diecidue and Van De Ven (2008), Mookherjee, Ray, and Napel (2010), and Ray (2006). 18. Because the distribution of the wages of the top 60 is likely skewed to the right by the top earners, while the bottom 40 is truncated, the average wage may be misleading. So, we use the median wage. The trends hold for average wage as well, although the gaps are larger because the average wage among the top 60 is higher than the median wage. 19. See “Topics in Development: Migration, Remittances, and Diaspora,” World Bank, Washington, DC, http://go.worldbank.org/0IK1E5K7U0. 20. The net effect of changes in food prices needs to be further investigated in light of the fact that poorer households are also more likely to be food producers. For instance, Cuesta et al. (2010) study the distributive repercussions of the 2008 food price crisis in the Andean countries and find substantive poverty impacts ranging from 2 to 6 percentage points, although these results are sensi- tive to the net consumer (or producer) position of the households. 21. The estimate includes health costs (actual and loss of productivity), the costs of security and judicial procedures in the public sector and among households and firms, and the associated material costs (property loss). 22. Although the asset-based framework and its interaction with policy variables is presented statically, it is important to recognize that the interaction between policies and the elements that define the income generation capacity of house- holds is dynamic. Moreover, government interventions that are associated with these policy areas and that are implemented today may have positive effects on the accumulation and use of assets, the returns to assets, consumer prices, and risk management in the future. References Acevedo, Carlos. 2008. “Los costos económicos de la violencia en Centroamérica.” National Council of Public Security, Presidency of the Republic, San Salvador, El Salvador. Arias, J., and G. Esquivel. 2012. “A Note on the Side Effects of the War on Drugs: Labor Market Outcomes in México.” Unpublished working paper. Attanasio, Orazio, and Miguel Székely, eds. 2001. Portrait of the Poor: An Assets- Based Approach”, Washington, DC: Inter-American Development Bank; Balti- more: Johns Hopkins University Press. Azevedo, João Pedro, Viviane Sanfelice, and Minh Cong Nguyen. 2012. “Shapley Decomposition by Components of a Welfare Measure.” Unpublished working paper, World Bank, Washington, DC. 48 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Báez, Javier E., Alejandro de la Fuente, and Indhira Santos. 2010. “Do Natural Disasters Affect Human Capital? An Assessment Based on Existing Empirical Evidence.” IZA Discussion Paper 5164, Institute for the Study of Labor, Bonn, Germany. Barro, Robert J. 2006. “On the Welfare Costs of Consumption Uncertainty.” Schol- arly Articles 3224745, Department of Economics, Harvard University, Cam- bridge, MA. Basu, Kaushik. 2013. “Shared Prosperity and the Mitigation of Poverty: In Practice and Precept.” Policy Research Working Paper 6700, World Bank, Washington, DC. Becker, Gary S. 1968. “Crime and Punishment: An Economic Approach.” Journal of Political Economy 76 (2): 169–217. Bruns, Barbara, and Javier Luque. 2015. Great Teachers: How to Raise Student Learning in Latin America and the Caribbean. With Soledad De Gregorio, David Evans, Marco Fernández, Martin Moreno, Jessica Rodriguez, Guillermo Toral, and Noah Yarrow. Latin American Development Forum Series. Washing- ton, DC: World Bank. Bussolo, Maurizio, and Luis F. López-Calva. 2014. Shared Prosperity: Paving the Way in Europe and Central Asia. Europe and Central Asia Studies. Washington, DC: World Bank. Cárdenas, Mauricio, Rafeal E. de Hoyos, and Miguel Székely. 2014. “Out of School and Out of Work Youth in Latin America: A Persistent Problem in a Decade of Prosperity.” Working paper, Brookings Institution, Washington, DC. Cárdenas, Mauricio, and Sandra Rozo. 2008. “Does Crime Lower Growth: Evi- dence from Colombia.” Commission on Growth and Development Working Paper 30, World Bank, Washington, DC. Carter, Michael R., and Christopher B. Barrett. 2006. “The Economics of Poverty Traps and Persistent Poverty: An Asset-Based Approach.” Journal of Develop- ment Studies 42 (2): 178–99. Carter, Michael R., Peter D. Little, Tewodaj Mogues, and Workneh Negatu. 2007. “Poverty Traps and the Long-Term Consequences of Natural Disasters in Ethio- pia and Honduras.” World Development 35 (5): 835–56. Caudillo, Mónica L., and Florencia Torche. 2014. “Exposure to Local Homicides and Early Educational Achievement in Mexico.” Sociology of Education 87 (2): 89–105. Cavallo, Eduardo, and Ilan Noy. 2011. “Natural Disasters and the Economy: A Survey.” International Review of Environmental and Resource Economics 5 (1): 63–102. Cord, Louise, Oscar Barriga Cabanillas, Leonardo Lucchetti, Carlos Rodríguez- Castelán, Liliana D. Sousa, and Daniel Valderrama. 2014. “Inequality Stagna- tion in Latin America in the Aftermath of the Global Financial Crisis.” Policy Research Working Paper 7146, World Bank, Washington, DC. Cuesta, José, Suzanne Duryea, Fidel Jaramillo, and Marcos Robles. 2010. “Distrib- utive Impacts of the Food Crisis in the Andean Region.” Journal of International Development 22 (7): 846–65. Datt, Gaurav, and Martin Ravallion. 1992. “Growth and Redistribution Com- ponents of Changes in Poverty Measures: A Decomposition with Applications Chapter 1: Overview 49 to Brazil and India in the 1980s.” Journal of Development Economics 38 (2): 275–95. de la Torre, Augusto, Eduardo Levy Yeyati, Guillermo Beylis, Tatiana Didier, Car- los Rodríguez-Castelán, and Sergio Schmukler. 2014. “Inequality in a Lower Growth Latin America.” Semiannual Report (October), Latin America and Caribbean Region, World Bank, Washington, DC. Dell, Melissa. 2014. “Trafficking Networks and the Mexican Drug War.” Working paper, Department of Economics, Harvard University, Cambridge, MA. Demombynes, Gabriel, and Aaron Thegeya. 2012. “Kenya’s Mobile Revolution and the Promise of Mobile Savings.” Policy Research Working Paper 5988, World Bank, Washington, DC. Dercon, Stefan, and Luc Christiaensen. 2011. “Consumption Risk, Technology Adoption, and Poverty Traps: Evidence from Ethiopia.” Journal of Develop- ment Economics 96 (2): 159–73. Diecidue, Enrico, and Jeroen Van De Ven. 2008. “Aspiration Level, Probability of Success and Failure, and Expected Utility.” International Economic Review 49 (2): 683–700. Dinkelman, Taryn. 2011. “The Effects of Rural Electrification on Employment: New Evidence from South Africa.” American Economic Review 101 (7): 3078–3108. Elbers, Chris, Jean O. Lanjouw, and Peter Lanjouw. 2003. “Micro-Level Estima- tion of Poverty and Inequality.” Econometrica 71 (1): 355–64. Enamorado, Ted, Luis F. López-Calva, and Carlos Rodríguez-Castelán. 2014. “Crime and Growth Convergence: Evidence from Mexico.” Economics Letters 125 (1): 9–13. Enamorado, Ted, Luis F. López-Calva, Carlos Rodríguez-Castelán, and Hernán Winkler. 2014. “Income Inequality and Violent Crime: Evidence from Mexico’s Drug War.” Policy Research Working Paper 6935, World Bank, Washington, DC. Fajnzylber, Pablo, and J. Humberto López, eds. 2008. Remittances and Develop- ment: Lessons from Latin America. Latin American Development Forum Series. Washington, DC: World Bank. Ferreira, Francisco. H. G., Julian Messina, Jamele Rigolini, Luis F. López-Calva, María Ana Lugo, and Renos Vakis. 2013. Economic Mobility and the Rise of the Latin American Middle Class. Washington, DC: World Bank. Gasparini, Leonardo, Guillermo Cruces, Leopoldo Tornarolli, and Mariana Mar- chionni. 2008. “A Turning Point? Recent Developments on Inequality in Latin America and the Caribbean.” Working Paper 81, Center for Distributive, Labor, and Social Studies, Facultad de Ciencias Económicas, Universidad Nacional de La Plata, La Plata, Argentina. Gasparini, Leonardo, Sebastián Galiani, Guillermo Cruces, and Pablo Acosta. 2011. “Educational Upgrading and Returns to Skills in Latin America: Evidence from a Supply-Demand Framework, 1990–2010.” Working Paper 127, Center for Distributive, Labor, and Social Studies, Facultad de Ciencias Económicas, Universidad Nacional de La Plata, La Plata, Argentina. Hanushek, Eric A., and Ludger Woessmann. 2012. “Schooling, Educational Achievement, and the Latin American Growth Puzzle.” Journal of Development Economics 99 (2): 497–512. 50 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Keefer, Philip, and Norman V. Loayza. 2010. Innocent Bystanders: Developing Countries and the War on Drugs. Washington, DC: World Bank. Lagos, Marta, and Lucía Dammert. 2012. “La Seguridad Ciudadana: El problema principal de América Latina.” Corporación Latinobarómetro, Santiago, Chile. Levy, Santiago. 2008. Good Intentions, Bad Outcomes: Social Policy, Informality, and Economic Growth in Mexico. Washington, DC: Brookings Institution Press. López-Calva, Luis F., and Nora Lustig. 2010. “Explaining the Decline in Inequality in Latin America: Technological Change, Educational Upgrading, and Democ- racy.” In Declining Inequality in Latin America: A Decade of Progress?, edited by Luis F. López-Calva and Nora Lustig, 1–24. New York: United Nations Development Programme; Baltimore: Brookings Institution Press. López-Calva, Luis F., and Carlos Rodríguez-Castelán. 2014. “Pro-Growth Equity: A Policy Framework for the Twin Goals.” World Bank, Washington, DC. Lustig, Nora, Luis F. López-Calva, and Eduardo Ortiz-Juárez. 2013. “Declining Inequality in Latin America in the 2000s: The Cases of Argentina, Brazil, and Mexico.” World Development 44: 129–41. Maldonado, René, Natasha Bajuk, and María Luisa Hayem. 2012. “Remittances to Latin America and the Caribbean in 2011: Regaining Growth.” Multilateral Investment Fund, Inter-American Development Bank, Washington, DC. Michaelsen, Maren M., and Paola Salardi. 2013. “School’s Out: The War on Drugs and Educational Performance in Mexico.” Working paper, Department of Eco- nomics, Ruhr University Bochum, Bochum, Germany. Mookherjee, Dilip, Debraj Ray, and Stefan Napel. 2010. “Aspirations, Segregation, and Occupational Choice.” Journal of the European Economic Association 8 (1): 139–68. Narayan, Ambar, Jaime Saavedra-Chanduvi, and Sailesh Tiwari. 2013. “Shared Prosperity: Links to Growth, Inequality, and Inequality of Opportunity.” Policy Research Working Paper 6649, World Bank, Washington, DC. Powell, Benjamin, G. P. Manish, and Malavika Nair. 2010. “Corruption, Crime, and Economic Growth.” In Handbook on the Economics of Crime, edited by Bruce L. Benson and Paul R. Zimmerman, 328–41. Cheltenham, United King- dom: Edward Elgar. Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press. Ravallion, Martin. 2003. “Measuring Aggregate Welfare in Developing Countries: How Well Do National Accounts and Surveys Agree?” Review of Economics and Statistics 85 (3): 645–52. ———. 2012. “Poor, or Just Feeling Poor? On Using Subjective Data in Measuring Poverty.” Policy Research Working Paper 5968, World Bank, Washington, DC. Ray, Debraj. 2006. “Aspirations, Poverty, and Economic Change.” In Understand- ing Poverty, edited by Abhijit Vinayak Banerjee, Roland Bénabou, and Dilip Mookherjee, 409–22. New York: Oxford University Press. SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad de La Plata, La Plata, Argentina; World Bank, Washington, DC. http://sedlac.econo.unlp.edu.ar/eng/ index.php. Chapter 1: Overview 51 SNSP (Mexico, Secretariado Ejecutivo de Sistema Nacional de Seguridad Pública). 2012. “Estadísticas y Herramientas de Análisis de Información de la Incidencia Delictiva (Fuero Común, Fuero Federal, 1997–actual).” SNSP, Mexico City. Székely, Miguel, and Andrés Montes. 2006. “Poverty and Inequality.” In The Long Twentieth Century, edited by Victor Bulmer-Thomas, John H. Coatsworth, and Roberto Cortés-Conde, 585–646. The Cambridge Economic History of Latin America, vol. 2. Cambridge: Cambridge University Press. UNDP (United Nations Development Programme). 2013. Informe Regional de Desarrollo Humano 2013–2014, Seguridad Ciudadana con Rostro Humano: diagnóstico y propuestas para América Latina. New York: UNDP. Velásquez, Andrea. 2014. “The Economic Burden of Crime: Evidence from Mexico.” Job Market Paper, Department of Economics, Duke University, Dur- ham, NC. World Bank. 2011a. “Food Price Watch.” February, World Bank, Washington, DC. ———. 2011b. “Crime and Violence in Central America: A Development Chal- lenge.” Poverty Reduction and Economic Management Unit, Sustainable Devel- opment Department, Latin America and the Caribbean Region, World Bank, Washington, DC. ———. 2012a. “The Effect of Women’s Economic Power in Latin America and the Caribbean.” Poverty and Labor Brief (August), Latin America and Caribbean Region, World Bank, Washington, DC. ———. 2012b. “Costs and Impacts of Crime and Violence in Mexico.” Brief, World Bank, Washington, DC. ———. 2013a. “Shifting Gears to Accelerate Shared Prosperity in Latin America and the Caribbean.” Poverty and Labor Brief, Report 78507 (June), Latin Amer- ica and Caribbean Region, World Bank, Washington, DC. ———. 2013b. World Development Report 2014: Risk and Opportunity, Manag- ing Risk for Development. Washington, DC: World Bank. ———. 2014a. “Social Gains in the Balance: A Fiscal Policy Challenge for Latin America and the Caribbean.” Poverty and Labor Brief, Report 85162 rev (Feb- ruary), Latin America and Caribbean Region, World Bank, Washington, DC. ———. 2014b. “When Prosperity Is Not Shared: The Weak Links between Growth and Equity in the Dominican Republic.” Report 85760, World Bank, Washing- ton, DC. ———. 2014c. “An Assessment of the Poverty Methodology in Jamaica: A Consoli- dated Technical Note.” World Bank, Washington, DC. ———. 2014d. “Avoiding Crime in Latin America and the Caribbean.” Enterprise Surveys, Latin America and the Caribbean Series Note 7, World Bank, Wash- ington, DC. ———. 2015a. A Measured Approach to Ending Poverty and Boosting Shared Prosperity: Concepts, Data, and the Twin Goals. Policy Research Report. Wash- ington, DC: World Bank. ———. 2015b. Global Monitoring Report 2014/2015: Ending Poverty and Sharing Prosperity. Washington, DC: World Bank. ———. 2015c. Global Economic Prospects, January 2015: Having Fiscal Space and Using It. Washington, DC: World Bank. 52 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean ———. Forthcoming. “The Aftermath of the 2008 Global Financial Crisis in the Eastern Caribbean: The Impact on the St. Lucian Labor Market.” Report, World Bank, Washington, DC. World Bank and ONPES (Haiti, National Observatory of Poverty and Social Exclu- sion). 2014. “Investing in People to Fight Poverty in Haiti, Overview: Reflections for Evidence-Based Policy Making.” World Bank, Washington, DC. Zapata Martí, Ricardo. 