Riding the Wave An East Asian Miracle for the 21st Century WORLD BANK EAST ASIA AND PACIFIC REGIONAL REPORT Riding the Wave WO RLD BA NK E A S T A SIA A N D PACI FIC REGIO N A L REP O RT S Known for their economic success and dynamism, countries in the East Asia and Pacific region must tackle an increasingly complex set of challenges to continue on a path of sustainable devel- opment. Learning from others within the region and beyond can help identify what works, what doesn’t, and why, in the search for practical solutions to these challenges. This regional flagship series presents analyses of issues relevant to the region, drawing on the global knowledge and experience of the World Bank and its partners. The series aims to inform public discussion, policy formulation, and development practitioners’ actions to turn challenges into opportunities. T I T L ES IN T H E SE R IES Riding the Wave: An East Asian Miracle for the 21st Century Live Long and Prosper: Aging in East Asia and Pacific East Asia Pacific at Work: Employment, Enterprise, and Well-Being Toward Gender Equality in East Asia and the Pacific: A Companion to the World Development Report Putting Higher Education to Work: Skills and Research for Growth in East Asia All books in this series are available for free at https://openknowledge.worldbank.org​ /­handle/10986/2147 World Bank East Asia and Pacific Regional Report Riding the Wave An East Asian Miracle for the 21st Century © 2018 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 20 19 18 17 This work is a product of the staff of The World Bank with external contributions. The findings, interpre- tations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2018. Riding the Wave: An East Asian Mira- cle for the 21st Century. World Bank East Asia and Pacific Regional Report. Washington, DC: World Bank. doi:10.1596/978-1-4648-1145-6. License: Creative Commons Attribution 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 content ­ contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned indi- vidual 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 reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse 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 World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1145-6 ISBN (electronic): 978-1-4648-1176-0 DOI: 10.1596/978-1-4648-1145-6 Cover design: Debra Naylor, Naylor Design, Washington, DC Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Analytical approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Securing upward mobility: removing the constraints to inclusive growth. . . . . . . . . . . . . . . . 4 The elements of a new social contract to underpin inclusive growth . . . . . . . . . . . . . . . . . . . 8 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Structure of the report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 New thinking for a new era: economic classes, country groupings, and inclusive growth. . . . 14 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1 The Building Blocks of East Asia’s Economic Miracle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Policies that fostered rapid, labor-intensive growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Policies that built up basic human capital. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Limited social protection and redistributive policies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2 Ready for the Challenges of the 21st Century?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Concerns over the inclusiveness of the region’s development model are growing. . . . . . . . . 29 v vi   C O N T E N T S Factors that were once opportunities are turning into challenges. . . . . . . . . . . . . . . . . . . . . 39 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3 Looking beyond the Poverty Line: Economic Class in Developing East Asia and Pacific. . . 51 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 A picture of the income distribution by economic class . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 The distribution of economic classes across countries . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Dramatic income gains across economic classes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Different countries’ pathways to prosperity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Looking ahead: extrapolating past trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 4 Economic Class and the Challenges of Inclusive Growth. . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Securing upward mobility: removing constraints to inclusive growth. . . . . . . . . . . . . . . . . . 61 Eliminating extreme poverty. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Transitioning to economic security. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Matching economic security with high-quality public services. . . . . . . . . . . . . . . . . . . . . . . 71 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 5 A Policy Agenda for Inclusive Growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77  The three pillars: economic mobility, economic security, and institutions for inclusive growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 A new social contract: priorities across different country groups. . . . . . . . . . . . . . . . . . . . . 80 Fostering economic mobility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Increasing economic security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Strengthening the institutions for inclusive growth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 A Statistical Appendix: Poverty data availability for developing East Asia and Pacific. . . . . 105 B Statistical Appendix: Methodology for the estimation of internationally comparable poverty estimates for East Asia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 C Statistical Appendix: Supplementary tables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 D Statistical Appendix: A profile of economic class in developing East Asia and Pacific . . . . 121 Boxes 1.1 Socioeconomic progress in developing East Asia and Pacific. . . . . . . . . . . . . . . . . . . . . 22 1.2 Growth and poverty reduction in the Pacific Islands and Papua New Guinea. . . . . . . . 24 1.3 East Asia lags behind other middle-income regions in providing social protection. . . . . 25 2.1 Inequality can be damaging—and so can perceptions of inequality . . . . . . . . . . . . . . . 30 2.2 Beyond monetary inequalities: disparities in access to services and curtailed opportunities for future generations across developing East Asia and Pacific . . . . . . . . 32 2.3 China as a driver of inequality trends across developing East Asia and Pacific . . . . . . . 34 C O N T E N T S    vii 2.4 Growth of the super rich across developing East Asia and Pacific: economic dynamism or elite capture? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 2.5 Regulatory barriers to migration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.1 Partitioning the income distribution into classes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 3.2 Country-level economic class dynamics: some examples. . . . . . . . . . . . . . . . . . . . . . . . 58 4.1 Sources used to analyze mobility. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 4.2 Shocks in developing East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 4.3 Diverse barriers make escaping extreme poverty difficult . . . . . . . . . . . . . . . . . . . . . . . 66 5.1 Diverse countries, different challenges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 5.2 A policy agenda for the Pacific Island countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 5.3 Starting off strong through community-focused early childhood development in Indonesia. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 5.4 Creativity plus a little behavioral economics: rural savings in Mindanao . . . . . . . . . . . 88 5.5 National household targeting system in the Philippines (Listahanan). . . . . . . . . . . . . . 89 5.6 Guarding against health risks: Thailand’s success. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 5.7 (Almost) Ready for Winston . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 5.8 Institutions for effective urban governance: lessons from Malaysia. . . . . . . . . . . . . . . . 98 Figures O.1 Developing East Asia and Pacific, 2002–15: a wave of prosperity and the evolution of economic class in the region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 O.2 Population distribution by economic class and country, 2002 and 2015. . . . . . . . . . . . . 3 O.3 Class mobility transition matrix, short run. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 O.4 Extreme poverty in selected countries in East Asia and Pacific, US$1.90-a-day purchasing power parity (PPP), 2015. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 O.5 Nonmonetary poverty by class, 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.1 Real gross domestic product per capita in developing regions. . . . . . . . . . . . . . . . . . . . 18 1.2 Real gross domestic product per capita in developing East Asia and Pacific . . . . . . . . . 18 1.3 Poverty rates, across developing regions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 1.4 Poverty rates, across developing East Asia and Pacific. . . . . . . . . . . . . . . . . . . . . . . . . . 20 B1.1.1 Socioeconomic indicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.1 Inequality across developing regions of the world. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.2 Growth incidence curve and absolute consumption gains for aggregate developing East Asia and Pacific, 1988–2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.3 Regional composition of developing East Asia and Pacific distribution of consumption expenditure, 1988 and 2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 B2.3.1 Developing East Asia and Pacific region distribution of consumption expenditure, by subregion, 1988–2012. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 B2.3.2 Contribution of within- and between-country differences to inequality across developing East Asia and Pacific, 1988–2012. . . . . . . . . . . . . . . . . . . . . . . . . . . 36 2.4 Inequality across developing East Asia and Pacific, 2002–14. . . . . . . . . . . . . . . . . . . . . 37 2.5 Welfare disparities between rural and urban areas for select East Asia and Pacific countries: characteristics versus returns, 2000 and 2010. . . . . . . . . . . . . . . . . . 37 2.6 Income share of the top 1 percent and of the top 0.1 percent: selected economies of East Asia compared with the United States, 1980–2015 . . . . . . . . . . . . . . . . . . . . . . . . 38 2.7 Population aging across developing East Asia and Pacific . . . . . . . . . . . . . . . . . . . . . . . 42 B2.5.1 Number of migrants in Vietnam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 viii   C O N T E N T S 3.1 Population distribution by economic class in developing East Asia and Pacific, 2002–15. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.2 Distribution of extreme poverty across developing East Asia and Pacific, 2015. . . . . . . 53 3.3 Population distribution by economic class and country in developing East Asia and Pacific, 2002 and 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 3.4 Poverty projections, developing East Asia and Pacific, US$1.90 a day and US$3.10 a day PPP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.5 Population distribution by economic class in developing East Asia and Pacific, 2015–30. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 B3.2.1 Population distribution by economic class in China, Indonesia, the Philippines, and Vietnam, 2002–15. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.1 Short-run class mobility, aggregate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.2 Class mobility, by initial income group, short run. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 4.3 Nonmonetary poverty by class, developing East Asia and Pacific 2012. . . . . . . . . . . . . 71 4.4 Public and private health care facility use by economic class, developing East Asia and Pacific, 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.5 Private education by economic class, developing East Asia and Pacific, 2012 . . . . . . . . 73 Tables B1.3.1 Performance of social protection programs in different developing regions. . . . . . . . . . 25 3.1 Population distribution by economic class and country in developing East Asia and Pacific, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 B4.1.1 Data used to analyze short- and long-term mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 A.1 Available household surveys for countries in developing East Asia and Pacific . . . . . . 106 C.1 Growth rates in developing East Asia and Pacific, 1960–2015 . . . . . . . . . . . . . . . . . . 115 C.2 Tax revenue in developing East Asia and Pacific, percentage of GDP, 2013. . . . . . . . . 116 C.3 Access to basic services in developing East Asia and Pacific, 2015. . . . . . . . . . . . . . . . 117 C.4 Access to sanitation and improved water in developing East Asia and Pacific, rural versus urban, 2012 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 C.5 Percentage of self-reported membership in different organizations, by economic security status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 C.6 Percentage of self-reported lack of confidence in public institutions, by economic security status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 C.7 Indicators of financial inclusion, percentage of population age 15 and above, 2014. . . . 120 D.1 Key socioeconomic characteristics of the various economic classes. . . . . . . . . . . . . . . 122 Foreword D eveloping East Asia and Pacific has the income distribution, reflecting the grow- led the way in showing how rapid ing concentration of income and wealth at the and broadly shared growth can lift top, as well as limited access to basic services, millions out of poverty. Over the past quar- often of poor quality, even among the middle ter century, the wave of prosperity that has class. Newer challenges—such as rapid aging spread across the region has meant that in countries such as China, Thailand, and almost a billion people are no longer mired in Vietnam—and uncertain growth prospects, extreme poverty. Sixty percent of its citizens especially among the region’s lower-middle- are now economically secure in that they do income countries, are making it more difficult not face the risk of falling back into poverty. to ensure economic security for all. In most countries, a sizable middle class has Riding the Wave discusses how countries emerged and is growing rapidly. across the region can effectively confront However, there is no guarantee that inclu- these challenges to inclusive growth. In doing sive growth—growth that reduces extreme so, it looks at those who live in the region poverty while delivering upward mobility and through the lens of economic class, rang- economic security for all—is assured, even ing from the extreme poor at the bottom of for the countries of this successful region. the distribution to the middle class (which While the region is now composed entirely includes the rich) at the top. This partition, of middle-income countries, its diversity has in turn, helps provide a nuanced view of how increased. Progress in reducing poverty and the income distribution has evolved in various vulnerability and in promoting mobility has countries so as to highlight their policy priori- varied across countries. Extreme poverty, ties in ensuring inclusive growth. although limited, is now increasingly concen- Riding the Wave underscores how, beyond trated among specific groups, such as ethnic reducing poverty, upward economic mobil- minorities and those in remote locations. ity (improving one’s lot in life) and economic Roughly one-fifth of the region’s population security (being able to hold on to the gains remains at risk of falling into poverty, with made) are key features of inclusive growth. this proportion especially high in Indonesia The report illustrates how the constraints to and the Philippines. Prospects for upward securing upward mobility differ across eco- mobility are seen as increasingly elusive across nomic classes. For some segments, such as the ix x   F O R E W O R D remaining extreme poor, the key challenge is more economic security will require a focus very much about achieving mobility or mov- on strengthening social assistance systems, ing up the distribution. For other classes, such expanding social insurance, and increasing as the vulnerable, the constraint that needs resilience to systemic shocks. For the state to to be addressed is about enhancing economic be better able to deliver on this agenda for security so that they do not fall back into pov- inclusive growth will require relying more on erty. For those who are economically secure progressive taxation to raise public resources, or part of the middle class in these coun- promoting fairer competition, making inclu- tries, enhancing their access to quality public sive spending programs more effective, ­ services is the key. adapting to the pressures of rapid aging, and The report outlines how policy makers in fostering efficient urbanization. countries across East Asia need to approach This policy agenda, while ambitious, could these challenges in a selective and priori- form the basis for a new social contract that tized manner. The three pillars it highlights will allow governments in this region to for policy are fostering economic mobility, address the needs of all groups in their popu- enhancing economic security across economic lations while remaining fiscally responsible in classes, and strengthening the capacity of the the face of an uncertain global environment. state to deliver. Success in ensuring inclusion will mean that Although the relative emphasis across developing East Asia and Pacific can lay claim ­ pillars and the policy details will differ across to another miracle well into the 21st century. countries, many elements will be common. To promote economic mobility, addressing Victoria Kwakwa gaps in access to jobs and services, improv- Regional Vice President ing the quality of jobs, and deepening finan- East Asia and Pacific Region cial inclusion are important. To provide The World Bank Acknowledgments T his report is a joint product of the Office Uochi, Daniel Walker, Darendhra Wardhana, of the Chief Economist, East Asia and Laura Wijaya, and Judy Yang. Pacific Region and the Poverty and Additional background notes were Equity Global Practice. It was authored by ­ produced by Travers Barclay Child of Vrije Caterina Ruggeri Laderchi, Nikola Spatafora, Universiteit Amsterdam, Carol Graham of Sudhir Shetty, and Salman Zaidi, with the the Brookings Institution and the University guidance of Axel van Trotsenburg and of Maryland, College Park, and Soumya Victoria Kwakwa, Regional Vice Presidents, Chattopadhyay of the Overseas Development East Asia and Pacific, and Ana Revenga and Institute in London; Erwin Tiongson and Carolina Sanchez-Paramo, Senior Directors of Madita Wiese from Georgetown University; the Poverty and Equity Global Practice at the and Hai-Anh Dang and Sebastian James from time of the report preparation. the World Bank. Mildred Gonsalvez and The report synthesizes the work of a larger Cecile Wodon provided invaluable adminis- World Bank team that included Andrew Beath, trative and logistical assistance. Wendy Cunningham, Reno Dewina, Carolina The team benefited from valuable sugges- Diaz-Bonilla, Samuel Freije-Rodriguez, tions provided throughout the production Christoph Lakner, Nancy Lozano-Gracia, process of this report, by colleagues both Ririn Salwa Purnamasari, Matthew Grant within and outside the World Bank Group. Wai-Poi, and Andrea Woodhouse. In addi- Special thanks for their participation in an tion, major contributions to the background inception workshop goes to Caroline Freund work were made by Sarah E. Antos, Brendan of the Peterson Institute, Ricardo Fuentes- Brian Brady, Dorothee Buhler, Imogen Cara Nieva of Oxfam International, Ravi Kanbur Halstead, Leander Heldring, Yumeka Hirano, of Cornell University, Homi Kharas of the Alejandro Huertas, Taufik Indrakesuma, Brookings Institution, and Erwin Tiongson La-Bhus Fah Jirasavetakul, Vera Kehayova, of Georgetown University, together with Jae Kyun Kim, Chun Kuan, Barbara Parker, Shubham Chaudhuri, Uwe Deichmann, Obert Pimhidzai, Manohar Sharma, Yan Sun, Indermit Gill, Andrew D. Mason, Ana Pham Thu Trang, Michele Tuccio, Ikuko Revenga, Carolina Sanchez-Paramo, Carlos xi xii   A C K N O W L E D G M E N T S Silva-Jauregui, and Renos Vakis from the James Anderson, Rocio Castro, Ted Chu, World Bank. Several World Bank colleagues Chorching Goh, Camilla Holmemo, and Tara provided guidance and advice at different Vishwanath from the World Bank Group. stages of the process, including Judy Baker, The team is also very grateful to Alejandro John Giles, Tobias Haque, Steve Jaffee, Cedeno-Ulloa and Livia Pontes-Fialho from Somik Lall, Branko Milanovic, Nataliya the World Bank’s East Asia and Pacific Mylenko, Ambar Narayan, Philip O’Keefe, External Relations group for their assistance Douglas Pearce, Bob Rijkers, Emmanuel with the outreach and dissemination. Mary Skoufias, Mauro Testaverde, and Rogier van Fisk and Patricia Katayama provided excel- den Brink. lent advice and guidance on the publication Professor Hal Hill of the Australian process. National University, Dr. Somchai Jitsuchon of The cover was designed by Debra Naylor, the Thailand Development Research Institute, Washington, DC. The graphics were designed Professor Shantong Li of the Development by Kristen Dennison. Andrew Johnston and Research Center, State Council, China; and Publications Professionals edited and copy- Gabriel Demombynes and Aart C. Kraay edited the report, respectively. from the World Bank were peer reviewers. The team also thanks others who have Comments were also received during the helped prepare this report and apologizes to review process from Valerie Kozel of the those who may have been overlooked in these University of Wisconsin–Madison; and from acknowledgments. Abbreviations BPS Central Bureau of Statistics (Indonesia) EAP East Asia and Pacific ECA Europe and Central Asia ECD early childhood development ECED early childhood education and development FDI foreign direct investment FIBOS Fiji Island Bureau of Statistics FIES Family Income and Expenditure Survey (the Philippines) GDP gross domestic product GEP growth elasticity of poverty HIES Household Income and Expenditure Surveys HIS Household Income Survey (Malaysia) HSES Household Socio-Economic Surveys (Mongolia) IMF International Monetary Fund LAC Latin America and the Caribbean LECS Lao Expenditure and Consumption Surveys LICs lower-income countries LMICs lower-middle-income countries LPI Logistics Performance Index MNA Middle East and North Africa MS Master Sample MSF Master Sampling Frame NSO National Statistics Office OECD Organisation for Economic Co-operation and Development PICs Pacific Island countries PNGHS Papua New Guinea Household Survey xiii xiv   A B B R E V I A T I O N S PPP purchasing power parity PPS Probability Proportion to Size SAR South Asia s.d. standard deviation SEED Save, Earn, and Enjoy Deposits SES Socio-Economic Survey (Thailand) SMEs small and medium enterprises SSA Sub-Saharan Africa SUSENAS National Socio-Economic Survey (Indonesia) TLSLS Timor-Leste Survey of Living Standards (second national) TLSS Timor-Leste Living Standards Survey (first national) TOD transit-oriented development UCS Universal Coverage Scheme UMICs upper-middle-income countries VHLSS Vietnam Household Living Standards Surveys VIP ventilated improved pit VLSS Vietnam Living Standards Survey VoIP voice over IP WLL wireless local loop Overview Introduction growth will continue to deliver unprecedented improvements in welfare. Current trends sug- The East Asian experience has come to sym- gest that more deliberate policy efforts will be bolize how growth that is both rapid and needed to meet such expectations. broadly shared can improve the lives of mil- lions of people.1 Over the past two decades, a wave of rising prosperity lifted more than Analytical approach 40 percent of the region’s population out of To identify a policy agenda to support inclu- poverty. This remarkable experience reflected sive growth in the region, this report adopts a growth model that has spread through an analytical framework with three main fea- much of the region over the past half century. tures. First, it goes beyond a binary classifica- Policies that aimed to promote labor-­ intensive tion of the population into poor and nonpoor, growth and investments in human capi- offering instead a more nuanced discussion tal reduced poverty rapidly with little or no of the entire income distribution. Specifically, increase in ­inequality—the East Asian “mira- it separates households into five economic cle” of growth with equity. This approach was classes (figure O.1): the extreme poor, the popularized by the 1990 World Development moderate poor, the economically vulnerable Report as the “two-and-a-half-point strategy” (those at high risk of falling into poverty), the for poverty reduction (World Bank 1990). economically secure, and the middle class.2 It emphasized two pillars—labor-intensive Second, this analysis of economic class and growth and the accumulation of basic human its evolution is used to highlight similarities capital—while putting limited emphasis on a and differences between countries. Countries third pillar, social protection, which was seen are grouped into categories that reflect how as needed only by those too disadvantaged to their income distributions have evolved, to participate in the growth process. identify their policy priorities. Finally, the Riding the Wave is about sustaining this report adopts a specific definition of inclusive progress so that the lives of millions in the East growth. Economic growth is characterized as Asia and Pacific region continue to improve inclusive if it reduces poverty and enhances despite the challenges that remain. Past suc- economic mobility and security across all cess means that expectations are high that parts of the income distribution.3 1 2   RIDING THE WAVE FIGURE O.1  Developing East Asia and Pacific, 2002–15: a wave of to the regional average. Third, the share of the prosperity and the evolution of economic class in the region economically vulnerable has remained almost constant over this period, at roughly a quarter 100 of the total population of the region. The pop- 90 ulations in this class in China, Indonesia, the Percentage of total population 80 Philippines, and Vietnam are large. Indonesia 70 and the Philippines stand out because their 60 share of economically vulnerable in the popu- 50 lation is much higher than average, whereas 40 China’s and Vietnam’s are very close to the 30 regional average. 20 The country groupings that emerge from looking at the evolution of economic 10 classes between 2002 and 2015 include the 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 following: Year •  Progressive prosperity countries (Malaysia Extreme poor (less than PPP US$1.90 per day) and Thailand), which have eliminated Moderate poor (PPP US$1.90–US$3.10 per day) income poverty while substantially increas- Economically vulnerable (PPP US$3.10–US$5.50 per day) ing the share of middle-class households Economically secure (PPP US$5.50–US$15.00 per day) Middle class (PPP US$15.00 and higher per day) •  Out-of-poverty-into-prosperity countries (China, Mongolia, and Vietnam), where Note: PPP = purchasing power parity. most people are at least economically secure and the middle class is growing •  O u t - o f - e x t r e m e - p o v e r t y c o u n t r i e s A look at the evolution of economic classes (Cambodia, Indonesia, and the Philippines), at the country level between 2002 and 2015 where extreme poverty levels are low, but highlights three key facts (figure O.2). First, so are the shares of the middle class in 2015, the economically secure and the mid- •  Lagging-progress countries (Lao People’s dle class in developing East Asia and Pacific Democratic Republic and Papua New accounted for nearly two-thirds of the region’s Guinea), which still have high levels of population, a significant increase from their extreme poverty share of just over one-fifth of the population •  Pacific Island countries, which are a het- in 2002. Because of its size, China accounts erogeneous group and are distinct from for the largest proportion of the two classes. the rest of the region Compared with the regional average, China, along with Malaysia and Thailand, also has a The definition of inclusive growth used proportionately larger share of its population in this report allows it to focus on the con- in these two classes. This contrasts with the straints that are relevant to each economic larger middle-income countries—Indonesia, class. In turn, this focus helps delineate the the Philippines, and Vietnam—which all have elements of a new social contract that could a proportionately lower share of their popula- allow countries in the region to sustain inclu- tions in this group. Second, reflecting the rapid sive growth. Such a focus on inclusive growth progress in reducing poverty across much of is motivated by the realization that the well- the region, the shares of the extreme and mod- tested approach to “growth with equity” that erate poor have fallen dramatically—from the region has relied on will be more diffi- more than half the population in 2002 to less cult to implement than in the past and will than an eighth in 2015. China, Indonesia, and prove less successful in addressing the needs the Philippines together account for the bulk of many in the population. of this group, with the populations in the lat- For sustaining growth, export-oriented ter two disproportionately represented relative manufacturing appears likely to be a less O verview  3 FIGURE O.2  Population distribution by economic class and country, 2002 and 2015 100 90 80 Percentage of country’s population 70 60 50 40 30 20 10 0 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 cambodia china Indonesia lao PDR malaysia mongolia Papua Philippines Thailand Vietnam New Guinea Extreme poor moderate poor Economically vulnerable Economically secure middle class Source: World Bank EAP Team for Statistical Development. powerful engine than in the past. For the first by 2025, the region’s working-age population time since the early 2000s, growth in global is expected to begin to decline. From a policy trade has slowed below the pace of growth point of view, the key concern is that much of countries’ gross domestic product, and the of the region is “getting old before g ­ etting potential for the region to integrate further rich” (World Bank 2016, 9): social insurance through the development of new value chains systems have not developed enough and accu- is not clear. Lower-income countries also face mulated enough resources to cover the aged continued competition from China, while population with sufficient incomes to avoid they do not have access to some of the export- poverty. Even where specific programs for promoting measures used by countries that the poorest elderly are provided, such as in industrialized successfully in the past. Higher- China, program coverage will have to expand income countries, despite their comparatively significantly given demographic trends. low wages, see manufacturing shift toward Urbanization, which underpinned broad- skilled labor and automation—developments based improvements in living standards for that are unlikely to support broad-based the early developers in the region—such as growth unless they are well managed. the Republic of Korea and Taiwan, China— Adapting to longer-term trends such as the and helped support an expansion of basic aging of the population and urbanization, services, has now turned into a challenge which have been sources of opportunity in for other economies. Rapid urban growth is the past, might require further adjustments being accompanied by growing congestion, to the region’s growth model. In the past, a environmental strains, and urban squalor large share of people of working age pro- because market and policy failures remain vided a demographic dividend that helped unaddressed. Urbanization can also increase support the region’s growth. In most coun- exposure to natural disasters, because most tries, this demographic dividend will turn large cities in the region are in low-lying into a demographic tax on growth because coastal areas. Further, urbanization may 4   RIDING THE WAVE erode traditional support networks, com- security (being able to hold onto the gains pounding the challenges posed by the aging made) are defining features of inclusive populations and reducing living standards growth. Indeed, perceptions of growing on both sides of the urban-rural divide. As inequality are closely related to the sense that people move to cities to take advantage of avenues for moving up are closing for those the opportunities and services that cities pro- who are not already well placed. Although vide, and as the poor become increasingly systematic evidence on mobility is hard to concentrated in those urban areas, a key come by, this report builds on new longitu- policy priority will be investing in cities to dinal analysis based on data obtained follow- make them livable and effective sources of ing households over time, either panel data opportunities for all. or approximations of panel data (synthetic Recent developments have made groups panels).4 This evidence covers the period from such as migrants, the elderly, and the unskilled the early 2000s though the years covered vary, more vulnerable to shocks, while other groups and the lengths of the panels vary by country. are experiencing unprecedented affluence. In Given developing East Asia and Pacific’s combination with challenges to the region’s high growth and rapid poverty reduction, it export-oriented development model, the trend can be expected that most households expe- helps to explain why inequality is seen to be rienced increases in income and that many rising and is indeed high in many countries. experienced upward mobility across classes. Those perceptions, even if not always borne The analysis confirms this upward mobil- out by the data, coincide with a sense that ity but also shows a lot of heterogeneity by income and wealth are now becoming more country and economic class (figure O.3). concentrated in the hands of the wealthiest. Overall, households other than those in the As a result, upward income mobility—a hall- middle class experienced increases in con- mark of the East Asian development expe- sumption. In addition, for at least one in rience—is now seen to be more elusive, a five people, consumption increased so much perception that is reinforced by the gaps in that they moved to a higher economic class. service delivery, including in higher education That upward mobility appears to be higher in and in water and sanitation. the longer term than in the short term, and A final challenge to the sustainability of it is generally higher than in Latin American the existing model of broad-based growth is countries for which comparable analysis that the region has become much more differ- exists. Nevertheless, some countries such as entiated than it was even three decades ago. the Philippines saw little mobility. Economic success has transformed it from a Despite this good news, most households region largely of poor people in low-income did not move across economic classes, and in countries to a region of middle-income all classes, some people moved downward. countries, each characterized to a varying ­ The downward mobility reflects both ex ante extent by economic classes that range from exposure to risks and limited ex post ability the extreme poor to the middle class and the to manage adverse shocks, despite significant very wealthy. That reality makes it very hard resilience. Shocks include economic shocks, to generalize about one model that can bring which affect prices and returns to household inclusive growth for all. assets such as labor or land or skills, personal shocks such as illness, and natural shocks. The moderate poor and the economically Securing upward mobility: vulnerable are most likely to move down the removing the constraints to class ladder when affected by shocks. Even inclusive growth the middle class, despite its relative affluence, Riding the Wave makes the case that beyond faces risks. Conversely, the moderate poor reducing poverty, upward economic mobil- who do not receive a shock are more likely ity (improving one’s lot in life) and economic than those who do to move up one or two O verview  5 FIGURE O.3  Class mobility transition matrix, short run a. Lao PDR, transition matrix, 2010–12 b. Indonesia, transition matrix, 2008–10 2010 2012 2008 2010 1% 3% 1% 2% 7% 9% 15% 22% 25% 27% 28% 31% 41% 36% 32% 28% 26% 25% 23% 17% c. Philippines, transition matrix, 2007–09 d. Vietnam, transition matrix, 2010–12 2007 2009 2010 2012 5% 5% 16% 18% 27% 28% 45% 47% 29% 30% 23% 25% 24% 24% 11% 9% 14% 8% 3% 4% Extreme poor Moderate poor Economically vulnerable Economically secure Middle class Source: Cunningham 2016. 