2005. “The 2004 Hurricanes in the Caribbean and the Tsunami in the Indian Ocean: Lessons and Policy Challenges for Development and Disaster Reduction.” Estudios y Perspectivas 35, United Nations Economic Commission for Latin America and the Caribbean, Mexico City. CHAPTER 2 Shared Prosperity and Poverty Reduction in Urban Argentina Santiago Garriga, Emmanuel Skoufias, and Liliana D. Sousa Introduction A rgentina rebounded following the severe crisis of 2001–02. The pov- erty rate fell sharply, from 31.0 percent living on less than $4.00 a day in 2004 to 10.8 percent in 2012; inequality narrowed; and incomes among the bottom 40.0 percent of the income distribution in the population (the bottom 40) expanded appreciably. As of 2011, more than half the popu- lation could be counted among the middle class, and, by 2012, the share of the population living on less than $2.50 a day was below 5 percent. Measured according to the Gini coefficient, income inequality was at 0.43, substantially lower than the 0.52 in the Latin America and Caribbean region. While Argentina has a significant social safety net, the impressive gains of the past decade have been largely driven by improved labor market outcomes. Greater labor earnings and a higher level of employment explain nearly 75 percent of the drop in poverty between 2004 and 2012. These gains were mainly generated by increases in earnings among men, particu- larly among the low skilled, and enhancements in the quality of jobs: the informality rate among wage earners fell from 58 percent in 2004 to 46 percent in 2012. Argentina has had a strong recovery, developed a considerable social safety net, and made meaningful progress in poverty reduction, but inequal- ity is still evident in the outcomes between men and women in the labor market, across the regions of the country, and in the access of children to essential goods and services, especially sanitation and good-quality edu- cation. Large dividends have been achieved through greater earnings and higher employment levels, but labor market outcomes among women lag the 53 54 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean corresponding outcomes among men, and women are experiencing higher unemployment rates. Signs of strain are also visible in the nation’s broaden- ing social safety net. For example, even while more households are receiving benefits, the neediest recipients of public transfers are still living in extreme poverty. There are also telling reasons to question the sustainability of the accomplishments in the face of deteriorating macroeconomic conditions. Changes in policy are needed to protect the advances that have been realized. Thus, for instance, additional investment is essential to boost the quality of schooling, which has important long-term implications for equity and growth. Meanwhile, weaknesses in innovation and competition may be limiting market resiliency and development, particularly in the labor market and the credit market. The fiscal health of the nation is paramount in pro- tecting the population from dramatic declines in welfare such as those expe- rienced during the crisis of 2001–02. Yet, expenditure growth has outpaced the growth of revenue over the past decade so that the commodity-fueled surpluses of the postcrisis period have now become deficits. Addressing this crucial issue will require revisiting many of the public spending choices of the past decade. Background Between 2004 and 2012, Argentina underwent a period of strong and inclusive growth, yielding substantial declines in poverty (reducing urban poverty from 31.0 to 10.8 percent) and a notable narrowing in inequality. These breakthroughs came on the heels of the country’s powerful macro- economic crisis of 2001–02, which resulted in a reduction of welfare on the order of 25 percent of gross domestic product (GDP) and led to a one- year increase of 56 percent in extreme poverty and 34 percent in moderate poverty (Sandleris and Wright 2014) (figure 2.1).1 Subsequent growth and policy changes resulted in a recovery not only from the crisis, but also from the rise in inequality of the 1990s (Gasparini and Cruces 2009). Following three years of economic recession, the 2001–02 crisis led to a sovereign default, a severe currency devaluation, and political instabil- ity. It generated serious job destruction and falling real wages; more than 6 in 10 households suffered from a fall in real income of more than 20 percent (McKenzie 2004). Though unemployment expanded widely dur- ing this period, nearly three-quarters of the reduction in household labor income was caused by declining real wages rather than fewer earners in households (McKenzie 2004). The crisis impacted people at the bottom of the income distribution disproportionately. Private sector workers with less educational attainment, whether wage workers or the self-employed, were the most vulnerable to job loss (Corbacho, Garcia-Escribano, and Inchauste 2007). Evidence suggests that the crisis caused food insecurity among people who had not completed secondary schooling (Bozzoli and Quintana-Domeque 2014). Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 55 Figure 2.1 Poverty Rates and the Share of Income Held by the Bottom 40, Argentina, 1991–2012 50 16 45 14 40 12 Income share (%) 35 Poverty rate (%) 30 10 25 8 20 6 15 4 10 2 5 0 0 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Extreme poor ($2.50 a day) Moderate poor ($4.00 a day) Share of income of the bottom 40 (right axis) Source: Calculations based on data in Socio-Economic Database for Latin America and the Caribbean (SEDLAC). However, beginning in 2003 and largely because of favorable macroeco- nomic conditions, including high commodity prices and a weaker peso, GDP rebounded and grew at an annualized rate of 7.9 percent between 2003 and 2006.2 While the recovery was largely spurred by economic growth and poverty reduction arose mainly from improvements in the labor market, the postcrisis period has also been characterized by strengthened labor institu- tions and the implementation of more redistributive policies (Gasparini and Cruces 2009). The minimum wage was raised multiple times beginning in July 2003 and surpassed the precrisis value in September 2004 (Khamis 2013). Public transfers were also augmented, notably through the Jefes y Jefas de Hogar Program (a public transfer program introduced in 2002 and aimed at unemployed household heads), the universal child allowance, and the pension moratorium. These policy changes were accompanied by a jump in public spending: total government spending rose from 30 percent of GDP in 2003 to 43 percent by 2009, and social spending accounted for half the government spending (Lustig and Pessino 2014). This expansion in spending was largely financed through tax collection, which increased by 10 percentage points of GDP between 2003 and 2009, mainly from three sources: a tax on financial transactions, taxes on primary exports, and employee contributions to the social security system (Lustig and Pessino 2014). 56 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean As commodity prices have fallen and economic growth has slowed in the region, the poverty gains in Argentina have diminished and now arise more from changes in income distribution than from income growth. While Argentina’s poverty and shared prosperity indicators remain strong relative to the regional average, it is unclear whether the progress can be continued or even preserved over the medium term. Diagnostics Argentina has witnessed large advances in shared prosperity over the past decade. The urban poverty rate dropped from 31.0 percent in 2004 to 10.8 percent in 2012, and the middle class—people with incomes of between $10 and $50 a day—had expanded to more than half the urban popula- tion by 2011.3 The decline in poverty and the growth of the middle class reflected a strong recovery from the severe 2001–02 crisis and the inclusive social policies enacted over the past decade. However, despite the achieve- ments, 10.8 percent of the urban population is still living in poverty (less than $4 a day), and another 33.0 percent is vulnerable to the risk of falling back into poverty in the event of an adverse shock because they are living on only $4 to $10 a day (López-Calva and Ortiz-Juárez 2011) (figure 2.2). While monetary poverty rates are only available for the 60 percent of the population that lives in larger metropolitan areas, nonmonetary indicators Figure 2.2 Poverty Headcounts, Urban Areas, Argentina, 2004–12 60 50 40 Percent 30 14.0 20 12.5 10.3 10.8 17.0 9.0 8.3 10 13.3 7.9 10.3 6.9 6.2 8.8 8.2 8.0 6.1 4.6 4.7 0 2004 2005 2006 2007 2008 2009 2010 2011 2012 Year Moderate poor ($4.00 a day) Extreme poor ($2.50 a day) Vulnerable ($4.00–$10.00 a day) Middle class ($10.00–$50.00 a day) Source: Based on data in SEDLAC. Note: Poverty lines are represented in 2005 purchasing power parity (PPP) U.S. dollars. Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 57 Figure 2.3 Annualized Income Growth Ratio, the Bottom 40, Urban Argentina vs. Region, 2003–12 2.5 Ratio of income growth, bottom 40 to overall income growth 2.0 1.5 1.0 0.5 0 2003–12 2003–08 2008–12 Years Region Argentina Source: Based on data in SEDLAC. Note: The growth of per capita household income is calculated in 2005 purchasing power parity (PPP) U.S. dollars. Because of data limitations, the 2003 values for Argentina are based on 2004 data. from the population census suggest that poverty rates are higher outside the larger urban areas.4 Among the population, 12 percent had at least one unsatisfied basic need in 2010, including over 20 percent of the residents of Chaco, Corrientes, Formosa, Salta, and Santiago del Estero, provinces located in regions in the northwest and northeast of the country. The cor- responding shares are also particularly high among people living in small towns (with less than 2,000 inhabitants) or rural areas, which, together, accounted for 11 percent of the population in 2001. While data are unavail- able for 2010, the 2001 census showed that 36 percent and 24 percent of rural residents and inhabitants of small towns, respectively, had at least one unsatisfied basic need, far greater than the 16 percent of the rest of the population that had at least one unsatisfied basic need (World Bank 2010). Argentina’s strong economic growth since the 2001–02 crisis has been more propoor than the average growth in the Latin America and Caribbean region. Relative to the mean income, the income of the bottom 40 in urban areas in Argentina grew more quickly than the income of the bottom 40 in the region (1.7 times more quickly in Argentina versus 1.5 times in the region) (figure 2.3). Even within the bottom 40, income growth in Argen- tina since 2004 has substantially and consistently favored the poorest: the annualized growth rates were more than twice as high among the bottom decile than among the top decile. The differential in growth rates among the bottom 40 between Argentina and the region was particularly pronounced 58 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 2.4 Trends in Inequality, Urban Argentina and the Region, 2004–12 a. Gini coefficient b. 75/25 Income share 0.56 3.9 3.7 0.55 3.7 Ratio of income share 0.52 0.52 Gini coefficient 0.50 3.5 3.6 3.3 3.3 0.48 3.1 0.44 2.9 3.0 0.43 2.7 0.40 2.5 04 05 06 07 08 09 10 11 12 04 05 06 07 08 09 10 11 12 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Year Region Argentina Source: Based on data in SEDLAC. in the second half of the period: relative to the respective growth rate in overall income, the income growth rate among the bottom 40 in Argentina was more than double (2.3 times the mean income growth) compared with the income growth rate among the bottom 40 in the region (1.6 times the mean income growth). As a result, the narrowing of inequality among the urban population in Argentina since 2004 has eclipsed the strong performance of the region (fig- ure 2.4). In 2012, the Gini coefficient in urban Argentina was 0.43, much lower than the 0.52 in the region. During the period, income inequality fell by 15 percent in urban Argentina, substantially greater than the 5 per- cent decline in the region. Similarly, large reductions in inequality are evi- dent in the ratio of average household income among the bottom quartile and among the top quartile of the population (the 75/25 income share). In 2004, the top 25 percent of the urban population in Argentina had an average income 3.7 times that of the bottom quartile, larger than the gap in the region. Since then, however, Argentina has strongly outperformed the region: in 2012, the average income of the top quartile was 3.0 times that of the bottom quartile, while the ratio was 3.3 in the region. While some of the gains since 2004 reflect a continuation of the recovery from the crisis, including adjustments associated with the process of unpeg- ging the peso from the dollar, and the expansion of the social safety net, the impressive performance in urban poverty reduction over the past decade has been largely driven by improved labor market outcomes (Gasparini and Cruces 2009). The depression in real wages following the crisis, combined with the strong economic recovery, led to more job creation; changes in the relative price of labor benefited unskilled labor-intensive industries, thus Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 59 Figure 2.5 Employment Profile, Argentina, 2004–12 a. Employment growth (2004 = 0) b. Share of employment, by type of employer 50 40 35.1 35 30.8 40 30 26.1 Index, 2004 = 0 Index, 2004 = 0 30 25.0 25 19.7 20 20 17.5 15 18.4 17.3 10 10 0 4.9 5.1 5 –10 0 2004 2006 2008 2010 2012 2004 2006 2008 2010 2012 Year Year Wage (large firm) Wage (small firm) Public sector Self-employment Other employment Source: Based on data in SEDLAC. Note: Data are based on the main employers of employed individuals between the ages of 16 and 65. Employers and unpaid workers are only reported in panel b; these two groups accounted for 5 percent of employment in 2004 and 2012. Small firms = firms with five or fewer employees. generating more unskilled jobs; and slower technological upgrading rela- tive to the surge in the adoption of new technologies in the 1990s (partially caused by the higher relative cost of imports and uncertainty because of the crisis and social unrest) led to expansion in labor-intensive industries (Gasparini and Cruces 2009). Between 2004 and 2012, improved labor outcomes (in both earnings and the level of employment) accounted for nearly 75 percent of the drop in total poverty; higher earnings alone explain 54 percent of the poverty reduction.5 The postcrisis period saw a sharp rise in the quantity and the quality of jobs. Overall, the number of employed adults was 18 percent higher in 2012 than in 2004. Employment expansion was accompanied by enhance- ment in the quality of jobs, particularly evident in the decline in the rate of informality. While employment grew across all firm types, the number of adults whose primary jobs were in firms with more than five workers each rose the most, increasing by 34 percent between 2004 and 2012 and accounting for 35 percent of all employed adults in 2012 (figure 2.5). The public sector also played a key role in job creation, especially after 2008, when there was a steep recovery in hiring in the sector following the cuts earlier in the decade. Because of the greater employment in large firms and in the public sector, the informality rate among wage earners, measured as the share of wage earners without the right to pensions or retirement 60 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 2.6 Average Monthly Earnings, by Gender and Educational Attainment, Argentina, 2004 and 2012 170 165 Index, earnings of men with 150 secondary schooling = 100 140 130 104 109 110 78 90 68 70 63 70 46 34 50 30 Primary Postsecondary Primary Secondary Postsecondary Men Women Educational attainment 2004 2012 Men with secondary schooling Source: Based on data in SEDLAC. Note: The figure reports average monthly earnings for all employed individuals aged 15 years or older. benefits, fell from 58 to 46 percent between 2004 and 2012. In Argentina, obtaining a formal job is three times more likely than obtaining informal employment to bring a family out of poverty (Beccaria et al. 2013). While earnings grew at all skill levels, the boost was particularly strong among low-skilled men because the earnings premium of education had fallen among men (figure 2.6). The earnings of men who had not completed secondary school rose to 78 percent of the earnings of men with second- ary schooling. Similarly, the monthly earnings of women across all skill- groups increased more quickly than the earnings of highly skilled men even as the gender hourly wage gap widened slightly between 2004 and 2012 for all women except those with tertiary education. The climb in earnings among the low skilled was associated with a rise in the minimum wage, which resulted in wage increases among low-skilled formal and informal workers, as well as changes in the sector of employment (Khamis 2013). Between 2004 and 2012, low-skilled labor shifted to construction, which augmented its share of low-skilled employment by 3 percentage points to reach 16 percent of employment among workers who had not completed secondary school, as well as to the hospitality sector, transportation, and private households.6 Since the crisis, the government has engaged in a considerable expansion of the social safety net. Three programs are worth highlighting: • The Jefes y Jefas de Hogar Program was a critical source of income for lower-income households during the crisis and in the early period of the recovery. Spending on this program was approximately 1 Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 61 percent of GDP in 2003, but fell appreciably as the unemployment rate declined (Lustig and Pessino 2014). • A leading source of poverty reduction since 2007 has been the widen- ing access to pensions, primarily through the pension moratorium instituted in the mid-2000s.7 This program significantly expanded access to pensions by introducing a mechanism by which pensioners who did not contribute the full 30 years to the national pension sys- tem could still receive a pension. It had provided pensions to approxi- mately 2.2 million beneficiaries by 2009 at an estimated cost of 2.4 percent of GDP (Lustig and Pessino 2014). • The coverage of conditional cash transfers was broadened to include the children of parents in the informal sector through the introduction of the universal child allowance program in 2009. The previous condi- tional cash transfer was limited to low-income formal sector workers, with the exception of the Jefes y Jefas program, which was a tempo- rary program more akin to unemployment insurance. In 2010, the transfer program cost 0.6 percent of GDP (Lustig and Pessino 2014). Social spending helped cut poverty largely through positive changes in pensions, especially after 2007, the year the pension moratorium was fully implemented. Despite their significance, cash transfer programs did not lead to additional poverty reduction between 2004 and 2012. While the coverage rate of public transfers expanded, there was a drop in the share of household income from transfers. Among the bottom quintile, for example, the share of household income from transfers fell from 23 percent in 2004 to 12 percent in 2012. However, some of the poverty reduction attribut- able to labor force increases may be partially attributable to public transfer programs. For example, the Jefes y Jefas de Hogar Program required that recipients engage in training or community service or work for a private company benefiting from an employment subsidy, potentially leading to better employment outcomes among low-income households.8 Not all groups have benefited equally in the gains in poverty reduction and shared prosperity. Outcomes in the north of the country and among rural residents, women, and children lag along some dimensions. Sub- stantial regional disparities persist. Thus, the northeast and the northwest trailed in several indicators of well-being. Although the extreme poverty rate among the urban population in the northeast was cut by three-quarters between 2004 and 2012 (from 34.1 to 7.7 percent), it was still higher than the national rate and double the rate in sparsely populated Patagonia, the region with the lowest extreme poverty rate (3.4 percent). At 5.6 percent, the extreme poverty rate in the northwest was also higher than the national aver- age. While half the bottom 40 among the urban population lives in Greater Buenos Aires, and another fifth in the Pampeana region, the most populous parts of the country, the majority of the urban population in the northeast and northwest are in the bottom 40. There are also disparities in health out- comes. In Jujuy Province, 165 deaths per 100,000 live births are attributed to maternal-related causes, while the rate is only 18 in Buenos Aires.9 62 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 2.7 Improved Sanitation: Disparities in the HOI and Coverage, by Location and Region, Argentina, 2012 100 90 Human opportunity index (diamonds = coverage) 80 70 60 50 40 30 Ca al a m lá án L Sa a – a R lta fi oja Re s jo Co sten as ie a r s an a sa G an Ce a n os ri G ran ta o n ar e M Gr C ór ná el Sa co ba ta R ia lla R – B sa n G ns ua n is n M tuc o ity Gr l C ndo n or Gr o an ho za va r en J o n qu Ra Ai s én da res a ío lo ly rm aw ío lleg r en son Gr os Pa Tre e ta lew s Ca R – R Ga ttie hí Gr Fo nte eu – s e ne n P d c rr ci r – t ra P F d G Sa ari Co ío C atá Lu ra ti rt E e ó Ri te u n ill da Bu os ua ra R r ta pa Pl nta rd ua R – P Til n Po Vie G ca Pla ar si ad Bl n L mo o N via no Air ra – an C a ar an o do de – an i Ta i o ea f B Sa rr go n G juy a B n Ju – L e a a o – er ra m a i st G d Vi cu lE Ba sh – Tu C de Sa ás U n ol go – ra ic a ia G m N od nt ed n Sa m Sa Vi Co Area Northwest Pampeana Greater Buenos Aires Coverage Northeast Cuyo Patagonia Source: Based on data in SEDLAC. Note: These areas are defined using aglomerados (metropolitan areas that may include more than one city). HOI = human opportunity index. Notwithstanding the public transfer programs targeting them, children are disproportionately poor. Almost one in five children (19.0 percent) under the age of 15 was living in poverty in 2012, nearly double the overall poverty rate of 10.9 percent. Although more than half the urban population is in the middle class, only 37 percent of children were living in middle-class households. Households are more likely to fall into poverty if they include children, and, once poor, such households are less likely to exit poverty (Beccaria et al. 2013). This means households with children suffer from longer spells of poverty. An analysis of the human opportunity index (HOI) and the relevant coverage rates indicates that inequalities persist in access to good-quality schooling and improved sanitation because of circumstances at birth, such as parental educational attainment and parental income.10 Disparity in access to improved sanitation is wide across the country; the northeast, the Pampeana region, and the northwest lag. Access to improved sanitation among urban children varies from a low of 41.5 percent in the Gran Santa Fe area to universal coverage in Río Gallegos in Patagonia (figure 2.7). Nonetheless, access to improved sanitation is more prevalent in urban areas. Based on 2010 census data that include rural households and households in Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 63 smaller towns, a starker picture emerges: only 18.6 percent of households in Misiones in the northeast and 21.9 percent of households in Santiago del Estero in the northwest had access to improved sanitation.11 Entries into poverty and exits out of poverty are largely determined by labor market events. Job loss is the primary driver of households falling into poverty (Beccaria et al. 2013). Though all groups have experienced declines in unemployment since 2004, women continue to exhibit higher unemploy- ment rates, including women heads of household. Unemployment rates plunged across the board, but women face higher levels of unemployment than men in all groups defined according to educational attainment except for the lowest skilled, that is, people who have not completed primary school. In 2012, women in the middle of the skills distribution—those who had completed primary school, but had not pursued postsecondary educa- tion—showed an unemployment rate of 12 percent, double the unemploy- ment rate of similarly educated men. Women with postsecondary education also had higher unemployment rates than similarly educated men: 7 versus 4 percent in 2012. This pattern holds among heads of household: women household heads with primary and secondary schooling had unemploy- ment rates of 8 and 6 percent, respectively, in 2012. Similarly educated men household heads had unemployment rates of only 3 percent. Female unem- ployment poses a serious challenge not only because it puts households at higher risk of poverty, but also because it has implications for the economic independence and agency of women. Policy Discussion While Argentina has experienced a strong economic recovery, developed a broad social safety net, and made notable progress in poverty reduction, inequality in outcomes is still evident in the labor market, across the coun- try’s regions, and in the access of children to improved sanitation and good- quality education. More significant still is the question of sustainability: to what extent are the gains sustainable in the face of deteriorating macro- economic conditions? Two key sources of income underlie the progress in poverty and shared prosperity witnessed between 2004 and 2012: improvements in labor mar- ket outcomes, especially among low-skilled labor, and the greater coverage of pensions and public transfer programs. This section takes a closer look at the policies and risks influencing each and assesses the prospects for con- tinued improvements in shared prosperity.12 Drivers of labor market outcomes: productivity growth and human capital Sustaining the reduction in poverty requires continued resilience and growth in the labor market. While meaningful advances in shared prosperity have been achieved through higher earnings and better employment levels, par- ticularly among the low skilled, the labor market outcomes of women lag those of men. The gender wage gap has not changed among people without 64 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean tertiary education, and women are facing higher unemployment rates. Additionally, 18 percent of youth between the ages of 15 and 24 years are neither working nor in school.13 Improving labor market outcomes requires short-term strategies, such as addressing higher unemployment among women and younger workers. It also requires long-term strategies, especially investing in productivity growth through increased human capi- tal and the creation of an environment of well-functioning labor and credit markets to feed productivity growth. Long-term labor market development depends on productivity growth, an area in which Argentina’s performance has historically lagged, but has recently shown much improvement. The growth in employment in larger firms since 2004 suggests that workers have been moving toward more pro- ductive activities. Indeed, beginning in the 1990s, total factor productivity rose and, in 2010, was higher in Argentina than in Brazil or Colombia and approximately the same as in Chile. Despite these dividends, total factor productivity in Argentina was only 60 percent that of the United States in 2010.14 A barrier to more productivity growth is the weak business climate. Business managers cite high tax rates (19.6 percent of respondents), poor access to finance (15.1 percent), excessive labor regulations (14.3 percent), and political instability (13.9 percent)—four areas directly influenced by the government—as the top obstacles impeding enterprise growth.15 Access to credit is a particular challenge faced by the private sector. While the volume of private sector credit represents 50 and 80 percent of GDP in Brazil and Chile, respectively, it is only 13 percent of GDP in Argentina, where it is lower now than it was before the 2001–02 crisis.16 A lack of access among firms to technology and financing and weak market competition, especially in nonexport sectors, has resulted in low investment in research and devel- opment in Argentina.17 The high labor productivity—$23,000 in value added per worker—is attributable to high capital intensity rather than efficiency improvements. The median firm uses $10,000 in capital (more than the corresponding average in Brazil, Chile, Mexico, or Uruguay).18 Adjusted for capital use and sector, firms in Chile and Uruguay show similar labor productivity lev- els.19 Meanwhile, in Argentina, labor costs have climbed from 37 percent of the value added in the median firm in 2006 to 48 percent in 2010 even as capital-adjusted productivity has remained steady.20 Between 1995 and 2012, the value added per worker grew twice as quickly in the manufactur- ing, utilities, transportation, and communication sectors as in the overall economy although, measured by educational attainment, the most highly skilled workers are found disproportionately in services (public adminis- tration and defense, education, and social and health care services) (fig- ure 2.8).21 Combined, these findings suggest that a continued expansion in labor income and employment may not be sustainable without greater productivity and efficiency in the labor force. The future of labor productivity will depend on the quality of human capital generated among youth today. Yet, the childhood opportunities Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 65 Figure 2.8 Sector of Employment, by Educational Attainment, Argentina, 2012 100 Private households 90 Other social services Social and health 80 care services Sector of employment (%) Education 70 Public administration 60 and defense Finance, insurance, 50 and real estate Transportation 40 Trade and hospitality 30 services Construction 20 Manufacturing 10 Mining and utilities 0 Agriculture Primary Secondary Postsecondary Educational attainment Source: Based on data in SEDLAC. Note: Data are based on the main employers of employed individuals between the ages of 16 and 65. essential for human capital creation, particularly good schooling and access to housing with proper sewerage (important for childhood health), con- tinue to lag. In 2012, over a quarter (28 percent) of children in urban areas did not have access to improved sanitation in their homes (World Bank 2014).22 Moreover, though public spending on education grew from 3.4 percent in 2003 to 5.6 percent in 2009 (Lustig and Pessino 2014), this did not lead to improved education outcomes. Only about half the students who took the Program for International Student Assessment (PISA) tests in 2006 or 2009 showed a basic ability to apply the subject matter to real- world situations in reading or science, and only 40 percent were able to do so in mathematics.23 The 2012 PISA scores indicate that little progress has been made in the quality of schooling outcomes since then: the rates have remained about the same as in previous years. Access to these two important childhood opportunities—sanitation and good-quality schooling—is unequal. Access to improved sanitation is rela- tively lower among rural residents and the population in the poorer north- ern provinces. Moreover, international test scores reveal that the quality of schooling among children is largely determined in Argentina by parental socioeconomic background. PISA scores adjusted for equity using the HOI methodology indicate that unequal access to good-quality schooling is a 66 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean significant problem in the region, and Argentina is no exception. For exam- ple, while 52 percent of Argentine students scored a 2 or higher on the PISA reading test, the HOI for reading is only 44, indicating that there is a sub- stantial penalty associated with unequal outcomes across groups identified according to their circumstances (World Bank 2014).24 Differences in the educational attainment of parents and the occupations of fathers account for more than half the difference in test scores across groups of children. This is also reflected in the significant variations in the outcomes among public and private school students. Students in private schools are two times more likely to achieve a passing score relative to students in public schools. Differences in childhood access to basic services such as good-quality schooling and sanitation have long-term effects on inequality because they reinforce limited intergenerational upward mobility in Argentina. Because children in households with lower socioeconomic status exhibit worse educational outcomes, they can expect to have less success in the labor market as adults, all else being equal. Upward mobility has been substan- tial in Argentina in recent years; at least 42 percent of the poor in 1994 had escaped poverty by 2009. However, mobility across generations is less significant (Ferreira et al. 2013). Thus, while the income distribution has shifted toward less poverty, the outcomes of each generation continue to be highly correlated with the outcomes of the previous generation. Aside from issues of fairness, the lack of access to such opportunities can also hurt a nation’s growth prospects because potential human capital is left untapped and underutilized. Public spending: fiscal health and household resiliency Social spending directly benefited 44.6 percent of the urban population in 2009, including 91.9 percent of the extreme poor and 78.8 percent of the moderate poor (Lustig and Pessino 2014). Public spending in Argentina is largely progressive; that is, it leads to a reduction in income inequality. All social spending programs in 2009 were progressive. Additionally, over a third of public spending on food, direct household transfers, and noncon- tributory pensions benefit people with incomes below $2.50 a day (Lustig and Pessino 2014). However, indirect subsidies, which largely favor the middle class, dou- bled in value from 2.5 percent of GDP in 2003 to 5.6 percent in 2009, a year in which nonpension cash transfers accounted for 0.8 percent of GDP and noncontributory and moratorium pensions accounted for an estimated 5.3 percent (Lustig and Pessino 2014) (figure 2.9). Social spending pro- grams—both direct transfers and indirect transfers through education and health care spending—are progressive; that is, lower-income groups receive a disproportionate share of the benefits, thereby leading to a reduction in inequality. However, subsidies are not