6   RIDING THE WAVE classes over a two-year period. More detailed The key constraints to securing upward analysis, however, highlights a lot of resilience mobility for each segment of the income dis- in the face of shocks, with even households tribution are discussed below, focusing on affected by shocks capable of moving to a one class at a time. For some classes, such higher economic class. Ultimately what makes as those stuck in extreme poverty, the chal- a difference is not whether households experi- lenge is very much about achieving mobility, ence shocks, but whether they have the asset whereas for others, such as the economi- base they need to avoid sliding into poverty. cally vulnerable, being able to secure what The policy tools countries need to secure they have is as important as eliminating con- upward mobility include helping households straints to moving up. accumulate assets (for example, by investing in human capital or providing access to credit to The challenge of eliminating invest in physical capital) and preventing the extreme poverty erosion of assets at times of crisis (for exam- ple, preventing children from being taken out The “last mile” challenge to eradicate extreme of school to help sustain the family). Social poverty can be seen as a special example of protection can play a key role in this respect, low mobility. Although extreme poverty is no by reducing poverty and offering all income longer a challenge for the region as a whole groups a way of managing risks. Countries in and for most countries, it remains concen- the region differ widely in the range of social trated in the poorer countries and, in some protection programs they have in place, and richer ones, among specific population groups, most are starting relatively late to put in place such as ethnic minorities and those in geo- a modern social protection system. Thus they graphically remote areas. In 2015, the extreme have an opportunity to learn from interna- poor accounted for less than 3 percent of tional experiences and to avoid pitfalls such developing East Asia and Pacific’s population. as disincentives for beneficiaries to graduate However, the share of the extreme poor ­ varies from the programs (so-called welfare traps) considerably across countries (figure O.4). and excessive taxation of labor incomes. In Lao PDR, Papua New Guinea, and the Solomon Islands, the extreme poor still make up a sizable share of the population. In other FIGURE O.4  Extreme poverty in selected countries in East Asia and Pacific, US$1.90-a-day purchasing power parity (PPP), 2015 countries, specific groups or locations account for the bulk of the extreme poor. In Vietnam, for instance, ethnic minorities now account Solomon Island for more than two-thirds of the remaining Papua New Guinea extreme poor, most of them concentrated in remote and mountainous areas. The same Lao PDR is true in China, and the remaining poor in Indonesia tend to be concentrated in the Fiji remote regions of Papua and Kalimantan. China In East Asian countries where economic progress has been lagging, as well as in most Vietnam Pacific Island countries, broad-based growth that creates demand for low-skilled labor Philippines and is complemented by investments to make Indonesia people more productive will still be effective in eradicating extreme poverty. But where 0 5 10 15 20 25 30 extreme poverty is associated with character- Percentage of households istics such as ethnicity and remote location, Source: World Bank EAP Team for Statistical Development. growth alone is unlikely to reduce poverty in Note: See appendix A for details on the sources. all its dimensions. O verview  7 Addressing the last-mile challenge of eradi- muted over a 10-year span, compatible with cating extreme poverty requires identifying an overall upward trend in incomes over time. and tackling the barriers that prevent spe- The economically vulnerable also show cific groups in the population from benefiting many signs of dynamism. For example, they from the economic opportunities available to have the ability to save to purchase small the rest. Such barriers can differ by context assets, such as cell phones or motorcycles. And and group and can include physical barriers, most of this group’s employment is through such as when communities are not served by small and medium-size enterprises, where an all-weather road; administrative barriers, they are often the entrepreneurs. This group is such as when isolated communities face high clearly resourceful in the face of missing pub- costs to obtain documents such as birth reg- lic goods: in Indonesia’s slums, for example, istrations; legal barriers, such as when unreg- they have organized early warning systems istered migrants cannot access services; and for flooding to make up for the deficiencies in cultural barriers, such as when there is preju- those provided by the government. dice or outright discrimination toward ethnic The lack of services and of systematic minorities. mechanisms to manage risks or cope with the consequences of shocks hinders this group’s progress, however. They struggle with a low The challenge of transitioning to asset base, including limited education and economic security for those living scarce productive assets, because they lack close to the poverty line mechanisms to save for bigger items. Where Despite the wave of prosperity that has they are concentrated in rural areas, but reshaped the region since 2002, about one- often also in urban ones, they suffer from quarter of the population continues to be poor health because of the lack of quality economically vulnerable (figure O.1). Such health services and limited access to public households face a high likelihood of falling infrastructure. Weaknesses in the business back into poverty, and they have only limited climate, including insecure property rights ability to move to economic security, espe- and limited access to electricity and other cially over the short run. infrastructure, constrain their businesses This group is not static, however, because from expanding. And in many countries, a lot of movement occurs across the moderate women’s participation in the labor force is poverty line in both directions. For example, particularly low among the moderate poor in Indonesia, 21 percent of those considered and economically vulnerable, in part because economically vulnerable in 2008 were found women’s domestic labor compensates for the to be in moderate poverty in 2010, whereas lack of services such as clean water, child- 35 percent of the moderate poor climbed care, and eldercare. into the economically vulnerable class. In the Philippines, between 2007 and 2009, The challenge of matching 22 percent of the economically vulnerable economic security with fell into moderate poverty, while 17 percent high-quality public services of the moderate poor climbed above the pov- erty line to become economically vulnerable. The economically secure and middle classes, In Vietnam, between 2010 and 2012, only which have both grown considerably across 7 percent of the economically vulnerable slid the region in numbers as well as in popula- into moderate poverty, and 46 percent of the tion share, are at low risk of sliding back into moderate poor became economically vulnera- poverty. In many countries, however, their ble (Cunningham 2017). Longer-term mobil- economic security has not been accompanied ity estimates for Indonesia and Vietnam show by appropriate provision of public services, that slippages from the economically vulner- let alone high-quality services, in contrast able group to the moderate poor were more with what has happened in other parts of 8   RIDING THE WAVE the world at similar income levels. Despite national policy dialogue on service delivery improved income, health, and education out- at a time when their higher aspirations could comes, one-third of the region’s economically spur improved quality of services. Across secure and 15 percent of its middle class (out- countries and at different levels of develop- side of China) lack access to one or more of ment, there are already signs that such opting the following services: good quality housing, out may be beginning. The middle classes, in clean water, and to a lesser degree, modern particular, are relying significantly on private sanitation (figure O.5). health care and to a lesser extent on private Similarly, in countries where risk man- primary and junior secondary education. agement tools are underprovided, even the On some issues, however, the middle middle classes are at risk of sliding down classes have proved to be powerful catalysts the income ladder. Although the middle for change. Their push for better consumer classes benefit from the social protection mea- safety regulations and better goods and ser- sures such as health insurance and pensions vices can benefit all households. For instance, that often come with secure jobs, they are recent food safety scandals in China have led not immune to systemic shocks. In Thailand’s to stronger regulations and better products floods in 2011, for example, 70 percent of for all. In addition, given these groups’ own the estimated US$47 billion of the damages vulnerability to shocks, they could have a and losses were in the manufacturing subsec- stake in developing an appropriate system for tor, in which many middle-class workers are managing risks. employed. Countries that fail to deliver on some of these classes’ needs and priorities risk their The elements of a new social turning to an insular world of private service contract to underpin inclusive provision. They may be disengaging from the growth An extraordinary wave of prosperity has improved the lives of millions across East FIGURE O.5  Nonmonetary poverty by class, 2012 Asia over the past half century. To sustain that success and address the new challenges 90 that have emerged, governments need to Percentage of households in each income group 80 address the constraints that different groups 70 in the population now face. Riding the Wave 60 identifies a selective policy agenda with three pillars: fostering economic mobility, 50 providing greater economic security, and 40 strengthening the institutions required for 30 inclusive growth. 20 Fostering economic mobility will require an agenda that focuses primarily on three areas: 10 0 •  Addressing existing gaps in access to jobs No access to No access to Poor-quality Poor on at least and services by reducing barriers to female clean water proper sanitation housing one dimension labor force participation and simplify- Extreme poor Moderate poor Economically vulnerable ing or eliminating regulations for obtain- Economically secure Middle class ing identity documents that are needed to access services or government programs. Source: Wai-Poi and others 2016. Note: Access to clean water denotes piped water, a protected well, or water sold by a vendor. Access •  Improving the quality of jobs available by to proper sanitation denotes a flush toilet or an improved pit latrine. Good housing quality means defining property rights over agricultural the roof is concrete, cement, brick, stone, wood, or tiles; the walls are concrete, cement, brick, or stone; the floor is nonsoil. The sample includes Cambodia, Indonesia, Lao PDR, Mongolia, the land more clearly and securely, reducing Philippines, Thailand, and Vietnam. regulatory barriers to rural-urban mobility, O verview  9 improving rural connectivity, and easing property taxes; and by enhancing tax the process of formalization for microen- administration. terprises and small-scale firms. •  Increasing the effectiveness of pro-inclusion •  Furthering financial inclusion by strength- spending programs by creating better man- ening legal and regulatory frameworks agement tools, including social registries that balance financial stability, integrity, (that is, databases of individuals and their and inclusion; improving information socioeconomic situations) and improved flows needed to assess creditworthiness targeting mechanisms. and by expanding financial literacy, espe- •  Leveling the playing field by boosting cially for those at the lower end of the eco- competition, reducing opportunities for nomic spectrum. capture in the more regulated sectors of the economy, and tackling nepotism and Providing greater economic security will corruption. call for a focus on strengthening social assis- •  Adapting to population aging by support- tance and social insurance and on increasing ing lifelong learning to upgrade workers’ resilience to large systemic shocks: skills throughout the life cycle, eliminating •  Strengthening social assistance by intro- incentives for older workers to retire too ducing cash transfer programs (condi- early, opening up aging labor markets to tional or not) in the lower-income and greater inflows of young immigrants, and lower-capacity countries where they do adapting the health sector to the challenges not yet exist and better targeting exist- of aging, including the increased incidence ing programs to the population groups of noncommunicable diseases. that remain overrepresented among the •  Promoting efficient urbanization by sup- extreme and moderate poor. porting efficient land use in cities of all •  Strengthening social insurance by expand- sizes and by strengthening urban gov- ing more systematically the coverage of ernance to improve service delivery measures to help households across the and urban infrastructure, including by income distribution insure against idio- improving coordination between differ- syncratic risks (such as through pension ent levels of government and different schemes) while reducing their fragmenta- jurisdictions. tion to promote labor mobility; and by The policy agenda will need to be tailored expanding the revenue base to sustain the to each country’s circumstances because not system’s expansion. all the pillars are important for each of the •  Increasing resilience to large systemic economic classes. For the previously identi- shocks by complementing household- fied country typology, the following priorities level risk management measures with may be identified: country-level mechanisms for managing ­ risks ex ante (such as market-based crop •  For lagging-progress countries, acceler- insurance) and coping with them ex post ating poverty reduction is the priority. (such as adaptive or scalable safety nets). Key policy actions in countries such as Lao PDR and Papua New Guinea should Strengthening the institutions required focus, first, on reducing poverty more for inclusive growth will involve focusing on quickly—by strengthening basic education several aspects: and financial inclusion, removing barri- •  Mobilizing resources through progressive ers to labor reallocation from agriculture taxation by strengthening personal income to more productive sectors, and tackling taxation, including by broadening the tax obstacles that prevent the marginalized base, reviewing marginal tax rates, and from accessing public services. Second, eliminating the preferential treatment of policies should focus on helping the vul- capital income; by establishing effective nerable manage the impact of shocks by 10   RIDING THE WAVE strengthening social assistance and invest- These priorities require improving the ing in making economies more resilient access of migrants to public services, to disasters. enhancing pension systems in a sustainable •  For out-of-extreme-poverty countries, the manner, and ensuring ­ high-quality, cost- priority is securing progress and foster- effective health care and education. In ing upward mobility, while tackling the China, it is also critical to reform urban remaining pockets of extreme poverty. planning to promote efficient land use and Cambodia, Indonesia, and the Philippines environmental sustainability. have reduced levels of extreme poverty to •  For progressive prosperity countries, the single digits. Unlike other countries that priority is to satisfy the middle classes’ now enjoy low levels of extreme poverty, growing aspirations while mobilizing however, the size of the middle class is resources to address remaining disparities. also in the single digits. The priorities for The richer countries, which include these countries, therefore, should reflect Malaysia, Mongolia (at a lower income the needs and aspirations of a diverse level), and Thailand, have continued to population, spanning those in moder- move large numbers of the economically ate poverty, the vulnerable, and those in vulnerable into economic security, and economic security. Integrating existing they are now characterized by large mid- social assistance programs into a stron- dle (and rich) classes. The first priority is ger social protection pillar is a key pri- to satisfy the middle classes’ growing aspi- ority in these countries, with strengthen- rations and demands for quality public ing mobility a close second. Improving services, including health and education, mobility will be particularly important which will support continued upward in the Philippines, where the distribu- mobility. The second, and related, prior- tion has been static until recently. Given ity is to implement reforms that narrow the size of their populations, however, rural-urban and cross-regional dispari- Indonesia and the Philippines cannot ties in economic and social development neglect the large numbers of extreme in these three countries, which in turn poor to which they are still home. These will require increased or more efficient countries need to identify and remove the resource mobilization. very specific constraints the extreme poor •  For the Pacific Island countries, given their face in accessing economic opportunities specificities, the inclusive growth agenda and service delivery. needs to be especially tailored. Because •  For out-of-poverty-into-prosperity coun- human resources will remain their princi- tries, the priority is removing the remaining pal resource, these countries’ priorities for vulnerabilities while meeting the needs and fostering economic mobility include using aspirations of the middle class. In China existing opportunities better, seeking access and Vietnam, sustained poverty reduction to international job markets, support- has led to the emergence of a middle class. ing managed urbanization, and improv- The first priority is to reduce residual vul- ing agricultural productivity. Priorities nerabilities, including among rural house- to strengthen economic security include holds and groups with special needs (in investing in disaster loss prevention, China, the elderly and children left behind mitigation, and managing and improv- when their parents migrate; in Vietnam, ing health and other social programs. ethnic minorities). The second priority, Finally, delivering on inclusive growth which is increasingly important, is to meet will require improving the effectiveness of the needs and aspirations of the middle public expenditures and service provision, class. Simultaneously, these econo- addressing the challenges of remoteness, mies must prepare for rapid aging. and mobilizing development aid. O verview  11 Implementing this policy agenda will not poor—between US$1.90 and US$3.10; be easy for any of these groups of countries, economically vulnerable—between US$3.10 particularly in a global context that is likely to and US$5.50; economically secure—between remain challenging. Nevertheless, this agenda US$5.50 and US$15.00; and middle class— more than US$15.00. could form the basis for a new social contract 3. The report does not refer explicitly to the if governments address the priority needs of concept of shared prosperity because the each economic class while they remain fiscally term is generally defined on a country-specific responsible. To accomplish that, countries basis. The analysis focuses on the entire must mobilize additional public revenues in region rather than on country-by-country an efficient and equitable manner. In so doing, developments. Nevertheless, the concept of the countries of developing East Asia and the inclusive growth captures the broad intuition Pacific could ride the wave of prosperity and behind shared prosperity. continue to improve the lives of their citizens 4. Synthetic panels rely on a methodology well into the 21st century. of matching households by time-invariant characteristics and their distributions across two comparable cross-sectional data sets. Notes In other words, rather than following the same household over time as in a real panel, 1. Although this statement refers to a broad they match observed data in one survey definition of East Asia and Pacific—comprising with predictions based on the information the developing countries of East Asia and contained in the second survey. Pacific, Japan, and the newly industrialized economies—this report’s main area of interest is developing East Asia and Pacific that comprises Cambodia, China, Indonesia, Lao References People’s Democratic Republic, Malaysia, Cunningham, W. 2016. “Class Mobility.” Mongolia, Myanmar, Papua New Guinea, the Background paper for Riding the Wave: An Philippines, Thailand, Timor-Leste, Vietnam, East Asian Miracle for the 21st Century . and the Pacific Island countries. Throughout Washington, DC: World Bank. this report, the Pacific Island countries Wai-Poi, M., R. Purnamasari, T. Indrakesuma, comprise Fiji, Kiribati, the Marshall Islands, the I. Uochi, and L. Wijaya. 2016. “East Asia’s Federated States of Micronesia, Nauru, Palau, Rising Middle Classes.” Background paper Samoa, the Solomon Islands, Tonga, Tuvalu, for Riding the Wave: An East Asian Miracle and Vanuatu. Data are not available, however, for the 21st Century. Washington, DC: for Kiribati, the Marshall Islands, Nauru, World Bank. Palau, Samoa, and Tuvalu. World Bank. 1990. World Development Report 2. The economic classes are defined (in per 1990: Poverty. Washington, DC: World Bank. capita daily consumption expenditures, ———. 2016. Live Long and Prosper: Aging in 2011 purchasing power parity) as follows: in East Asia and Pacific. Washington, DC: extreme poor—less than US$1.90; moderate World Bank. Introduction Overview the region’s population remains at risk of fall- ing into poverty. Upward mobility is increas- East Asia’s progress has come to symbolize how ingly elusive for people in all income groups. economic growth that is rapid and broadly Such a decrease in mobility reflects a grow- shared can improve the lives of millions.1 Since ing concentration of income and wealth, as the late 1950s and early 1960s, a succession well as limited access to basic social services, of economies has experienced the East Asian which are often of poor quality. miracle—rapid economic growth that reduces Riding the Wave is about sustaining the poverty significantly, with little or no increase momentum of the East Asian miracle and con- in income inequality. The initial success of the tinuing to improve the lives of millions—across newly industrialized economies—Hong Kong all income groups—while taking into account SAR, China; Republic of Korea; Singapore; their different needs. The report’s main area and Taiwan, China2—spread in the 1980s to of interest is developing East Asia and Pacific, Indonesia, Malaysia, and Thailand, as well taken to comprise Cambodia, China, Indonesia, as, most dramatically, to China. The poorer Lao PDR, Malaysia, Mongolia, Myanmar, Southeast Asian economies of Cambodia, Lao Papua New Guinea, the Philippines, Thailand, People’s Democratic Republic, and Vietnam Timor-Leste, Vietnam, and the Pacific Island have followed suit since the early 1990s. countries. Past performance has created expec- A wave of prosperity has spread across tations that growth will continue to deliver East Asia. Today, three of five people can be unprecedented improvements in welfare and considered economically secure in that they economic security for all. But global and spe- face a very low risk of falling into poverty. cific challenges—including less certain growth A solid middle class has emerged in most prospects, population aging, and urbaniza- countries. But those successes do not guaran- tion—will make that task more difficult. tee inclusive economic growth—growth that reduces poverty and enables upward mobil- ity and economic security for all. Economic Structure of the report progress varies from country to country, and This introduction presents methods and con- extreme poverty is increasingly concentrated cepts used throughout the report, including a among specific groups. Roughly one-fifth of definition of inclusive growth, a classification 13 14   RIDING THE WAVE of the East Asian population into economic countries’ ability to deliver inclusive growth. classes, and indicative groupings of countries For instance, every income group faces the risk that are likely to face similar challenges in of sliding back in regard to living standards but pursuing inclusive growth. with different implications for different classes. Chapter 1 examines the region’s success That risk may intensify, given the regional and in promoting broad-based growth. East Asia global trends discussed in this report. and Pacific’s development model has resulted Chapter 5 outlines a package of policies in spectacular improvements in living stan- designed to tackle the emerging challenges to dards and human welfare, as well as prodi- inclusive growth in developing East Asia and giously rapid economic growth compared Pacific. The package rests on the two pillars with other regions. The chapter shows how of fostering economic mobility and increasing that success was underpinned by policies that economic security. In addition, building the provided the enabling environment for rapid capacity to deliver inclusive growth emerges labor-intensive growth, combined with public as a cross-cutting priority. Implementing such spending on basic human capital—education, a policy package will help meet the needs of health, and family planning services. all economic classes, although each country Chapter 2 discusses concerns about the will need to customize a package to its own ability of East Asia and Pacific’s development circumstances. model to continue enhancing the welfare of all income groups. Inequality is high or ­rising in many countries. Large shares of the popu- New thinking for a new era: lation increasingly perceive the deck to be economic classes, country stacked against them, with reduced oppor- groupings, and inclusive growth tunities for upward mobility. And emerging Many of the challenges that East Asia and global and regional trends are making it more Pacific faces are a by-product of its success. difficult to achieve and sustain high rates of A region of mostly poor countries and poor broad-based growth built on export-oriented people in the 1980s has become a region of manufacturing. middle-income countries with a diversity of Chapter 3 characterizes the region and economic class—from pockets of extreme its countries in different economic classes, poverty to a significant number of middle- thereby offering a more detailed picture of class households. The specific needs of past progress in improving living standards. today’s more economically diverse popula- This characterization also provides a useful tion require a different way of thinking about tool to describe the heterogeneity of countries policy than what served the region so well in in the region and the changing geography of the past. extreme poverty. Across much of the region, To capture the region’s heterogeneity and extreme poverty is no longer a significant how it has evolved, this report first introduces challenge. However, some countries, includ- the partition of households into e ­ conomic ing Lao PDR, Myanmar, Papua New Guinea, classes: extreme and moderate poor; eco- and the Pacific Island countries, have not nomically vulnerable, defined as those at yet emulated the success of their peer coun- high risk of falling into poverty; economically tries in generating and sustaining rapid and secure; and the middle class.3 This approach broadly shared growth so as to reduce pov- explicitly aims to avoid a simple dichotomy erty significantly. between poor and nonpoor, which would Chapter 4 builds on the characterization blur some of the most important develop- of economic classes to analyze the emerging ments the region has witnessed over the past challenges to inclusive growth faced by each two decades. From a methodological view- income group. The chapter draws attention point, the classification aims to strike a bal- to several challenges that are common across ance between relying on well-established classes and that will increasingly weigh on indicators, such as the international poverty I ntrod u c tion  15 line of $1.90 per day (based on purchasing it is underpinned by a much broader char- power parity) and grounding the economic acterization of human welfare, which is classes in observed patterns, such as the likeli- reflected in the report’s use of available data. hood of falling into poverty (see chapter 3). In particular, the definition assumes that pov- Characterizing the region by economic erty is a multidimensional concept, so equal class has the advantage of throwing into access to social services is needed both directly stark relief some of the features that coun- and instrumentally, to reduce poverty, favor tries share, and that is the second aspect of upward mobility, and increase economic this report’s analysis. The report presents five security. The report documents systematic groupings of East Asia and Pacific countries, differences in access to social services across based on common features. These groupings different areas (for example, rural and urban) are not expected to do justice to all country- and population groups (for example, ethnic level specificities but rather to introduce more minorities) even if it does not suggest any syn- nuanced messages than could emerge from thetic measure to capture such differences. a simple regional overview, especially with By focusing on the challenges and oppor- regard to policies. The five groupings are as tunities for inclusive growth as they emerge follows: in relation to every economic class, this report highlights those elements of a new social con- •  The progressive prosperity countries tract that could reinforce the region’s abil- (Malaysia and Thailand), which have ity to continue delivering inclusive growth. eliminated income poverty while substan- Each country can tailor the elements of such tially increasing the share of middle-class a social contract to its own circumstances, households including its development level and its com- •  The out-of-poverty-into-prosperity coun- position by different classes. tries (China, Mongolia, and Vietnam), where most people are at least eco- nomically secure and the middle class is growing Notes •  The out-of-extreme-poverty coun- 1. This statement refers to a broad definition tries (Cambodia, Indonesia, and the of East Asia and Pacific, the region that Philippines), where extreme poverty levels comprises developing East Asia and are low, but so are the shares of the middle Pacific, Japan, and the newly industrialized class economies. Throughout this report, •  The lagging-progress countries (Lao PDR geographic groupings are defined as follows: and Papua New Guinea), which still have the newly industrialized economies comprise high levels of extreme poverty Hong Kong SAR, China; the Republic of •  The Pacific Island countries, which make Korea; Singapore; and Taiwan, China. The up a heterogeneous group discussed sepa- Pacific Island countries comprise Fiji, Kiribati, the Marshall Islands, the Federated States of rately here because those countries are dis- Micronesia, Nauru, Palau, Samoa, Solomon tinct from the rest of the region. Islands, Tonga, Tuvalu, and Vanuatu. The third element that characterizes the Data are not available for some of those analytical approach of the report, together countries—Kiribati, the Marshall Islands, with economic classes and country groupings, Nauru, Palau, Samoa, and Tuvalu. East Asia is used as shorthand to refer to the countries is a definition of inclusive growth. Economic of developing East Asia and Pacific without growth is regarded as inclusive if it reduces the Pacific Island countries. poverty and enhances economic mobility 2. Japan also relied on many of the policies and economic security across all parts of the described below. It is not included here income distribution.4 Although this defini- because by most measures it was already a tion of inclusive growth is narrowly described relatively advanced economy even before with respect to monetary welfare measures, World War II. 16   RIDING THE WAVE 3. By focusing explicitly on different classes, Alternative approaches are possible, such this report takes into account the needs and as adoption of a social welfare function that expectations of the whole population. This would attach explicit weights to different reflects the realization that, by focusing on parts of the distribution, but the advantages of the dichotomy of the poor and nonpoor, rigorous ranking of different policy outcomes crucial developments in the region over would come at the cost of a significant loss of the past 20 years are missed, such as the simplicity and flexibility. rise of the middle class and the fact that the 4. The report does not refer explicitly to the majority of people in developing East Asia concept of shared prosperity because that and Pacific now face only a slim probability concept is generally defined on a country- of falling back into poverty; those people specific basis. The analysis here focuses on are economically secure. This approach the entire East Asia and Pacific region rather also reflects policy makers’ concerns about than on country-by-country developments. growing inequality (Kanbur, Rhee, and Zhuang Nevertheless, the concept of inclusive growth 2014), as well as the public’s concerns, as captures the broad intuition behind shared evidenced by the perceptions data discussed prosperity. in chapter 2. It is understood that, on the basis of their specific country situation and preferences, policy makers would want to pay Reference attention to different segments of the income Kanbur, R., C. Rhee, and J. Zhuang, eds. 2014. distribution. For example, extreme poverty in Inequality in Asia and the Pacific: Trends, Lao PDR is likely to feature in policy debates Drivers, and Policy Implications. New York: differently from Malaysia or Thailand. Asian Development Bank and Routledge. The Building Blocks of East Asia’s Economic Miracle 1 Introduction redistribution per se2 or on insuring against risks, other than providing limited public East Asia’s development model, which has assistance to the severely disadvantaged. come to be seen as the best way to deliver This model was the basis for the “two-and- growth with equity, has resulted in improve- a-half-point strategy” for poverty reduction ments in living standards and human welfare put forward by the 1990 World Development that are nothing less than spectacular. A wave Report (World Bank 1990). This paradigm of prosperity has spread throughout the region has since been broadened into a three-pronged as a result of prodigiously rapid growth com- approach to reducing poverty, “Grow, Invest, pared with other regions ( ­figure 1.1), driven and Insure,” where “grow” refers to the since the 1980s not only by China but also by ability to promote labor-intensive growth, other countries (figure 1.2; see also table C.1 “invest” to the provision of quality social ser- in appendix C). Extreme and moderate pov- vices, and “insure” to the ability to manage erty have declined rapidly and steadily across risks, both ex ante and ex post.3 In terms of the region (figures 1.3 and 1.4). Those reduc- this approach, the unprecedented success of tions in income poverty were accompanied by East Asia was built primarily on the growth improvements in nonmonetary dimensions of and investment pillars, with the systematic welfare, such as health outcomes (box 1.1). provision of social assistance and insurance The experience of the Pacific Island countries, only a supplement to them. The next sections however, has differed from that of East Asia discuss in more detail the key elements of (box 1.2). these pillars. The development model that underpinned this success combined policies that promoted rapid labor-intensive growth with public spending on basic human capital—educa- Policies that fostered rapid, tion, health, and family planning services. labor-intensive growth The result was growth that reduced pov- East Asia and Pacific economies promoted erty ­ rapidly, without significantly increasing labor-intensive growth to take advantage of income inequality in most cases.1 In con- their most abundant resource: unskilled labor. trast, there was little emphasis on pursuing By using their labor more intensely than in 17 18   RIDING THE WAVE FIGURE 1.1  Real gross domestic product per capita in developing Although policies varied from country to regions country, reflecting differences in initial politi- cal and economic conditions, the successful economies shared several policy and insti- 4,800 tutional characteristics (see, for example, Birdsall and others 1993; Commission on Growth and Development 2008). Foremost 2,400 among those characteristics were a commit- 2010 US$ ment to macroeconomic stability; a primary 1,200 (but not exclusive) reliance on markets to allocate resources; and committed, credible, and capable governance. Market orientation 600 did not prevent governments from interven- ing to spur development or favor specific 300 industries (Birdsall and others 1993). Rather, 1980 1985 1990 1995 2000 2005 2010 2015 such interventions (now commonly termed Year industrial policy) were subject to monitoring and modified if they were deemed ineffective. East Asia and Pacific Europe and Central Asia Governments also ensured that the measures Latin America and the Caribbean Middle East and North Africa South Asia Sub-Saharan Africa did not become too costly, particularly in fis- cal terms (Birdsall and others 1993, 6–7). Source: World Bank World Development Indicators (database), https://data.worldbank.org/data- This emphasis on evaluating and modifying catalog/world-development-indicators. policies was feasible, in turn, because gov- ernments were able to build competent civil FIGURE 1.2  Real gross domestic product per capita in developing services that were largely insulated from polit- East Asia and Pacific ical interference. Most of the successful East Asian economies also established mechanisms 9,600 for consultation between the government and the private sector. Trade openness was a second key ele- 4,800 ment. In general, East Asian economies aban- doned import substitution early on in favor 2010 US$ 2,400 of export promotion, in sharp contrast with Latin America up to the 1990s. The result was 1,200 swift growth in exports—especially of manu- factured goods—linked to growing absorp- 600 tion of new technology. East Asia moved rapidly to establish a competitive, pro-export 300 regime, which included duty-free imports for 1980 1985 1990 1995 2000 2005 2010 2015 exporters and their suppliers; export credits Year and export marketing institutions; incentives Cambodia China Indonesia Lao PDR Malaysia for foreign direct investment (FDI) and the Mongolia Philippines Thailand Vietnam associated technology transfers; and competi- Source: World Bank World Development Indicators (database), https://data.worldbank.org/data- tive exchange rates. Selective interventions, catalog/world-development-indicators. as observed in the Republic of Korea, and Taiwan, China, were based on clearly estab- traditional agriculture, mostly poor popu- lished and well-monitored economic perfor- lations were able to take advantage of the mance criteria (usually exports), were time expansion of opportunities that came with bound, and were quickly halted when their rapid growth, thereby benefiting from higher costs rose. In China, this trade opening was wages and accumulating assets. initially confined to selected special economic T he B u i l ding B l o c k s o f E a s t A s ia ’ s E c ono m i c Mira c l e   19 zones, whose number and size was gradually FIGURE 1.3  Poverty rates, across developing regions expanded.4 A third common feature was that most of a. US$1.90 a day (purchasing power parity) these economies (particularly the early devel- 80 opers and the larger economies of Southeast 70 Asia) were able to achieve high rates of pri- vate investment and domestic financial sav- 60 ings. The mobilization of domestic savings 50 was supported by measures such as institut- Percent ing and supporting banks and development 40 banks; lowering transaction costs and pro- 30 viding incentives for savings; and discour- aging consumption (through high interest 20 rates for consumer items and high taxes on 10 luxury goods). Investment was promoted by establishing low tariffs for importing 0 81 84 87 90 93 96 99 02 05 08 10 11 12 13 14 15 capital goods and increasing public invest- 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 ment in physical infrastructure, such as Year roads and power, which helped attract pri- b. US$3.10 a day (purchasing power parity) vate investment by raising returns to it (see 100 Commission on Growth and Development 2008, 34–35). 