progressive (Lustig and Pessino 2014). Subsidies going to agriculture, manufacturing, and airlines are regressive; that is, the benefits accrue disproportionately to higher-income households, while transportation and energy subsidies are progressive only Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 67 Figure 2.9 Government Spending as a Share of GDP, Argentina, 2003 and 2009 8 7.2 7 6.2 6 5.6 5.6 Share of GDP (%) 5 4 3.4 3 2.9 2.6 2.4 2.5 2 1.9 1.3 1 0.8 0.7 0.4 0 s io ry io m ki tion ki are id ect io ry er ns to ns riu ns to ns ns ) ) s ns bs ir sf (in th c nd nd ie (in ca pe ribu pe ribu su Ind pe ato an u l tr Ed ea or nt nt sh H co M Co Ca on N Spending category 2003 2009 Source: Lustig and Pessino 2014. in relative terms because, while they favor higher-income households, they are more progressive than the distribution of income. As a result, the gov- ernment has begun taking steps to cut subsidies by announcing reductions in utility subsidies (in March 2014). Further subsidy cuts may be one way for it to trim public spending without diminishing the support for the poor and vulnerable. The receipt of pensions was appreciably enlarged in 2006 and 2007 because of the pension moratorium, which granted pensions to beneficia- ries who had not contributed the full 30 years of contributions into the system. The share of urban households with at least one member eligible for pensions—age 60 or above among women and age 65 or above among men—who was not receiving some income from pensions halved between 2006 and 2012, from 28 to 13 percent (figure 2.10, chart a). Bosch and Guajardo (2012) find that, while the moratorium generated overall employ- ment declines among older men and women of 4.5 and 5.0 percentage points, respectively, it may also have led some older workers to switch to informal employment so as to continue receiving pensions. As a result of the broadening in pension access and the frequent adjust- ments in pension benefits, poverty rates have continued to decline among households with pensioners (figure 2.10, chart b). Between 2006 and 2007, poverty among households with no pension income fell by 5 percent as poor households began to receive pensions and move out of poverty. In 2007, the extreme poverty rate began to fall among households with pensioners, 68 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 2.10 The Impact of the Pension Moratorium on Pension Coverage and Poverty, Argentina, 2004–12 a. Households receiving pensions b. Extreme poverty, by pension income 35 25 30 Extreme poverty rate (%) Share of households (%) 20 25 20 15 15 10 10 5 5 0 0 2004 2006 2008 2010 2012 2004 2006 2008 2010 2012 Year Year Share receiving pensions: Share of pensions in household income: 0% > 80%– ≤ 100% 0% > 60%– ≤ 80% > 80%– ≤ 100% Source: Based on data in SEDLAC. Note: The data refer to households with members of pension age. such that fewer than 1 percent of households receiving at least 60 percent of their income from pensions were in extreme poverty in 2012. The relatively high extreme poverty rate among households not receiving pensions indi- cates that some pension-aged individuals are without access to pensions. A new pension moratorium program was announced in June 2014, expand- ing coverage to anyone who made pension contributions between 1994 and 2003. Care should be taken to address the high transaction costs associated with accessing these pensions. These costs may inhibit a share of the older population from obtaining the benefit, particularly individuals with lower educational attainment. While the rise in pension access has led to a notable reduction in pov- erty, the growing coverage of public spending has not. The share of house- holds receiving income from public sources increased from 42 to 50 percent between 2004 and 2012.25 However, this has been largely limited to house- holds receiving less than 20 percent of their incomes from public sources, that is, households with less need. Meanwhile, beginning in 2010, benefits across various types of income from public sources have not risen at the same rate; thus, the minimum pension has climbed more quickly than the minimum wage or the minimum universal child allowance.26 Even though public spending has increased, some of the neediest recipi- ents are still in extreme poverty. Consider the households most likely to need public transfers: households with low educational attainment and child dependents.27 The majority of such households were not recipients of Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 69 Figure 2.11 Public Transfers, Households with Children and Low Educational Attainment, Argentina, 2004–12 a. Share of households b. Income relative to the receiving transfers average nonbeneficiary 50 80 45 by income from transfers (%) Income: beneficiary relative Share receiving transfers, 40 to nonbeneficiary (%) 60 35 30 25 40 20 15 20 10 5 0 0 2004 2006 2008 2010 2012 20–40 40–60 60–80 80–100 Year Share of income from transfers (%) 1–20% 20–40% 40–60% 2012 2004 60–80% 80–100% Source: Based on data in SEDLAC. Note: The data refer to households (beneficiary and nonbeneficiary) with low educational attainment and children under the age of 18. A household has low educational attainment if none of the adult members has completed secondary schooling. direct public transfers, while transfers accounted for less than 20 percent of the incomes of two-thirds of the households that did receive transfers (figure 2.11, chart a).28 However, among households characterized by low educa- tional attainment, dependents, and a heavy reliance on public transfers, the extreme poverty rate is still high; 94 percent of households receiving more than 80 percent of their incomes from social transfers were living in extreme poverty in 2012, along with more than 70 percent of households receiving at least 40 percent of their incomes from social transfers: public transfers are not sufficient to keep these households from living in extreme poverty. As a result, while poverty rates have fallen significantly among some ben- eficiaries of social spending, notably pensioners, the incomes of the recipi- ents of public transfers with low income from other sources have declined relative to the incomes of nonbeneficiaries. Households receiving more than 80 percent of their incomes from transfers in 2004 reported per capita incomes equivalent to 29 percent of the corresponding incomes of similar nonbeneficiary households; by 2012, the share was only 13 percent (figure 2.11, chart b). Beccaria et al. (2013) find that the exits from poverty related to nonlabor income between 2003 and 2008 arose primarily because of the receipt of pension income, while the receipt of public transfers did not translate into transitions out of poverty. This suggests a closer look at the 70 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean nation’s social spending programs is needed: more antipoverty gains might be possible if spending were reallocated from households with less need to households living in extreme poverty. Because of the size and scope of Argentina’s social programs, maintain- ing many of the gains in poverty and shared prosperity realized since 2004 must rely on the continued fiscal sustainability of these programs. Yet, this sustainability is being weakened by rising fiscal deficits as expenditure growth outpaces GDP growth (Lustig and Pessino 2014). Since 2011, the International Monetary Fund has been reporting and predicting annual pri- mary budget deficits on the order of between −0.5 and −0.9 percent of GDP through 2019 (IMF 2014).29 To accommodate spending increases, tax collection and social security contributions were rising throughout the period, reaching 29.5 percent of GDP in 2012. Nonetheless, expenditure growth has outpaced revenue over the past decade as the commodity-fueled surpluses of the postcrisis period became deficits more recently. Without spending cuts, especially to regressive subsidies, the current fiscal position means there is minimal flexibility to address shocks. Protecting the postcrisis advances and investing in future progress require that macroeconomic and fiscal conditions must not deteriorate. Several cru- cial macroeconomic and fiscal challenges face the country in the near to medium term. Primary among these is the decline in growth relative to the past decade because of weak global demand, slowing growth in Brazil and China, Argentina’s two largest trading partners, and restrictive domestic measures. Facing tighter access to international capital markets, the gov- ernment’s dollar reserves are the only source of financing for external debt servicing. These reserves have deteriorated quickly, dwindling from $52.2 billion in 2010 to $30.6 billion in 2013. High inflation, estimated at 16.4 percent for the first six months of 2014, also continues to be a challenge (INDEC 2014). Additionally, because of the importance of labor income in poverty reduction, a business climate and labor market outlook that are dimmed by restrictive policies can restrict future poverty cuts. According to the World Development Report 2014, a variety of financial tools are necessary for effective household risk management (World Bank 2013b). Among these are savings instruments and a reliable and accessible banking sector that allows individuals to save in good times to smooth out consumption during bad times. However, banking and savings rates are low in Argentina. The country trails other upper-middle-income countries and other countries in the region in the share of adults with accounts at for- mal financial institutions, particularly among adults who have completed secondary schooling (37 percent in Argentina compared with 46 percent in the region) (Demirgüç-Kunt and Klapper 2012). Saving is severely dis- couraged by high inflation rates; hence, maintaining a low inflation rate can be an important risk management tool for raising household savings. Another important tool is access to credit, which, in the absence of savings, can also be used to smooth consumption and boost investments in human capital (such as through educational loans) and productivity (such as the Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 71 purchase of a vehicle or business input). Also, access to emergency public transfer programs, such as the Jefes y Jefas de Hogar program, is a key safety net; this program accounted for nearly 40 percent of all household income among the bottom quintile in October 2002 (McKenzie 2004). Above all, protecting the social gains and insuring against another cri- sis require prudent macroeconomic and fiscal management. The significant currency devaluation during the 2001–02 crisis, combined with the freez- ing of bank accounts, showed that household savings and access to credit are insufficient in the face of a major crisis (Gasparini and Cruces 2009). Argentina’s vulnerability to fiscal and macroeconomic shocks is evident in the frequency of crises experienced by the country since the 1980s and the severity of the 2001–02 crisis (Gasparini and Cruces 2009). Even prior to the 2001–02 crisis, Argentina surpassed all countries in the region in vola- tility, as well as the regional averages in East Asia and Sub-Saharan Africa (Fatás and Mihov 2003). This turbulent history, along with the bleak mac- roeconomic prospects confronting the region, suggests a need for cautious spending policies and extra care in fortifying the risk mitigation tools avail- able to households. Notes 1. In this chapter, all data attributed to calculations based on the Socio-Economic Database for Latin America and the Caribbean (SEDLAC) rely on a harmo- nized version of the urban-only household survey, the Encuesta Permanente de Hogares-Continua. The survey is collected quarterly by the Instituto Nacional de Estadística y Censos (National Institute of Statistics and Censuses, INDEC), though the results included in this chapter rely only on the last two quarters of each year. The survey is representative of the 61 percent of the population living in the 31 largest urban areas in the country. The harmonization undertaken for the database increases the comparability of household surveys across countries in the Latin America and Caribbean region, allowing for internationally com- parable indicators. All monetary measures, including poverty rates, are adjusted to 2005 purchasing power parity (PPP) U.S. dollars using official inflation esti- mates prior to 2007 and private estimates in later years. Because the microdata have been harmonized, poverty is reported using only international poverty lines. See SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad Nacional de La Plata, La Plata, Argentina and World Bank, Washington, DC, http://sedlac.econo.unlp.edu.ar/eng/statistics.php. 2. Information based on tabulations using data from WDI (World Development Indicators) (database), World Bank, Washington, DC, http://data.worldbank .org/data-catalog/world-development-indicators. 3. Because of changes in survey methodology, data from before 2004 are not strictly comparable with data from later years. As a result, much of the analysis included in this chapter covers a period beginning in 2004. 4. Information based on INDEC tabulations using the National Census of Popu- lation, Households, and Housing 2010 (“Resultados definitivos,” Serie B, N.2, Tomo 1). Five measures of deprivation rates (unsatisfied basic needs indicators) 72 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean were calculated from the census: (a) overcrowding: more than three people per room; (b) housing conditions: unsuitable or precarious housing; (c) sanitation: lack of a bathroom; (d) education: at least one school-aged child (6–12 years of age) not attending school; and (e) high dependency ratio: four or more people per employed household member, and the household head has less than three years of primary schooling. 5. The calculation is based on a Shapley decomposition of poverty changes using data in SEDLAC. Also see Azevedo, Sanfelice, and Nguyen 2012; Barros et al. 2006. Extreme poverty is measured at $2.50 a day (2005 PPP U.S. dollars). 6. Tabulations based on data in SEDLAC. 7. Argentina has a noncontributory pension program that covers elderly people who are ineligible for contributory pensions. 8. Pi Alperin (2009) finds that the Jefes y Jefas de Hogar Program had an unclear impact on job creation, while Galasso and Ravallion (2003) find that employ- ment rose among participants through flows from both unemployment and inactivity. 9. See “Advierten que reducir la mortalidad materna es uno de los desafíos centrales en salud reproductiva,” Pan American Health Organization, Buenos Aires, June 2008, http://www.paho.org/arg/index.php?option=com_ content&view= article&id=107&Itemid=259. 10. The differences across the coverage rate, the proportion of children who have access to a particular good or service, and the human opportunity index (HOI) are the penalty for unequal access across groups defined according to circum- stances. For example, if a service were evenly distributed across all groups so defined, the relevant HOI and the coverage rate would be equivalent. 11. Tabulations based on National Census of Population, Households, and Hous- ing 2010, Subsecretaría de Planificación Territorial de la Inversión Pública. Programa Argentina Urbana, Avance II, Plan Estratégico Territorial, 2011. 12. The areas examined here align closely with the four policy areas identified by the World Bank (2013a) as essential for boosting and sustaining shared pros- perity: (a) strengthening fair, transparent institutions that deliver high-quality goods; (b) enabling an environment of well-functioning and accessible markets; (c) maintaining equitable, efficient, and sustainable fiscal policy; and (d) devel- oping instruments to improve risk management at the macro and household levels. 13. Tabulations based on data in SEDLAC. 14. See “Total Factor Productivity Level at Current Purchasing Power Parities for Argentina,” FRED (Federal Reserve Economic Data) (database), Federal Reserve Bank of St. Louis, St. Louis, http://research.stlouisfed.org/fred2/series /CTFPPPARA669NRUG. See also Feenstra, Inklaar, and Timmer (2013). 15. Data for 2010 in Enterprise Surveys (database), International Finance Corpo- ration and World Bank, Washington, DC, http://www.enterprisesurveys.org. 16. Calculations based on 2011 data in FinStats (internal database), World Bank, Washington, DC. 17. Calculations based on 2010 data in Enterprise Surveys (database), Interna- tional Finance Corporation and World Bank, Washington, DC, http://www .enterprisesurveys.org. Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 73 18. Calculations based on data in Enterprise Surveys (database), Interna- tional Finance Corporation and World Bank, Washington, DC, http://www .enterprisesurveys.org. 19. Calculations of technical efficiency based on data in Enterprise Surveys (data- base), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. 20. Calculations of technical efficiency, value added, and labor costs based on 2010 data in Enterprise Surveys (database), International Finance Corporation and World Bank, Washington, DC, http://www.enterprisesurveys.org. Similarly, Frenkel and Rapetti (2012) decompose the rise in labor costs between 2002 and 2010 and find that wages grew more than productivity. 21. Calculations based on data of INDEC and the Groningen Growth and Devel- opment Center, Economics Department, University of Groningen, Groningen, Netherlands. 22. Access to running water is defined as the availability in the dwelling of piped water from a public water source. Access to sanitation is defined as the avail- ability in the dwelling or on the property of a bathroom or latrine that is con- nected to a sewerage system or a septic tank. 23. PISA is a worldwide study in member and nonmember nations carried out by the Organisation for Economic Co-operation and Development (OECD) among 15-year-olds to gauge their scholastic performance in mathematics, reading, and science. Passing in the text refers to achieving a score of 2 or higher, the threshold indicating a basic ability to apply the subject matter to real-world situations. In 2012, 77 percent of children in the OECD scored a 2 or higher on the mathematics section (OCED 2014). 24. The rates reported here are not identical to those reported in the OECD reports because, to calculate the HOI, the OECD observations involving incomplete data on the circumstances of the children have been dropped. The circum- stances used to calculate the HOI are gender of the child, parental education, school location (region), father’s occupation, and a household wealth index based on the composition of household assets. 25. Public income sources include direct cash transfer programs (unemployment insurance, the Jefes y Jefas de Hogar Program, the Programa Familias, the uni- versal child allowance, and scholarship programs), pensions (both contributory and noncontributory), and wages from public employment (the main employer only). 26. Conclusions based on analysis of Argentina’s published legal code. 27. Specifically, households with dependents in which no adult has completed sec- ondary education and in which there is no pension access. 28. The transfers include unemployment insurance, the Jefes y Jefas de Hogar Pro- gram, the Programa Familias, the universal child allowance, and scholarship programs. 29. The primary deficit excludes net interest payments. 74 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean References Azevedo, João Pedro, Viviane Sanfelice, and Minh Cong Nguyen. 2012. “Shapley Decomposition by Components of a Welfare Measure.” Unpublished working paper, World Bank, Washington, DC. Barros, Ricardo Paes de, Mirela De Carvalho, Samuel Franco, and Rosane Men- doça. 2006. “Uma Análise das Principais Causas da Queda Recente na Desigual- dade de Renda Brasileira.” Revista Econômica 8 (1): 117–47. Beccaria, Luis, Roxana Maurizio, Ana Laura Fernández, Paula Monslavo, and Mariana Álvarez. 2013. “Urban Poverty and Labor Market Dynamics in Five Latin American Countries: 2003–2008.” Journal of Economic Inequality 11 (4): 555–80. Bosch, Mariano, and Jarret Guajardo. 2012. “Labor Market Impacts of Non- Contributory Pensions: The Case of Argentina’s Moratorium.” IDB Working Paper IDB-WP-366, Inter-American Development Bank, Washington, DC. Bozzoli, Carlos, and Climent Quintana-Domeque. 2014. “The Weight of the Crisis: Evidence from Newborns in Argentina.” Review of Economics and Statistics 96 (3): 550–62. Corbacho, Ana, Mercedes Garcia-Escribano, and Gabriela Inchauste. 2007. “Argentina: Macroeconomic Crisis and Household Vulnerability.” Review of Development Economics 11 (1): 92–106. Demirgüç-Kunt, Asli, and Leora Klapper. 2012. “Measuring Financial Inclusion: The Global Findex Database.” Policy Research Working Paper 6025, World Bank, Washington, DC. Fatás, Antonio, and Ilian Mihov. 2003. “The Case For Restricting Fiscal Policy Discretion.” Quarterly Journal of Economics 118 (4): 1419–47. Feenstra, Robert C., Robert Inklaar, and Marcel P. Timmer. 2013. “The Next Generation of the Penn World Table.” University of Groningen, Groningen, Netherlands. http://www.rug.nl/research/ggdc/data/pwt/. Ferreira, Francisco H. G., Julian Messina, Jamele Rigolini, Luis F. López-Calva, María Ana Lugo, and Renos Vakis. 2013. Economic Mobility and the Rise of the Latin American Middle Class. Washington, DC: World Bank. Frenkel, Roberto, and Martin Rapetti. 2012. “External Fragility or Deindustrial- ization: What Is the Main Threat to Latin American Countries in the 2010s?” World Economic Review 1: 37–57. Galasso, Emanuela, and Martin Ravallion. 2003. “Social Protection in a Crisis: Argentina’s Plan Jefes y Jefas.” World Bank Economic Review 18 (3): 367–99. Gasparini, Leonardo, and Guillermo Cruces. 2009. “A Distribution in Motion, the Case of Argentina: A Review of the Empirical Evidence.” Research for Public Policy, Inclusive Development ID-06–2009, Regional Bureau for Latin America and the Caribbean, United Nations Development Programme, New York. IMF (International Monetary Fund). 2014. “Public Expenditure Reform: Mak- ing Difficult Choices.” Fiscal Monitor, World Economic and Financial Surveys, IMF, Washington, DC. INDEC (Instituto Nacional de Estadística y Censos). 2014. “Índice de Precios al Consumidor Nacional urbano, base IV trimestre 2013=100.” August 15, INDEC, Buenos Aires. http://www.indec.mecon.ar/desaweb/uploads/informes- deprensa/ipc_08_14.pdf. Chapter 2: Shared Prosperity and Poverty Reduction in Urban Argentina 75 Khamis, Melanie. 2013. “Does the Minimum Wage Have a Higher Impact on the Informal Than on the Formal Labour Market? Evidence from Quasi- Experiments.” Applied Economics 45 (4): 477–95. López-Calva, Luis F., and Eduardo Ortiz-Juárez. 2011. “A Vulnerability Approach to the Definition of the Middle Class.” Policy Research Working Paper 5902, World Bank, Washington, DC. Lustig, Nora, and Carola Pessino. 2014. “Social Spending and Income Redistri- bution in Argentina in the 2000s: The Rising Role of Noncontributory Pen- sions.” In “Analyzing the Redistributive Impact of Taxes and Transfers in Latin America,” ed. Nora Lustig, Carola Pessino, and John Scott, special issue, Public Finance Review 42 (3): 304–25. McKenzie, David J. 2004. “Aggregate Shocks and Urban Labor Market Responses: Evidence from Argentina’s Financial Crisis.” Economic Development and Cul- tural Change 52 (4): 719–58. OECD (Organisation for Economic Co-operation and Development). 2014. “PISA 2012 Results in Focus: What 15-Year-Olds Know and What They Can Do with What They Know.” Overview, OECD, Paris. Pi Alperin, María N. 2009. “The Impact of Argentina’s Social Assistance Program Plan Jefes y Jefas de Hogar on Structural Poverty.” Estudios Económicos (spe- cial issue): 49–81. Sandleris, Guido, and Mark L. J. Wright. 2014. “The Costs of Financial Crises: Resource Misallocation, Productivity, and Welfare in the 2001 Argentine Crisis.” Scandinavian Journal of Economics 116 (1): 87–127. SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad de La Plata, La Plata, Argentina; World Bank, Washington, DC. http://sedlac.econo.unlp.edu.ar/eng /index.php. World Bank. 2010. “The Invisible Poor: A Portrait of Rural Poverty in Argentina.” Report 54035, World Bank Country Study. World Bank, Washington, DC. ———. 2013a. “Shifting Gears to Accelerate Shared Prosperity in Latin America and the Caribbean.” Poverty and Labor Brief, Report 78507 (June), Latin Amer- ica and Caribbean Region, World Bank, Washington, DC. ———. 2013b. World Development Report 2014: Risk and Opportunity, Manag- ing Risk for Development. Washington, DC: World Bank. ———. 2014. “Social Gains in the Balance: A Fiscal Policy Challenge for Latin America and the Caribbean.” Poverty and Labor Brief, Report 85162 rev (Feb- ruary), Latin America and Caribbean Region, World Bank, Washington, DC. CHAPTER 3 Poverty and Shared Prosperity in Brazil: Where to Next? Javier E. Báez, Aude-Sophie Rodella, Ali Sharman, and Martha Viveros Introduction B razil has succeeded in significantly reducing poverty in the last decade. It has nearly eliminated extreme poverty, which fell from a rate of almost 10 percent in 2001 to 4 percent in 2013. About 60 percent of Brazil- ians climbed to a higher economic group, that is, a higher level of income, between 1990 and 2009. Overall, approximately 25 million Brazilians escaped extreme or moderate poverty; this represented one in every two people who escaped poverty in the Latin America and Caribbean region during the period. The evolution of monetary and nonmonetary poverty across the states of Brazil has been a systematic process of poverty conver- gence: poverty is falling more rapidly in those states that had higher poverty rates before 2001. Brazil has also shown strong income growth among the bottom 40 per- cent of the national income distribution (the bottom 40), indicating that economic progress has been leading to shared prosperity. The income growth among the bottom 40 averaged 6.1 percent annually from 2002 to 2012, well above the growth of mean income in the country (3.5 percent). In light of the positive evolution of the shared prosperity indicator (SPI), it is not surprising that income inequality has declined rapidly in Brazil. The Gini coefficient, a standard measure of income or consumption concentra- tion, fell from 0.59 in 2001 to 0.52 in 2013, similar in magnitude to the reduction across the region. What is behind these positive trends? At least three forces stand out as the main explanatory factors. First, Brazil enjoyed relatively stronger and more stable growth after 2001 than in the two preceding decades. At an average real annualized growth rate of 2.3 percent per year from 1999 to 77 78 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean 2012, per capita income grew more rapidly in Brazil than in the region (1.8 percent) and more rapidly than in previous decades in Brazil (0.18 and 0.80 percent in the 1980s and 1990s, respectively). Overall, a standard decom- position analysis of the changes in poverty because of growth and redistri- bution suggests that economic growth explains two-thirds of the drop in poverty in Brazil from 2001 to 2012. The second force that enhanced a growth process that favored the poor is the stronger policy focus on poverty. The government reinvigorated pov- erty and inequality reduction through the active use of redistributive poli- cies. Reforms in social assistance transfers resulted in the establishment of large-scale noncontributory unconditional and conditional cash transfer programs targeted at low-income families that helped accelerate poverty reduction. The third force is the dynamic labor market. Largely as an outcome of strong growth, the labor market has performed at record levels in the last decade (annex 3A). Healthy job creation has been accompanied by a rise in labor force participation and employment rates. The quality of jobs has also improved significantly. In 2012, nearly 60 percent of all jobs were in the formal sector, superseding the share of informal employment for the first time. Additionally, the economy has seen a large expansion in real wages, partly fueled by periodic boosts in the minimum wage. While Brazil has made laudable progress in reducing poverty and inequality and in fostering economic and social inclusion, the task has not yet been carried to completion. Around 18 million Brazilians are still liv- ing in poverty, and over one-third of the population has not yet joined the middle class, remaining instead in a condition of economic vulnerabil- ity and lacking the assets, skills, and employability necessary to abandon vulnerability permanently. Inequality in Brazil is still above the average in Latin America and the Caribbean, a region that is already associated with substantial income disparities. The richest 1 percent of the population in Brazil receives 13 percent of total income, more than the income accrued by the bottom 40 (11 percent). Sustaining and deepening the inclusive growth agenda will require chal- lenges to be addressed in fiscal matters, service delivery, and productivity. Bringing prosperity to the less well off and sustaining the gains that have been achieved will demand policy action on at least three fronts. Key to this agenda will be enhancements to the progressivity of the fiscal system to ensure that public resources continue advancing social goals. There also needs to be a focus on improving the quality of basic ser- vices. Despite the expansion in the coverage of and equitability of access to a range of services in the last decade, quality remains low and uneven across the parts of the country and across population groups. Poor quality is affecting low-income households disproportionally. Finally, bolstering inclusive and sustainable growth will require a boost in productivity, especially among the poor and vulnerable so that they are able to contribute to and benefit from the growth process. The country has Chapter 3: Poverty and Shared Prosperity in Brazil: Where to Next? 79 seen practically no gain in labor productivity since the late 1990s, and most of the growth has been fueled by an increase in labor supply, itself boosted by a demographic trend toward a larger share of the population of working age. Underlying the stagnation in productivity is a low rate of investment, underdeveloped infrastructure, skill shortages and mismatches, rigidities in the labor market, financial exclusion, and a business environment that is not entirely conducive to private sector development and to competition. The Impressive Pace of Poverty Reduction In line with global and regional trends, Brazil made considerable progress in reducing poverty between 1999 and 2013. Based on poverty lines derived from the Bolsa Família (family allowance, BF) conditional cash transfer program and the Brasil sem Misería Plan (Brazil without Misery, BSM), estimates show that poverty fell from 24.7 to 8.9 percent in 2001–13 (box 3.1). Extreme poverty also declined sharply during the period, dropping from 9.9 to 4.0 percent (figure 3.1, chart a). By 2013, over 17 million and 8 million people were counted among the poor or the extreme poor, respec- tively, corresponding to 23.5 million fewer individuals in poverty relative to 2001. Poverty fell more quickly in Brazil than in the Latin America and Caribbean region, and this contributed substantially to poverty reduction regionally. Calculations based on internationally comparable poverty lines uncover the same trends observed in the national lines and also reveal that both moderate and extreme poverty declined more quickly in Brazil than in the region.1 In 1999, the extreme poverty rates of Brazil and the region were similar, at around 26.0 percent. While the rate in the region had fallen to 12.0 percent by 2012, the drop in Brazil was to 9.6 percent. Addition- ally, while the region and Brazil shared similar moderate poverty rates in 1999 (about 43.0 percent), the rate in Brazil had declined to 20.8 percent by 2012, which was below the regional rate, at 25.0 percent (see figure 3.1, chart b). Given the size of the country and the speed of the poverty reduc- tion there, Brazil has contributed substantially to the progress in poverty in the region, where the population living in poverty narrowed from 120 million to 67 million people during the period. According to internation- ally comparable methodologies, the 27 million Brazilians who rose out of poverty in 1999–2012 accounted for half of the people who abandoned poverty in the region. Location is a key element to understanding poverty and equity in the country. The incidence of poverty has traditionally shown a strong correla- tion with geographical borders. Thus, for example, trends in income pov- erty have been heterogeneous across the five macroregions of Brazil. The states in the north and northeast macroregions face levels of poverty above those at the national level. In 2012, poverty rates (measured using the BF- BSM poverty lines) in the south and southeast macroregions