90 Finally, these economies all adopted broad- 80 based agricultural policies early in their 70 ­ development process. As low-income econo- 60 mies, they began with a mostly agricultural Percent 50 production base, and a dynamic agricultural sector proved critical to sustaining growth 40 that reduced poverty. Agricultural growth 30 benefited large shares of the population, both 20 directly and by facilitating schooling, migra- 10 tion, urbanization, and expansion of employ- 0 ment in food processing and manufacturing. 81 84 87 90 93 96 99 02 05 08 10 11 12 13 14 15 The outcome was lower rural poverty, less 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 inequality, and containment of urban-rural 20 Year income gaps. East Asia and Pacific Europe and Central Asia Latin America and the Caribbean Middle East and North Africa This progress was made possible by South Asia Sub-Saharan Africa Developing world policies such as low levels of agricultural ­ taxation, including implicit taxes through Sources: World Bank PovCalNet (database) http://iresearch.worldbank.org/PovcalNet/povOn​ price regulation (Ahuja and others 1997). Demand.aspx; for 2002–15, World Bank East Asia and Pacific Team for Statistical Development. Note: See appendix C for details on the methodology. Agricultural subsidies, such as rice and fer- tilizer price policies in Indonesia, were used to raise rural incomes. And large-scale pub- lic support to agricultural extension helped Between 1978 and 1983, China dismantled to absorb the new technologies of the Green its system of collective agricultural produc- Revolution. These measures proved particu- tion, granting individual farmers land-use larly pro-poor when they were accompanied rights, although not the right to sell their land. by significant land reform and redistribution, Likewise, Vietnam introduced a new land law as occurred from the 1960s in Japan, Korea, in 1993 that dismantled the country’s agricul- and Taiwan, China (Acemoglu and Robinson tural cooperatives and collectives, allowing 2002; Boyce, Rosset, and Stanton 2005). transactions in land. In both countries, the 20   RIDING THE WAVE FIGURE 1.4  Poverty rates, across developing East Asia and Pacific a. Larger economies, US$1.90 a day (purchasing power parity) 40 30 Percent 20 10 0 Cambodia China Indonesia Lao PDR Malaysia Mongolia Philippines Thailand Vietnam b. Smaller economies, US$1.90 a day (purchasing power parity) 50 40 30 Percent 20 10 0 Fiji Kiribati Micronesia, Papua Samoa Solomon Timor-Leste Tonga Tuvalu Vanuatu Fed. Sts. New Guinea Islands 2002 2007 2012 Source: World Bank East Asia and Pacific Team for Statistical Development. measures significantly boosted agricultural scale, including jobs for unskilled workers. productivity and reduced poverty (Ravallion And they enabled economies based on agri- and Chen 2007; Ravallion and van de Walle culture to undergo a rapid transformation 2008).5 The poor especially benefited because into economies dominated by manufacturing the initial land allocation in both countries and services. Millions of rural workers moved was relatively equitable. 6 In Vietnam, the into manufacturing jobs. As rapid economic introduction of a market for land-use rights expansion increased the demand for services, led to additional productivity gains, as non- rural workers moved into service jobs too. farm opportunities expanded rapidly. Between the early 1980s and 2012, agricul- Those policies helped the region’s econo- ture’s share of the labor force declined in mies create formal sector jobs on a large aggregate across developing East Asia, from T he B u i l ding B l o c k s o f E a s t A s ia ’ s E c ono m i c Mira c l e   21 more than 50 percent to around 35 percent. At 3 percent of gross domestic product Meanwhile, industry’s share rose from (GDP), public spending on education in 1985 15 percent to 20 percent, and services’ share was lower in Korea than in the República de grew from 25 percent to nearly 45 percent. Venezuela (as the country was named before With these structural changes came rapid pro- 1999), where it accounted for 4.3 percent of ductivity growth across sectors. Increasing GDP. At the time, the República de Venezuela returns to labor and the expanding number allocated 43 percent of that spending to of jobs outside of agriculture helped spread higher (­tertiary) education, whereas Korea the benefits of growth to large sections of the spent only 10 percent on tertiary education. population. Consequently, Korea was able to spend almost twice as much (2.5 percent of GDP) on basic education as did the República de Venezuela Policies that built up basic (1.3 percent). Similarly, Bolivia and Indonesia human capital had roughly the same per capita income in The successful East Asian economies also the mid-1980s and spent about the same invested in the skills, knowledge, and expe- (as a share of GDP) on e ­ ducation. However, rience of workers—their human capital—by Bolivia devoted only about 40 percent of its expanding the provision of basic services, education budget to basic education, whereas p articularly primary education, health ­ Indonesia’s share was almost 90 percent. By and nutrition, and family planning. Those 1987, Indonesia had enrolled over 90 percent investments, which also benefited the poor, of rural children in primary school, while complemented the policies that promoted the Bolivian school system covered only labor-intensive growth by giving the poor the 60 percent of all children, with fewer than means to take advantage of the expansion in half of all rural schools providing instruction employment opportunities. through fifth grade.7 From the 1960s onward, many East Asian East Asian economies similarly expanded countries made intensive, sustained invest- efforts to prevent and treat diseases afflict- ments in human capital. The initial focus ing the poor. As early as the 1960s, Korea on the widespread provision of basic skills, established a large number of health centers including through universal primary edu- to provide primary health care and fam- cation, was followed by an increase in the ily planning services. It also introduced a availability of secondary education in both national health insurance scheme for urban urban and rural areas. As growth proceeded, workers in the 1970s. The scheme was the resulting demographic transition meant extended to rural areas in the mid-1980s. that the number of school-age children grew Across the region, public health expenditure more slowly, reducing the need for additional increased markedly from the 1980s onward, resources. At the same time, improvements in with a focus on expanding service provision agricultural productivity and greater urban- in rural areas and reducing the cost of treat- ization reduced the opportunity cost of chil- ment for the poor. Improving access to fam- dren attending school—in other words, such ily planning was also emphasized by many factors alleviated the pressure on families to successful East Asian economies. Thailand, put children to work—thus increasing the for instance, implemented one of the world’s demand for education. Meanwhile, provision most effective and innovative family plan- of tertiary education was left largely to the ning programs with the help of public-­ private sector, with little public support. private partnerships. The examples of Korea, an early developer, Those investments in education, health, and Indonesia, a later developer, illustrate this and family planning brought a range of ben- East Asian pattern of human capital invest- efits (box 1.1). School enrollment rose rapidly ment, which contrasts with patterns seen in many countries. Overall, the educational in comparable countries in Latin America. attainment of the work force in the region 22   RIDING THE WAVE increased fivefold between 1950 and 2010, Similarly, basic health services and access with rapid increases in labor force skills of the poor to health care were improved. among broad segments of the population. Together with efforts to boost child nutrition Efforts were also made to improve the qual- and improve access to clean water and sani- ity of education, which in several countries tation, greater access to health care reduced reached levels comparable with those found sharply the incidence of disease and of indebt- in high-income economies (World Bank, edness stemming from catastrophic health forthcoming). Reflecting those trends, stu- expenditures. Family planning programs dents in many economies achieved learning coupled with rapid growth helped reduce outcomes unmatched anywhere.8 fertility rates sharply: the average number of BOX 1.1  Socioeconomic progress in developing East Asia and Pacific The countries in developing East Asia and Pacifica child mortality, which overwhelmingly affects the have seen significant improvements in various non- poor, fell particularly sharply. Already in the 1970s monetary dimensions of welfare (figure B1.1.1, pan- and 1980s, much of the region performed better on els a and b). School enrollment rose rapidly in many most indicators than did most of the world’s lower- countries, including China, Indonesia, Thailand, middle-income countries; since then, its performance and Vietnam, as did educational attainment (that is, has improved further to the values observed in upper- average years of schooling). The access of the poor middle-income countries. to health services, as demonstrated, for instance, by the share of births attended by skilled health work- a. Developing East Asia and Pacific is defined as Cambodia, China, Indonesia, Lao People’s Democratic Republic, Malaysia, Mongolia, Myanmar, Papua New ers, increased sharply. Likewise, access to clean water Guinea, the Philippines, Thailand, Timor-Leste, Vietnam, and the Pacific Island and sanitation significantly increased. As a result, countries and is the central focus of this report. FIGURE B1.1.1  Socioeconomic indicators a. 1975, or earliest available Births attended by skilled health staff (% of total) 250 200 Average educational 150 Improved sanitation facilities attainment (% of population with access) (years of schooling) 100 50 0 Adjusted net Improved water source enrollment rate, (% of population with access) primary, both sexes (%) Mortality rate, children under 5 (per 1,000 live births) box continues next page T he B u i l ding B l o c k s o f E a s t A s ia ’ s E c ono m i c Mira c l e   23 BOX 1.1  Socioeconomic progress in developing East Asia and Pacific (continued) FIGURE B1.1.1  Socioeconomic indicators (continued) b. 2015, or latest available Births attended by skilled health staff (% of total) 100 80 Average educational 60 Improved sanitation facilities attainment (% of population with access) (years of schooling) 40 20 0 Adjusted net Improved water source enrollment rate, (% of population with access) primary, both sexes (%) Mortality rate, children under 5 (per 1,000 live births) Developing EAP LICs UMICs LMICs Sources: World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world​ -development-indicators; Barro-Lee Educational Attainment Dataset, http://datatopics.worldbank.org/education​ /­wProjection/bpopmodel; World Bank staff calculations. Note: EAP = East Asia and Pacific; LICs = lower-income countries; LMICs = lower-middle-income countries; UMICs = upper-middle-income countries. LICs, LMICs, and UMICs include all relevant countries outside developing East Asia and Pacific with available data. All values refer to unweighted group means; country samples are balanced over time. Data for panel a refer to 1975 or the earliest available year after then; data for panel b refer to 2015 or the latest available year before then. For average educational attainment, all values are multiplied by 10 for presentational purposes. births per woman in Thailand, for instance, social protection broadly reflected three fac- fell from six in the mid-1960s to fewer than tors. First, there was a perception that the two by 1996. Similar successes had been seen rising tide from rapid growth would lift all earlier in Korea and Taiwan, China (Gill, boats. Second, governments sought to mini- Revenga, and Zeballos 2016). mize the risks facing the population by pur- suing macroeconomic stability rather than by systematically providing insurance against Limited social protection and adverse idiosyncratic shocks. As a corol- redistributive policies lary, when sharp economic fluctuations did Overall, East Asia’s development model occur, as during the Asian financial crisis of placed less emphasis on redistribution than the late 1990s, the accompanying decrease on growth per se. in real wages and increase in unemployment Social protection systems were underde- significantly affected prospects for inclusive veloped, with the state playing a smaller role growth. Third, the region relied on strong than in other developing regions at com- traditional family structures to mitigate parable income levels.9 The limited role of the impact of shocks. For instance, early in 24   RIDING THE WAVE BOX 1.2  Growth and poverty reduction in the Pacific Islands and Papua New Guinea The growth and poverty reduction story of the Pacific disasters has increased the risks already inherent in Island countries (PICs) differs markedly from the the narrow production base. For example, Cyclone headline-grabbing success of many of their mainland Winston caused total damages in Fiji of around counterparts. With the possible exception of Papua F$2 billion, equivalent to 25 percent of the country’s New Guinea, which shares many of the characteris- gross domestic product (GDP). tics of the PICs despite not being one, many of these Although extreme poverty is widespread in Papua differences have been dictated by the PICs’ geography New Guinea, the Solomon Islands, Vanuatu, and and the limited size of their populations and econo- some atoll countries in the north Pacific, extreme mies. Their remoteness from large markets and the poverty in most of the smaller island countries dispersion of their populations over vast areas of remains concentrated in smaller communities, which ocean have impeded broad-based growth through are far from the main cities, are deprived of access labor-intensive and export-oriented manufacturing, to basic services, and are highly vulnerable to natu- because high transport costs make it difficult for the ral shocks. The limited employment opportunities PICs to compete with countries closer to major inter- outside of subsistence farming and of the smaller national markets. government, tourism, and fishing sectors have made Overall, therefore, the economies of most PICs international migration an important economic still have a narrow production base, with semisub- option for many Pacific islanders. In countries such sistence agriculture playing a major role, especially in as Samoa and Tonga, where remittances account the livelihoods of the poor. Foreign exchange, which for more than 20 percent of GDP, migration helps is much needed, comes from tourism (in Fiji, Palau, many households to avoid slipping into poverty. Samoa, Tonga, and Vanuatu); mining and minerals Large amounts of foreign aid have similarly contrib- (in Papua New Guinea); logging (in the Solomon uted to poverty reduction. Strong traditional family Islands); fisheries and commercial agricultural prod- and community sharing networks also help to keep ucts (coffee in Papua New Guinea and Vanuatu, extreme poverty at bay, especially in the Polynesian copra in several of the smaller countries, and sugar countries. However, the adverse impacts of climate in Fiji). The high costs of service delivery to a very change, the frequent and systemic effects of natu- dispersed population have made it difficult to invest ral disasters, and the effects of monetization and in human capital; to expand economic opportuni- urbanization are putting stress on those traditional ties for the population; to support a more diversified networks. economy; and to provide basic needs such as water, sanitation, and electricity. High exposure to natural Sources: World Bank 2016b, 2016c, 2017. Korea’s development process, social assis- of decades, several countries have introduced tance programs that offered tax incentives to elements of a social protection system as people who took care of their aged parents their circumstances have changed. Social were clearly designed to complement and not pensions (noncontributory old-age transfers) replace the role of the family. And although have been widely adopted in recent years, a social assistance program was instituted in though their age thresholds and generosity 1965, it covered only about 5 percent of the vary. Such pensions range from those that population. It was not until after the Asian are universal (Timor-Leste and several Pacific financial crisis that Korea broadened the Island countries), to those that are available scope and coverage of its social protection to everyone without a formal sector pension programs. Taiwan, China, followed a similar (Thailand), to those that are more tightly course.10 targeted (Malaysia and the Philippines). This situation has changed recently in Similarly, in China, pension schemes for rural some parts of the region. Over the past couple and urban residents, which involve large T he B u i l ding B l o c k s o f E a s t A s ia ’ s E c ono m i c Mira c l e   25 public subsidies, have brought coverage to an of national identification schemes, and sig- additional 350 million working-age people. nificant initiatives in areas such as payment More generally, countries across the region systems have produced positive results (see that had delayed putting in place a modern chapter 5). In several countries, including social protection system (box 1.3) have the Indonesia and Malaysia, recent measures to opportunity to learn from the experiences reduce spending on subsidies for food and of other countries in the region and beyond, fuel are creating the fiscal space for a more thereby avoiding pitfalls such as the creation appropriate social protection system. welfare traps and excessive labor taxation of ­ Just as they had relied little on social pro- (see chapter 5). tection to redistribute wealth, most economies Better-targeted social assistance pro- in the region have tended to use redistributive grams offering cash transfers have also taxes less frequently than developed econo- been gradually established in the region. mies do. For instance, direct taxes account Examples include the conditional cash trans- for only 41 percent of total taxes in develop- fer schemes in Cambodia, Indonesia, and the ing East Asia, compared with 65 percent in Philippines;11 the programs that were con- member countries of the Organisation for solidated in Fiji; and China’s dibao program, Economic Co-operation and Development which now reaches around 70 million peo- (OECD), and capital income accounts for ple. In addition, several countries, including only around 5 percent of personal income Indonesia and the Philippines, have made sub- tax revenues (World Bank 2016a; see also stantial investments in improving and mod- table C.2 in appendix C). ernizing social protection delivery systems. In addition, tax policy and tax administra- Greater use of technology, the introduction tion both suffer from significant shortcomings. BOX 1.3  East Asia lags behind other middle-income regions in providing social protection East Asia still has fewer and less effective social pro- program coverage is generally narrower than found grams than do other mostly middle-income regions in, for example, countries in Eastern Europe and (table B1.3.1). Except in Malaysia and Thailand, Central Asia or in Latin America and the Caribbean. TABLE B1.3.1  Performance of social protection programs in different developing regions Extent of benefits, Beneficiary Benefit incidence, Coveragea Extent of poorest quintileb incidence, poorest poorest quintiled (%) benefitsb (%) (%) quintilec (%) (%) East Asia and Pacific 49.8 17.5 11.9 28.1 3.1 Eastern Europe and Central Asia 62.1 27.1 32.1 24.2 7.8 Latin America and the Caribbean 60.9 29.8 31.8 20.1 5.4 Middle East and North Africa 51.4 12.8 11.2 21.9 7.6 South Asia 27.5 3.2 2.5 21.2 1.8 Sub-Saharan Africa 19.4 14.6 28.7 18.3 4.0 Sources: Atlas of Social Protection Indicators of Resilience and Equity (database), http://datatopics.worldbank.org/aspire; World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world-development-indicators. Note: Regional aggregates denote population-weighted means. Data for China and Lao PDR are not available. a. Coverage denotes the percentage of population participating in social assistance programs (including direct and indirect beneficiaries). b. Extent of benefits denotes the total transfer amount received as a share of the total welfare of beneficiaries. c. Beneficiary incidence, poorest quintile denotes the percentage of program beneficiaries belonging to the poorest quintile (of the posttransfer welfare distribution). d. Benefit incidence, poorest quintile denotes the percentage of total program benefits received by the poorest quintile. 26   RIDING THE WAVE The tax base for both direct and indirect these reduced rates in absolute terms. Within taxes is small, reflecting the large size of the indirect taxes, there is high reliance on trade informal sector and the numerous exemp- taxes, particularly in the Pacific, even though tions, deductions, and incentives allowed by these taxes are inefficient and regressive. tax legislation in the region. Personal income taxes are particularly low and display little de Notes facto progressivity, owing to the widespread 1. As noted in chapter 2, China constituted an exemptions allowed, to the small share of the exception to this trend, although the initial population to which the maximum marginal level of inequality was much lower than in rate applies,12 and to minimum threshold many other developing East Asian countries. levels that exclude significant proportions of 2. There were, however, notable exceptions. even well-off households. In most countries, A few economies implemented radical land a limited number of workers are registered reform (Taiwan, China; the Republic of as taxpayers, further restricting the pool to Korea; and Vietnam), and a few pursued which taxes are applied.13 Compliance is fur- extensive nationalization (China, Indonesia, ther reduced by the restricted use of withhold- and Vietnam). Whatever their immediate effects (nationalization, for example, boosted ing taxes on capital income.14 equality but reduced incomes), these policies Personal capital income and wealth largely can be seen more as one-off interventions than escape taxation, thereby reducing the overall as careful efforts to redistribute the benefits progressivity of the tax system. Interest, divi- of growth. The most notable exception is dends, and capital gains are typically taxed at Malaysia, which for decades has had in place much lower rates than is labor income, 15 ethnically targeted programs. which creates the potential for wealthier 3. See Gill, Revenga, and Zeballos (2016) for a households to reduce their tax burden by dis- detailed exposition of this point. guising labor income as capital income. 4. After the Shenzhen Special Economy Zone Recurrent property taxes are very limited, was established in 1980, a further 14 coastal despite being efficient16 and equitable;17 even cities (including Fuzhou, Guangzhou, with the low existing taxes, compliance is and Shanghai) were opened to overseas investment in 1984, and the Hainan Island poor.18 The lack of compliance reflects the Special Economic Zone was created in 1988. considerable political and social resistance to 5. Agricultural total factor productivity growth such taxes, as well as their significant admin- was particularly high in China, reflecting also istrative requirements. Countries are more significant public investment in agricultural reliant on corporate income taxes, including research and development. those paid by multinational companies.19 6. Even now, inequality in land distribution is However, the growing international mobility much lower in China’s eight most important of capital threatens this revenue source, owing agricultural provinces than in India or Brazil to rising tax competition among developing (the respective Gini coefficients equal 17, 75, countries, and implies that much of the inci- and 89 percent). dence of corporate income taxes falls on 7. See Birdsall and others (1993, 199–201), for wages, reducing its progressivity. more details about these cases. In general, revenue relies heavily on indi- 8. For example, Vietnamese students, including rect taxation, which is easy to administer but those from poor backgrounds, regularly outperform peers from member countries of less progressive (as measured by incidence of the Organisation for Economic Co-operation taxes as a proportion of the income or con- and Development. See also Hanushek and sumption of a given group) than is direct tax- Woessmann (2011, 2015). ation. Equity concerns have led to some of the 9. Social protection refers to measures that lowest indirect tax rates in the world, limiting help reduce poverty and vulnerability tax yields, and to reduced rates on sensitive by (a) diminishing people’s exposure to commodities. However, better-off house- economic risks, (b) enhancing their capacity holds appropriate most of the benefits of to manage such risks, and (c) promoting T he B u i l ding B l o c k s o f E a s t A s ia ’ s E c ono m i c Mira c l e   27 efficient labor markets. There are three 16. An efficient tax is one that generates limited common types of social protection. The distortions. first type, social assistance , consists of 17. Revenues are particularly low in Indonesia noncontributory transfers in cash or kind, (0.35 percent of GDP), Mongolia (0.15 percent conditional or nonconditional, aimed at the of GDP), and Thailand (0.45 percent of GDP). poor and the economically vulnerable. The In contrast, in China, property taxes account second type, social insurance , comprises for 1.5 percent to 2.0 percent of GDP, a level programs that minimize the impact of that is comparable to those observed in OECD economic shocks on individuals and families countries (Norregaard 2015). by providing benefits and services in 18. In the Philippines, the collection rate for recognition of contributions to an insurance property taxes stands at 50 percent. scheme. Social insurance schemes include 19. Commodity producers, including Indonesia publicly provided or mandated insurance and Malaysia, derive a substantial share of schemes that provide coverage for old age, revenue from the hydrocarbon sector. disability, and death of the main household provider; maternity leave and sickness cash benefits; and health insurance coverage. References The third type, labor market intervention, is Acemoglu, D. and J. A. Robinson. 2002. “The designed to promote employment, efficient Political Economy of the Kuznets Curve.” operation of labor markets, and protection Review of Development Economics 6 (2): of workers. 183–203. 10. See Gill, Revenga, and Zeballos (2016) for Ahuja, V., B. Bidani, F. Ferreira, and M. Walton. more details. 1997. Everyone’s Miracle? Revisiting Poverty 11. In the Philippines, population coverage is now and Inequality in East Asia. Washington, DC: around 20 percent. In Indonesia, the target is World Bank. to triple coverage. Birdsall, N., J. E. Campos, C.-S. Kim, W. M. 12. The top marginal personal income tax rate Corden, H. Pack, J. Page, and J. E. Stiglitz. is applied at a threshold level exceeding 1993. The East Asian Miracle: Economic 20 times GDP per capita in China, Myanmar, Growth and Public Policy. New York: Oxford and Thailand and 10 times GDP per capita in University Press. Indonesia. In contrast, the average threshold Boyce, J. K., P. Rosset, and E. A. Stanton. 2005. level in the OECD is four times GDP per “Land Reform and Sustainable Development.” capita (Claus, Martinez-Vazquez, and Vulovic PERI Working Paper 98, Political Economy 2012; Kanbur, Rhee, and Zhuang 2014; Lam Research Institute, University of Massachusetts and Wignender 2015). at Amherst. http://works.bepress.com​ 13. For instance, in China, the effective share /­james_boyce/15. of individuals who pay personal income Claus, I., J. Martinez-Vazquez, and V. Vulovic. tax is less than 3 percent of the working 2012. Government Fiscal Policies and population. Redistribution in Asian Countries. Manila: 14. In Indonesia, there are no withholding Asian Development Bank. taxes on capital gains, even though such Commission on Growth and Development. gains are subject to personal income tax. 2008. The Growth Report: Strategies for As a result, only 5 percent of personal Sustained Growth and Inclusive Development. income tax revenues in Indonesia stem from Washington, DC: World Bank. http://hdl​ capital income, with the rest coming from .handle.net/10986/6507. withholding on salaries. A similar situation Gill, I. S., A. Revenga., and C. Zeballos. 2016. exists in Thailand. Grow, Invest, Insure: A Game Plan to End 15. Tax rates on dividends range from 5 percent Poverty by 2030. Washington, DC: World Bank. in Vietnam to 20 percent in China and Korea. Hanushek, E., and L. Woessmann. 2011. “The In Indonesia, tax rates on dividends and Economics of International Differences in interest equal 10 percent and 20 percent, Educational Achievement.” In Handbook of respectively—substantially less than the the Economics of Education, vol. 3, edited by 30 percent top marginal tax rate on salary E. Hanushek, S. Machin, and L. Woessmann. income. Amsterdam: North-Holland. 28   RIDING THE WAVE ———. 2015. “The Economic Impact of ———. 2016a. Reducing Vulnerabilities: East Asia Educational Quality.” In Handbook of and Pacific Economic Update, October 2016. International Development and Education, Washington, DC: World Bank. edited by P. Dixon, S. Humble, and C. ———. 2016b. Pacific Islands—Systematic Counihan. Cheltenham, UK: Edward Elgar. Country Diagnostic for the Eight Small Pacific Kanbur, R., C. Rhee, and J. Zhuang, eds. 2014. Island Countries: Priorities for Ending Poverty Inequality in Asia and the Pacific: Trends, and Boosting Shared Prosperity. Washington, Drivers, and Policy Implications. New York: DC: World Bank. http://documents.worldbank​ Routledge and Asian Development Bank. .org/curated/en/313021467995103008/Pacific​ Lam, W. R., and P. Wignender. 2015. China: How -Islands-Systematic-country-diagnostic-for-the​ Can Revenue Reforms Contribute to Inclusive -eight-small-Pacific-Island-Countries-priorities​ and Sustainable Growth? Washington, DC: -for-ending-poverty-and-boosting-shared​ International Monetary Fund. -prosperity Norregaard, J. 2015. “Taxing Immovable Property: ———. 2016c. “Papua New Guinea Systematic Revenue Potential and Implementation Country Diagnostic Concept Note.” Unpublished Challenges.” In Inequality and Fiscal Policy. paper, World Bank, Washington, DC. Washington, DC: International Monetary Fund. ———. 2017. “Kiribati, Nauru, Marshall Islands, Ravallion, M., and S. Chen. 2007. “China’s Micronesia, Palau, Samoa, Tonga, Tuvalu and (Uneven) Progress against Poverty.” Journal of Vanuatu Regional Partnership Framework,” Development Economics 82 (1): 1–42. Report 100997-EAP, World Bank, Washington, Ravallion, M., and D. van de Walle. 2008. D C . h t t p : / / p u b d o c s ​ . w o r l d b a n k . o r g​ Land in Transition: Reform and Poverty in /­en/118921488323842915​/­R2017-0028.pdf. Rural Vietnam. Washington, DC: Palgrave ———. Forthcoming. Growing Smarter: Learning Macmillan and World Bank. and Growth in East Asia and the Pacific . World Bank. 1990. World Development Report Washington, DC: World Bank. 1990: Poverty. Washington, DC: World Bank. Ready for the Challenges of the 21st Century? 2 Introduction slowing growth, with much of the region “getting old before getting rich.” Rapid urban Developing East Asia and Pacific has long growth risks entrenching inequalities and can been synonymous with a development model lead to congestion, environmental strains, and of economic growth with equity, a model that poor living conditions. has delivered better lives for millions. To con- tinue to benefit from the wave of prosperity that has changed the region, however, govern- Concerns over the inclusiveness ments need to invest now more evenly than of the region’s development they did previously across all three pillars model are growing of the grow-invest-insure paradigm. Such a Income inequality is high1 or rising in most shift is needed because the evolving regional countries in the East Asia and Pacific region, and global contexts pose several challenges particularly in the large ones. As a result, the to the region’s development model. In many average person in the region lives in a coun- countries in the region, inequalities are ris- try where inequality has been increasing over ing, jeopardizing growth and raising the risk the past couple of decades, which is quite a of wasted production potential and of inef- different experience from what has happened ficient allocation of resources. Perceptions of in most other parts of the world, particularly inequality, whether justified or not, influence in Latin America (figure 2.1). In addition people’s behaviors, choices, and investments, to any other concerns that high inequality with wider effects on societies and economies. may generate (box 2.1), the rising inequality Achieving or sustaining high levels of growth means that economic growth is becoming less may become more difficult because trade is effective in reducing poverty. expanding more slowly, the expansion of Perceptions that inequalities are too high global value chains is uncertain, and countries and still growing have been on the rise across need to navigate structural, technological, and the East Asia and Pacific region and are now institutional transitions. Furthermore, popu- very visible in policy debates. An informal lation aging and urbanization, which were survey of policy makers identified inequal- sources of opportunity for growth, are now ity as a key concern for the region (Kanbur, turning into potential threats. Aging is Rhee, and Zhuang 2014). Much data also 29 30   RIDING THE WAVE FIGURE 2.1  Inequality across developing regions of the world suggests that several countries in the region are experiencing an increasing concentration 60 of wealth in the hands of the richest, albeit 55 from a generally low base and to an extent that is still lower than in other parts of the 50 world. This development contributes to per- Mean national Gini index, population-weighted 45 ceptions that for many people the opportuni- ties for upward mobility may have declined. 