were 3.4 and 80 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Box 3.1 Poverty Measurement in Brazil Brazil does not have an official poverty line. Most poverty measurements are derived from an abso- lute poverty line constructed using monthly household income. Several unofficial lines exist. They include lines constructed as a fraction of the official minimum wage (one-quarter or one-half, for example), regionalized monetary lines that reflect variable costs of living in different areas of the country, and a food basket price index based on minimum calorie-intake recommendations of the Food and Agriculture Organization of the United Nations and the World Health Organization.a The lines produced by the Institute for Applied Economic Research (IPEA) were long considered de facto poverty lines in Brazil and were used as such in the World Development Indicators database of the World Bank.b In recent years, R$70 (extreme poverty: indigência) and R$140 (poverty: pobreza) per capita per month, which are administrative poverty lines for the Bolsa Família (family allowance, BF) program and the Brasil sem Misería Plan (Brazil without Misery, BSM) plan, are increasingly taking the place of official poverty lines. Monitoring poverty rates using these administrative lines is crucial, particularly in studies of trends in poverty in the country. According to an agreement with Brazilian authorities, these lines are now applied by the World Bank in data on Brazil in the World Development Indicators database. The international $1.25-a-day extreme poverty line is also used on occasion in Brazil, notably in relation to the Millennium Development Goals. Indeed, complementary to the lines set in Brazil, the lines applied by the World Bank—$1.25, $2.50, and $4.00 a day at purchasing power parity (PPP) U.S. dollars—serve to harmonize the measurement and comparison of poverty and the identification of trends in poverty across countries. The choice to use one or another of these lines may reflect the objectives of an analysis or international comparison or the definition of a public policy. As a result of methodological differences in the computation of lines and income aggregates, there are sometimes small differences between government and World Bank estimates. However, the poverty trends revealed in Brazil are broadly consistent across methodologies. Whenever possible, this chapter reports poverty rates using the Brazilian administrative poverty lines. In international comparisons, the analysis relies on the Socio-Economic Database for Latin America and the Caribbean, which includes a compilation of harmonized household survey data on 24 countries in the region and data on the international poverty lines applied by the World Bank and described above.c a. Based on consumption baskets established for each of the nine metropolitan areas and Brasília, respective values are also derived for 15 urban and rural areas in various parts of the country, thereby establishing a total of 25 extreme poverty lines and poverty lines. The monetary amounts are adjusted relative to a reference date each year according to the varying prices for each product in the basket, based on the national consumer price index set by the Brazilian Institute of Geography and Statistics. Concerning the regional poverty lines, see Rocha (2006). b. In December 2013, IPEA updated its extreme poverty and poverty numbers for the period ranging from 2009 to 2012, but no updated data on the regional lines relied on are available. For 2012, IPEA has put the extreme poverty rate at 5.3 percent and the overall poverty rate at 15.9 percent. See the IPEA website, at http://www .ipeadata.gov.br. See also WDI (World Development Indicators) (database), World Bank, Washington, DC, http:// data.worldbank.org/data-catalog/world-development-indicators. c. See SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad Nacional de La Plata, La Plata, Argentina and World Bank, Washington, DC, http://sedlac.econo.unlp.edu.ar/eng/statistics.php. 4.0 percent, respectively, while, in the north and northeast macroregions, the corresponding rates were 15.6 and 18.4 percent, respectively. Despite the significant heterogeneity in poverty headcounts across states, poverty convergence has been systematic across Brazil. For the most part, poverty rates have fallen more rapidly in states that had higher poverty rates before 2001. This may be observed in figure 3.2, where the vertical Chapter 3: Poverty and Shared Prosperity in Brazil: Where to Next? 81 Figure 3.1 Poverty Lines, Brazil, 1999–2013 a. BF-BSM poverty lines b. International poverty lines 30 50 24.7 45 43.0 25 40 Poverty rate (%) Poverty rate (%) 20 35 30 26.3 25.0 15 25 9.9 10 8.9 20 20.8 15 12.0 5 4.0 10 9.6 0 5 01 02 03 04 05 06 07 08 09 10 11 12 13 20 9 20 0 20 1 02 20 3 20 4 05 20 6 20 7 20 8 20 9 20 0 20 1 12 9 0 0 0 0 0 0 0 0 1 1 20 20 20 20 20 20 20 20 20 20 20 20 20 19 20 20 Year Year Extreme poverty rate (5,000 inhabitants (left-side axis) Extreme poverty, urban areas w/>5,000 inhabitants (left-side axis) Source: INE 2013. Note: Poverty rates correspond to Montevideo and urban areas with 5,000 or more inhabitants. The poverty headcount series is broken in 2002 because of a methodologi- cal survey change introduced by the National Statistics Institute. The poverty data for 2000 and 2001 are not comparable with the rest of the series. GDP = gross domestic product. external economic conditions, such as buoyant demand for the country’s main export products and a booming regional economy, all contributed to Uruguay’s slow, but successful recovery. The real growth of gross domes- tic product (GDP) averaged over 5 percent beginning in 2003, and open unemployment fell from 17 percent in 2002 to close to 11 percent in 2006. Both private consumption and investment recovered, contributing to a sub- stantial rise in imports that offset the positive contribution to growth of exports. The most rapidly growing sectors behind this growth were trade, transport, and communications. Additionally, increased coverage in social assistance programs softened the impact of the crisis on the welfare of the poor. The Asignaciones Familiares family allowance, initially a program for formal sector employees, was expanded in 2004 to include households with incomes at a threshold of three times the minimum wage. In 2005–07, the Plan de Atención Nacional a la Emergencia Social (PANES), a new cash transfer program, was implemented to offer more assistance to the poor. By 2007, the share of the poor was still high, at 30.8 percent, though this was 9 percentage points lower than the peak in 2004. During the recovery, incomes across the distribution grew at a similar rate, around 4.8 percent, Chapter 9: Poverty and Shared Prosperity in Uruguay 307 Figure 9.3 Growth Incidence Curves of per Capita Household Income, Urban Areas, Uruguay, 2000–07 10 5 Annual growth rate, % 0 –5 –10 –15 2 12 22 32 42 52 62 72 82 92 Percentile of per capita household income 2003–07 2000–03 Source: Data of the 2000, 2003, and 2007 Continuous Household Survey. Note: Poverty rates correspond to Montevideo and urban areas with 5,000 or more inhabitants. although the rate was slightly lower among the lowest deciles. As a result, income inequality in Montevideo and urban areas with at least 5,000 inhabitants was almost unchanged. The next phase, from 2007 to 2013, was an expansionary period, with average annual GDP growth rates of around 5.7 percent. The favorable external environment was characterized by strong external demand, high commodity prices, and high global liquidity. High commodity prices and abundant international liquidity contributed to strong investment in Uru- guay; foreign direct investment averaged 5.0 percent of GDP and accounted for about a third of total investment. In addition, public investment rose appreciably as the government undertook significant investment projects. Rapid economic growth was accompanied by substantial job creation, and unemployment declined to historically low levels. Health care benefits and several other government transfers were monetized gradually and became part of calculated household income. Meanwhile, the successor of the PANES program, the Plan de Equidad (equity plan), lowered the incidence and intensity of poverty noticeably. Thus, poverty rates declined rapidly after 2007. By 2012, only 12.4 percent of the population was living with a per capita income below the poverty line, about one-third the rate seven years earlier. The reduction in extreme poverty was even more dramatic in relative terms: the rate dropped 308 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.4 Growth Incidence Curves of per Capita Household Income, Uruguay, 2007–12 a. Nationwide b. Urban and rural areas 12 14 10 12 Annual growth rate, % Annual growth rate, % 10 8 8 6 6 4 4 2 2 0 0 2 12 22 32 42 52 62 72 82 92 2 12 22 32 42 52 62 72 82 92 Percentile of per capita household income Percentile of per capita household income Rural Urban Source: Data of the 2007 and 2012 Continuous Household Survey. to around 0.5 percent, less than one-sixth the rate seven years earlier (see figure 9.2). The drop in poverty in 2007–12 was accompanied by strong income growth among the bottom 40. Between 2007 and 2011, the real per capita income of the bottom 40 rose by more than 9.7 percent annually, while mean income growth was closer to 6.0 percent (figure 9.4). In contrast, households in the top 20 percent of the income distribution experienced the smallest increase in income. A similar pattern was evident in both urban and rural areas. The improvements at the bottom of the distribution were reflected in narrowing income inequality beginning in 2007 and the more rapid tight- ening starting in 2010 (figure 9.5, chart a). In 2007–12, the nationwide Gini coefficient fell an impressive 7 percentage points, from 0.45 to 0.38. This reduction was evident in urban areas and in rural areas (where the inequality gap has traditionally been smaller). Despite its strong standing regionally, however, Uruguay continues to exhibit greater inequality than any member of the OECD not in the region (figure 9.5, chart b). As poverty fell, the size of the middle class rose steadily and currently represents the largest socioeconomic group in Uruguay. In 2002, 46 percent of the population was living on incomes of between $10 and $50 per person a day (in 2005 purchasing power parity [PPP] U.S. dollars), which is the World Bank monetary definition of the middle class (Ferreira et al. 2013) (figure 9.6).5 Because of the noteworthy recovery after the crisis, more than 65 percent of the population belonged to the middle class by 2011. Another Chapter 9: Poverty and Shared Prosperity in Uruguay 309 Figure 9.5 Inequality, Uruguay, the Region, and the OECD, 2006–13 a. Gini coefficient, Uruguay b. Income Gini, the region, and the OECD 50 Slovenia Norway Czech Republic Sweden France OECD 46 Germany Greece 45.5 Spain Canada Gini coeficient Italy 42 United Kingdom Portugal Israel United States Country Turkey 38 Argentina 38.4 Uruguay Peru El Salvador Nicaragua 34 Ecuador Dominican Republic Mexico Costa Rica Bolivia Region 30 Panama Chile 08 09 11 12 06 13 07 10 Region 20 20 20 20 20 20 20 20 Paraguay Brazil Year Colombia Guatemala Nationwide Honduras Montevideo Urban (+5,000 inhabitants) 0 0.1 0.2 0.3 0.4 0.5 0.6 Rural Gini coefficient Sources: Chart a: Data of the 2006–13 Continuous Household Survey. Chart b: Uruguay and the region: data from SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad de La Plata, La Plata, Argentina; World Bank, Washington, DC, http://sedlac .econo.unlp.edu.ar/eng/index.php; OECD countries: OECD.StatExtracts, http://stats.oecd.org/. Note: OECD = Organisation for Economic Co-operation and Development. 16 percent, accounting for the second largest group in society, was the vul- nerable, who have incomes above the official poverty line, but below the $10 a day threshold. Drivers of the Reductions in Poverty and Inequality Uruguay has made remarkable progress over the past decade in improving well-being and narrowing inequalities among the population. What are the main drivers behind the gains? This section explores the importance of both labor and nonlabor income sources in reducing poverty and enhancing the distribution of incomes. In particular, the advances in well-being are the result of a combination of favorable economic conditions and growth pat- terns that have led to greater employment opportunities, as well as key pol- icy reforms in the labor market and in social protection to ensure that the less well off are able to contribute to growth and better living conditions. 310 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.6 Socioeconomic Groups, by Poverty Status, Uruguay, 2002–11 70 65 60 Share of the population, % 50 46 42 40 35 37 30 19 20 16 16 10 14 3 2 5 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year Moderate poor Vulnerable Middle class ($10–$50 per person a day) Rich (more than $50 per person a day) Source: Data of the 2002–11 Continuous Household Survey. Note: Data correspond to Montevideo and other urban areas with 5,000 or more inhabitants. The moderate poverty line is equivalent to $8.57 per person per day (in 2005 purchasing power parity U.S. dollars) for people living in families of average size. Vulnerable represents people above the official poverty line, but below the middle class. Declines in the share of the population that is living below a poverty threshold (the poverty line) can be decomposed into two parts: rising incomes (economic growth reflected in shifts in income distribution) and improve- ments in income distribution (income redistribution reflected in a narrowing of dispersion in income distribution) (see Datt and Ravallion 1992). In Uruguay, economic growth is more significant than greater equal- ity in the distribution of income in explaining the drop in urban poverty between 2003 and 2012; however, beginning in 2007, the narrowing of the inequality gap played an equally important role. In 2003–12, around 19.7 percentage points in the 26.4 percentage point decline in the moderate pov- erty headcount (from 39.4 to 13.1 percent) is explained by income growth, while the remaining 6.7 percentage points were the result of improve- ments in equitable income distribution (figure 9.7). Yet, this outcome arose from events over two distinct periods: in 2003–07, growth in the mean was the only driving force behind the observed drop in the share of the poor, whereas, over the next five years, in 2007–12, the almost 10 percent decline in the poverty headcount that was driven by economic growth was accompanied by a similar contraction associated with improved income distribution. Similar decompositions of the poverty reduction in the region Chapter 9: Poverty and Shared Prosperity in Uruguay 311 Figure 9.7 Decomposition of Shifts in Moderate Poverty, Urban Areas, Uruguay, 2003–12 5 Change in the moderate poverty rate, % 0 1.4 –5 –9.8 –9.6 –10 –19.7 –8.3 –15 –20 –6.7 –25 –30 2003–12 2003–07 2007–12 Years Growth Distribution Source: Data of the Continuous Household Survey. Note: The figure shows a Datt-Ravallion (1992) decomposition based on official moderate poverty lines. Data correspond to Montevideo and other urban areas with more than 5,000 inhabitants. during these years reveal analogous trends: economic growth accounted for two-thirds of the drop in poverty between 2003 and 2012, while shifts in income distribution explain the remaining third.6 As in Uruguay, changes in income distribution had a more substantial role in the region in 2007–12, representing nearly half of the reduction in poverty there. The redistribution component should not be considered as representing the government’s fiscal and social policies; rather, it reflects the fact that the incomes of less well-off households were rising at a more rapid pace than the incomes of the rest of the population because of increases in labor incomes (through higher earnings or higher employment rates) or upsurges in nonlabor incomes, including capital gains, private transfers, and public transfers. Labor markets Through greater earnings and the rising number of the employed, labor markets are fundamental in explaining the improvements in well-being in Uruguay. At the peak of the 2001–02 crisis, 17 percent of the economically active population was unemployed, and over 40 percent of the employed were not covered by social security. The situation began to improve signifi- cantly only after 2004. Unemployment fell rapidly and reached