40 35 Changing perceptions about 30 inequality and mobility 25 Today, large majorities of the region’s popula- 1988 1993 1998 2003 2008 2013 tion (more than 90 percent in China, Republic Year of Korea, and Taiwan, China; around East Asia and Pacific Eastern Europe and Central Asia 80 percent in Japan; and more than 50 Latin America and the Caribbean Middle East and North Africa percent in the Philippines) think that income South Asia Sub-Saharan Africa Industrialized economies differences in their own country are too Source: World Bank 2016c. large.2 In Indonesia, 89 percent of the popu- Note: Mean national Gini index, population-weighted. The national Gini indexes are estimated from lation thinks that it is quite or very urgent household surveys of consumption expenditure or income. BOX 2.1  Inequality can be damaging—and so can perceptions of inequality In the past decade, inequality has returned to the cen- and wealth—and that typically are also bad for the ter of policy debates in both advanced and develop- investment, innovation, and risk-taking that under- ing economies. Apart from the philosophical focus pin long-term growth,” (World Bank 2006, 8–9). on redistributive justice and the moral case against Elite capture—the processes by which small minori- inequality, an important strand in this discussion is ties appropriate large shares of resources—and crony the instrumental impact of inequality in outcomes. capitalism—an economic system where public offi- Specifically, inequality can harm development in two cials and business leaders collude for their mutual major ways. First, “with imperfect markets, inequali- benefit—for example, may shape institutions that ties in power and wealth translate into unequal serve only the needs of the few, hindering a society’s opportunities, leading to wasted productive potential ability to solve widespread problems cooperatively and to an inefficient allocation of resources” (World and to manage common resources and public ser- Bank 2006, 7). Such inequality can reduce growth vices. Those citizens who are left behind may resort if the poor do not earn enough, lack sufficient sav- to violence to affirm their claims or, on a smaller ings, or are unable to borrow to invest in the acqui- scale, may resort to crime to improve their lot in life. sition of human capital or productive physical capi- A key element of such impacts is that for many tal (Banerjee and Newman 1993; Galor and Zeira adverse behavioral responses to materialize, the 1993), a point indirectly supported by the evidence perceptions of inequality have economic implica- that countries with a larger middle class tend to grow tions, whatever the objective reality. For example, if faster (Easterly 2001).a social stereotyping exists, inequality can affect moti- Second, “economic and political inequalities are vation and choices with negative effects on lifelong associated with impaired institutional development. opportunities.b […] unequal power leads to the formation of insti- Analysis of data on perceptions of inequality high- tutions that perpetuate inequalities in power, status, lights what objective inequality means to people in box continues next page R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    31 BOX 2.1  Inequality can be damaging—and so can perceptions of inequality (continued) at least two respects. First, it helps identify whether significantly correlated with the demand for redis- people see inequality as a source of opportunity tribution (Gimpelson and Treisman 2015; Tiongson (that is, there is a big payoff to doing well in life) and Weise 2016). A wrong (lower) perception of or as a sign that opportunities for upward mobil- the levels of inequality and a negative correlation ity are limited (that is, economic advantages are between perceiving (high) inequality and turning out limited to those already at the top). Different inter- to vote may have contributed to countries in the pretations can affect people’s choices about invest- region not having adopted a more redistributive pol- ments, for example, and therefore their life chances icy stance (Tiongson and Weise 2016). and outcomes. Second, perceptions of inequality are important drivers of public policy preferences, a. A large body of literature has examined the relationship between inequal- ity and subsequent economic growth both theoretically and empirically by particularly regarding redistribution. Recent work using cross-country data. The results have generally been inconclusive and in the East Asia and Pacific region found that the have tended to vary depending on country samples, measures of inequality, actual level of inequality is generally not significantly econometric methods, and time period (World Bank 2016b). b. For example, junior-high school students in villages in India, when their related to the preference for redistribution, whereas status as belonging to a disadvantaged group was made noticeable, would the perceived level of inequality is positively and put in less effort in performing a task (Hoff and Pandey 2004). to address inequality (World Bank 2016a). people are increasingly able to meet their In Vietnam, “a majority of those surveyed basic everyday needs, longer-term ambitions and eight in ten urban residents worry about that are harder to achieve—such as improving disparities in living standards” (World Bank one’s health or job prospects—play a greater 2014b, 47), with young people more likely role and lead to frustration—the paradox of to worry about inequalities in all spheres of the happy peasant and frustrated achiever life—health, ­education, incomes—and by geo- (Graham and Pettinato 2002). In this sense, graphical ­levels (local, national). the East Asia and Pacific region may be a vic- These perceptions are more marked in East tim of its own success: citizens who become Asia and Pacific than in other regions of the better off also become increasingly aware world3 and are common across subgroups that they may be falling short of the new identified by age, gender, rural or urban resi- standards they set for themselves (Graham dence, and education level. The same percep- 2017; Graham, Chattopadhyay, and Zhang tions are even more marked for lower-income 2016). Again, to the extent that economic groups and for those who identify themselves growth has resulted in a very visible emerg- as being in the lower class. ing ­ m iddle class, poorer households may In addition, there has been a significant question whether they have a fair chance of decline since the 1990s in the number of reaching similar levels of economic security or people who see income inequality as provid- whether the odds of catching up are becom- ing desirable incentives for individual effort.4 ing slimmer. Similarly, perceptions of inequal- The reversal in attitudes has been particularly ity, and of the link between effort and upward sharp in China,5 although a belief in hard work mobility, partly reflect the increased visibility remains stronger there than in richer countries of various aspects of inequality, including outside developing East Asia and Pacific, such conspicuous consumption by the rich. Those as Japan and the Republic of Korea. perceptions are accompanied by evidence that Perceptions of rising inequality are likely large accumulations of wealth reflect not only driven by several factors, including rising aspi- the payoff for innovation and creativity but rations, which may be only weakly correlated also significant persistence and concentration with objective measures of inequality. When in sectors where economic rents are high. 32   RIDING THE WAVE among all its residents, ignoring national Rising inequality, high inequality, boundaries) as captured by the Gini coef- and both ficient has been increasing since 1988. 6 Increased concerns about economic inequality Relatedly, growth over the past three decades are justified in general by trends in objective has favored those at the top of the income inequality in the East Asia and Pacific region, distribution more than those at the bot- as captured by data from household surveys. tom (figure 2.2, panel a), 7 particularly in Overall inequality has been rising, driven China.8 Small differences in growth rates have by increases in large populous countries, resulted in substantial differences in absolute although over time China’s growth moderated (figure 2.2, panel b). Between 1988 and gains ­ its contribution to regional inequality, while 2012, the consumption expenditure of the the opposite happened in the Philippines. top 5 percent of East Asia and Pacific citizens Few countries experienced large drops in increased by almost US$400 per annum, com- inequality of the kinds witnessed over the past pared with less than US$30 for the bottom decade in other parts of the world (especially 20 percent and around US$60 for the median Latin America). In addition, inequality at the group (all figures in 2011 terms of purchasing top is rising, although starting from a low power parity [PPP]). basis. Detailed country-level evidence shows Changes in the aggregate regional distribu- that monetary disparities are underpinned by tion reflect different underlying growth expe- disparities in nonmonetary welfare indicators riences across countries. The cross-country (box 2.2). composition of the different income groups Inequality among residents in develop- has changed significantly. China, for exam- ing East Asia and Pacific (that is, inequality ple, doubled its share in the top 5 percent BOX 2.2  Beyond monetary inequalities: disparities in access to services and curtailed opportunities for future generations across developing East Asia and Pacific Monetary inequalities are underpinned and often results from protracted, rather than acute, malnutri- reinforced by persistent disparities in access to edu- tion and is correlated with poorer achievement later cation and other services, as well as in the quality in life in many areas, from cognitive skills to health of such services (see appendix C, table C.4 for the to income. case of disparities between rural and urban areas in Evidence from Indonesia also highlights the dif- access to services). In Indonesia, for example, more ficulties of reversing overlapping disparities. The than one-quarter of inequality can be explained by influence of birth circumstances (that is, a set of fac- differences in educational attainment across groups. tors over which an individual had no control, among Similarly, the existence of a social gradient in health, which education plays the largest role, followed by with mortality rates declining significantly across being born in an urban area and the province of economic classes, has long been established in the birth—all of which are likely to determine the oppor- international literature (Marmot 2015). These con- tunities for an individual to have access to services) siderations become all the more important when declined from 39 percent for people born in the disparities in access to services (such as education or 1950s to 34 percent for people born in the 1970s, as perinatal care) curtail the opportunities of the disad- access to services increased over the years. Yet despite vantaged, limiting their future economic mobility. In continued expansion of services since the 1980s, the Indonesia, the incidence of children who are stunted, decline in the influence of birth circumstances has or short for their age, is higher than in comparable stalled or even reversed. countries in the region and is particularly high for children of less educated parents. Stunted growth Source: World Bank 2016a. R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    33 FIGURE 2.2  Growth incidence curve and absolute consumption gains for aggregate developing East Asia and Pacific, 1988–2012 a. EAP growth incidence curve, b. Absolute gains in consumption expenditure, 1988–2012 1988–2012 6.5 400 Annual growth rate (2011 PPP US$), % Gain in annual expenditure per year 6.0 300 (2011 PPP US$) 5.5 200 5.0 4.5 100 4.0 0 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Consumption expenditure distribution Consumption expenditure distribution for percentile groups for percentile groups Source: Jirasavetakul and Lakner 2017. Note: The horizontal axis in panel a shows a growth rate in mean consumption of 5.54 percent. The horizontal axis in panel b shows an absolute gain in developing East Asia and Pacific mean consumption of approximately US$88 (2011 PPP adjusted). EAP = East Asia and Pacific; PPP = purchasing power parity. For each fractile group, panel a shows the annualized growth rate in average consumption expenditure and panel b shows the absolute gain. The horizontal lines show the respective values for the regional mean: a growth rate in mean consumption of 5.6 percent in panel a, and an absolute gain in mean consumption of approximately US$88 (2011 PPP adjusted). The distribution is evaluated at 20 ventile groups (for instance, the bottom 5 percent). Because of churning, these growth incidence curves must be interpreted with caution as, say, the bottom 5 percent of the population at the beginning of the period might be composed of a different set of individuals at the two points in time for which the growth incidence curve is constructed. Population weights are used throughout. The database includes microdata and grouped data in combination with a parametric Lorenz curve. of the distribution between 1988 and 2012 panels a and b). In Indonesia, inequality ­(figure 2.3). increased significantly after 2000, after remain- The increase in aggregate regional ing steady for the previous 15 years. Inequality inequality between 1988 and 2012 was in China remained stable since the early 2000s driven by increased within-country inequal- but at a much higher level than at the begin- ity in large countries, particularly in China ning of its expansion in the early 1980s. and Indonesia.9 Conversely, the impact of Likewise, over the same period, inequality between-country inequality has declined, rose significantly in Lao People’s Democratic largely reflecting China’s rapid convergence Republic and remained high in Malaysia and toward regional average consumption levels the Philippines. Thailand and—based on lim- (box 2.3). That pattern contrasts with pat- ited data—Papua New Guinea were the only terns in other regional analyses.10 countries where inequality displayed a signifi- Analysis at the country level enables a more cant downward trend. detailed understanding of the rise of inequal- Spatial disparities—between regions ity at the regional level. The unweighted mean and between rural and urban areas—have of country-level Gini coefficients, which treats been a key factor behind inequality at the each country as one observation, remained country level, especially in China (World approximately unchanged from 2002 to 2014. Bank 2015a; World Bank and Development Yet in many countries in the region, particu- Research Center of the State Council, larly some of the most populous ones, inequal- People’s Republic of China 2014). The rural- ity increased or remained high (figure 2.4, urban divide now accounts for 45 percent of 34   RIDING THE WAVE FIGURE 2.3  Regional composition of developing East Asia and overall inequality in China (Kanbur, Rhee, Pacific distribution of consumption expenditure, 1988 and 2012 and Zhuang 2014). In other countries, rural- urban disparities are much more contained, a. 1988 accounting for 10–20 percent of economy- 1.0 wide inequality.11 Similar results also broadly Share of ventile population, % apply to inequalities between regions. Those 0.8 spatial inequalities and the unequal distribu- 0.6 tion of economic activity may reflect positive or neutral factors, such as higher productiv- 0.4 ity in parts of a country owing to economies of agglomeration12 or the size distribution 0.2 of cities.13 Or they may reflect institutional constraints such as significant de jure and de 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 facto restrictions on internal migration, or Ventile of PPP-adjusted 2011 US$ distortions in land markets14 and barriers to connectivity. b. 2012 Overall, spatial disparities within coun- 1.0 tries are compatible with, and indeed may be Share of ventile population, % instrumental to, the achievement of contin- 0.8 ued development and long-run convergence 0.6 in living standards (World Bank 2009). Such convergence is not automatic, however. It 0.4 requires policies that focus on tackling the obstacles hampering lagging regions. A key 0.2 question is whether poor regions are poor because of their intrinsic characteristics or 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 because of the characteristics of those who Ventile of PPP-adjusted 2011 US$ (choose to) live there (Ravallion and Wodon 1999). A useful way to analyze the issue is China Indonesia Philippines Other East Asia to frame it in terms of whether spatial dis- Source: Jirasavetakul and Lakner 2017. parities reflect differences in household Note: PPP = purchasing power parity. For ease of interpretation, the figure does not show Pacific characteristics, or rather the returns to such countries separately, although they are included in the calculations. Population weights are used throughout. characteristics. The former would suggest BOX 2.3  China as a driver of inequality trends across developing East Asia and Pacific China accounts for 70 percent of the population of Another way to visualize the impact is to sepa- East Asia and Pacific, so its income dynamics strongly rate regional inequality into its within-country and shape regional inequality. Between 1988 and 2012, between-country components (figure B2.3.2). By China’s average consumption moved from below 2012, China was reducing between-country inequal- mean consumption to above mean consumption for ity, and thus overall regional inequality, but rising the region (figure B2.3.1). That change helped equal- inequality in China had made a major contribution ize the regional distribution of consumption. to inequality across the East Asia and Pacific region. box continues next page R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    35 BOX 2.3  China as a driver of inequality trends across developing East Asia and Pacific (continued) FIGURE B2.3.1  Developing East Asia and Pacific region distribution of consumption expenditure, by subregion, 1988–2012 a. 1988 b. 1992 Density of log consumption Density of log consumption .8 .6 .6 .4 .4 .2 .2 0 0 0 0 0 400 500 0 1,050 00 00 4,000 6,0 0 8, 00 10 000 15 00 20 00 30 0 0 0 0 0 400 500 0 1,050 00 00 4,000 6,0 0 8, 00 10 000 15 00 20 00 30 0 0 10 15 20 30 0 ,00 ,00 10 15 20 30 0 ,00 ,00 7 2,0 3,0 ,0 ,0 7 2,0 3,0 ,0 ,0 PPP-adjusted 2011 US$ (annual) PPP-adjusted 2011 US$ (annual) c. 1997 d. 2002 Density of log consumption Density of log consumption .6 .5 .4 .4 .3 .2 .2 .1 0 0 0 0 0 400 500 0 1,050 00 00 4,000 6,0 0 8, 00 10 000 15 00 20 00 30 0 0 0 0 0 400 500 0 1,050 00 00 4,000 6,0 0 8, 00 10 000 15 00 20 00 30 0 0 10 15 20 30 0 ,00 ,00 10 15 20 30 0 ,00 ,00 7 2,0 3,0 ,0 ,0 7 2,0 3,0 ,0 ,0 PPP-adjusted 2011 US$ (annual) PPP-adjusted 2011 US$ (annual) e. 2007 f. 2012 Density of log consumption Density of log consumption .5 .5 .4 .4 .3 .3 .2 .2 .1 .1 0 0 0 0 0 400 500 0 1,050 00 00 4,000 6,0 0 8, 00 10 000 15 00 20 00 30 0 0 0 0 0 400 500 0 1,0 0 00 00 4,000 6,0 0 8,000 10 00 15 00 20 00 30 0 0 10 15 20 30 0 ,00 ,00 10 15 20 30 75 0 ,00 ,00 7 2,0 3,0 ,0 ,0 2,0 3,0 ,0 ,0 PPP-adjusted 2011 US$ (annual) PPP-adjusted 2011 US$ (annual) China Indonesia Philippines Other East Asia Source: Jirasavetakul and Lakner 2017. Note: PPP = purchasing power parity. Population weights are used throughout. box continues next page 36   RIDING THE WAVE BOX 2.3  China as a driver of inequality trends across developing East Asia and Pacific (continued) FIGURE B2.3.2  Contribution of within- and between-country differences to inequality across developing East Asia and Pacific, 1988–2012 1988 1992 1997 Year 2002 2007 2012 −0.1 0 0.1 0.2 0.3 0.4 Mean log deviation China, Indonesia, Philippines, Other East Asia, Pacific countries, between-group between-group between-group between-group between-group China, Indonesia, Philippines, Other East Asia, Pacific countries, within-group within-group within-group within-group within-group Source: Jirasavetakul and Lakner 2017. Note: Figure shows the components of a country-level decomposition of East Asia and Pacific–wide inequality (as measured by the mean log deviation ). that public policy should focus on investing Cambodia, Indonesia, Lao PDR, and many in people, the latter that tackling disparities Pacific Island countries, differences in returns may require area-specific interventions that have increased over time. would equalize returns across geographi- cal areas, such as improving infrastructure Income and wealth are increasingly and connectivity and reducing the cost of concentrated at the top migration. In general, characteristics and returns As significant as measured increases in both play a role in the East Asia and Pacific inequality in several countries may be, those region, although their relative importance increases are based on data from household varies by country (figure 2.5). Much of the surveys, which tend to underestimate inequal- rural-urban welfare gap, at the level of both ity; also, new research suggests that wealth countries and their individual subregions, is is even more concentrated in the hands of due to differences in characteristics, particu- the richest than previously thought.15 For larly occupation and education. In Thailand example, combining data on the top tail from and Vietnam, the impact of differences in tax records with data on the rest of the dis- returns has diminished over time even if it tribution from a 2013 household survey for remains significant, suggesting that there has Indonesia raises the estimated Gini coefficient been progress in integrating markets. But in from 41 percent to 46 percent (Wai-Poi and R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    37 FIGURE 2.4  Inequality across developing East FIGURE 2.5  Welfare disparities between rural and urban areas for Asia and Pacific, 2002–14 select East Asia and Pacific countries: characteristics versus returns, 2000 and 2010 a. Higher-income economies 50 70 60 Percentage of urban-rural 45 50 welfare gap 40 40 Gini index 30 35 20 30 10 0 25 2000 2010 2000 2010 2000 2010 2000 2010 2000 2010 2002 2005 2008 2011 2014 Cambodia Indonesia Lao PDR Thailand Vietnam Year Characteristics Returns China Fiji Indonesia Malaysia Mongolia Philippines Thailand Sources: Lozano-Gracia, Walker, and Antos 2016. Note: Welfare is measured as the log of (consumption versus regional poverty line). The rural-urban b. Lower-income economies welfare gap is measured as (urban welfare minus rural welfare). This gap is decomposed, using the Blinder–Oaxaca procedure, into the contribution of differences between rural and urban 50 areas in (a) the characteristics of households, such as education, versus (b) the returns to such characteristics, such as the returns to education. 45 than in the United States, even though most 40 Gini index countries for which data are available have seen an increase in top-income shares since 35 the Asian financial crisis in 1997. The evo- lution of the income of the top 1.0 percent 30 figure 2.6, panel a) is broadly similar to that (­ of the income of the top 0.1 percent (figure 25 2.6, panel b). 2002 2005 2008 2011 2014 Wealth inequality at the top is high in sev- Year eral East Asia and Pacific economies. China Cambodia Lao PDR Papua New Guinea and Hong Kong SAR, China, recorded the Timor-Leste Vietnam largest increases since 2000 among all econ- Sources: World Bank EAP Team for Statistical Development, for EAP omies in the region (Credit Suisse Research 2002–14; World Bank PovCalNet (database), http://iresearch.worldbank​ Institute 2015). In contrast, the wealth .org/PovcalNet​/povOnDemand.aspx; World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world- share of the top percentile fell in Japan development​-indicators. and Singapore (low- to middle-inequality countries), as well as in Malaysia and the others, forthcoming). Although such detailed Philippines (high-inequality countries). evidence is hard to come by, a variety of data East Asia and Pacific is the region of the sources beyond standard household sur- world where billionaire wealth has been veys provide useful pointers on the extent of increasing most rapidly—30 percent annu- wealth concentration at the top for the East ally between 2002 and 2014.16 Despite the Asia and Pacific region. strong increase, the region still accounts for According to the World Wealth and only around 10 percent of global billionaire Income Database, inequality at the top of the wealth. As in almost all regions, billionaire distribution tends to be lower in East Asia wealth in East Asia and Pacific has been 38   RIDING THE WAVE FIGURE 2.6  Income share of the top 1 percent and of the top SAR, China; the Malaysia; the Philippines; 0.1 percent: selected economies of East Asia compared with the Singapore; Taiwan, China; and Thailand—are United States, 1980–2015 among the top 20 global economies in billion- aire wealth compared with GDP. In Thailand, a. Income share of top 1 percent since 1980 the combined wealth of the country’s billion- 20 aires equals around 10 percent of GDP. Although these different sources point to rising concentrations of wealth, albeit from a 15 low base, there is significant debate on what Income share (%) to make of this finding. A growing class of bil- lionaires, for example, may be seen as the result of innovation and dynamism being rewarded. 10 If that dynamism solidifies over time into elite capture and wealth concentrated in sectors from which high rents could be extracted, however, the growth-with-equity model may 5 face a worrisome challenge (box 2.4). This lat- 1980 1985 1990 1995 2000 2005 2010 2015 ter concern is particularly strong with regard Year to large family firms or family conglomerates China Japan Korea, Rep. Singapore with strong political connections—one of the Taiwan, China United States key channels for concentration of wealth. Such firms may have been able to solve serious coor- b. Income share of top 0.1 percent since 1980 dination failures in underdeveloped markets when many East Asia and Pacific economies 8 took off, but they may no longer be appropri- ate for the new challenges ahead. Without adjudicating whether the concen- tration of high incomes at the top is good or Income share (%) 6 bad, especially given that the situation most likely varies by country and sector of activ- ity, two main elements of this concentration 4 of wealth are worth highlighting. First, it is difficult to draw a clear picture of overall inequality without considering both informa- tion on top incomes and standard inequality 2 measures. Yet in none of the 10 economies for which both household survey estimates and 1980 1985 1990 1995 2000 2005 2010 2015 Year top-income data are available do all measures point toward decreasing inequality.17 Second, China Japan Korea, Rep. Singapore Taiwan, China United States the growing concentration of wealth is likely to shape perceptions of overall inequality Source: World Wealth and Income Database, http://www.wid.world. (accessed March 17, 2017) when it fuels very conspicuous consumption. Note: Estimates are typically derived from tax records data. The United States is included for comparison. Indonesia was excluded because of wide fluctuations in top income. For example, data on sales of high-perfor- mance Italian cars in China, Malaysia, and Thailand show that spending grew faster than growing faster than gross domestic product GDP between 2000 and 2014.18 Similarly, (GDP). Today, billionaire wealth is equivalent consumer spending on Swiss watches in East to almost 9 percent of the GDP of the East Asia and Pacific grew faster than GDP over the Asia and Pacific economies that have billion- same period, and the average value of Swiss aires. Six East Asian economies—Hong Kong watches sold in the region almost doubled. R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    39 BOX 2.4  Growth of the super rich across developing East Asia and Pacific: economic dynamism or elite capture? The growth of a very rich segment of the popula- may generate inefficiencies and create a potential for tion may be seen as a sign of economic dynamism. elite capture. Evidence from a broader group of rich In China, a new class of high-net-worth individuals people than only the billionaires analyzed by Freund with investable assets in excess of RMB10 million (or (2016)a suggests that even companies founded by approximately US$1.6 million) has emerged, and this entrepreneurs show a tendency to become family segment is distributed along a wide spectrum of areas dynasties and that large family firms are more likely of the country. These individuals made their fortunes to enjoy more plentiful political connections than in modern sectors such as information technology equally large publicly held firms. Similarly, regarding and biotechnology (Zeng, Shang, and Chishty 2015). the potential for elite capture, some indication can be They appear to have a different outlook than that of obtained by classifying billionaires’ sources of wealth their more established older counterparts, thereby on the basis of whether they are primarily in sectors being more keen to invest in new industries rather linked to rent extractionb or whether they became than traditional ones and paying close attention to rich through sectors that benefit from globalization. “government policies regarding domestic reforms, In the average developing country in East Asia and market openings and innovation” (Zeng, Shang, and Pacific, 42 percent of billionaire wealth is generated Chishty 2015, 11). A similar assessment is made of in rent-seeking sectors. Further, on the basis of the the new billionaires on the Forbes list; increasingly, size of rent-seeking sectors as a share of GDP, 6 of the they are wealth creators—company founders outside top 10 economies in the world are in East Asia. Firms of finance and company executives or owners—as that operate in rent-seeking industries are more likely opposed to individuals who are politically connected to have family members who hold political office, as or who inherited their wealth (Freund 2016). In that shown by evidence for Thailand (Bunkanwanicha light, as of 2014, emerging East Asia and Pacific was and Wiwattanakantang 2008). doing particularly well, with 48 percent of billion- aires identified as company founders (nonfinancial a. This analysis was based on a sample selected from the 200 largest firms listed sectors) or as owners and executives. on the main stock exchanges of nine East Asian economies in 1996 and 2008 (see Carney and Child 2013; Claessens, Djankov, and Lang 2000). Yet there are concerns that concentrated owner- b. Based on the definition of rent-seeking sectors used by the Economist’s 2014 ship structures, even if initially spurred by innovation, index of crony capitalism. Factors that were once by features of the East Asia and Pacific opportunities are turning region’s past development model, including into challenges the lack of a systematic approach to social protection in many countries, low levels of Several global and regional developments resource mobilization, and regressive taxa- suggest that it will be difficult for countries tion (see chapter 1). Unless countries address in developing East Asia and Pacific to sus- those challenges, their abilities to support tain (or achieve) high levels of growth fol- long-term inclusive growth will progressively lowing their recent period of broad-based, narrow. oriented development. labor-intensive, export-­ Manufacturing-based growth is threatened by a slowed expansion in trade and u ­ ncertain Can export-oriented manufacturing prospects for global value chains. The demo- still support growth with equity? graphic transition and urbanization, which have been sources of opportunity in the Although openness and absorption of new past, are now turning into potential threats. technologies continue to be important for These challenges are likely to be compounded the region, particularly for lower-income 40   RIDING THE WAVE countries such as Cambodia, Lao PDR, and no longer be feasible under international trade Myanmar, the global environment appears rules. If lower-income countries are not able less conducive to manufacturing-based to rely on export-oriented manufacturing as export-led development. an engine for growth with equity, they might Current forecasts suggest a reversal in the find themselves relying on exports from other patterns observed in the past 15 years, with sectors, such as natural resources and services, growth in trade slowing below expected GDP which tend not to create significant opportu- growth (World Bank 2017b). Slower growth nities for broad segments of the population. appears to be driven by the continuation of While dark clouds gather on the horizon factors that have manifested themselves over for lower-income countries, richer countries the past few years, both cyclical (such as the face threats of their own that will require lack of buoyancy in the world economy) ­ g overnment policies designed to support and structural (such as China’s shift from an growth. Technological innovation is disrupting export- and investment-led growth model to production patterns and business models world- a consumption-led one). Growing protec- wide, which could lead to what Rodrik (2015) tionism also is part of the picture, although termed “Premature Deindustrialization.” it has not yet played a major role in driving Thus, while low wages remain a source of global trends. comparative advantage, manufacturing is Other factors are less clear, such as the becoming more skill intensive and increasingly potential for further expansion of global automated. Those changes may offer opportu- supply chains to countries that so far have nities for some workers, particularly those with not been able to participate. Although the better skills, but may render many jobs out of decline in the costs of doing business across reach of others, unless properly managed. The borders has likely stalled at the global level, sustainment of high economic growth in upper- lower-income countries in East Asia and middle-income countries will involve navigating Pacific may be able to reap the benefits of three key transitions to avoid a middle-income China shifting some of these activities such trap (Gill and Kharas 2015): as electronics, as its labor costs continue to •  The structural transformation transition increase, to neighboring countries with lower tightens the market for unskilled labor wages. Given current infrastructure and skill when the supply of surplus labor from constraints, however, a concerted policy effort agriculture starts to disappear. may be needed to realize these opportunities •  The technological upgrading transi- (World Bank 2015b). In particular, global tion centers on the need to move up the value chains increase the importance of both value chain. This transition involves the efficient logistics and a favorable business intraindustry reallocation of resources, environment. While Malaysia has been able and it differs from earlier phases of the to improve its logistics performance and cur- growth process, when moving into new rently ranks 32nd in the World Bank’s global industries was key to growing productiv- benchmarking tool, the Logistics Performance ity. A few countries in the region, such as Index (LPI), other countries have found it dif- Singapore, have managed this transition ficult to keep up. Indonesia’s ranking fell from well by upgrading their skill base. 43rd in 2007 to 63rd in the 2016 LPI. •  The institutional transition involves the These developments are a concern for development of effective and responsive lower-income countries in the East Asia and government bureaucracies, particularly Pacific region, not least as China continues as the weight of government structures to be a significant source of competition for becomes sizable. would-be industrializers. In addition, some of the export-promoting measures adopted Together, these challenges suggest that by previous industrializers, including export lower- and higher-middle-income countries subsidies and currency undervaluation, may in the East Asia and Pacific region may face R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    41 lower overall economic growth as they seek young, and it will take much longer for those to adjust to the new global environment and countries to experience significant increases in seize its opportunities. In addition, two long- the old-age dependency ratio.20 term global trends that are particularly salient In addition to slowing economic growth, for the region and that have been a source of population aging is likely to increase house- opportunity in the past will complicate the holds’ vulnerability to shocks, particularly picture and necessitate further adjustments health shocks.21 This effect is especially found to the economic growth model: population in rural areas, where population aging tends aging and urbanization. to accompany shrinking family size and migration, as well as a weakening of tradi- tional support networks.22 The lack of pen- Population aging could slow economic sion coverage and insufficient accumulated growth and increase household savings mean that the elderly, especially those vulnerability who live in rural areas, tend to continue The populations of developing East Asia working until they are very old. In addition, and Pacific are aging rapidly—a reflection the lack of well-developed social insurance of sharp declines in fertility and increased systems diminishes households’ ex post ability longevity—and the topic is a major chal- to manage health and income shocks in both lenge worthy of its own report (World Bank urban and rural areas, particularly where 2016b). China alone is already home to more informal employment represents a large share people age 65 and older than any developing of total employment. region in the world. In many countries, popu- From a policy viewpoint, the key concern is lation aging is occurring at low-income levels. that much of the East Asia and Pacific region Although in the past a large share of working- is getting old before getting rich. Social insur- age people provided a demographic dividend ance systems are not sufficiently developed, that helped support the region’s growth, the nor have enough resources accumulated, to demographic dividend in most countries is provide the elderly population with suffi- about to reverse. By 2025, it is expected that cient income to avoid poverty.23 Even where the size of the region’s working-age popula- specific programs for the poorest elderly are tion will have begun to decrease. The impact in place, such as in China, in the near term, of that decrease may be intensified by the fact program coverage will need to be expanded that most countries in East Asia and Pacific significantly to address demographic trends. lack the requisite mechanisms to adjust their policies to accommodate longer work lives Urbanization may entrench inequalities (World Bank 2014a). Both the pattern and the pace of demo- Developing East Asia and Pacific is urbanizing graphic change differ widely by ­ c ountry rapidly, though there is significant variation (figure 2.7). A group of countries is com- across countries.24 Indeed, the increase in the posed of the rich, already aged economies of region’s urban population from 2000 to 2010 Hong Kong SAR, China; Japan; Korea; and was enough to constitute the equivalent of Singapore (panel a); in these economies, the the sixth-largest country in the world (World old-age dependency ratio will continue to rise Bank 2015a). Urbanization is expected to to unprecedented levels (Chomik and Piggott continue.25 In many countries, such as Japan, 2015). A second group comprises ­ rapidly in the newly industrialized economies, and in aging middle-income economies—China, many advanced economies, urbanization has Indonesia, Malaysia, Mongolia, Thailand, and been an opportunity for inclusive economic Vietnam (panel b).19 A third group of gener- growth, rather than a threat to it. The rapid ally poorer countries—Cambodia, Lao PDR, growth of cities has generated significant Myanmar, Papua New Guinea, the Philippines, agglomeration economies, enabling finer spe- and Pacific Island countries (panel c)—is still cialization and greater competition in labor 42   RIDING THE WAVE FIGURE 2.7  Population aging across developing East Asia and Pacific (Old-age dependency ratio [population age 65 and older to population age 25 to 64]) a. High-income economies 90 80 Old-age dependency ratio population aged 25–64) (population aged 65+ to 70 60 50 40 30 20 10 0 50 55 60 65 70 75 80 85 90 95 00 05 10 15 20 25 30 35 40 45 50 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 Year Hong Kong SAR, China Japan Korea, Rep. Singapore b. Medium-income economies 70 60 Old-age dependency ratio population aged 25–64) (population aged 65+ to 50 40 30 20 10 0 50 55 60 65 70 75 80 85 90 95 00 05 10 15 20 25 30 35 40 45 50 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 Year China Indonesia Malaysia Mongolia Thailand Vietnam c. Lower-income economies 30 25 Old-age dependency ratio population aged 25–64) (population aged 65+ to 20 15 10 5 0 50 55 60 65 70 75 80 85 90 95 00 05 10 15 20 25 30 35 40 45 50 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 Year Cambodia Fiji Kiribati Lao PDR Micronesia, Fed. Sts. Myanmar Papua New Guinea Philippines Solomon Islands Tonga Vanuatu Source: United Nations, United Nations Department of Economic and Social Affairs, Population Division, 2014. Note: Medium variant population projection. R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    43 and product markets, as well as increased the region’s urban poor are vulnerable to knowledge spillovers. In turn, that growth economic shocks and especially vulnerable has enhanced productivity and has spurred to natural disasters, which are frequent in innovation and economic diversification. As East Asia and Pacific and cause significant a result, urbanization has been accompanied economic disruption. 27 The vulnerability by broad-based improvements in living stan- of the urban poor is compounded by their dards and has helped support the expanded dependence on cash income, the higher- provision of basic services across all areas. risk locations where they often live (includ- Yet in the East Asia Pacific region, there ing hillsides and flood-prone areas), and are growing concerns over the pattern of their limited access to social services and growth in urban areas and the role growth infrastructure. can play in entrenching inequalities (Baker Certain groups, such as women, the elderly, and Gadgil 2017). These concerns are not and migrants, face specific sources of exclu- just because the concentration of the poor in sion and vulnerability (World Bank 2017). urban areas is increasing with urbanization, Migrants might remain trapped in urban as people are attracted to urban areas by the poverty if they are not fully integrated into opportunities and services that cities provide. modern urban labor markets. In some cases, Rapid urban growth can be accompanied by particularly in China, regulations on migra- congestion, environmental strains, and urban tion and residence permits prevent migrants squalor. Given typical urbanization patterns from gaining access to the same types of jobs in the region—as elsewhere—the less urban- and social services as other urban residents ized, lower-middle-income countries of today (box 2.5). Migrants earn systematically less are likely to see large increases in the popula- than nonmigrants in urban centers because tion of their large urban agglomerations. As of such restrictions.28 Urbanization may also a result, more people will be living in con- weaken extended family structures and tra- gested spaces, with poor access to services. In ditional support networks, compounding the contrast, upper-middle-income countries will challenges posed by population aging29 and see large expansions in built-up areas as cit- thus reducing living standards on both sides ies spread to accommodate the needs of their of the urban-rural divide. Furthermore, where large populations. urbanization is associated with highly vis- It is of particular concern that an esti- ible intraurban and urban-rural disparities, it mated 250 million people in East Asia and is likely to reinforce perceptions of growing Pacific today still live in slums.26 In addition, inequality. BOX 2.5  Regulatory barriers to migration Internal migration has played a key role in reducing and emotional costs of moving, as well as by regula- poverty in many East Asia and Pacific countries by tory barriers. Registration requirements that assign reallocating resources from agriculture to higher- people to specific rural or urban locations and limit productivity sectors, such as export-driven, low-cost their access to social services to those locations are manufacturing in China. Yet the movement of peo- the most obvious barriers. ple to different parts of a country and the creation Since 1958, China has enforced a household of new jobs can be hindered by the monetary, social, registration (hukou) system that assigns individuals box continues next page 44   RIDING THE WAVE BOX 2.5  Regulatory barriers to migration (continued) an agricultural (rural) or nonagricultural (urban) FIGURE B2.5.1  Number of migrants in Vietnam location, along with either farmland or jobs, hous- ing, food, and state-sponsored benefits (Fan 2008). 3.0 80 Similar registration requirements are in force in 70 Percentage of population 2.5 Vietnam. Although strict application of the original 60 Millions of people legislation made it essentially impossible for rural 2.0 50 migrants in China to survive in urban areas, reforms 1.5 40 of the hukou since the mid-1980sa have resulted in 30 1.0 an estimated 260 million people moving to the cit- 20 ies (National Bureau of Statistics of China 2010). In 0.5 10 fact, “in 2011, the State Council encouraged the reg- 0 0 istration of temporary populations in cities [and] the Binh HCMC Hanoi Da Nang adoption of a residence permit system, and [it] for- Duong mulated hukou reform policy that was linked to the city’s administrative level” (World Bank 2017b, 92). People (left y-axis) Percentage of population (right y-axis) This led to large regulatory disparities between cities. The current system in China appears to have cre- Source: World Bank and Ministry of Planning and Investment of ated a group of second-class citizens who benefit Vietnam 2016. from accessing better earning opportunities in the Note: HCMC = Ho Chi Minh City. cities than they would have at home, but who have problems accessing (or must pay more to access) urban services. On average, migrant workers work paid holidays, and some did not even have week- nine hours longer per week than their urban peers, ends off (Wang and Chen 2010). but their mean monthly wage is only 76 percent A similar situation exists in Vietnam under the ho of that of other urban workers (Li, Li, and Chen khau system. More than 5 million Vietnamese do not 2010). Migrants and urban citizens are treated dif- have what is called permanent registration where they ferently in terms of political, economic, and social live (figure B2.5.1). As in China, citizens in major rights, particularly in accessing education and Vietnamese urban centers who do not enjoy perma- health services. Thus, the net subsidy for migrants nent registration face higher living costs and difficul- can be negative, as any public transfers they receive ties accessing health care, education, social protection, are offset by the taxes and fees they are required to and utilities, and they experience discrimination in pay (Khan and Riskin 2005). An analysis using data employment and social connections. To obtain per- for 2002 from the China Household Income Project manent registration, it is often necessary for people to (Gao and Riskin 2009) revealed that only 5 percent make large unofficial payments to local officials. of migrant workers were covered by the country’s a. Changes introduced at the city level included high fees for migrants from pension system, less than 5 percent were covered by rural areas to obtain a city hukou and to access urban services (Fan 2008). unemployment insurance, only 3 percent were cov- Later, a different type of hukuo, known as blue-stamp hukou, was introduced ered by medical insurance, and less than 10 percent in Shanghai and Shenzhen for migrants who met high skill requirements and who had sizeable resources that could be invested (Wong and Wai-po 1998). benefited from public housing. Moreover, migrants The blue-stamp hukou could then be converted into a permanent urban hukou in Beijing, Guangzhou, Shanghai, and Tianjin were after a specified period of time. In 2001, the state council approved a scheme to discriminated against in obtaining unemployment grant urban hukou to migrants from rural areas who held stable jobs and who had resided in small cities or towns for more than two years (Fan 2008). Later, insurance, pensions, medical insurance, and work- in 2003, the state council issued a directive that affirmed the rights of rural place injury insurance. Most did not receive any migrants to work in cities (Cai 2003). R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    45 Notes reasonable parameter values have all found increasing inequality. For a fuller discussion, 1. As in much of the literature, here inequality see Jirasavetakul and Lakner (2017). is defined as being high if the Gini index 7. These patterns have changed over time. exceeds 40. Between 2007 and 2012, the growth 2. Among developing regions, East Asia and incidence curve for the region was almost flat Pacific has the second-highest share (after (Jirasavetakul and Lakner 2017). Latin America and the Caribbean) of 8. These findings also broadly hold over different respondents who strongly perceive income subperiods, with the exception that those at inequality as being too high in their country the very top of the income distribution saw (in the International Social Survey Program, large income losses immediately following the ISSP, respondents were asked whether they 1997–98 Asian financial crisis. agree or disagree on a 5-point scale that 9. Together, these two countries account for “differences in income are too large,” Weise most of the region’s population and therefore 2016, 2) and whether incomes in their country receive a relatively high weight in calculations should be made more equal. In the World of regional inequality. Value Survey, respondents were asked to 10. A similar analysis for Africa (Jirasavetakul identify on a 10-point scale whether “incomes and Lakner 2017) shows that, in contrast should be made more equal” or whether “we to East Asia and Pacific, the within-country need larger income differences as incentives component gradually and continuously for individual effort” (Weise 2016, 2). declined, although, similarly to East Asia 3. When shown different types of distribution, and Pacific, it still explains most of overall East Asians preferred distributions that Africa-wide inequality—around 60 percent. would result in Gini indexes of 20 percent to Findings for East Asia and Pacific are 25 percent (Tiongson and Weise 2016, using also in contrast with global findings that ISSP data). The preference for greater equality between-country differences remain the is not rooted in a correct assessment of current dominant source of inequality even if the inequality levels, as respondents consistently overall weight of within-country inequality underestimated true inequality (although less increased (Lakner and Milanovic 2016). It is so in richer countries such as Japan). perhaps not surprising that between-country 4. Results from the World Values Survey differences matter more at the global level (2005–14) show that respondents gravitate compared with an analysis at the regional toward “incomes should be made more equal” level such as the present one (countries’ rather than “we need large income differences average income or consumption levels may as incentives for individual efforts.” However, be expected to be more similar within a in some cases, such as China, respondents region). For details, see Jirasavetakul and display a full reversal of preferences toward Lakner (2017). greater equality, between 2005 and 2014 11. This is the case in, for example, Cambodia, (Weise 2016). Lao PDR, the Philippines, Thailand, 5. Relatedly, the increases in inequality that and Vietnam. These results are based on were part and parcel of China’s growth boom decompositions of the Theil T index (that is, in the 1990s and early 2000s are, in part, GE(1)). thought to be the cause of the sharp declines 12. This is the case when, for example, certain in life satisfaction at the time that growth industries locate close to one another to took off (Graham, Chattopadhyay, and exploit the economies of scale and scope that Zhang 2016, 10). can result from proximity to one another and 6. As the Lorenz curve for 2012 crosses to the to large markets. left of the one for 1988 at around the 95th 13. For instance, an urban distribution character­ percentile, this increase in inequality appears ized by many small towns, closely integrated to be driven by an increase in inequality at the with the rural economy, would result in much bottom of the distribution. Although there smaller urban-rural disparities than the same was no strict Lorenz dominance between rate of urbanization concentrated in a few 1988 and 2012, inequality measures with large urban agglomerations. 46   RIDING THE WAVE 14. For example, in China, inadequate property of such an Italian car was around 80 times rights for farmers, low levels of compensation GDP per capita, but the cost fell to around for land requisition, and nonmarket-based 30 times GDP per capita by 2014. land pricing have exacerbated urban-rural 19. For example, in China, low fertility and inequalities of income and wealth and have declining mortality are translating into a reduced the efficiency of both urban and rural rapidly aging society. In 2013, there were land use. For details, see World Bank and 202 million people over the age of 60, Development Research Center of the State accounting for 15 percent of the country’s Council, People’s Republic of China (2014). total population. By 2030, the number 15. Household surveys tend to underestimate is expected to double in absolute terms, overall inequality for several reasons. Rich accounting for 24 percent of the total people are less likely to open the door to population (World Bank 2016b). an interviewer or to respond directly, or 20. In the first group, the old-age dependency truthfully, to a questionnaire. In addition, ratio (that is, the ratio of people ages 65 such surveys are likely to fail to capture and older to those ages 15 to 64) will reach rare income events or income (and wealth 58 percent to 70 percent by 2050. This is from it) that is obtained illegally or hidden well above the value for European countries offshore. Furthermore, the sampling frame (47 percent) or the United States (37 percent). of a typical household survey is not designed In the second group, the ratio will reach to be representative for the extreme upper 30 percent to 43 percent, with China at the tail, so any estimates obtained from a typical upper end. In the third group, the ratio will household survey would be imprecise. It broadly remain below 25 percent. is worth underscoring that, because rich 21. Older workers, even with high human people actually spend only a small fraction capital, are relatively less likely to find work of their income, household surveys that following negative shocks, especially to their use consumption expenditure data report health. lower estimates of inequality than if income 22. For instance, coresidence with children has were measured. Entrepreneurial and capital been declining rapidly in China, Korea, and incomes, which are important for those at Thailand (Giles and Huang 2015). the top, also are difficult to capture, even in 23. For an analysis of policy issues related to a well-designed household survey. See Lakner pension systems and old-age protections in the and Ruggeri Laderchi 2016. region, see World Bank (2016b). 16. See the 2016 Forbes billionaires list, https:// 24. The most urbanized countries in the region www.forbes.com/billionaires/list/. are Malaysia (73 percent) and Mongolia 17. The East Asia and Pacific economies that (71 percent). Urbanization rates are below show an increase in both the Gini index and 30 percent in 8 of 19 countries for which data all available top income or wealth measures are available. are China; Indonesia; Korea; and Taiwan, 25. For example, in Latin America, a region China. For all other economies for which of mostly middle-income countries, the both household surveys and top-income data urbanization rate equals approximately are available, the conclusions are mixed, with 80 percent, suggesting the urbanization rate household surveys and (at least some) top- of the East Asia and Pacific region is lower income data going in opposite directions. than would be expected given its level of For instance, for Thailand, the Gini measure development. China, Indonesia, and the declined by 0.7 percent annually between Philippines are projected to reach urbanization 2002 and 2012, whereas billionaire wealth rates between 60 percent and 70 percent by in GDP grew 16.9 percent annually between 2050, which implies that those countries will 2002 and 2014. For more details, see Lakner experience further rural-urban migration and Ruggeri Laderchi 2016. of around 350 million people. Among the 18. This spending on high-performance Italian other regional economies, only Thailand is cars has resulted in a rising share of luxury projected to reach an urbanization rate above cars’ total import value to GDP, even though 50 percent by 2050 (United Nations, United rising incomes have made the cars much more Nations Department of Economic and Social affordable. In China, the average cost in 2000 Affairs, Population Division 2014). R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    47 26. The share of the urban population living in Carney, R. W., and T. B. Child. 2013. “Changes slums has declined, but it remains higher to the Ownership and Control of East Asian than in large Latin American countries. Corporations between 1996 and 2008: The Between 1950 and 2009, the share declined Primacy of Politics.” Journal of Financial in China from 41 percent to 29 percent, in Economics 107 (2): 494–513. Indonesia from 43 percent to 23 percent, Chomik, R., and J. Piggott. 2015. “Population in the Philippines from 50 percent to Ageing and Social Security in Asia.” 40 percent, and in Vietnam from 54 percent Asian Economic Policy Review 10 (2): to 35 percent. For comparison, the share of 199–222. the urban population that lives in slums is Claessens, S., S. Djankov, and L. H. P. Lang. 2000. estimated to be only 14 percent in Mexico, “The Separation of Ownership and Control in 21 percent in Argentina, and 27 percent in East Asian Corporations.” Journal of Financial Brazil (Baker and Gadgil 2017). Slums are Economics 58 (1–2): 81–112. defined by residents’ access to improved Credit Suisse Research Institute. 2015. Global water, improved sanitation, durable housing, Wealth Report 2015. Zurich: Credit Suisse and sufficient living area (UN-Habitat 2013, Research Institute. table 2). Easterly, W. 2001. “The Middle Class Consensus 27. In 2011, one of the most expensive years on and Economic Development.” Journal of record for natural disasters, losses in East Economic Growth 6 (4): 317–35. Asia and Pacific from natural disasters totaled Fan, C. C. 2008. “Migration, Hukou, and the City.” US$259 billion. Most costs accrued in the first In China Urbanizes: Consequences, Strategies, nine months of the year and accounted for and Policies, edited by S. Yusuf and T. Saich, 80 percent of global losses that were due to 65–89. Washington, DC: World Bank. http:// natural disasters that year (Jha and Stanton- siteresources.worldbank.org/INTEAECOPRO​ Geddes 2013). / R e s o u r c e s / 3 0 8 7 6 9 4 - 1 2 0 6 4 4 6 4 7 4 1 4 5​ 28. In China, migrants in urban centers earn up /­Chapter_3_China_Urbanizes.pdf. to 40 percent less hourly (Frijters, Kong, and Freund, C. 2016. Rich People Poor Countries: Meng 2011). The Rise of the Emerging-Market Tycoons 29. Urbanization makes it more difficult and Their Mega Firms. New York: Peterson for people to rely on family members or Institute for International Economics. neighbors for help. And weaker ties to rural Frijters, P., T. Kong, and X. Meng. 2011. “Migrant areas make it more difficult for people to Entrepreneurs and Credit Constraints under migrate back to family farms in the wake of Labour Market Discrimination . ” IZA urban job losses. Discussion Paper 5967, IZA (Institute for the Study of Labor), Bonn. http://ftp.iza.org​ /­dp5967.pdf. References Galor, O., and J. Zeira. 1993. “Income Distribution Baker, J., and G. Gadgil. eds. 2017. East Asia and and Macroeconomics.” Review of Economic Pacific Cities, Expanding Opportunities for the Studies 60 (1): 35–52. Urban Poor. Washington, DC: World Bank. Gao, Q., and C. Riskin. 2009. “Market versus Banerjee, A. V., and A. F. Newman. 1993. Social Benefits: Explaining China’s Changing “Occupational Choice and the Process of Income Inequality.” In Creating Wealth and Development.” Journal of Political Economy Poverty in Postsocialist China , edited by 101 (2): 274–98. D. S. Davis and W. Feng, 18–34. Stanford, CA: Bunkanwanicha, P., and Y. Wiwattanakantang. Stanford University Press. 2008. “Big Business Owners in Politics.” Giles, J., and Y. Huang. 2015. “Are the Elderly Review of Financial Studies 22 (6): 2133–68. Left behind in a Time of Rapid Demographic Cai, F., ed. 2003. Zhongguo Renkou Yu Laodong and Economic Change? A Comparative Study Wenti Baogao: Zhuangui Zhong Di Chengshi of Poverty and Well-Being of East Asia’s Pinkun Wenti (Report on China’s Population Elderly.” Background paper for the East Asia and Labor: Urban Poverty in Transitional and Pacific regional report on aging, World China). Beijing: Shehui Kexue Wenxian Bank, Washington, DC. Chubanshe (Social Sciences Documentation Gill, I. S., and H. Kharas. 2015. “The Middle- Press). Income Trap Turns Ten . ” Policy Research 48   RIDING THE WAVE Wo r k i n g P a p e r 7 4 0 3 , Wo r l d B a n k , Li, L., S. Li, and Y. Chen. 2010. “Better City, Washington, DC. Better Life, But for Whom?: The Hukou and Gimpelson, V., and D. Treisman. 2015. Resident Card System and the Consequential “Misperceiving Inequality.” NBER Working Citizenship Stratification in Shanghai.” City, Paper 21174, National Bureau of Economic Culture and Society 1 (3): 145–54. Research, Cambridge, MA. Lozano-Gracia, N., D. Walker, and S. E. Antos. Graham, C. 2017. Happiness for All? Unequal 2016. “Spatial Disparities in EAP.” Background Hopes and Lives in Pursuit of the American paper for Riding the Wave: An East Asian Dream. Princeton, NJ: Princeton University Miracle for the 21st Century. Washington, DC: Press. World Bank. Graham, C., S. Chattopadhyay, and J. Zhang. Marmot, M. 2015. The Health Gap. The 2016. “Perceptions of Inequality in Southeast Challenge of an Unequal World. London and Asia: Some Novel Insights from Metrics New York: Bloomsbury Publishing. of Well-Being.” Unpublished background National Bureau of Statistics of China. 2010. paper for this report, Brookings Insitution, “National Economy Maintained Stable Washington, DC. Performance and Progress with Good Graham, C., and S. Pettinato. 2002. “Frustrated Momentum for Growth in August.” Retrieved Achievers: Winners, Losers, and Subjective December 10, 2015 from http://www.stats.gov​ Well-Being in New Market Economies.” .cn/english/. Journal of Development Studies 38 (4): 100–40. Ravallion, M., and Q. Wodon. 1999. “Poor Areas, Hoff, K., and P. Pandey. 2004. “Belief Systems or Only Poor People?” Journal of Regional and Durable Inequalities: An Experimental Science 39 (4): 689–711. Investigation of Indian Caste . ” Policy Rodrik, D. 2015. “Premature Deindustrialization.” Research Working Paper 3351, World Bank, NBER Working Paper 20935, National Bureau Washington, DC. of Economic Research, Cambridge, MA. Indonesia Ministry of Finance and World Bank. Tiongson, E. R., and M. Weise. 2016. “The Forthcoming. “Estimating Top Incomes in Misperception of Inequality and Political Indonesia.” World Bank, Jakarta. Participation: Summary of Emerging Jha, A. K., and Z. Stanton-Geddes, eds. 2013. Findings.” Unpublished background note for Strong, Safe, and Resilient: A Strategic Policy this report, Georgetown University School of Guide for Disaster Risk Management in Foreign Service, Washington, DC. East Asia and the Pacific. Washington, DC: UN-Habitat. 2013. State of the World’s Cities World Bank. 2012/2013: Prosperity of Cities. Nairobi: Jirasavetakul, L.-B. F., and C. Lakner. 2017. UN-Habitat. “Distribution of Consumption Expenditure United Nations, United Nations Department in East Asia.” Policy Research Working Paper of Economic and Social Affairs, Population 7968, World Bank, Washington, DC. Division. 2014. World Urbanization Prospects: Kanbur, R., C. Rhee, and J. Zhuang, eds. 2014. The 2014 Revision. New York: United Nations. Inequality in Asia and the Pacific: Trends, Wai-Poi, M., M. Wihardja, M. Mervisiano, and Drivers, and Policy Implications. New York: I. Setiawan. Forthcoming. Estimating Top Asian Development Bank and Routledge. Incomes in Indonesia. Jakarta: World Bank. Khan, A. R., and C. Riskin. 2005. “China’s Wang, H., and Y. Chen. 2010. “An Analysis on the Household Income and Its Distribution, 1995 Work Welfare Discrimination against Migrants.” and 2002.” China Quarterly 182 (June): 356–84. Chinese Journal of Population Science 2: 47–54. Lakner, C., and B. Milanovic. 2016. “Global Weise, M. 2016. “Key Trends in Perceptions Income Distribution: From the Fall of the of Income Inequality in WVS and ISSP . ” Berlin Wall to the Great Recession.” World Unpublished background report for this report, Bank Economic Review 30 (2): 203–32. World Bank, Washington, DC. Lakner, C., and C. Ruggeri Laderchi. 2016. “What Wong, L., and H. Wai-po. 1998. “Reforming the Do We Know about the Super-Rich in East Household Registration System: A Preliminary Asia?” Background paper for Riding the Wave: Glimpse of the Blue Chop Household An East Asian Miracle for the 21st Century. Registration System in Shanghai and Shenzhen.” Washington, DC: World Bank. International Migration Review 32 (4): 974–94. R ead y f or the Cha l l enge s o f the 2 1 s t Cent u r y ?    49 World Bank. 2006. World Development Report ———. 2016b. Live Long and Prosper: Aging 2006: Equity and Development. Washington, in East Asia and Pacific. Washington, DC: DC: World Bank and Oxford University Press. World Bank. ———. 2009. World Development Report 2009: ———. 2016c. Poverty and Shared Prosperity Reshaping Economic Geography. Washington, 2016: Taking on Inequality. Washington, DC: DC: World Bank. World Bank. ———. 2014a. East Asia Pacific at Work: ———. 2017. Global Economic Prospects, Employment, Enterprise, and Well-Being. January 2017: Weak Investment in Uncertain Washington, DC: World Bank. Times. Washington, DC: World Bank. ———. 2014b. Taking stock: an update on World Bank and Development Research Center Vietnam’s recent economic development. of the State Council, People’s Republic of Washington, DC: World Bank. http:// China. 2014. Urban China: Toward Efficient, d o c u m e n t s . w o r l d b a n k ​ . o r g / c u r a t e d ​ / e n​ Inclusive, and Sustainable Urbanization. / 6 4 1 2 4 1 4 6 8 1 3 3 5 4 4 4 5 1 / Ta k i n g ​ - s t o c k​ Washington, DC: World Bank. -an-update-on-Vietnams-recent-economic​ World Bank and Ministry of Planning and -development. Investment of Vietnam. 2016. Vietnam 2035: ———. 2015a. East Asia’s Changing Urban Toward Prosperity, Creativity, Equity, and Landscape: Measuring a Decade of Spatial Democracy. Washington, DC: World Bank. Growth. Washington, DC: World Bank. doi:10.1596/978-1-4648-0824-1. ———. 2015b. East Asia and Pacific Economic Zeng, J., A. Shang, and S. Chishty. 2015. Update, April 2015: Adjusting to a Changing “The Evolution of China’s Private Wealth World. Washington, DC: World Bank. Market.” Business Insights, Bain & Company, ———. 2016a. Indonesia’s Rising Divide: Why Washington, DC. http://www.bain.com​ Inequality Is Rising, Why It Matters, and What /­publications/articles/the-evolution-of-chinas​ Can Be Done. Washington, DC: World Bank. -private-wealth-market.aspx. Looking beyond the Poverty Line: Economic 3 Class in Developing East Asia and Pacific Introduction A picture of the income The rapid economic growth in developing East distribution by economic class Asia and Pacific1 over the past few decades This report divides the income distribution has brought an unprecedented reduction in into five economic classes based on levels poverty. To ensure that growth is inclusive in of consumption per capita. The thresholds the face of emerging challenges, however, it is are derived from a combination of well- vital to move beyond the simple poor–nonpoor established international norms and country-­ dichotomy and to obtain a richer understand- specific evidence (box 3.1): ing of the region’s transformation. Doing so •  The extreme poor, living on less than requires taking a closer look at trends across US$1.90 a day the income distribution through the lens of eco- •  The moderate poor, living on US$1.90 to nomic class. As poverty decreased dramatically US$3.10 a day in East Asia and Pacific, a significant middle •  The economically vulnerable, living on class emerged, although a sizable proportion of US$3.10 to US$5.50 a day the region’s population remains vulnerable to •  The economically secure, living on US$5.50 falling back into poverty as a result of shocks. to US$15.00 a day Meanwhile, gaps in the delivery of services, •  The middle class, living on more than such as higher education, housing, and water US$15.00 a day. and sanitation, result in limited access and poor quality even for those in higher economic In 2015, the extreme poor accounted for classes. The lens of economic class highlights 2.5 percent of the region’s population, the how, over the past few decades, developing moderate poor for almost 10 percent, and the East Asia and Pacific has been transformed economically vulnerable for approximately from a region of mostly poor people in low- one-quarter (figure 3.1). The economically income countries to a region of middle-income secure were the largest group in the region, countries with a diversity of economic classes. with almost half of the population. The mid- Such economic diversity implies that efforts to dle class accounted for 17 percent. deliver inclusive growth must take into account Not surprisingly, households in higher a range of specific needs and circumstances. economic classes do better according to 51 52   RIDING THE WAVE BOX 3.1  Partitioning the income distribution into classes A key theme of this report is that the challenges •  The economically secure, living on US$5.50 to standing in the way of more inclusive growth in US$15.00 a day. These households are not at sig- developing East Asia and Pacific are different for nificant risk of falling into poverty (see above) but each income group. The relevant groups may be cannot yet be considered middle class (see below). defined in two ways. First, one can define economic •  The middle class, living on more than US$15.00 classes on the basis of some distinguishing charac- a day. This threshold is broadly consistent with teristic that emerges from an empirical analysis— the values used by other studies.b In principle, for instance, calculating the probability of different another class could be identified at the top end parts of the income distribution to fall into poverty of the distribution—for instance, those living over time, on the basis of panel data (also known on more than US$50.00 a day, following Lopez- as longitudinal data—­ observations obtained over Calva and Ortiz-Juarez (2011). However, the several time periods for the same individuals). This household survey data are not well suited for was the approach followed for Latin America and measuring this group (see chapter 2). the Caribbean in a recent report (Ferreira and others 2013). Second, one can identify economic class with In practice, this report is not dogmatic about the reference to well-understood and widely supported definition of classes per se, and it doesn’t shy away thresholds that capture some broadly held views on from identifying common challenges across classes. discontinuities across classes. This report strikes a This report also recognizes that although qualita- balance between these two approaches, integrating tive differences are clearly identifiable between the detailed country evidence and adopting preexist- two ends of the distribution, a considerable degree ing lines, reflecting also the limited availability of of continuity exists between classes. In particular, panel data across the region. Specifically, the report strong continuities exist across the poverty line, with focuses on the following segments of the expendi- a lot of movement between the moderate poor and ture distribution, where all expenditure cutoffs are the economically vulnerable classes. Such continuities expressed in terms of 2011 U.S. dollars (purchasing are highly relevant from a policy perspective. power parity [PPP]): a. The mean estimate of this expenditure level, across countries with available •  The extreme poor, living on less than US$1.90 a data and using alternative methodologies, is adopted. Specifically, the probabil- day. This threshold is the World Bank’s interna- ity of falling into poverty is less than 10 percent, based on panel estimates using a Lowess (locally weighted scatterplot smoothing) or a logistic model, for incomes tional poverty line. above the following levels in the following countries: Indonesia, US$8.25 •  The moderate poor, living on US$1.90 to US$3.10 (Lowess) or US$6.01 (logistic); the Philippines, US$6.20 (Lowess) or US$5.40 (logis- a day. The upper threshold is the moderate pov- tic); and Vietnam, US$4.40 (Lowess) or US$4.20 (logistic). Data availability does not allow for the reliable estimation of country-specific vulnerability lines. erty line traditionally used by the World Bank b. For instance, Wilson and Dragusanu (2008) define members of the middle class in analyzing trends in developing East Asia and as those who earn an income of US$6,000 a year to US$30,000 a year (2007 PPP), Pacific. and Kharas (2010) uses US$10 a day to US$100 a day (2005 PPP). Focusing on Latin America, Ferreira and others (2013) adopt a minimum threshold of US$10 a •  The economically vulnerable, living on US$3.10 day (2005 PPP); likewise, respondents to the Latinobarómetro perception survey to US$5.50 a day. The cutoff to assess vulner- become more likely to self-identify as middle class than as poor at an income level of US$10 a day (2005 PPP). Adjusting for the intervening inflation and ability is based on a 10 percent or higher chance adopting a mean value yield a lower threshold of approximately US$15 a day of falling into poverty in the next measurement (2011 PPP). Similarly, Pritchett (2003, 2006) argues that the upper bound for the period, using panel data (following the methodol- global poverty line should be based on the lowest poverty line in the wealthy OECD (Organisation for Economic Co-operation and Development) countries; the ogy of Lopez-Calva and Ortiz-Juarez 2011).a United States has the lowest poverty line, equivalent to US$15.80 a day (2011 PPP). several indicators that go beyond income and and more likely to be employed in formal consumption, including access to services sector salaried jobs, thereby highlighting the (appendix D, table D.1), asset ownership, and importance of structural transformation and employment. Members of these households of better jobs in supporting higher living are less likely to be employed in agriculture standards. For instance, a large majority of Loo k ing be y ond the P overt y Line   53 the middle class are civil servants, are profes- FIGURE 3.1  Population distribution by economic class in sional or technical workers, or are in service developing East Asia and Pacific, 2002–15 and sales jobs. Percentage of total population Some patterns are less obvious because they go beyond a simple association with 100 greater expenditure as economic class 90 Percentage of total population improves. Three stand out: 80 70 •  The extreme poor are a heterogeneous 60 group along several dimensions, including 50 location and ethnicity. 40 •  Groups that are more well off, including the economically secure and the middle 30 class, still have limited access to some 20 basic services, including good-quality 10 housing. 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 •  Large differences exist in female labor Year force participation rates across eco- nomic classes, as measured by household Extreme poor (less than PPP US$1.90 per day) Moderate poor (PPP US$1.90–US$3.10 per day) surveys.2 These rates are very low for the Economically vulnerable (PPP US$3.10–US$5.50 per day) two poor classes: less than 40 percent for Economically secure (PPP US$5.50–US$15.00 per day) the extreme poor and about 43 percent Middle class (PPP US$15.00 and higher per day) for the moderate poor, compared with 55 percent for the economically vulnerable Source: World Bank EAP Team for Statistical Development. Note: PPP = purchasing power parity. Appendix B provides details on the methodology underlying and more than 68 percent for each of the this figure. higher economic classes. The distribution of economic classes across countries FIGURE 3.2  Distribution of extreme poverty across developing East Asia and Pacific, 2015 The distribution of economic classes across Percentage of region’s extreme poor the countries of the region looks very differ- ent today than it did three decades ago. Most 0.3 of the extreme poor are now spread across 1.0 0.2 1.9 the three larger middle-income countries: 2.9 Indonesia, China, and the Philippines together 3.8 account for more than 80 percent of the total 5.2 (figure 3.2). Indonesia accounts for two-fifths of the extreme poor in East Asia and Pacific. 13.8 40.2 In contrast, in 1981, China alone accounted for more than 80 percent of the extreme poor in East Asia and Pacific. The incidence of extreme poverty is very 30.5 heterogeneous and remains high in a few countries. Rates of extreme poverty as a share of the population are highest in Papua Indonesia China Philippines Vietnam New Guinea; the Lao People’s Democratic Papua New Guinea Lao PDR Timor-Leste Republic; Timor-Leste; and the Pacific Island Solomon Islands Cambodia Rest of the region countries of Kiribati, the Federated States of Micronesia, the Solomon Islands, and Source: World Bank EAP Team for Statistical Development. 