a record low of 6.1 percent in 2012, while labor informality (proxied by eligibility for 312 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.8 Unemployment and Formal Employment, Uruguay, 2006–12 a. Unemployment rate b. Formal employment 16 75 Share of the employed eligible for 14 70 Unemployment rate, % 12 retirement benefits, % 10 65 8 60 6 4 55 2 0 50 2006 2007 2008 2009 2010 2011 2012 2006 2007 2008 2009 2010 2011 2012 Year Year Total Men Women Sources: Chart a: National Statistics Institute. Chart b: Data of the 2006–12 Continuous Household Survey. retirement benefits) also declined as a consequence of better macroeconomic conditions and the enhanced collection of social security contributions (fig- ure 9.8). The higher employment rates and better employment outcomes were felt across the skills distribution: unemployment fell both among workers with only primary education or less and workers with only some secondary schooling (figure 9.9). Meanwhile, the shares of workers in wage employment (as opposed to self-employment or unpaid status) rose among both groups, from 38 to 41 percent among workers with primary education and from 53 to 56 percent among workers with some secondary schooling. Employment declined during the crisis, especially in manufacturing, con- struction, transport, storage, and communications. During the recovery, employment expanded by about 4 percent annually. Employment growth accelerated in primary sector activities, the retail trade, and the hospitality industry, and recovered in manufacturing, construction, transport, storage, and communications. During the recent economic expansion, employment increased rapidly in construction and at a more moderate pace in transport, storage, and communications and in other services. Utilities and transport, storage, and communications were the most rapidly growing sectors in terms of output in 2007–13 (table 9.1). The rise in the employment share of trade, tourism, and transport, combined with the dynamic growth of these sectors, meant that these sectors were the larg- est contributors to GDP and the largest sectors of employment, accounting for 29 percent of all employment in 2012. Chapter 9: Poverty and Shared Prosperity in Uruguay 313 Figure 9.9 Employment and Participation Rates, by Skill Level, Uruguay, 2007–12 100 22 22 20 19 22 23 80 60 Rate, % 38 41 53 56 59 63 40 6 4 7 5 20 34 33 4 3 20 20 15 14 0 2007 2012 2007 2012 2007 2012 Completed primary Completed some Completed secondary school or less secondary school school Year Other employment Wage worker Unemployed Not in the labor force Source: Data of SEDLAC. Note: Data refer to individuals age 15–70 years who reported they were not in school. Other employment = self-employment, employers, and unpaid workers. Table 9.1 Sectoral Output and Employment Growth, Uruguay, 2003–13 average annual rate, % Output Employment Sector 2003–06 2007–13 2007–12 Gas, electricity, water −9.4 15.5 6.7 Transport, storage, and communications 9.5 15 4.1 Trade and hospitality 4.2 7.2 4.6 Construction 6.3 4.9 1.1 Other services 0.5 3.8 4.0 Primary sector activities 6.6 3.5 −1.0 Manufacturing 7.7 3.1 0.2 Source: Based on data of the Banco Central del Uruguay. The lack of growth in employment in the primary sector and in manu- facturing resulted in a shift of low-skilled labor away from these sectors and toward construction, trade, and hospitality–transport. Among workers with only primary schooling, for example, the share of the employed in the primary sector fell by 1.7 percentage points between 2007 and 2012, while the share in manufacturing fell by 1.9 percentage points (figure 9.10, chart a). Even so, the primary sector continued to be the largest sector of employment among workers with only primary schooling or less, account- ing for 20 percent of employment in 2012 (figure 9.10, chart b). In contrast, 314 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.10 Sector of Low-Skilled Employment, Uruguay, 2007–12 a. Shifts in sector of employment, 2007–12 b. Sector of employment, 2012 100 Private households 16 17 25 26 Contribution to employment, % Hospitality, 80 transport, and 15 15 Sector of employment communications 10 9 7 8 Wholesale and 60 10 11 retail trade 17 17 Construction 23 25 40 10 11 13 7 Manufacturing 12 7 20 17 15 22 20 Primary sector 8 6 0 –3 –2 –1 0 1 2 3 2007 2012 2007 2012 Shift in employment Primary school Some share in each sector, % or less secondary school Some secondary school Skill level, by education Primary school or less Primary sector Hospitality, Manufacturing transport, Construction communications Private households Wholesale and retail trade Other Source: Data of SEDLAC. Note: Data refer to adults age 15 years or over who reported they were employed. employment in construction and hospitality–transport grew at an annual rate of 1 percent among the lowest skilled workers, while trade was the largest employer of workers with some secondary schooling. In addition to expanded employment opportunities, earnings rose throughout the decade. Real wages experienced a major and sudden drop during the 2001–02 crisis, losing about a fourth of their value. Real wages began recovering thereafter and, by 2009, had surpassed the precrisis level (figure 9.11). Alves et al. (2012) argue that the recent narrowing in earnings inequality was driven largely by rising returns to education and declines in skill premi- ums. Indeed, earnings grew more quickly among the less skilled. Between 2007 and 2012, average monthly earnings increased by 27 percent among workers with only primary education or less and by 24 percent among workers with only some secondary schooling (figure 9.12). Meanwhile, the earnings of more highly skilled labor expanded by 10 percent among workers who had only finished secondary school and by 4 percent among Chapter 9: Poverty and Shared Prosperity in Uruguay 315 Figure 9.11 Real Wage Index, Uruguay, 2000–13 130 120 Index (July 2008 = 100) 110 100 90 80 70 60 00 01 02 03 04 05 06 07 08 09 10 11 12 13 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Source: Data of the National Statistics Institute. workers with some tertiary education. Earnings growth rates were strong across the four skill groups between 2007 and 2009, but diverged there- after. While the least skilled saw their earnings rise each year, workers with some tertiary education experienced annual reductions in mean monthly earnings on the order of 4 percent a year between 2010 and 2012. This drop in real returns to skills suggests that the skills premium was falling in Uruguay, while the binding minimum wage was leading to higher earnings among the lowest skilled. The recent narrowing in earnings inequality occurred in a context of major institutional changes including increases in the minimum wage, which rose 200 percent between 2004 and 2010, the restoration of collec- tive bargaining in 2005, and passage of a new law on wage negotiations in 2009 (Levy and Schady 2013).7 In addition, there was an expansion in the rate of formality among private sector workers (Amarante et al. 2011). The role of public spending While labor markets played a fundamental role in the enhancement of liv- ing conditions in Uruguay, the introduction of income transfer schemes targeted at the bottom of the income distribution, as well as reforms in the personal income tax code, contributed toward a more equal society. Public spending in the country is generally prodigious and effective. It accounted for 21.7 percent of GDP in 2009. The three largest components of public spending were contributory pensions (8.5 percent of GDP), health care (4.7 percent), and education (3.7 percent) (Bucheli et al. 2014). Prior to the 2001–02 crisis, the family allowance program Asignaciones Familiares provided monthly cash benefits to formal sector workers with 316 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.12 Growth in Labor Earnings, by Skill Level among Workers, Uruguay, 2007–12 10 Growth in real mean monthly earnings, % 8 6 4 2 0 –2 –4 –6 2007–08 2008–09 2009–10 2010–11 2011–12 Years Primary school or less Completed secondary school Some secondary school Some postsecondary Source: Data of SEDLAC. Note: Data refer to adults age 15 years or over who reported they were employed. children (Amarante and Vigorito 2012). A social pension scheme for the elderly and disabled that was implemented in 1919 targeted the socially vulnerable. Though these transfers were protected during the crisis, they had limited impact on preventing people from falling into poverty because the value of the transfers was not raised. Asignaciones Familiares was expanded in 2004 to include households with incomes below $39 a month (three times the national minimum wage), but was still too low to have a significant effect on poverty (Amarante and Vigorito 2012). Recognizing the need for more assistance for the poor and to facilitate social inclusion, PANES, an emergency social plan (see above), was car- ried out by the new Ministry of Social Development from April 2005 to December 2007 to target the bottom 20 percent of households living below the poverty line (8 percent of the total population). The plan had four main components: (1) a cash transfer, (2) a food card, (3) educational and social reinsertion programs, and (4) housing subsidies and public works. Of the households that were selected to participate, almost all obtained the cash transfer; 80 percent received the food card; and 20 percent participated in the latter two programs. PANES covered 83,000 households (5 percent of all households and 10 percent of the population), and the benefits represented an average of 30 percent of household incomes among the beneficiaries. Chapter 9: Poverty and Shared Prosperity in Uruguay 317 The program cost $80 million or 0.41 percent of GDP annually (Amarante and Vigorito 2012). In 2007, PANES was refashioned into the Plan de Equidad Program (see above). The new program included tax and health care reforms, continued the $8–$16 per child cash family allowance transfers among households that did not receive more than the national minimum wage, maintained the food cards, expanded the coverage of early childhood services, and lowered the retirement age. By 2009, there were 364,000 beneficiaries, including 76 percent of all destitute children, 68 percent of children living in pov- erty, and almost all households in the poorest quintile. Under the program, noncontributory transfers accounted for an average of approximately 20 percent of household income among households in the bottom decile (Ama- rante et al. 2011). According to Dean and Vigorito (2011), the new transfer scheme significantly reduced extreme poverty, but had a limited effect on moderate poverty. Despite the importance of current social programs, poverty persists in the country, particularly among households with low educational attain- ment, especially in Montevideo, and among households that are likely to be excluded from the transfer system. Moreover, even if they receive trans- fers, poor households with children are more likely to remain poor (Bucheli et al. 2014). This suggests that there is room for improvement in social spending. On the revenue side, the progressive tax policy is estimated to have nar- rowed income inequality by 0.03 points of the Gini coefficient, a reduction of 6.0 percent after taxes (figure 9.13). Using the incidence analysis frame- work developed by Commitment to Equity, Bucheli et al. (2014) find that inequality prior to any taxation or public transfer (market income) was 0.49 in 2009.8 Direct taxation, including personal income taxes and employ- ment taxes, reduces inequality to 0.48, and public transfers further reduce the Gini coefficient to 0.46. Although indirect taxes, such as consumption taxes, are regressive or inequality increasing, inequality after these taxes have been taken into account (postfiscal income) remains at 0.46. Bucheli et al. (2014) extend the analysis one step further. By adding the cost of publicly provided educational and health care services to the incomes of households that used these services, they estimate that the Gini coefficient would narrow to 0.39. This is the result of the greater take-up of these public services among lower-income households. It suggests that dif- ferences in service quality may also play a role given that households able to afford private services seem to opt out of public services at a higher rate. Key Challenges Vulnerability to external shocks A small open economy, Uruguay is vulnerable to regional contagion effects, particularly from neighboring Argentina and Brazil. The correlation in GDP 318 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.13 The Gini Coefficient and Pre- and Postfiscal Incomes, Uruguay, 2009 0.60 Gini coefficient 0.55 0.50 0.49 0.48 0.46 0.46 0.45 0.39 0.40 0.35 Market Net market Disposable Postfiscal Final income income income income income Type of income Source: Bucheli et al. 2014. growth is 0.94 between Uruguay and Argentina and 0.83 between Uruguay and Brazil (IMF 2011). The correlation with the economic cycle in Argen- tina has likely declined in recent years as Uruguay has increased its trade with other partners. Argentina nonetheless remains one of the most impor- tant markets for service exports from Uruguay, albeit net tourism export revenues have declined markedly in recent years because of the difficult situation in Argentina. Uruguay has successfully diversified export destina- tion markets; China and the Russian Federation have become important markets. China now accounts for more than 20 percent of Uruguay’s mer- chandise exports (including reexports from free trade zones in China). Only 5.6 percent of merchandise exports went to Argentina in 2012–13, down from 14.5 percent in 1990–98. Brazil accounts for a much more significant share of Uruguay’s exports, around 19.0 percent, and continues to be one of the main destinations of merchandise exports from Uruguay. Barriga et al. (2014) have conducted a simulation exercise suggesting that, as a result of the policies implemented since the 2001–02 crisis, Uru- guay is now in a better position to weather a severe crisis.9 The predicted impact on poverty would be considerably smaller; inequality would not change significantly; and household incomes would only fall by 8 percent (figure 9.14). A large contributing factor in the greater resilience is the improved and expanded social safety nets, as well as the larger role of social transfers and nonlabor components in household incomes. While the overall effect of a crisis on poverty would be relatively mild, the average per capita income of the vulnerable (those between the pov- erty line and the middle-class cutoff) would be 25 percent less in the event of a crisis (figure 9.15). The simulation predicts that younger individuals, Chapter 9: Poverty and Shared Prosperity in Uruguay 319 Figure 9.14 The Impact of a Crisis on Poverty and Inequality, Uruguay, 2011–14 a. Impact on poverty b. Impact on inequality 18 0.45 15.4 16 Poverty rate or poverty gap, % 15.2 Gini coefficient or Theil index 13.7 13.5 0.40 0.382 0.388 0.392 0.391 14 12 0.383 0.382 12.5 11.7 0.35 10 8 0.30 0.284 0.280 6 0.261 0.277 4.2 4.2 3.7 3.6 4 0.25 0.266 0.262 2 3.3 3.1 0 0.20 2011 2012a 2013 2014 2011 2012a 2013 2014 Year Year Poverty rate, no crisis Gini, no crisis Poverty rate, crisis Gini, crisis Poverty gap, no crisis Theil, no crisis Poverty gap, crisis Theil, crisis Source: Barriga et al. 2014. a. Baseline year. woman-headed households, larger households, and people who have not completed secondary education would be relatively more vulnerable to the risk of falling into poverty were a crisis to strike again. The likelihood of falling back into poverty is also higher in Montevideo than in rural areas; although this effect would be mostly driven by a lack of growth, changes in income distribution would also play a part. Even in the absence of local economic shocks such as the Argentine crisis of 2001–02, it is important that government policy in Uruguay be designed with an eye toward protecting the gains in poverty and shared prosper- ity. Policy should reflect, for example, insight into how Uruguay might weather a slowdown in growth in the region. Though the simulation results of Barriga et al. (2014) are encouraging and suggest that Uruguay has suc- cessfully built up a resistance to regional contagion, they also highlight that some groups, particularly the less highly skilled and woman-headed house- holds, are still vulnerable. Service delivery Education Hanushek and Woessmann (2012) find that disparities in human capital can account for half to two-thirds of the income variations between Latin America and the rest of the world. In large part, this is driven by differences in both educational attainment and school quality. 