54   RIDING THE WAVE TABLE 3.1  Population distribution by economic class and country in developing East Asia and Pacific, 2015 Extreme Moderate Economically Economically Middle poor (%) poor (%) vulnerable (%) secure (%) class (%) Cambodia 0.7 14.0 49.6 34.9 0.7 China 1.1 6.7 21.8 51.0 19.4 Indonesia 7.5 24.6 35.9 27.7 4.3 Lao PDR 13.8 27.9 36.0 20.1 2.2 Malaysia 0.0 0.3 2.6 31.3 65.7 Mongolia 0.2 2.4 15.9 64.8 16.6 Papua New Guinea 24.0 22.5 28.7 22.3 2.5 Philippines 6.6 18.7 30.8 34.7 9.2 Thailand 0.0 0.8 10.1 53.6 35.4 Timor-Leste 40.3 15.3 29.0 14.3 1.1 Vietnam 2.7 7.1 23.7 57.0 9.5 Pacific Island countries Fiji 1.0 5.4 36.4 46.1 11.2 Kiribati 13.3 19.7 36.8 27.5 2.6 Micronesia, Fed. Sts. 15.4 20.1 33.0 28.4 3.0 Samoa 0.8 8.3 28.0 49.0 13.9 Solomon Islands 24.4 31.7 28.4 14.1 1.4 Tonga 1.0 5.4 22.4 54.8 16.4 Tuvalu 1.6 9.4 27.9 49.1 12.0 Vanuatu 15.0 25.0 34.1 23.9 2.0 Source: EAP Team for Statistical Development. Note: Population shares within any given country may not add exactly to 100 because of rounding errors. Vanuatu (all above 10 percent; see table 3.1). cities in Indonesia, where poverty alleviation Conversely, the incidence of extreme poverty programs have led to especially sharp poverty is negligible in Malaysia and Thailand, both reduction in smaller towns and rural areas). upper middle-income countries. As a result, in most countries in East Asia The extreme poor are also unequally and Pacific, including the Philippines and distributed within countries along the Vietnam, the vast majority of the poor live rural-urban spectrum, with rural areas hav- in smaller towns and rural areas. As urban- ing much higher rates of extreme poverty. ization proceeds, however, the share of poor People living in remote areas that lack con- people living in cities, particularly in mega- nective infrastructure or services, or that are cities, is rising. agro-ecologically or otherwise geographi- In several countries, extreme poverty cally disadvantaged, also lack the services is concentrated among ethnic minorities, available to their more urban counterparts particularly when it is associated with geo- (see appendix C, table C.4), as well as access graphical remoteness. In Vietnam, although to jobs and markets. This means that people only 3 percent of the ethnic majority Viet in remote areas either have a low asset base population live below the extreme poverty to begin with (because of low education and line, 37 percent of those belonging to ethnic poor health care) or cannot use their assets minorities live below that same threshold, effectively (because of lack of access to jobs accounting for 69 percent of the remaining and markets). Conversely, poverty rates are extreme poor population (World Bank 2012). generally lower in larger cities than in smaller The differential poverty incidence across eth- towns.3 Poverty rates have also decreased nic groups, which has been increasing over faster in larger cities (with the exception of time, is strongly correlated with location: Loo k ing be y ond the P overt y Line   55 ethnic minorities are concentrated in rural, and nearly 10 times that of Malaysia), and remote, and mountainous areas. In China, the Philippines and Vietnam for approxi- ethnic minorities, which account for approxi- mately 10 percent each. mately 10 percent of the population, have lower incomes and are more likely to be poor than are members of the Han ethnic major- Dramatic income gains across ity (Gradín 2015; Gustafsson and Sai 2014).4 economic classes This statement generally holds even at the Adopting the lens of economic class helps level of individual provinces, although ethnic underscore how the income distribution at minorities fare better in provinces where eth- the regional level in East Asia and Pacific nicity is more heterogeneous (such as Guizhou has changed dramatically. The incidence of or Yunnan). These differences largely reflect extreme poverty has fallen sharply since the disadvantaged, including mountainous, loca- early 2000s (figure 3.1). Adopting a longer tions with inadequate services. perspective, East Asia and Pacific has expe- Substantial variation exists across East rienced an unprecedented decline in extreme Asia and Pacific in the incidence of moder- poverty over the last three decades, steeper ate poverty and economic vulnerability, two than in all other developing regions of the groups that largely overlap in their charac- world—from 81.0 percent of the population teristics and that experience a high degree in 1981 to 2.5 percent in 2015. The num- of churning between categories. The mod- ber of extreme poor has decreased by more erate poor and the economically vulnerable than 1 billion over the same period, although account for 64 percent of the population there are still more than 50 million people in in Cambodia and Lao PDR, and for more extreme poverty. than half in Indonesia; Papua New Guinea; The share of the population classified as and the Pacific Island countries of Kiribati, moderate poor has also fallen sharply, from the Federated States of Micronesia, the 26 percent in 2002 to less than 10 percent in Solomon Islands, and Vanuatu; however, 2015. This decline has been particularly pro- these groups account for less than 3 percent nounced in rural areas, where the number of of the population in Malaysia. Again, there moderate poor declined by 92 million people are significant rural-urban disparities. 5 between 2002 and 2012, and by 20 million The economically vulnerable are larger in excluding China. In contrast, the share of the number than the moderate poor across the economically vulnerable has remained stable, region, with the exception of the Solomon at around one-quarter of the population. Islands. Almost two-thirds of all households in The size of the economically secure and East Asia and Pacific now either are econom- middle classes also varies sharply across the ically secure or have joined the middle class, region. In absolute terms, China accounts reflecting the sustained growth that most of for approximately three-quarters of both East Asia and Pacific has seen since 2002.6 groups (and of the region’s total population). The expansion of these classes has been Almost all the remainder of these economic checked only by the end of the commodi- classes live in four countries with large ties boom. The economically secure grew populations—Indonesia, the Philippines, from approximately 350 million in 2002 Thailand, and Vietnam—and in smaller but to 975 million in 2015, or by 8.2 percent wealthier Malaysia. Indonesia accounts for per year. Over the same period, more than one-quarter of the non-China economically 300 million people joined the middle class, secure in East Asia and Pacific. Of the non- an increase of nearly 17 percent per year. The China middle class, Malaysia accounts for economically secure were heavily affected more than one-third, Thailand for more than by the Asian financial crisis of 1997–98; the one-quarter, Indonesia for 13 percent (despite middle class in the region is almost entirely a having 4 times the population of Thailand phenomenon of the 2000s. 56   RIDING THE WAVE Different countries’ pathways Asia and Pacific, as elsewhere, the incidence to prosperity poverty remains higher in rural settings. of ­ A less-expected finding is that, whereas in Regional growth trends have led the distri- most countries it is cities that bring greater bution of people across classes to change in economic security, in several countries different ways in different countries. Between (including Lao PDR, Thailand, and Vietnam), 2002 and 2015, the majority of East Asian economic security has also been rising sharply countries reduced extreme poverty rates in rural areas. In part, such economic secu- to single digits, with the notable exceptions of rity is attributable to urbanization, which Papua New Guinea and Lao PDR (­ figure 3.3). has reduced the supply of rural labor and However, the relative sizes of other classes thus raised rural wages while also increasing varied significantly from country to country. the remittances that rural areas receive from While the Philippines’ economic class struc- urban migrants. ture remained roughly the same over this An analysis of the changes in the economic period, Indonesia has reduced sharply the class structure at the country level, combined incidence of both extreme and moderate pov- with the current differences across countries, erty since 2002 (see also box 3.2). suggests the following broad typology of These trends have changed not only pathways to prosperity that East Asia and the concentration of classes across coun- Pacific countries have followed: tries but also the distribution of economic classes within countries, particularly across •  Progressive prosperity : Malaysia and the rural-urban spectrum. As urbanization Thailand, the richer countries in the region, advances, the share of the poor living in have eliminated extreme poverty. Much of urban areas also increases. However, in East the population (approximately two-thirds FIGURE 3.3  Population distribution by economic class and country in developing East Asia and Pacific, 2002 and 2015 Percentage of country’s population 100 90 Percentage of country’s population 80 70 60 50 40 30 20 10 0 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 2002 2015 cambodia china Indonesia lao PDR malaysia mongolia Papua Philippines Thailand Vietnam New Guinea Extreme poor moderate poor Economically vulnerable Economically secure middle class Source: EAP Team for Statistical Development. Loo k ing be y ond the P overt y Line   57 in Malaysia and one-third in Thailand) New Guinea with respect to growth has now belongs to the middle class. proved relatively low. As a result, the •  Out of poverty into prosperity: In China, income distributions of these countries Mongolia, and Vietnam, as a result of resemble Indonesia back in 2002. sustained poverty reduction, most peo- •  The Pacific Island countries: These are a ple are at least economically secure, and heterogeneous group. In terms of current the middle class is growing. Overall, the composition of economic classes, Samoa, income distributions of these countries Tonga, and Tuvalu resemble the Out-of- now resemble Malaysia and Thailand poverty-into-prosperity group; Kiribati, back in 2002. the Federated States of Micronesia, and •  Out of extreme poverty : Cambodia, Vanuatu look similar to the Philippines, Indonesia, and the Philippines enjoy low and the Solomon Islands are poorer even levels of extreme poverty, but the mid- than Lao PDR (table 3.1). In terms of dle class in these countries is also small. changes over time, most of these countries A ­sizable share of the population is still have performed poorly, though some mod- either moderately poor or vulnerable to est progress has been made in reducing falling back into poverty. extreme poverty in the Solomon Islands, •  Lagging progress: Lao PDR and Papua Tuvalu, and Vanuatu. New Guinea face more serious challenges, although each has distinctive features. These countries have made progress in Looking ahead: extrapolating reducing extreme poverty but still expe- past trends rience much higher incidence of extreme Projections based on a simple mechanical poverty than most other countries in the extrapolation of past trends would suggest region. Put differently, the elasticity of that the goal of eradicating extreme poverty is extreme poverty in Lao PDR and Papua well within reach for the region (figure 3.4); FIGURE 3.4  Poverty projections, developing East Asia and Pacific, US$1.90 a day and US$3.10 a day PPP 20 118 112 107 15 101 96 Poverty rate (%) 91 86 81 77 10 73 69 173 65 62 156 142 129 118 5 108 99 25 91 85 79 22 20 74 69 65 18 16 14 31 27 13 12 11 10 10 24 21 19 17 15 14 13 12 9 9 0 11 10 10 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 Year East Asia and Pacific, US$1.90 a day East Asia and Pacific, excluding China, US$1.90 a day East Asia and Pacific, US$3.10 a day East Asia and Pacific, excluding China, US$3.10 a day Source: World Bank EAP Team for Statistical Development. Note: The numbers in the figures represent millions of poor people according to the different poverty lines. PPP = purchasing power parity. Appendix B provides details on the methodology underlying this figure. 58   RIDING THE WAVE FIGURE 3.5  Population distribution by economic class in in such a framework, more than 40 percent developing East Asia and Pacific, 2015–30 of the region’s population would be part of Percentage of total population the middle class by 2030 (figure 3.5). Even if such projections were to come about, how- 100 ever, more than 60 million people would still 90 be living in poverty (extreme and moderate) Percentage of total population 80 in this increasingly upper-middle-income 70 region, and more than 200 million would 60 remain highly vulnerable to falling into 50 poverty. 40 More fundamentally, recent trends show growth becoming less inclusive, as captured 30 by indicators of growing real and perceived 20 inequality. Regional and global trends sug- 10 gest that sustaining high levels of inclusive 0 growth will become more difficult. Different 30 23 25 27 28 29 26 15 16 17 18 19 20 21 22 24 challenges to inclusiveness are likely to 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year emerge for different parts of the income Extreme poor (less than PPP US$1.90 per day) distribution; partitioning the income distri- Moderate poor (PPP US$1.90–US$3.10 per day) bution by classes allows us to characterize Economically vulnerable (PPP US$3.10–US$5.50 per day) these challenges more fully, as discussed in Economically secure (PPP US$5.50–US$15.00 per day) chapter 4. Middle class (PPP US$15.00 and higher per day) Source: World Bank EAP Team for Statistical Development. Note: PPP = purchasing power parity. Appendix B provides details on the methodology underlying this figure. BOX 3.2  Country-level economic class dynamics: some examples The aggregate class dynamics shown in figure 3.1 have be some movement here both at the low end (elimi- been shaped by country-specific patterns. This box nation of poverty) and at the top (rise in economic presents evidence from four large countries in East security). Second, Indonesia and even more China Asia and Pacific for which, despite some problems with and Vietnam in the past several years have seen slow- comparability of surveys over time (see appendix A), ing progress in eliminating extreme poverty (as the long-term evidence appears to be reliable. target has been virtually reached) and more progress Figure B3.2.1 highlights some interesting con- in moving households over the threshold of economic trasts in country-level economic class dynamics. First, security. China and Vietnam also stand out as hav- the Philippines stands out for having shown very ing made a serious dent in the extent of economic little dynamism until recently; now there appears to vulnerability. box continues next page Loo k ing be y ond the P overt y Line   59 BOX 3.2  Country-level economic class dynamics: some examples (continued) FIGURE B3.2.1  Population distribution by economic class in China, Indonesia, the Philippines, and Vietnam, 2002–15 Percentage of total population a. China b. Indonesia 100 100 90 90 Percentage of total population Percentage of total population 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 11 12 13 15 11 12 13 15 10 10 03 05 06 07 08 09 03 05 06 07 08 09 04 14 04 14 02 02 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Year c. Philippines d. Vietnam 100 100 90 90 Percentage of total population Percentage of total population 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 11 13 12 15 11 12 13 15 10 10 03 05 06 07 08 09 06 08 03 05 07 09 04 02 14 04 14 02 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Year Extreme poor (less than PPP US$1.90 per day) Moderate poor (PPP US$1.90–US$3.10 per day) Economically vulnerable (PPP US$3.10–US$5.50 per day) Economically secure (PPP US$5.50–US$15.00 per day) Middle class (PPP US$15.00 and higher per day) Source: World Bank EAP Team for Statistical Development. Note: PPP = purchasing power parity. Country series might include breaks in comparability, though overall trends are deemed to be robust; see appendix A for details. Appendix B provides details on the methodology underlying this figure. 60   RIDING THE WAVE Notes between Poverty and City Size in a Selection of Developing Countries.” World Bank, 1. Developing East Asia and Pacific is defined Washington, DC. doi:10.1093/wber/lhs007. as Cambodia, China, Indonesia, Lao PDR, Ferreira, F. H., J. Messina, J. Rigolini, L. F. Lopez- Malaysia, Mongolia, Myanmar, Papua Calva, M. A. Lugo, and R. Vakis. 2013. New Guinea, the Philippines, Thailand, Economic Mobility and the Rise of the Latin Timor-Leste, Vietnam, and the Pacific Island American Middle Class. Washington, DC: countries. World Bank. 2. Typically, respondents appear to underestimate Gradín, C. 2015. “Rural Poverty and Ethnicity women’s unpaid work both in the care in China.” In Measurement of Poverty, economy and in family businesses, as those are Deprivation, and Economic Mobility, edited by not seen as real “jobs” as captured by time-use T. I. Garner and K. S. Short, 221–47. Bingley, surveys. Although a shift to paid employment U.K.: Emerald Group Publishing. often does not lead to a parallel decline in time Gustafsson, B., and D. Sai. 2014. “Why Is There spent in unpaid activities, the data show that No Income Gap between the Hui Muslim participation in jobs that are defined as such Minority and the Han Majority in Rural (most likely “paid jobs”) sets apart the poor Ningxia, China?” The China Quarterly 220: from the rest of the distribution. 968–87. 3. This pattern is not unique to East Asia and Kharas, H. 2010. The Emerging Middle Class Pacific. Poverty map data for eight countries in Developing Countries. Paris: OECD in six developing regions also find an inverse Publishing. relationship between poverty rates and city Lopez-Calva, L. F., and E. Ortiz-Juarez. 2011. size (Ferré, Ferreira, and Lanjouw 2012). “A Vulnerability Approach to the Definition 4. This does not apply to the Manchu minority of the Middle Class.” Policy Research Working and applies relatively less to the Hui minority. Paper No. WPS 5902, World Bank, Washington, 5. In China in 2015, the rate of moderate DC. https://openknowledge​.worldbank.org​ poverty was only 2.2 percent in urban areas /­handle/10986/3669. but 15.0 percent in rural areas. Pritchett, L. 2003. “Who Is Not Poor? Proposing 6. The economically secure increased from a Higher International Standard for Poverty.” 18 percent of the population in 2002 Working Paper 33, Center for Global to 47 percent in 2015. The middle class Development, London, U.K. increased from 3 percent of the population ———. 2006. Who Is Not Poor? Dreaming of a in 2002 to 17 percent in 2015; this is likely World Truly Free of Poverty. Oxford, U.K.: an underestimate given the difficulties for Oxford University Press. household surveys to capture the top tail of Wilson, D., and R. Dragusanu. 2008. The the distribution (see chapter 2). Expanding Middle: The Exploding World Middle Class and Falling Global Inequality. New York: Goldman Sachs. References World Bank. 2012. Malaysia Economic Monitor Ferré, C., F. H. Ferreira, and P. Lanjouw. 2012. “Is 2012: Unlocking Women’s Potential. There a Metropolitan Bias? The Relationship Washington, DC: World Bank. Economic Class and the Challenges of Inclusive Growth 4 Introduction behind, involves two elements in addition to poverty reduction: (a) economic mobil- Inclusive growth involves reducing poverty, ity, defined as the ability to improve one’s lot fostering upward economic mobility, and in life and move to a higher economic class, providing economic security for all income and (b) economic security, which enables groups. Since the beginning of the century, an individual to hold on to his or her gains. countries in developing East Asia and Pacific1 Comprehensive evidence on these dimen- have experienced unprecedented poverty sions is hard to come by for the East Asia and reduction. Yet perceptions of rising inequal- Pacific region, but such evidence provides an ity and of shrinking opportunities for upward important complement to the familiar story mobility, together with a high exposure to of the region’s success in reducing extreme shocks, suggest the need for scaled-up pub- poverty (box 4.1). lic policy actions to secure upward mobility Given the East Asia and Pacific region’s throughout the income distribution. Such recent high growth and significant reduc- actions might involve a focus on different tion in poverty, it would be expected that a constraints for each segment of the income longitudinal analysis would show that most distribution, as discussed in this chapter. For households experienced increases in income some residents of the region, such as people and that much upward mobility had occurred trapped in extreme poverty, the challenge is across classes. The analysis here confirms very much about upward mobility, whereas those hypotheses, with some nuances. First, for other people, such as the economically in each class other than the middle class, a vulnerable, being able to secure what they majority of people experienced a positive have is as important as public policy designed increase in consumption over a two-year to reduce barriers to upward mobility. period,2 and the increases were proportionally larger for poorer classes. Second, although at Securing upward mobility: least one in five people moved to a higher eco- removing constraints to nomic class, some people from all economic inclusive growth classes experienced downward transitions. In this report, the concept of inclusive growth, Despite those movements, the largest share where poverty is reduced and no one is left of households remained in the same class 61 62   RIDING THE WAVE BOX 4.1  Sources used to analyze mobility TABLE B4.1.1  Data used to analyze short- and long-term mobility Country Analysis Survey instrument Survey years Cambodia Long run 2004–12 Indonesia Short run National Socio-Economic Survey 2008–10 Long run 2000–10 Lao PDR Short run Lao Consumption and Expenditure Survey 2008–12/13 Philippines Short run Philippine Family Income and Expenditure Survey 2006–09 Vietnam Short run Vietnam Household Living Standards Survey 2010–12 Long run 2004–12 This chapter’s analysis of mobility relies on several characteristics and their distributions across cross- sources of data. To discuss short-term regional trends sectional data sets. b The methodology required in mobility, data were used for Indonesia, the Lao choosing countries for which there are cross-sectional People’s Democratic Republic, the Philippines, and comparable data spanning a long period, defined as Vietnam. These four countries were selected because at least eight years. In addition, an attempt was made of the availability of panel—or longitudinal—data to compare long- and short-term mobility to under- for each, which allow for individual households to stand the extent to which short-term trends repre- be followed over time. These four countries are suf- sent permanent transitions. Using these criteria, the ficiently different from each other in economic devel- long-term analysis was conducted on Indonesia and opment, class composition, and growth rates that it is Vietnam, for which short-term mobility estimates possible to begin to sketch a picture of class dynam- also were available, and on Cambodia. ics for the region as a whole. The most recent sam- ple years available were selected for use here, with a. All transitions have been adjusted to the shortest period for which true a two-year minimum break between observations. transitions can be measured for the greatest number of countries (that is, two years). Vietnam has a two-year panel, and the annual Indonesia data can Whereas Lao PDR, the Philippines, and Vietnam also be explored in a two-year panel. The three-year Philippines data and the have followed the same households for many years, five-year Lao PDR data are extrapolated to two-year periods. These adjust- Indonesia has a very short panel. To facilitate com- ments to estimate two-year transitions rates allow comparison of mobility across countries, although at the cost of strong assumptions about household parisons of countries, two-year trends were extrapo- income growing linearly. By definition, therefore, short-term volatility cannot lated from the longer panel periods. a The data are be captured. Similarly, such comparisons cannot take into account the possibil- ity that different countries might be at different points in their economic cycle drawn from expenditure surveys, which serve as a in the periods for which they are observed. However serious, such constraints proxy for consumption (table B4.1.1). could be addressed only by relying on higher-frequency (for example, yearly) The analysis of long-term trends does not use nationally representative panel data for all countries that are to be compared— data that are currently unavailable. true household panels and instead relies on a meth- b. See Dang and others (2014) for a discussion of the construction of odology of matching households by time-invariant synthetic panels. (figure 4.1). Third, upward mobility appears net result of the region’s high reduction in to be higher over the long run than the short poverty, supported by high growth, has been run, with more households moving up and mapped by very different household-level tra- fewer staying in the same class or experiencing jectories. And it is important to remember negative transitions. The share of households that mobility patterns are very country spe- that stayed in the same class over a 10-year cific, partly reflecting the different country period was also significantly less than in Latin situations over the period of analysis.3 America over a 15-year period (Ferreira and A close look at the characteristics of others 2013). Overall, therefore, it seems the the households that experienced different E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   63 trajectories helps explain what has enabled FIGURE 4.1  Short-run class mobility, aggregate upward mobility for different income groups: •  Climbers —those who improved their Lao PDR 29% 56% 16% class—accounted for one-fifth (in the Indonesia 37% 45% 18% Philippines) to one-third (in Indonesia and Lao People’s Democratic Republic) of Philippines 19% 69% 12% households. Households in poorer classes were more likely to improve their class Vietnam 26% 60% 14% figure 4.2) than were their richer coun- (­ 0 10 20 30 40 50 60 70 80 90 100 terparts. Country patterns differed, with Percentage of households economically secure households being the big winners in Indonesia, whereas extreme Climbers Stayers Sliders poor households experienced the larg- Source: Cunningham 2016. est gains in Lao PDR. Households that Note: “Climbers” denotes households that moved up in economic status, “stayers” are those that remained in the same economic class, and “sliders” are those that moved to a lower economic class. climbed out of moderate poverty or eco- Analysis for different countries covers different years, as detailed in box 4.1. nomic vulnerability tended to have better access to public services (education, health, and so forth) and more physical and finan- who received fewer transfers, those who cial assets, as well as a broader social net- lived in more rural settings, and asset-poor work than other households in the same households being particularly likely to have economic class. People who moved out slid in Vietnam. In the Philippines, sliders of economic security into the middle class were older and held more formal jobs. had more education, were living in an Downward mobility across classes reflects urban area, had access to better jobs and, both an ex ante exposure to risks and a lim- if women, were more likely to be in paid ited ex post ability to manage the impacts employment. of adverse shocks by some people, despite •  Stayers—those who did not change eco- significant resilience overall. Shocks include nomic class—accounted for between economic shocks that affect prices and assets, 45 percent of the population (in personal shocks such as illness, and natu- Indonesia) and 69 percent (in the ral shocks. It is challenging to learn more Philippines). The class composition of this about the effects of exposure to shocks and group varied by country, with households their impacts by class because of the paucity in extreme poverty being more likely not of relevant data. Nevertheless, for Vietnam, to have changed class in Lao PDR and some evidence is available from the coun- the Philippines and economically secure try’s nationally representative panel and from households more likely not to have other sources (box 4.2). Negative shocks changed class in Indonesia and Vietnam. as reported by households5 are associated In Vietnam, those who did not change with a greater probability of moving down class tended to be older, more urban, and the class ladder among the moderate poor more educated, compared with those who and the economically vulnerable. Even the did change class. In contrast, stayers in middle class faces risks. Conversely, mod- the Philippines tended to be younger and erate poor households that did not receive have greater household assets than those a shock were 50 percent more likely than who moved.4 those who did experience a shock to have •  Sliders—those who moved to a lower eco- moved up one class over a two-year period nomic class—represented about one-sixth and were 34 percent more likely to move up of all samples. Profiles of the backsliders— two classes.6 Although these findings under- those who moved to a lower class—­ varied line the significance of shocks, they also show by country, with less-educated people, those that even moderate poor households can 64   RIDING THE WAVE FIGURE 4.2  Class mobility, by initial income group, short run a. Lao PDR, transition matrix, 2010–12 b. Indonesia, transition matrix, 2008–10 2010 2012 2008 2010 1% 3% 1% 2% 7% 9% 15% 22% 25% 27% 28% 31% 41% 36% 32% 28% 26% 25% 23% 17% c. Philippines, transition matrix, 2007–09 d. Vietnam, transition matrix, 2010–12 2007 2009 2010 2012 5% 5% 16% 18% 27% 28% 45% 47% 29% 30% 23% 25% 24% 24% 9% 14% 11% 8% 3% 4% Extreme poor Moderate poor Economically vulnerable Economically secure Middle class Source: Cunningham 2016. E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   65 BOX 4.2  Shocks in developing East Asia and Pacific According to Vietnam’s nationally representative case because aggregate shocks—those that affect entire panel, moderate poor households are particularly communities—are very common in agricultural set- ­ susceptible to shocks, although the risk is rather tings. More than 80 percent of moderate poor house- similar across classes. Approximately 10 percent of holds in Cambodia and more than 60 percent in Lao a sample of urban and rural moderate poor in the PDR and Thailand experienced a shock in the year of region experienced a shock within the previous two the survey. Similarly, 70 percent of economically vul- years, whereas 8 percent of economically vulner- nerable households in rural Cambodia and around 60 able households were hit by a shock.a The type of percent in rural Lao PDR and Thailand were hit by at shock did not appear to differ by class. A grouping least one shock in the previous year. Those percent- of the available information into four broad catego- ages compare with 68 percent of economically secure ries—natural shocks that encompass all nature-related households in Cambodia having experienced at least shocks, economic shocks that include prices and use one shock in the previous year, although the differ- of assets, personal shocks such as illness or family ence in probabilities between classes is not statistically structure, and other shocks such as crime—shows that significant (at the 5 percent level) and there was no more than 40 percent of households that experienced difference with the middle classes in Lao PDR and a shock in the previous two years named an economic Thailand. shock, with low income and high food prices being In Cambodia, Lao PDR, and Thailand, illness of the primary shocks. The second most common shock a household member was by far the most frequent among all classes was a personal shock, typically sick- personal shock. The majority of shocks reported by ness or death of a household member. Natural disas- moderate poor households and economically vulner- ters emerged strongly for the poorer classes, although able households in rural areas of the three countries natural shocks constituted less than 10 percent of all were natural shocks and personal shocks. The differ- shocks experienced by economically secure house- ence in the propensity of having experienced one kind holds and were not listed at all by middle-class of shock or the other was not statistically significant. households.b The time period covered by these data Livestock disease was the most commonly cited natural (2010–12) had a low incidence of negative impacts shock, followed by floods (and droughts in Lao PDR). from natural disasters. For example, the number of These trends did not differ by economic class, with the people killed, the number of houses destroyed, and the exception of the middle class in rural Thailand, where cumulative damage from tropical cyclones (as a share personal shocks far outweighed the others. of gross domestic product [GDP]) during 2010–12 was among the lowest in the time period 1990–2013. a. The Vietnam Household Living Standards Survey asks households about Similar results emerged for a sample of rural house- the shocks they had experienced, up to three. Among households that experi- holds as for the reported national sample.c enced a shock, a high number of households reported the maximum number Data collected by small panels in marginalized rural (for an average of 2.6 shocks). b. Recent evidence using new high-resolution flood hazard maps and spatial regions of Cambodia, Lao PDR, and Thailand provide socioeconomic data found that, although poor households overall in Vietnam complementary insights, despite not being strictly com- were not particularly exposed to flooding, poor households living in slums in Ho Chi Minh City faced much higher chances of flooding than the rest of the parable with the findings for Vietnam. As in Vietnam, population (Bangalore, Smith, and Veldkamp 2016). the propensity to experience a shock was similar across c. The time period covered by these data (2010–12) had a low incidence of classes in rural areas of Cambodia, Lao PDR, and negative impacts from natural disasters. For example, the number of people killed, the number of houses destroyed, and the cumulative damage from Thailand. This last finding is less surprising in a rural tropical cyclones (as a share of GDP) during 2010–12 was among the lowest in setting than in the nationally representative Vietnam the time period 1990–2013. overcome negative circumstances if they have households’ exposure to shocks and in help- the appropriate assets base, because lack of ing households cope with the consequences assets rather than lack of shocks appears to of shocks without eroding the future profit- explain slides into poverty. ability of their assets. The effectiveness of Social protection programs and insur- those programs and insurance varies greatly, ance play an important role in reducing however, depending on the nature of the 66   RIDING THE WAVE shocks experienced. Shocks are often classi- aimed at supporting inclusive growth in the fied as aggregate shocks, which affect entire region must address the overall risk manage- communities (such as financial crises, natural ment agenda by covering both idiosyncratic disasters, global epidemics, or spikes in food and aggregate shocks and including both prices) or idiosyncratic shocks, which affect ex post and ex ante measures. individuals (such as bad health, job loss, unexpected family event). Aggregate shocks limit people’s ability to pool risks and tend to Eliminating extreme poverty result in large losses. The extreme poor accounted for 2.5 percent In developing East Asia and Pacific, aggre- of the region’s population in 2015, down gate shocks are common. In the past couple sharply from 18 percent a decade earlier and of decades, different parts of the region have significantly lower than the 13 percent aver- been affected by two major financial crises, age extreme poverty rate across the develop- several spikes in food prices, natural disas- ing world. This remarkable success means the ters ranging from earthquakes to tsunamis, East Asia and Pacific region as a whole has and repeated hits by massive cyclones.7 In already met the World Bank Group’s 2030 addition, millions of people in the region are target for the reduction of extreme poverty. exposed to the severe consequences of large Yet the incidence of extreme poverty remains natural shocks because populations are con- high in some countries, while in others the centrated in vulnerable low-lying coastal areas rate is still significant for small pockets of or on marginal and hazardous lands in cities. the population, particularly ethnic minori- In such areas, insecurity of tenure provides ties and people who live in remote areas. The little incentive for households to invest in risk implied “last mile” challenge of eliminating management, and poor provision of services extreme poverty means very different things such as health care compounds the effects of in these different contexts (box 4.3). disasters such as flooding (Baker and Gadgil In Lao PDR, Papua New Guinea, and 2017). Therefore, any public policy strategy several Pacific Island countries where the BOX 4.3  Diverse barriers make escaping extreme poverty difficult Recent research among poor, marginalized com- prominent barrier is location. Villages are remote and munities in Lao PDR and the Philippines illustrates often accessible by only a single dirt road and thus the challenges of eradicating extreme poverty in are difficult, expensive, and time consuming to reach. the region. In these communities, people living in Villagers lack access to markets, jobs, and public ser- extreme poverty found it extraordinarily difficult to vices. Although most villages have a primary school, escape poverty—but for different reasons. Examining few have a secondary school, health center, sufficient the specific, localized, and heterogeneous nature of electricity, or adequate sanitation. Residents can sel- the living circumstances of those people highlights dom afford to travel outside their villages to obtain why it is difficult to design policies that reach such needed services or to pay the costs for services. groups effectively. Children tend to drop out of school early, and dirt, The research in Lao PDR was conducted mostly disease, and ill health are common. Geographical in poor, ethnic minority villages in the highlands. The remoteness thus places clear limits on the ability of extreme poor in these villages faced almost insur- poor families to live healthy lives, acquire skills and mountable barriers to upward mobility. The most education, and invest in their futures. box continues next page E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   67 BOX 4.3  Diverse barriers make escaping extreme poverty difficult (continued) The locations of such villages also curtail resi- difficulties in getting jobs, because they had arrived in dents’ livelihoods. Nonfarm job opportunities are the city from rural areas without any skills. In Manila, too far away, so poor people depend on agricultural they lived in poor-quality, makeshift housing or lacked work to survive. Yet markets are also distant, so it housing altogether and thus were exposed to crime, is not worthwhile for villagers to diversify into cash natural disasters, floods, dirt, and disease. Like their crops that could enable them to earn a better liv- rural counterparts, these city dwellers are extremely ing. In some villages, poor families whose diets con- vulnerable to shocks and have limited capacity to sist mainly of rice and cassava reported not having manage risk. Finally, their lack of identity documents enough food for several months a year and that they is a key barrier to escaping poverty. In the Philippines, suffered from poor nutrition and bad health. those born to extreme poor families sometimes lack Families living in extreme poverty have little a birth certificate, which is needed to get a job, to capacity to manage risk. Their livelihoods are receive government benefits, and to go to school, but highly susceptible to drought, bad weather, natu- the only way for them to get a birth certificate is to ral disasters, and other shocks, such as livestock return to the town of their birth, which is not some- epidemics and disease. Health shocks were almost thing that they can afford to do. Thus, they are stuck always catastrophic and were one of the commonly in the informal economy, often scavenging. reported causes of impoverishment. To cope with The contrasting examples of remote villages of such shocks, families borrowed beyond their means, Lao PDR and urban slums in the Philippines illus- leading to cycles of indebtedness and sometimes trate the disparate characteristics of the remaining bankruptcy. pockets of extreme poor households in the East Asia Ethnicity, language, and gender also are deeply and Pacific region. In both settings, those house- connected to poverty in these villages. Most villagers holds face difficulties in their ability to benefit from belong to ethnic minorities, and they usually speak investments in job growth, agriculture, health, edu- a language that differs from the country’s dominant cation, and other public services—from efforts to language (as in Lao PDR). The language barrier lim- grow and invest—but for different reasons. In rural its villagers’ uptake of services and reduces their con- villages of Lao PDR, extreme poor households are fidence to seek work in other parts of the country. constrained by location, agroecological disadvantage, Cultural norms regarding gender also significantly lack of public services, ethnicity, and gender. In the limit the ability of girls to continue their schooling slums of the Philippines, extreme poor households and lead poor families to avoid investing in girls, thus are constrained by lack of appropriate skills, insecure significantly exacerbating intrahousehold inequality. housing, and lack of identity documents. In both set- In one remote village, only 4 of 44 students in the tings, extreme poor residents find it hard to manage primary school were girls. risk. These differences highlight the specific chal- Research in urban slums in Manila identi- lenges that governments around the region face in fied a set of different—and yet similarly almost fostering upward mobility for the remaining groups insurmountable—barriers to escaping poverty. The living in extreme poverty. poorest people in Manila were homeless or were squatters living in informal settlements. They reported Source: Beath, Parker, and Woodhouse 2016. incidence of extreme poverty still reaches Yet over the decade to 2013, household con- double digits, the policy priority is to provide sumption grew at a rate almost 4 percent- plenty of opportunities for unskilled workers, age points lower than overall gross domestic following the path of the successful countries product per capita. Had growth been more in the region. Lao PDR has created opportu- broadly shared, rather than led by the exploi- nities for poor people by improving human tation of natural resources, even greater prog- capital and access to land (Pimhidzai 2015). ress would have been achieved. 