320 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean Figure 9.15 Households in the Crisis Scenario in Year 2 (2014), Uruguay a. Socioeconomic class b. Educational attainment 90 84.5 45 Income change or population share, % 40 70 35 Population share, % 50 30 25 30 20 11.5 10 3.7 15 0 10 –10 –4.2 –2.1 5 –25.4 –30 0 is or or e e e e e e e on et et et et et et is po po pl pl pl pl pl pl cr N m m m m m m s er to ay co co co co co co ev e lw in in in bl N y ry ry A ar ra y ry ry da a ar im rti ne da a on im rti Te Pr l on Vu c Te Pr Se c Se Population: poverty status Educational attainment Population share, % Always poor Vulnerable to crisis Income change, year 2 (2014), % Never poor Source: Barriga et al. 2014. High secondary-school drop-out rates are a particular problem in Uru- guay. School-age children whose parents have lower levels of educational attainment are significantly less likely to attend school (Ferreira et al. 2013). The performance of Uruguay appears to be about average for the share of 11- to 12-year-olds who are in third grade or higher, but below average in the share of 15-year-olds in seventh grade or higher for the parental group with less than primary educational attainment. Children of parents with no education have only a 60 percent chance of attending school by age 15. This is especially concerning given the desire in the country for inclusive growth and shared prosperity. Many of these dropouts are not working either (figure 9.16). About 15 percent of all boys age 15–18 years and 17 percent of girls in the same age-group in 2012 were neither in school nor working. While the rate of youth neither in school nor working among girls in Uruguay is similar to the regional average, boys in Uruguay are significantly more likely to be out of school and out of work. This means these young people are not Chapter 9: Poverty and Shared Prosperity in Uruguay 321 Figure 9.16 15- to 18-Year-Olds Not in School and Not Working, by Gender, Uruguay and the Region, 2000–12 25 Share of age group by gender, % 20 15 10 5 0 0 02 03 04 05 06 07 08 09 01 12 10 11 0 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Region, girls Uruguay, girls Region, boys Uruguay, boys Source: Tabulations of Equity Lab, Team for Statistical Development, World Bank, Washington, DC, using data from SEDLAC and based on the methodology of De Hoyos, Popova, and Rogers 2015. accumulating human capital through either formal education or on-the-job experience. In addition, the quality of schooling in Uruguay is poor. Although test scores are better in Uruguay than in other countries in the region, they are well below the scores in OECD countries outside the region. Hanushek and Woessmann (2012) note that Uruguay has achieved the highest cognitive score in the OECD’s Program for International Student Assessment (PISA), 4.30, among all Latin American countries.10 In the student assessment tests in Uruguay in 2012, the share of 15-year-old students failing to achieve a score of 2 was 48 percent in science, 49 percent in mathematics, and 39 percent in reading.11 Furthermore, the World Bank (2014b) finds that there is a significant disparity between the scores of students who attend public school and the scores of students who attend private school. In mathemat- ics, for example, 79 percent of private school students achieved a grade of 2 or higher, double the rate of public school children, of whom only 35 percent scored 2 or higher. Similar gaps in achievement occur in sci- ence and reading. Because attendance at a private school is strongly cor- related with parental earnings and educational attainment, this gap in test scores suggests that the provision of good-quality education is inequitable 322 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean across socioeconomic groups. This will have significant impacts on inter- generational mobility and inequality and, potentially, the prospects for the nation’s economic growth. Expansion of health care and nutrition Health care expenditure is around 4.6 percent of GDP and covers direct public health care for people living in poverty and a subsidy available to contributors to the Fondo Nacional de Salud, an in-kind transfer to health care providers. Nutrition benefits (0.3 percent of GDP) target families in extreme poverty by providing lunches. Social workers evaluate the (renew- able) eligibility for the program, which lasts for up to 24 months. A food card program was launched in 2006 to offer households, pregnant women, and children under 18 years of age access to food and hygiene, conditional on school enrollment among children under 14, income below a threshold, and regular health care visits. Health care indicators show that Uruguay outperforms the region in key areas, though there is room for improvement. Indicators of access to repro- ductive and health services are generally positive. In 2009, almost all births (99.7 percent) were attended by skilled health staff, which is better than the average among countries at the same income level (96.0 percent) and coun- tries in the region (90.1 percent). Similarly, the maternal mortality ratio in 2010 of 29 per 100,000 live births was well below the regional average of 80. Adolescent fertility has also been declining; in 2012, it stood at 58.3 births per 1,000 15- to 19-year-old girls. Though lower than the regional average in 2011 of 68.1 births per 1,000, this is considerably higher than the world average (49.3 births per 1,000). Final Remarks Uruguay has demonstrated a remarkable ability to recover from a devastat- ing crisis that not only damaged exports and production, but also led to sev- eral years of high inflation, macroeconomic instability, and soaring poverty rates. With prudent monetary and fiscal policies and a broad expansion of social protection programs, the government engineered a return to precrisis growth and poverty rates, while reducing inequality and fostering conver- gence across population groups, particularly with regard to the bottom 40, and both pro-poor and inclusive growth. Uruguay leads other countries in the region in many indicators of social inclusion and prosperity and exhibits lower levels of corruption, crime, and environmental degradation. The substantive improvements in well-being, especially since 2007, are the result of a combination of favorable economic conditions and growth patterns that have led to more employment opportunities, as well as key policy reforms in the labor market and social protection that have ensured that the less well off have been able to contribute to the growth process and enhance living conditions. Chapter 9: Poverty and Shared Prosperity in Uruguay 323 Serious challenges must still be overcome. The country is exposed to internal and external risks that render certain population groups vulnerable to the risk of falling back into poverty. However, Uruguay has adapted well, as evidenced by the brevity of the drop in GDP associated with the recent global financial crisis. The government has demonstrated its willingness to respond to the needs of the population and has been efficient in developing organizational and administrative responses. Nonetheless, the expanded middle class is demanding better quality in services, especially in education. The quality of education is still significantly heterogeneous across the coun- try and correlates highly with the socioeconomic status among students. Secondary-school drop-out rates are high, especially among the less well off. Improving education is thus crucial not only to ensuring the sustain- ability of economic growth, but also to enhancing economic mobility across generations. The demographic transition associated with the aging popula- tion may yet test the sustainability of the fiscal system. Notes 1. Moderate poverty in urban areas dropped from 29.7 percent in 1990 to 17.8 percent in 2000. These rates are not comparable with the moderate poverty headcounts produced after 2001 because the methodology for measuring pov- erty was changed. See the note to figure 9.2. 2. At the beginning of the decade, Argentina and Brazil accounted for half of all Uruguayan trade. This situation has changed in recent years and China now tops the list, at 21 percent of the country’s trade. About one-third of Uruguay’s exports now go to Argentina and Brazil, and over half of tourism receipts and one-third of foreign direct investment originate from Argentina. These two countries affect Uruguay directly and indirectly by amplifying shocks from the rest of the world. (See IMF 2011.) 3. While the economy has certainly grown at a high rate and the country has been successful in reducing poverty and inequality, a more complete analysis of trends in the past decade requires an examination of the periods before and after the 2001–02 crisis to separate out growth spells, the impacts of the crisis, and the recovery. This is rendered difficult because of changes in household survey data both in methodology and coverage. Until 2005, the Encuesta Continua de Hogares (continuous household survey) excluded rural areas and towns of fewer than 5,000 inhabitants (around 20 percent of the total population), but, in 2006, it became representative at the national level (covering both urban and rural areas). This important methodological change was extrapolated only back to 2002; it would therefore be imprudent to compare poverty and inequality lev- els before and after that year. Additionally, because of the focus on urban areas prior to 2006, trends in crisis recovery can only be considered for urban areas. 4. These data correspond to Montevideo and urban areas with more than 5,000 inhabitants. 5. The thresholds correspond, respectively, to Ur$153 and Ur$765 per person a day in 2005 prices. 6. World Bank (2014b) and tabulations of Equity Lab, Team for Statistical Devel- opment, World Bank, Washington, DC, based on data of SEDLAC. The data 324 Shared Prosperity and Poverty Eradication in Latin America and the Caribbean on the regional poverty decomposition are based on the per capita interna- tional moderate poverty line of $4 a day. 7. However, the direct impacts of these changes on earnings inequality are unclear. Some estimates suggest that the rise in the minimum wage had only a minor effect on the distribution of earnings because of the low starting point of the minimum wage and the lack of compliance (Borraz and Pampillón 2011). Like- wise, the influence of centralized wage setting on earnings inequality has not been definitively established because of other confounding concurrent effects (Amarante et al. 2011). 8. Led by Nora Lustig, Commitment to Equity is a joint initiative of Tulane Uni- versity, New Orleans, and the Inter-American Dialogue, Washington, DC. The website is at http://www.commitmentoequity.org/. 9. In this exercise, the benchmark scenario assumes that real GDP will continue to grow, though at a slightly slower rate; inflation will remain high; and there will be no drastic change in employment (though unemployment may begin to decline). However, under a scenario replicating a crisis similar in importance to the crisis of 2001–02, real GDP would contract significantly, though less than in 2001–02; inflation would reach double digits; and the share of trade in GDP would decline, leading to a rise in the unemployment rate. 10. Set to the program’s test score scale, the cognitive score is the average test score in mathematics and science in all years in primary and secondary school. The highest score was 5.45 (Taiwan), and the lowest was 3.09 (South Africa). 11. A score of 2 is considered to represent the equivalent of a basic ability to apply the material to real-world situations. References Alves, Guillermo, Verónica Amarante, Gonzalo Salas, and Andrea Vigorito. 2012. “La desigualdad del ingreso en Uruguay entre 1986 y 2009.” Documentos de Trabajo 03–12, Instituto de Economía, Universidad de la República, Monte- video, Uruguay. Amarante, Verónica, Marisa Bucheli, Cecilia Olivieri, and Ivone Perazzo. 2011. “Distributive Impacts of Alternative Tax Structures: The Case of Uruguay.” Documentos de Trabajo 09–11, Instituto de Economía, Universidad de la República, Montevideo, Uruguay. Amarante, Verónica, and Andrea Vigorito. 2012. “The Expansion of Non- Contributory Transfers in Uruguay in Recent Years.” Policy Research Brief 29, International Policy Centre for Inclusive Growth, United Nations Development Programme, Brasília, Brazil. Barriga Cabanillas, Oscar, María Ana Lugo, Hannah Nielsen, Carlos Rodríguez- Castelán, and María Pía Zanetti. 2014. “Is Uruguay More Resilient this Time? Distributional Impacts of a Crisis Similar to the 2001/02 Argentine Crisis.” Policy Research Working Paper 6849, World Bank, Washington, DC. Borraz, Fernando, and Nicolas González Pampillón. 2011. “Assessing the Distrib- utive Impact of More than Doubling the Minimum Wage: The Case of Uru- guay.” Documentos de Trabajo 17–11, Instituto de Economía, Universidad de la República, Montevideo, Uruguay. Bucheli, Marisa, Nora Lustig, Máximo Rossi, and Florencia Amábile. 2014. “Social Spending, Taxes, and Income Redistribution in Uruguay.” In “Analyzing the Chapter 9: Poverty and Shared Prosperity in Uruguay 325 Redistributive Impact of Taxes and Transfers in Latin America,” ed. Nora Lustig, Carola Pessino, and John Scott, special issue, Public Finance Review 42 (3): 413–33. Datt, Gaurav, and Martin Ravallion. 1992. “Growth and Redistribution Com- ponents of Changes in Poverty Measures: A Decomposition with Applications to Brazil and India in the 1980s.” Journal of Development Economics 38 (2): 275–95. Dean, Andrés, and Andrea Vigorito. 2011. “La Población de Menores recursos en Uruguay. Una Caracterización a Partir del Panel MIDES-INE-UDELAR.” Unpublished working paper, Instituto de Economía, Universidad de la República, Montevideo, Uruguay. De Hoyos, Rafael E., Anna Popova, and F. Halsey Rogers. 2015. “Out of School and Out of Work: A Diagnostic of Ninis in Latin America.” Working paper, World Bank, Washington, DC. Ferreira, Francisco H. G., Julian Messina, Jamele Rigolini, Luis F. López-Calva, María Ana Lugo, and Renos Vakis. 2013. Economic Mobility and the Rise of the Latin American Middle Class. Washington, DC: World Bank. Hanushek, Eric A., and Ludger Woessmann. 2012. “Schooling, Educational Achievement, and the Latin American Growth Puzzle.” Journal of Development Economics 99 (2): 497–512. IMF (International Monetary Fund). 2011. “2011 Article IV Consultation.” Decem- ber, IMF, Washington, DC. https://www.imf.org/external/pubs/ft/scr/2011 /cr11375.pdf. INE (Uruguay, National Statistics Institute). 2013. “Estimación de la pobreza por el Método del Ingreso, Año 2012.” April, División Estadísticas Sociodemográficas, INE, Montevideo, Uruguay. Levy, Santiago, and Norbert Schady. 2013. “Latin America’s Social Policy Chal- lenge: Education, Social Insurance, Redistribution.” Journal of Economic Per- spectives 27 (2): 193–218. SEDLAC (Socio-Economic Database for Latin America and the Caribbean), Center for Distributive, Labor, and Social Studies, Universidad de La Plata, La Plata, Argentina; World Bank, Washington, DC, http://sedlac.econo.unlp.edu.ar/eng/ index.php. World Bank. 2014a. “Macro Monitoring Notes: Uruguay.” Country Management Unit for Argentina, Paraguay, and Uruguay, Poverty Reduction and Economic Management, Latin America and the Caribbean Region, World Bank, Washing- ton, DC. ———. 2014b. “Social Gains in the Balance: A Fiscal Policy Challenge for Latin America and the Caribbean.” Poverty and Labor Brief, Report 85162 rev (Feb- ruary), Latin America and the Caribbean Region, World Bank, Washington, DC. Environmental Benefits Statement The World Bank Group is committed to reducing its environmental foot- print. In  support of this commitment, the Publishing and Knowledge Division leverages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, green- house gas emissions, and waste. The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on FSC certified paper, with nearly all containing 50–100 percent recycled content. The recycled fiber in our book paper is either unbleached or bleached using Totally Chlorine Free (TCF), Pro- cessed Chlorine Free (PCF), or Enhanced Elemental Chlorine Free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4.