68   RIDING THE WAVE The example of Lao PDR also underscores challenges or vulnerabilities of ethnic minor- the role that risk management plays in fos- ities tend to vary significantly across groups. tering truly inclusive growth. Because most In Lao PDR, for example, minority groups of the poor work in agriculture, they remain (the non Lao-Thai) are more likely overall to very vulnerable to shocks to that sector of the be poor, but the Chine-Tibet group has done economy, as well as to idiosyncratic shocks significantly better than other minorities. such as catastrophic health expenditures. Such differences between ethnic minorities Such unmitigated shocks erode the improve- apply to both the incidence of poverty and ments that households are able to make in the likelihood of moving into or out of pov- good years. erty (Pimhidzai 2015). Extreme poverty is deeply intercon- Location can similarly define much of nected with ethnicity and location in ways what ethnic minorities have access to, includ- that are difficult to identify, measure, and ing basic rights such as being registered at tackle. These complex interactions exist in birth, lack of which may seriously hinder a all countries, including those with a high person’s ability to access social services such overall poverty incidence such as Lao PDR as education (Beath, Parker, and Woodhouse (box 4.3). But the interactions are par- 2016). Birth registration is particularly low ticularly relevant in countries with a lower among the poor in Cambodia, Indonesia, and overall incidence of poverty, where extreme Myanmar. poverty is concentrated in small pock- Location also defines the “last mile” ets of the general population. In Vietnam, challenge of eradicating extreme poverty in for instance, ethnic minorities account for urban areas, particularly in sprawling urban 15 percent of the total population but con- slums of several large cities in the East Asia stitute a majority of the country’s extreme and Pacific region (Baker and Gadgil 2017). poor. Ethnic minorities have experienced Although slums can be places of opportu- gains in welfare since the early 1990s, but nity for their inhabitants if there is access progress toward reducing poverty and child to jobs, that opportunity may come at the mortality has stalled in recent years. Poor cost of safety if slum dwellers live in precari- education, malnutrition, and low access ous dwellings on land at high risk of natural to sanitation exert mutually reinforcing disasters, such as coastal flooding. In addi- adverse effects on equality of opportunity tion, although many people migrate to cities for ethnic-minority children. for new economic opportunities, those who The influence of ethnicity on people’s do so to flee desperate or threatening circum- livelihoods, barring situations of outright stances may not be able to take advantage of discrimination, tends to manifest itself in such opportunities if they lack the right skills, either specific cultural practices or language, identification documents, or basic assets which affect what people do or what they (such as savings or housing) that would help can access. The way ethnicity shapes the them settle in and start anew (Beath, Parker, livelihoods of households living in extreme and Woodhouse 2016). Similarly, for all the poverty can therefore be both positive and opportunities that urban areas offer, slum negative. Whereas a strong sense of iden- dwellers may have to rely on the expensive tity and reliance on local support networks private provision of services that are normally can be strengths, for example, traditional available publicly, such as water and electric- agricultural practices can lead to very low ity, which in turn erodes the prosperity that productivity for communities of people liv- living in a city would otherwise bring. ing on marginal lands. Similarly, ethnic As with extreme poor households in minorities can find themselves excluded by remote Lao PDR, great heterogeneity can be dominant ethnicities when the use of certain found, even in the relatively homogeneous services or markets requires knowledge of definition of slum. In metro Manila, where the mainstream language. Also, the specific an estimated 3 million people, or 1 in 4 of E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   69 the population, live informally, a recent In other words, above the extreme poverty study (World Bank forthcoming b) identi- line, there is a large gray area of households fied 2,500 different informal settlements. that struggle to make ends meet and achieve The settlements were grouped into five economic security, especially in the short ­ different categories. Each group faced very term. For example, in Indonesia, 21 percent different challenges in terms of access to of households considered economically jobs and services and vulnerability to risks, vulnerable in 2008 were found in 2010 to so the policy solutions required to improve have fallen into moderate poverty, whereas the livelihoods of their inhabitants would be 35 percent of moderate poor households very different. had climbed to the economically vulnerable From a policy perspective, the heterogene- class. In the Philippines, between 2007 and ity of extreme poor households is a challenge 2009, 22 percent of economically vulner- in at least two ways. First, although growth able households fell into moderate poverty, patterns that produce large numbers of whereas 17 percent of moderate poor house- unskilled job opportunities can be trusted to holds rose to become economically vulner- improve the lives of many extreme poor, spe- able. In Vietnam, between 2010 and 2012, cific groups may face sources of exclusion or only 7 percent of economically vulnerable may be located in areas too remote or under- households slid into moderate poverty, and served to benefit from such opportunities. 46 percent of moderate poor households Second, countries that have already seen the moved up to the economically vulnerable benefits of broad-based growth in reducing class (Cunningham 2016). In Indonesia and extreme poverty now face the task of tailor- Vietnam, where estimates of long-term mobil- ing policies to meet the specific needs of the ity using synthetic panels are available, the remaining small pockets of extreme poverty. backward movement in class from economi- Although those countries may already have cally vulnerable to moderate poor was more the necessary resources and institutional muted over 10 years. capacity, the fragmentation of target groups, The dynamism of both moderate poor as well as their specific needs and constraints, and economically vulnerable households is may pose significant challenges in designing shown by their ability to save to purchase and implementing programs. small assets, such as cell phones or motor- cycles. They appear to lack mechanisms to save for larger productive assets, however. In Transitioning to economic some contexts, they have been able to make security up for the deficiencies in what is provided to Headline figures on the East Asia and Pacific them—for example, by organizing their own region’s extraordinary improvement in living early-warning system for flooding in one of standards focus on the virtual eradication of Indonesia’s slums (Baker and Gadgil 2017). extreme poverty. Yet the mobility analysis pre- But ad hoc solutions, however valuable, can- viously presented, and the panel analysis that not replace systematic mechanisms to manage has informed the identification of the economic risks ex ante and to help households deal with classes adopted in this report ­ indicate that shocks. Moderately poor households are also those who live just above the poverty line face significantly likely to fall into extreme pov- a large risk of falling back into p ­ overty. In erty or to enjoy only limited income growth addition, statistical analysis of the character- that does not allow them to rise above their istics of those living in moderate poverty and initial class. of those one class above—the economically In many ways, both moderate poor and vulnerable—highlights significant overlap in ­ economically vulnerable households face the characteristics of the two groups, suggest- challenges that are similar to, if smaller than, ing that the group a household falls into may those faced by extreme poor households. be significantly influenced by chance. They may struggle with a low asset base, 70   RIDING THE WAVE limited education, poor health owing to a in jobs as defined by household surveys is lack of quality health care services, scarce low.9 In addition, a large number of young large assets, and limited access to public people are not in school, are not employed, infrastructure. Many work on small farms, or are not receiving job training. High inactiv- at low-value jobs, or not at all, particularly ity rates suggest that there is limited demand many women in these classes. Because of their for these groups’ skills. Even when employed, low asset base, negative shocks are likely to these young people have low returns to their hit them harder, often resulting in a slide into human capital. Their participation in the for- moderate or extreme poverty. mal sector is also low, reducing policy mak- The primary shocks likely to affect moder- ers’ scope for using formal labor market ate poor and economically vulnerable house- payment systems to deliver cash transfers that holds relate to loss of income, as a result of might strengthen their livelihoods. drought or livestock disease, personal shocks For moderate poor and economically vul- (including illness and death), and price shocks. nerable households, most employment oppor- Furthermore, the largely rural locations where tunities are with small and medium-sized these people reside and their limited links to enterprises. Weaknesses in the business cli- the formal sector make it difficult to estab- mate constrain these businesses from expand- lish effective social protection networks. For ing employment opportunities, including for instance, both the limited availability of, and the moderate poor and economically vulner- coverage for, health insurance magnify the able households, and from making invest- impact of illnesses, putting households at ments that would increase labor productivity. risk of impoverishment because of the out-of- A particular challenge lies in labor market pocket payments for health care. policies that distort employer incentives in Other factors, such as insecure tenure or an effort to protect workers (World Bank limited access to finance, make it difficult 2014a). The East Asia and Pacific region has for people in these classes to pursue better some of the strictest (although some of the opportunities and thus diversify the risks to weakest) employment protection legislation their livelihood by engaging in additional, in the world, as well as extremely high mini- more productive activities. Skills are also mum wages.10 The outcome is weaker incen- a constraint; many people in these classes tives to hire workers, particularly women, have limited education. Because the major- youth, and less-skilled individuals. In several ity are employed in agriculture or fishing, or countries, women’s labor force participa- are otherwise self-employed, there is a need tion is further discouraged by nonlabor con- to strengthen their skills to make them more straints: moderate poor and economically productive in their present activities and to vulnerable households have particularly low give them access to other opportunities. Also, female labor force participation rates, at least their children have little access to early child- in the formal jobs captured by household sur- hood development services and experience veys, which reflects traditional gender-based significant barriers to progression through the household roles and social norms, as well as formal education system, reinforcing the inter- limited availability of clean water, childcare, generational transmission of inequality.8 and eldercare. Another characteristic that distinguishes In many countries, moderate poor and moderate poor and economically vulnerable economically vulnerable households are households from economically secure ones is concentrated in rural areas. Thus, they that they are less likely to be in paid employ- are particularly affected by official limi- ment. Only 60 percent of the moderate poor tations to geographical mobility, where and 72 percent of the economically vulner- such restrictions are in force (particularly able are employed, in contrast to 80 percent China and Vietnam) and by disparities in of the economically secure. That is largely the provision of services across the rural- because female labor market participation urban spectrum and across different zones E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   71 in the same city. As discussed, de facto and Access to high-quality services, including de jure barriers to migration (chapter 2, housing, clean water, sanitation, electricity, box 2.5), together with limited support health care, and education, remains a challenge services to match migrants with city jobs, for many economically secure households in all reduce the reallocation of labor across the region. One-third of economically secure space and sectors and depress productiv- households and 15 percent of the region’s mid- ity. Also, many people in these classes dle class (outside China) lack access to good- have limited access to public infrastruc- quality housing, clean water, and (to a lesser ture, including electricity, clean water, degree) sanitation (figure 4.3). These nonmon- and sanitation. Some of these limitations, etary deprivations reflect geographical divides such as low access to electricity and poor in service delivery, particularly between urban mobility within cities, are closely related and rural areas.12 Education quality, defined to residents’ low productivity and lack of by learning outcomes, varies significantly employment opportunities. by country. In fact, the region has some of the world’s best-performing and some of the world’s worst-performing education systems. Matching economic security with In particular, education quality is a challenge high-quality public services in Indonesia, Mongolia, and the Philippines, Economically secure and middle-class house- as well as in lower-income countries such as holds are identified by their low risk of sliding Cambodia and Lao PDR. Other countries, back into poverty. In most of the region, how- such as Malaysia and Thailand, have recently ever, risk management tools are underpro- been losing ground on internationally compa- vided,11 so even the middle classes are at risk rable tests of student performance.13 of suffering significant downward mobility. Although the middle class especially can ben- efit from good jobs and the social protection FIGURE 4.3  Nonmonetary poverty by class, developing East Asia measures that often come with such jobs, the and Pacific 2012 middle class is not immune from aggregate 90 shocks. For example, in Thailand, in 2011, Percentage of households in each income group 70 percent of the estimated US$47 billion 80 in damage and losses resulting from natural 70 disasters affected the manufacturing subsec- 60 tor (Global Facility for Disaster Reduction and Recovery, World Bank, and Rockefeller 50 Foundation 2016), which tends to be a source 40 of “good jobs.” 30 In addition, in many countries in East Asia 20 and Pacific, economic security has not been accompanied by an appropriate provision 10 of public services, in contrast with countries 0 elsewhere at similar income levels. This lack No access to No access to Poor-quality Poor on at least clean water proper sanitation housing one dimension of services brings the risk that the middle classes could turn to an insular world of pri- Extreme poor Moderate poor Economically vulnerable vate service provision and could disengage Economically secure Middle class from the national policy dialogue on service Source: Wai-Poi and others 2016. delivery and developmental outcomes. In Note: Access to clean water means piped water, protected well water, or water sold by a vendor. Access to proper sanitation means flush toilets or improved pit latrines. Good-quality housing turn, the failure of the economically secure means a roof that is made of concrete, cement, brick, stone, wood, or tiles; walls that are made and the middle class to keep government of concrete, cement, brick, or stone; and a floor that is nonsoil. The countries sampled here are Cambodia, Indonesia, Lao PDR, Mongolia, the Philippines, Thailand, and Vietnam. Poor on at least accountable risks making everyone less likely one dimension means suffering from at least one deprivation among no access to clean water, to receive quality services. no access to proper sanitation, or poor-quality housing. 72   RIDING THE WAVE The growing size and incomes of those that provide residents with private sanitation, living in economic security and of the mid- electricity, and security service. That would dle class have increased their political influ- reduce demand and accountability for higher- ence particularly in upper-middle-income quality public services, as well as reinforce countries and in some lower-middle-income and perpetuate inequalities of opportunity. countries. A key question is whether this Across countries at different levels of devel- growing economic and political clout will opment, there are already signs that opting benefit only these groups or whether it will out of publicly provided services may have make them successful advocates for a service begun, especially regarding significant use of delivery agenda that benefits the broader private health care services (figure 4.4) and, population. to a lesser extent, private education at the pri- Given their high aspirations, there is a risk mary and lower secondary levels, particularly that the people who make up these classes will among the middle class (figure 4.5).14 increasingly opt out of public services that are Surveys also indicate that economi- of low quality and instead turn to privately cally secure households in Malaysia, the provided services for health care, education, Philippines, and Thailand are already less clean water, and transportation. Some may engaged in a range of civic organizations, even choose to live in gated communities such as political parties and labor unions FIGURE 4.4  Public and private health care facility use by economic class, developing East Asia and Pacific, 2012 a. Indonesia b. Vietnam Extreme poor Extreme poor Moderate poor Moderate poor Economically Economically vulnerable vulnerable Economically Economically secure secure Middle class Middle class 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Percentage of households Percentage of households c. Cambodia d. Lao PDR Extreme poor Extreme poor Moderate poor Moderate poor Economically Economically vulnerable vulnerable Economically Economically secure secure Middle class Middle class 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Percentage of households Percentage of households Public Private, modern Private, traditional healers Other Source: Wai-Poi and others 2016. E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   73 and religious, sporting, cultural, educa- FIGURE 4.5  Private education by economic class, developing tional, environmental, professional, human- East Asia and Pacific, 2012 itarian, consumer, and self-help groups a. Private primary school, 2012 (appendix C, table C.5). In the Philippines Percentage of households in each income group 50 and Thailand, residents have less faith in public institutions, including the church, army, press, television, unions, police, 40 courts, political p ­ arties, parliament, civil service, universities, corporations, and 30 banks (appendix C, table C.6), although the opposite is true in China and Malaysia. 15 20 Similarly, the middle class may push for some problems to be simply moved to places with poorer populations and whose residents 10 have less voice. For example, in China, mid- dle-class protests against factories that pol- 0 lute and that are dangerous have resulted in Indonesia Lao PDR Mongolia Thailand stricter enforcement of environmental pro- b. Private lower secondary school, 2012 tections and in plants being moved to poorer Percentage of households in each income group 50 areas.16 The dangers are heightened by two factors. First, any expansion of public services and 40 social protection would have to be funded through increased tax revenue, including 30 higher personal income taxes. Those taxes would be paid mostly by economically secure 20 and middle-class households, and those households might be reluctant to support 10 public spending that brings them no returns. Second, because of the greater investment these two classes make in their children, as 0 Indonesia Lao PDR Mongolia Thailand well as the tendency of people to marry oth- ers from the same class, the middle class will Extreme poor Moderate poor Economically vulnerable become increasingly disconnected from the Economically secure Middle class rest of society. In turn, any growth of inequal- Source: Wai-Poi and others 2016. ity and an opting out of public services would weaken social cohesion and possibly lead to a political impasse. At the same time, as middle-class house- Notes holds care increasingly more about such 1. Developing East Asia and the Pacific is issues as consumer safety, they can push for defined as Cambodia, China, Indonesia, the better regulations and better goods and ser- Lao People’s Democratic Republic, Malaysia, vices, which would benefit all households. Mongolia, Myanmar, Papua New Guinea, the Philippines, Thailand, Timor-Leste, Vietnam, For instance, recent food safety scandals in and the Pacific Island countries China have led to stronger regulations and 2. Economically secure households in Indonesia better products for all residents. In addition, are the exception. In Indonesia, fewer than given these middle-class households’ own vul- 50 percent experienced a positive increase. nerability to shocks, they could have a stake 3. For example, in the Philippines, most people in developing an appropriate system for man- did not move, but the chances of backsliding aging risks. were also the least. This pattern is striking 74   RIDING THE WAVE because the Philippines panel data cover a 2012 found that several East Asia and Pacific period that includes both growth and crisis countries, including China and Thailand, were (2006–09). The lack of overall economic underinsured for natural disasters (Centre for dynamism seems to also have implied greater Economic and Business Research 2012). resilience (lack of backsliding) during the 12. Only one-half of rural economically secure period of economic stagnation. households but three-quarters of urban 4. Receipt of transfers, degree of job formality, economically secure households are free of and household wealth do not seem to affect nonmonetary poverty, defined as suffering the likelihood of being a so-called stayer in the from at least one of lack of access to clean household samples used here. water, lack of access to proper sanitation, or 5. Note that shocks are not identified poor-quality housing. endogenously in the survey data as drops in   These rural constraints are particularly incomes but are self-reported (through questions pronounced in Cambodia and Lao PDR, such as “Did you experience a shock?” and where most rural economically secure “What kind?”). Their identification, therefore, households are deprived with respect to at is not linked to income gains or losses or to a least one nonmonetary dimension. In contrast, specific magnitude of a shock (for example, a the share of rural economically secure drop of more than 10 percent in income) or to households that experience nonmonetary the specific duration of a shock. poverty is two-thirds in Indonesia, 56 percent 6. The qualitative results are broadly similar in Thailand, 40 percent in Vietnam, and only across countries. For instance, in rural 19 percent in the Philippines. marginalized Cambodia, 32 percent of the 13. For instance, in Indonesia and Malaysia, more moderate poor who do not experience a than 50 percent of 15-year-olds reportedly had shock move up one or two classes over a two- less than limited proficiency in reading (OECD year period, compared with 26 percent of 2012). In Lao PDR, two-thirds of rural test those who experience a shock. Conversely, the takers and one-third of urban test takers moderate poor who do experience a shock are could not comprehend basic text (World Bank more likely than those who do not to become 2014b; see also World Bank forthcoming, a). extreme poor; the probabilities are 32 percent 14. In Indonesia, health care and education and 3 percent, respectively. account for 8 percent of total spending 7. More rigorous analysis shows how seriously by economically secure households and East Asia and Pacific has been hit in the past 12 percent of spending by the middle class. decade. For example, in the first 9 months of Fully two-thirds of economically secure 2011, the region experienced 80 percent of households and three-quarters of the middle global losses that were due to natural disasters class use private health facilities. On average, (Jha and Stanton-Geddes 2013). across developing East Asia and Pacific 8. A few countries, such as Vietnam, are countries for which data were available, one- approaching universal access to early childhood third of middle-class households use private development opportunities, but this is a new primary schools. endeavor for many other countries. 15. Wai-Poi and others (2016) used the World 9. This tends to exclude unpaid family labor in Values Surveys to examine these attitudes in households’ entrepreneurial activities, often China, Malaysia, the Philippines, and Thailand. because they are combined, interspersed, or Because data on income are not available from both throughout the day with domestic work the surveys, these researchers define “secure” or with caring for children or elderly relatives. households as those that never go without As is the case with those care-giving activities, food, those that never or only rarely go without their economic contribution tends to be needed medical treatment, and those that never underestimated. have to dip into savings to support themselves. 10. For an analysis of minimum wages in member The proportions of secure and nonsecure countries of the Association of Southeast households using that definition are largely Asian Nations (ASEAN), see Del Carpio and consistent with those of economically secure Pabon (2014). and middle-class households in each country. 11. Following a year of major natural disasters There is little difference in civic engagement globally, estimates by commercial insurers in in China between secure and nonsecure E c ono m i c C l a s s and the Cha l l enge s o f I n c l u s ive G rowth   75 households, where civic engagement is very Ferreira, F. H. G., J. Messina, J. Rigolini, L.-F. low to begin with. López-Calva, M. A. Lugo, and R. Vakis. 2013. 16. Protests in Chinese cities have been found Economic Mobility and the Rise of the Latin to be associated with increased government American Middle Class. Washington, DC: imposition of fines for pollution, of plant World Bank. closures, and of suspensions of activities of Forbes. 2015. “Environmental Protests polluting factories (Ekman 2015). However, Expose Weakness in China’s Leadership.” there are also examples of urban middle- http://www.forbes.com/sites/forbesasia​ class protests that resulted in a new chemical /2015/06/22/environmental-protests-expose​ factory being relocated from a wealthy coastal -weakness-in-chinas-leadership. city to a poorer, smaller city (Forbes 2015). Global Facility for Disaster Reduction and Recovery, World Bank Group, and Rockefeller Foundation. 2016. “Toward a References Regional Approach to Disaster Risk Finance Baker, J., and G. Gadgil, eds. 2017. East Asia and in Asia.” http://pubdocs.worldbank.org​ Pacific Cities: Expanding Opportunities for /­en/622001465500604140/DRF-Rockefeller​ the Urban Poor. Urban Development Series. -FINAL-002.pdf. Washington, DC: World Bank. Jha, A. K., and Z. Stanton-Geddes, eds. 2013. Bangalore, M., A. Smith, and T. Veldkamp. 2016. Strong, Safe, and Resilient: A Strategic Policy “Exposure to Floods, Climate Change, and Guide for Disaster Risk Management in Poverty in Vietnam.” Policy Research Working East Asia and the Pacific. Washington, DC: Paper 7765, World Bank, Washington, DC. World Bank. Beath, A., B. Parker, and A. Woodhouse. 2016. OECD (Organisation for Economic Co-operation “Heterogeneity and Marginalization among and Development). 2012. Program for the Extreme Poor in East Asia and the Pacific.” International Student Assessment database. Background paper for Riding the Wave: An http://www.oecd.org/pisa/pisaproducts​ East Asian Miracle for the 21st Century . /­pisa2012database-downloadabledata.htm Washington, DC: World Bank. Pimhidzai, O. 2015. “Drivers of Poverty Reduction Centre for Economic and Business Research. 2012. in Lao PDR.” Lao PDR Poverty Policy Notes, “Lloyd’s Global Underinsurance Report.” World Bank, Washington, DC. Lloyd’s, London. Wai-Poi, M., R. Purnamasari, T. Indrakesuma, Cunningham, W. 2016. “Class Mobility.” I. Uochi, and L. Wijaya. 2016. “East Asia’s Background paper for Riding the Wave: Rising Middle Classes.” Background paper for An East Asian Miracle for the 21st Century Riding the Wave: An East Asian Miracle for the Washington, DC: World Bank. 21st Century. Washington, DC: World Bank. Dang, H.-A., P. Lanjouw, J. Luoto, and World Bank. 2014a. East Asia Pacific at Work: D. McKenzie. 2014. “Using Repeated Cross- Employment, Enterprise, and Well-being. Sections to Explore Movements into and out of Washington, DC: World Bank. Poverty.” Journal of Development Economics ———. 2014b. Lao Development Report 2014: 107 (C): 112–28. Expanding Productive Employment for Broad- Del Carpio, X., and L. Pabon. 2014. “Minimum Based Growth. Washington, DC: World Bank. Wage Policy: Lessons with a Focus on the ———. Forthcoming a. Education Quality and ASEAN Region.” 87864-EAP, World Bank, Economic Prosperity in East Asia and Pacific. Washington, DC. Washington, DC: World Bank. Ekman, A. 2015. “China’s Emerging Middle ———. Forthcoming b. “Navigating Informality: Class: What Political Impact?” Institut français Perils and Prospects in Metro Manila’s Slums, des relations internationales, Center for Asian Metro Manila Slum Study.” World Bank, Studies, Paris. Washington, DC. A Policy Agenda for Inclusive Growth 5 Introduction relevant to their conditions, governments can lay the foundations for a new social contract, Despite the huge economic progress in devel- which will adapt the region’s development oping East Asia and Pacific, some groups model to the challenges ahead. run the risk of being left behind in extreme poverty. Other groups enjoy only limited economic security and are finding that they The three pillars: economic cannot hold on to their gains for long. At the mobility, economic security, and same time, the rapid pace of development institutions for inclusive growth is changing the needs and aspirations of the Economic mobility population, generating demand for higher- Despite the remarkable economic dynamism quality public services. And inequalities (real in the region, many households find it dif- or perceived) are growing or remain high. ficult to improve their economic status. For To address the needs of the different eco- all groups, short- and long-term income gains nomic classes, this report identifies three pol- are not always sufficient to lift households icy priorities, or pillars: fostering economic into higher economic classes (chapter 4). In mobility, increasing economic security, and Indonesia, for example, 36 percent of the underlying those priorities, strengthening the extreme poor and 29 percent of the moderate institutions required for inclusive growth. poor did not improve their class over a 10-year Within each area, the different needs of the period. Although broad-based employment various classes may be best served by differ- creation has been central to the region’s abil- ent types of policy or program designs, so this ity to increase opportunities in the past, ensur- chapter focuses selectively on those that appear ing that everybody can access jobs and that to be a binding constraint for specific classes. the quality of those jobs improves should be This chapter summarizes key areas for fundamental elements of a strategy for foster- action under each pillar, then translates those ing economic mobility. Such an agenda is very areas into priorities across different country broad, but key priorities include the following: conditions (box 5.1). Finally, it describes spe- cific measures that might help achieve those •  Bridging gaps in access to jobs and priorities. By addressing the priorities most services. Although overall levels of labor 77 78   RIDING THE WAVE BOX 5.1  Diverse countries, different challenges This report adopts a classification of countries into Out-of-poverty-into-prosperity countries (China, five groupings, based on their common characteris- Mongolia, and Vietnam), where most people are at least tics in terms of class structures and their evolution. economically secure and the middle class is growing Such groupings are not expected to do justice to all Out-of-extreme-poverty countries (Cambodia, ­ country-level specificities but rather to introduce Indonesia, and the Philippines), where extreme pov- more nuanced messages than could emerge from a erty levels are low but the shares of the middle class simple regional overview, especially with regard to are also low policies. The five groupings are as follows: Lagging-progress countries (Lao People’s Progressive prosperity countries (Malaysia and Democratic Republic and Papua New Guinea), which Thailand), which have eliminated income poverty still have high levels of extreme poverty while substantially increasing the share of middle- Pacific Island countries, which are a heterogeneous class households group and are distinct from the rest of the region market participation are high in the are therefore key interventions to improve region, women experience large gaps in the quality of jobs. participation across all income classes. •  Furthering financial inclusion. Given the Another barrier that prevents people from region’s employment profile and the role accessing jobs includes cumbersome regu- played by micro and small firms as key lations for obtaining identity documents. sources of jobs (World Bank 2014a), Those barriers may compound the dif- improving the quality and productivity of ficulties lower-income households face the employment opportunities they offer in accessing services or government pro- is another important priority. Furthering grams, which include language and high financial inclusion can be very effective implicit costs, both for themselves and for in this respect. It calls for developing their children. sound legal and regulatory frameworks •  Improving the quality of jobs. Although that balance financial stability, integrity, agricultural employment has declined, the and inclusion; supporting the information pace of change has differed by country and environment needed for the financial sec- income level. Lower-income households tor to assess creditworthiness; and educat- constitute the majority of those engaged ing and protecting the users of financial in agriculture, with many trapped there by services to enable them to make financially insecure or ill-defined property rights that sound decisions. Such an agenda has ben- prevent investment or mobility to urban efits that extend well beyond those recipi- areas. Regulatory barriers to mobility and ents who have no option but to become low connectivity of remote rural areas also microentrepreneurs, because it supports a thwart moves to more productive jobs. In financial sector that is more efficient and addition, those who do move out of agri- transparent for all income classes. culture are unlikely to join the salaried sector; they generally resort to working Economic security as microentrepreneurs, with barriers to formalization that make it too onerous to Social protection systems have historically leave the informal sector. Removing these played a limited role in East Asia’s devel- constraints and securing property rights opment model, despite the fast pace of A P o l i c y A genda f or I n c l u s ive G rowth   79 innovation over the past decade. Poverty natural disasters. Such measures would reduction has been mostly fueled by job cre- involve both managing risks ex ante—for ation and growing productivity rather than example, by developing market-based crop redistribution, while risk management, both insurance mechanisms or by investing in at the household and at the country level, has early warning systems—and setting up ex not been a priority. To provide greater eco- post mechanisms for coping with nega- nomic security for all groups, governments tive shocks when they materialize, such as need to focus on at least three priorities: adaptive (that is, scalable) safety nets. •  Strengthening social assistance. Over the past decade, several cash transfer pro- Delivering inclusive growth grams (conditional or not) have been To create an environment that fosters inclu- introduced in the region, although they sive growth, governments need to increase remain absent in a few lower-income and their capacity to mobilize resources. They lower-capacity countries. Such programs also need to level the playing field to allow can play a key role in improving the eco- different economic actors to compete fairly. nomic opportunities of the poorest and Supportive institutional and policy reforms thus reducing poverty more quickly, par- are required to deal with the two key regional ticularly in the pockets of the population trends highlighted earlier: aging and urban- where extreme poverty remains high. ization. Key areas for action include the •  Strengthening social insurance. Although following: most countries have put in place measures to help households insure against idiosyn- •  Mobilizing resources through progres- cratic risks, such as pensions, their cov- sive taxation by strengthening personal erage tends to be limited. This area has income taxation, including by broaden- seen the large-scale adoption of innova- ing the tax base, reviewing marginal tax tive measures, such as social pensions in rates, and eliminating the preferential China, but the programs are fragmented, treatment of capital income; establishing hindering mobility within the labor mar- effective property taxes; and enhancing ket. Population aging and urbanization are tax administration likely to exacerbate demands for the pub- •  Increasing the effectiveness of pro-­ inclusion lic sector to provide social protection more spending programs through better manage- systematically across the income spectrum. ment tools, including social registries (data- The blurring of the distinction between con- bases of individuals and their socioeco- tributory and noncontributory programs nomic situations) and improved targeting offers a way forward toward universal mechanisms ­ coverage, provided such coverage follows •  Leveling the playing field by boosting a fiscally responsible growth path. Many competition, reducing opportunities for countries in the region have made prog- capture in the more regulated sectors of ress by moving away from universal but de the economy, and tackling nepotism and facto regressive measures such as food and corruption fuel transfers. Yet expanding the system in •  Adapting to population aging by support- a fiscally sustainable manner is likely to ing lifelong learning to upgrade workers’ require expanding the revenue base. skills throughout their lives, eliminating •  Increasing resilience to large systemic incentives for older workers to retire too shocks, such as natural disasters. To early, opening up aging labor markets to complement risk management measures greater inflows of young immigrants, and at the household level, country-level mea- adapting the health sector to the chal- sures are needed that increase resilience lenges of aging, including the increased to large systemic shocks, particularly incidence of noncommunicable diseases 80   RIDING THE WAVE •  Promoting efficient urbanization by sup- remaining pockets of extreme poverty. The porting efficient land use in cities of all sizes, incidence of extreme poverty in Cambodia, including by clarifying property rights; Indonesia, and the Philippines is now in the strengthening title registration; creating insti- single digits. Unlike other countries in the tutions that allow for transparent, market- region that have low levels of extreme pov- based land valuation; establishing effective erty, however, the three countries also have a urban planning to reduce red tape and to percentage of the middle class that is in the facilitate changing land-use allocations; and single digits. Policy priorities for this group strengthening urban governance to improve therefore need to reflect the needs and aspira- the delivery of services and provision of urban tions of a diverse population, spanning those infrastructure, including by improving coordi- in moderate poverty, the vulnerable, and those nation between different levels of government who are economically secure. Integrating and different jurisdictions. existing social assistance programs to provide greater economic security is a key priority, with strengthening mobility a close second. A new social contract: priorities Strengthening mobility will be particularly across different country groups important in the Philippines, where the dis- This report proposes a selective policy agenda tribution has been mostly static until recently. to address the needs of the different economic Given the size of their populations, however, classes, with a focus on three pillars: foster- Indonesia and the Philippines are still home ing economic mobility, increasing economic to large numbers of extreme poor. Having security, and strengthening the institutions such large numbers calls for identifying and required for inclusive growth. The assignment removing the very specific constraints those of priorities within this policy framework groups face in accessing economic opportuni- can be adapted to the region’s heterogeneous ties and service delivery. country and policy contexts and level of eco- For the out-of-poverty-into-prosperity nomic progress. The following priorities are countries, the priority is reducing the remain- proposed based on the country typology used ing vulnerabilities while meeting the needs in this report. and aspirations of the middle class. In China, For the lagging-progress countries, the pri- Mongolia, and Vietnam, sustained poverty ority is to accelerate poverty reduction. In the reduction has led to the emergence of a mid- Lao People’s Democratic Republic and Papua dle class. The priorities are, first, to reduce New Guinea, the incidence of poverty is still residual vulnerabilities, including among more than 40 percent, although progress rural households and groups with special has moved some of the population up into needs (in China, this includes the elderly higher economic classes. Key policy actions and children left behind when their parents should center, first, on accelerating poverty migrate; in Mongolia, those living in the ger reduction by strengthening basic education areas;1 and in Vietnam, ethnic minorities); and financial inclusion; removing barriers to and second, to meet the needs and aspira- labor reallocation from agriculture to more tions of the middle class while preparing for productive sectors; and tackling obstacles rapid aging. That second priority will require that prevent the marginalized population improving migrants’ access to public ser- from accessing public services. Second, poli- vices; enhancing pension systems in a sustain- cies should help the vulnerable manage the able manner; and ensuring good-quality and impact of shocks by strengthening social cost-effective health care and education. In assistance and making economies more resil- China, reforming urban planning to promote ient to disasters. efficient land use and environmental sustain- For the out-of-extreme-poverty countries, ability is also critical. the priorities are securing progress and fos- For the progressive prosperity coun- tering upward mobility, while tackling the tries, the priorities are to satisfy the middle A P o l i c y A genda f or I n c l u s ive G rowth   81 class’s growing aspirations while mobiliz- opportunities more effectively, seeking access ing resources to address remaining dispari- to international job markets, supporting man- ties. These richer countries, which include aged urbanization, and improving agricultural Malaysia and Thailand, have continued to productivity. Priorities to strengthen economic move large numbers of the economically security include investing in disaster loss pre- vulnerable into economic security, and they vention, mitigation, and management, and now have large middle (and rich) classes. improving health and other social programs. A first priority is to satisfy the middle class’s Finally, delivering on inclusive growth will not growing aspirations and demands for quality be possible without improving the effective- public services, including health and educa- ness of public expenditures and service provi- tion, which will support continued upward sion, addressing the challenges of remoteness, mobility. Second, reforms should help narrow and mobilizing development aid. Malaysia’s and Thailand’s rural-urban and Addressing such issues simultaneously can cross-regional disparities in economic and form the basis for a new social contract that social development. In turn, such reforms will will offer something to all groups. The lowest- require increased resource mobilization. income groups will be given the opportunity For the Pacific Island countries , the to escape poverty. The vulnerable will be inclusive growth agenda needs to be tai- protected from shocks that could push them lored to their idiosyncratic circumstances back into poverty. The economically secure (see box 5.2). Because human resources will and the middle class will enjoy access to remain their principal asset, priorities to fos- the improved services they need to continue ter economic mobility include using existing raising their incomes and standard of living. BOX 5.2  A policy agenda for the Pacific Island countries Their combination of extreme remoteness, small PICs invest in overarching and market-responsive size, geographic dispersion, and environmental fra- improvements in basic education. To those ends, the gility means that Pacific Island countries (PICs) face PICs need to invest in solid foundational education a considerably diminished set of opportunities com- that rests on early childhood development and qual- pared with most other countries. PICs, for example, ity schooling, as well as in higher-end technical and are unlikely to experience export-driven growth and managerial skills. associated employment creation on the scale seen in Improving access to international jobs markets. the broader East Asia and Pacific region. The policy Given current rates of population growth, it is likely agenda for PICs, in the context of the three pillars that domestic formal job markets will not be able used in this report, will have to recognize those spe- to absorb the growing number of job seekers in the cial constraints. PICs, even in the most optimistic scenario. Therefore, a policy to aggressively seek and expand access to Fostering economic mobility international job markets would be highly advanta- Investing in people. In both the short and long term, geous. Not only can PICs seek to expand ongoing human resources will remain the PICs’ principal low-skill seasonal employment arrangements with resource. Using existing opportunities in tourism and Australia and New Zealand, but they also can imple- international job markets, as well as creating fresh ment policies that help workers climb the skills lad- opportunities in new areas such as information and der and effectively compete with other international communication technology, fisheries, and deep sea workers who are drawn by the emerging skill gaps in mining, will greatly depend on the extent to which advanced and bigger economies in the area. box continues next page 82   RIDING THE WAVE BOX 5.2  A policy agenda for the Pacific Island countries (continued) Supporting managed urbanization in the PICs. for social programs that complement traditional By consolidating populations living in far-flung and systems in helping people reduce exposure to these hard-to-service remote locations, urbanization will types of shocks. Measures include strengthening allow PICs to harness some economies of scale, cre- tobacco controls and increasing excise duties; reduc- ate “thicker” markets, and increased specialization. ­ onsumption of unhealthy foods through targeted ing c This objective will require investments in connec- preventive measures, taxes, and better regulation; tive infrastructure and improved urban planning and and improving the efficiency and impact of the health administration to prevent congestion, conflict, and budget. urban unemployment. Delivering inclusive growth Improving agricultural productivity. Most people in the PICs rely on subsistence agriculture for their Improving the effectiveness of public expenditures livelihood. Especially in the short run, programs and services to improve development outcomes. and policies that improve agricultural productivity, Human development indicators remain low in several including the introduction and production of new PICs, in part because of low levels of public spending and healthier foods, can go a long way in improv- on basic human services. The scope for further rev- ing livelihoods of agricultural households. Improved enue mobilization is limited, however, so improving value chains that link urban and rural areas will ben- public services has to rest on improving the quality of efit both urban dwellers and agricultural producers. expenditure. By setting priorities for improvements in public financial management and for the quality Improving economic security of service provision, the PICs can do more with less. Managing shocks. Several PICs are among the coun- Addressing challenges of remoteness. Meeting the tries in the world most vulnerable to natural disas- needs and aspirations of population groups living ters, which leads to frequent and significant reversals in remote islands will continue to be highly chal- of economic gains. Hence, significant investments in lenging. Although orderly urbanization provides a disaster loss prevention, mitigation, and management potential solution, innovative thinking, including the are needed. Tackling aggregate shocks will mean fur- use of appropriate information and communication ther expanding and strengthening regional risk-sharing technology systems, will be required to extend basic initiatives such as the Pacific Catastrophe Risk Assess- health, education, and social services to such groups ment and Financing Initiative. It will also mean pursu- in a sustainable manner. ing prudent macroeconomic policies that build up sav- Mobilizing development aid. Even assuming best- ings in good times to invest in disaster prevention and case outcomes for revenues and expenditures, the management. Such measures need to be paired with Pacific Island countries will continue to need devel- cost-effective and well-targeted safety net programs opment aid to maintain sustainable fiscal positions that protect the poorest and the most vulnerable. while addressing development needs. It will therefore Improving health and other social programs. In be important for PICs to pursue both bilateral and spite of strong traditional systems of self-help, the multilateral engagement to mobilize international growing epidemic of noncommunicable diseases assistance. and unwanted pregnancies have increased vulner- ability across the PICs. More investments are needed Sources: Haque and Packard 2014; World Bank 2017a, 2017b. And the rich will benefit from a stable social Fostering economic mobility and political environment. All classes will therefore have an incentive to contribute to Broad-based employment creation has been and support reforms, which in turn will posi- central to the ability of developing East Asia tion the region to respond effectively to the and Pacific to increase opportunities. Yet the challenges ahead. region cannot take for granted continued A P o l i c y A genda f or I n c l u s ive G rowth   83 productivity gains of the kind it experienced Public provision of services that women in the past (World Bank 2014a) at a time are expected to provide, such as caring for when inequalities in access to jobs, and children and the elderly, could make it pos- especially good jobs, are becoming more sible for women to engage in paid work. salient. Although “qualitative accounts from However, care for the elderly is not widely Cambodia, Indonesia, the Philippines, and provided currently, and childcare provision Thailand indicate that finding employment could be expanded to cover urban areas more and starting new businesses are the most fre- fully (Baker and Gadgil 2017). In rural areas, quently offered explanations for households nontraditional delivery mechanisms could be moving out of poverty” (Narayan, Pritchett, explored, such as training local women to and Kapoor 2009, as quoted in World Bank provide childcare in their homes. Fetching 2014a, 50), many still living in extreme water also consumes a significant amount poverty might not be able to access these of poorer women’s (and children’s) time,3 so opportunities. Barriers to accessing services strengthening access to water could play an might further compound their challenges important role in improving female labor by diminishing human capital or limiting market participation. Where electrification is mobility. And unless existing and new jobs not yet universal, such as in Cambodia and result in growing productivity, those who Lao PDR, improving access can allow women can access them will not experience rising from moderately poor and economically vul- living standards. Therefore, the central ele- nerable households to use labor-saving appli- ments of a strategy for fostering economic ances to reduce time spent in unpaid work mobility must be ensuring that everyone (Goldin 2006); such appliances appear to be has opportunities for paid employment, within the range of items that their families that people can move to better opportuni- can save up to buy. ties if available, and that the quality of jobs improves over time. High shares of young people not in education, employment, or training Youth who are not in employment, educa- Addressing existing gaps in access to tion, or training represent a large share of the jobs, particularly better jobs moderate poor and economically vulnerable. This section focuses first on some of the most Much policy innovation has been devoted to glaring gaps in access to jobs that are revealed assisting them, as learned from experiences in the analysis of economic classes and then from other parts of the world. For example, on the regulatory barriers that appear to con- the Urban Youth Employment Program in tribute to those gaps. Papua New Guinea provides behavioral and socioemotional skills training, financial liter- Female labor market participation acy training, and public works opportunities Although women’s participation in the labor to low-skilled youth. The program also pro- market is high overall in the region, it differs vides more skilled youth a more professionally according to economic class (see chapter 4). oriented, longer-term module of behavioral Lower-income households’ economic poten- and socioemotional training, technical tial remains partly unexploited because skills training, and internship opportunities. women face a range of nonmarket con- Although the program has yet to complete an straints, reflecting gender-specific household impact evaluation, the second module (for the roles and social norms. Although the situa- more skilled) has been frequently evaluated in tion varies by country and economic class, the Latin American context, with consistently alleviating such constraints can increase labor positive results for employment, wages, and market participation for women,2 and thus formal sector employment, particularly for can facilitate the transition of households women (Ibarraran and Rosas 2009; Vezza into the middle class. and others 2014). 84   RIDING THE WAVE Regulatory barriers Specific barriers to accessing services, Several regulatory requirements might especially education make it difficult for certain groups to access Access to some services can be limited jobs, services, and government programs, for some groups, especially lower-income beyond the difficulties that migrants face in groups, by the way provision of those ser- accessing services (see chapter 2, box 2.5). vices is organized. Access is particularly Examples include cumbersome procedures important in the case of education. However, to obtain identity documents, such as in Lao both access and quality4 of education at the PDR, where birth registration processes and preschool, primary, secondary, and higher charges and other regulations make it diffi- levels vary greatly within countries and, in cult for those in extreme poverty to ­register general, are lower in rural areas (World Bank, their children. Appropriate responses could forthcoming a). In Vietnam, for example, include incentives for birth registration upper secondary education is not universal. (such as small cash or in-kind incentives Compulsory education ends after the ninth to cover transport and other costs) and the grade, and access to the upper secondary level waiving of registration fees. In addition, is allocated through entrance exams in many procedures should be established for births provinces. In larger urban areas, children are to be registered late and for replacement likely to find a place in an academic second- documents to be available without oblig- ary school if they are interested in pursuing ing parents to return to the municipality an academic rather than a vocational educa- of birth. Central governments should con- tion. However, in smaller urban areas, the duct one-time outreach campaigns in every rationing is more binding. Larger numbers of community (perhaps connected with cen- students, typically from a poorer background, suses) to ensure that households or indi- are barred from secondary school and are viduals who lack documents acquire them. streamed toward vocational education or Administrative fees should be waived during simply left without any options for further the campaign (World Bank and World schooling. Vocational education is typically Health Organization 2014). considered to be of lower quality and carries Although barriers such as registration some social stigma. To identify alternatives to requirements are easy to identify, other types this type of early streaming, the government is of hurdles are more elusive, such as cultural currently experimenting with a greater variety and linguistic barriers. In countries where of courses, including vocational courses, in extreme poverty is increasingly concentrated academic secondary schools. among ethnic minorities, such as Vietnam, Early childhood development (ECD) pro- the low status of such minorities and the grams are among the most cost-effective strat- difficulty in reaching them with antipoverty egies for improving learning and increasing programs reinforce traditional stereotypes equity of outcomes because they improve of the marginalized (Wells-Dang 2012). school readiness and significantly increase Specific factors that limit ethnic minorities’ children’s chances of enrolling in and success- access to jobs in urban areas, to social assis- fully completing primary education (Belfield tance programs, and to other basic services and others 2006; Heckman and others 2010). are therefore overlooked. Language barriers ECD programs are also a critical component and access to information play a key role, of any coordinated strategy to tackle malnu- particularly where service delivery is con- trition, which has severe, detrimental effects cerned. In Vietnam, current initiatives to on educational and labor market outcomes use teaching assistants who know the local (World Bank 2016a, 72–95). Malnourishment mother language in the first few years of is particularly prevalent among ethnic minori- primary school hold promise for boosting ties and poorer households. However, partici- learning among children who do not learn pation in ECD programs is very low in most Vietnamese at home. of the region, particularly among the poorer A P o l i c y A genda f or I n c l u s ive G rowth   85 economic classes, thereby putting young transfers conditional on school attendance, ­ children at a social and cognitive disadvan- although coverage could be expanded. In the tage even before starting school. Fortunately, Philippines, where the Pantawid Pamilya pro- there is a growing consensus about the ele- gram provides educational subsidies, identity ments of a high-quality ECD program for the documents are one of the constraints that poor. A good example of the implementation prevent the extreme poor from enrolling. In of such principles is Indonesia’s early child- general, countries with large or concentrated hood education and development (ECED) groups of the extreme poor should consider program, which has successfully reached less school feeding programs, along the lines of affluent children living in villages across the India’s Mid-Day Meals Scheme. Efforts are geographically dispersed nation by using com- also needed to address the constraints that munity-defined and -operated ECD services result in low tertiary education enrollment (box 5.3). Relatedly, measures to improve rates among poorer, rural, and ethnic minor- parents’ attitudes toward the education of ity students. For instance, in China, measures their children and to improve parent-teacher have been implemented since 2012 to pro- relationships can have a significant effect on mote equal access to higher education, with students’ literacy.5 some success.6 Going beyond ECD, policy efforts should include incentives for extremely poor house- Improving the quality of jobs available holds to invest in primary and secondary Much of the productivity growth experi- education, although those alone might not enced by the region over the past couple of be enough. In Vietnam, financial support decades has been driven by labor reallocation is provided to ethnic minorities to attend and structural transformation. China’s case upper secondary schools, including cash is well known, but Indonesia, Mongolia, and BOX 5.3  Starting off strong through community-focused early childhood development in Indonesia In 2008, the government of Indonesia began imple- Data collected in 2009, 2010, and 2013 showed menting a community-driven early childhood educa- how the program performed: tion and development (ECED) program to increase •  The program increased enrollment in ECED the supply of ECED services in poor communities. services by 7 to 22 percent. The program capitalized on the country’s strong •  It increased social competence by 0.25 of a stan- tradition of community-driven development. At the dard deviation (s.d.), language and cognitive time, only 33 percent of children ages 0–6 from the development by 0.13 s.d., and emotional maturity wealthiest quartile of families were receiving ECED by 0.24 s.d. for the poorest children. services, and only 8 percent of children from the •  It reduced emotional symptoms of children from poorest quartile of families accessed such services. the poorest households by −0.17 s.d. The program targeted poor communities. Trained •  The program’s effects were larger for children local facilitators helped village members identify their from the poorest households and for children ECED needs and prepare proposals for small grants from households with poor parenting skills. to finance up to two services per village for three years. Up to two teachers per service received an As of 2013, the ECED program had reached intensive 200 hours of training. Most communities 3,000 villages across 50 districts in the country. To used their resources to establish center-based play- date, more than 600,000 children have been reached. groups primarily serving children ages 3 to 6. Centers typically operated at least three days a week, usually Sources: Brinkman and others 2015; Hasan, Hyson, and Chang 2013; Jung and in sessions of two hours a day. Hasan 2014. 86   RIDING THE WAVE Vietnam have similarly benefited (World Bank and Qu 2006; Yan 2013). Similarly, in the 2014a). Those processes have meant that the Philippines, despite a protracted program quality of jobs has improved and that peo- of agrarian reform (de los Reyes, Librojo, ple have been able to move to take up jobs. and Capacio 2016), land-related investment For such reallocation to occur, policies must remains hampered by a complicated and improve the regulation of property rights on sometimes contradictory legal framework agricultural land. That regulation not only governing property rights and by weak land favors investment in agriculture, which raises records based on inaccurate information. productivity, but also allows people to move Property holders are discouraged from regis- out of agriculture if they see better opportu- tering their claims by high transaction costs. nities elsewhere. Similarly, addressing gaps The priority is to secure property rights for in infrastructure provision, particularly for the majority of the population through land broad connectivity, is crucial to raise produc- governance reforms, which include improve- tivity of jobs in both rural and urban areas ments in land information systems and the and to favor mobility. Finally, addressing the removal of bias against rural titles. constraints to formalization that many micro and small enterprises face can help reduce Addressing poor connectivity the pervasive informality in the economy and raise productivity. Connectivity is a particularly significant constraint to mobility and market access in remote rural agricultural areas, including in Investing in agricultural development Indonesia, the Philippines, and Vietnam. The and removing uncertainty from lack of access can be tackled in the short term property rights to land by enhancing the cost-effectiveness of trans- With three-quarters of the extreme poor and port, which includes increasing transport two-thirds of the economically vulnerable capacity using the existing level of infrastruc- employed in agriculture, agricultural devel- ture, for instance, by disseminating adapted opment remains a central priority for the vehicles through specific investment financ- region, even if the goal is to invest in agri- ing schemes. Longer-term improvements can culture to foster the kind of structural trans- address rural road or track conditions, espe- formation that moves workers out of the cially by strengthening all-weather road con- sector. Agricultural development directly ben- nections between poorer areas and potential efits some of the poorest segments of society, growth poles. Improvements in physical infra- attenuates pressures for rising inequality, and structure must be complemented by enhanced lays the foundations for the success of other value-chain services, including transport ser- reforms. vices, input provision, and access to credit, The way land is regulated is central to including credit for agricultural inputs. efforts to improve agricultural productivity. In many areas in China, for example, uncer- Removing barriers to formalization tainty about the security and duration of land-use rights deters farm investment by Workers who move out of agriculture are both reducing incentives and preventing unlikely to move to the salaried sector. They land from being used as collateral for loans. generally resort to work as microentrepre- It also hampers agricultural consolidation, neurs, because barriers to formalization make dulls incentives for off-farm work and entre- moving out of the informal sector too oner- preneurship, and discourages both migration ous. Although such informal sector jobs could and the productive use of land left behind by be the springboard to better opportunities in migrants. The result is lower agricultural pro- an urban setting, they rarely are. Evidence ductivity and increased rural-urban inequali- from Indonesia, for example, shows that few ties (Hou and Huo 2016; Tan, Heerink, individuals see their businesses grow from A P o l i c y A genda f or I n c l u s ive G rowth   87 petty self-employment activities to businesses agricultural and rural enterprises, and young that hire workers. Furthermore, some groups firms—which account for a majority of the of workers, such as young people, receive region’s employment—are particularly con- negative returns from time they spend in petty strained in their access to finance.10 business activities, thus reducing their human Supporting greater financial inclusion capital (Naidoo, Packard, and Auwalin requires three broad, complementary poli- 2015). Many regulatory factors also help cies (World Bank 2014b): developing sound sustain the labor market segmentation that legal and regulatory frameworks that bal- results in pervasive informality. Those fac- ance financial stability, integrity, and inclu- tors include, for example, rules on access to sion; supporting the information environment credit that might exclude the self-employed; needed for the financial sector to assess cred- different treatment of different types of work- itworthiness; and educating and protect- ers by the labor code; and rules on how social ing the users of financial services to enable protection benefits are designed, particularly them to make financially sound decisions. those that tie coverage of social insurance to Beyond these high-level interventions, new certain types of work and that are financed insights from the behavioral economics lit- by mandatory contributions from employ- erature are also contributing to the design of ers. As discussed later in this chapter, some schemes that effectively promote savings at of the innovations in the area of social pro- low cost, without requiring matching funds tection are starting to address some of these or other expensive incentives. In general, such constraints, but much remains to be done in schemes rely on easy deposits to encourage this area. frequent saving behavior, somewhat difficult withdrawal to discourage impulse buying, and clear savings goals for larger assets. For Furthering financial inclusion to instance, with commitment savings accounts, foster savings individuals make deposits and relinquish Developing countries in East Asia and access to the funds for a given period or Pacific have the highest levels of financial until a goal has been reached (box 5.4). Such inclusion compared with other develop- accounts can both shield funds from family ing regions, as measured by the share of or social network pressures and mitigate self- adults who have an account at a financial control issues. institution or use mobile banking. That per- centage reflects a great increase over the past few years, particularly among lower- Increasing economic security income groups.7 Yet an estimated 490 mil- To increase economic security, governments lion people remain “unbanked”—without not only need to invest in measures that help access to the services of a bank or similar accelerate poverty reduction but also need to financial organization (Demirguc-Kunt and help households and countries become more others 2015). Prohibitive transaction costs, resilient. At the household level, such social distance from financial providers, or legal protection systems would ideally both help and regulatory barriers, including onerous close the poverty gap and offer all income documentation requirements, help explain groups a way of managing risks. At the coun- this finding. Poor, rural or geographically try level, investment would focus on putting remote, unemployed, less educated, elderly, in place both ex ante mitigation measures and and young households are especially likely ex post coping measures. not to hold accounts at formal financial Social protection policies and programs institutions. 8 Limited financial literacy vary widely. Yet all countries have some form among households reduces the benefits of social insurance covering at least some seg- from whatever financial services are avail- ments of the formal sector (notably the civil able.9 Small and medium-size enterprises, service), and most have some form of targeted 88   RIDING THE WAVE BOX 5.4  Creativity plus a little behavioral economics: rural savings in Mindanao A small rural bank in the Philippines has shown that education expenses or capital for business. Second, a combination of nontraditional savings mechanisms participants took home a piggy bank that had a slot and behavioral economics can induce and increase for depositing money. The key to open the bank and personal savings. The Green Bank of Caraga in access the money was held by the bank. The partici- Mindanao offered women the SEED (Save, Earn, pants would take the piggy bank to the Green Bank and Enjoy Deposits) program. Program clients were periodically for the latter to transfer the savings into offered the same interest rates and deposited their the formal account. The clients were not required to money in the same way as anyone else, with a few make deposits into their account. adjustments. First, the clients signed a contract with By outsourcing to the bank the enforcement of the the bank that allowed withdrawal of funds only when savings goal and providing a mechanism for small a trigger, set by the client herself, was met. The client savings to be easy to deposit but hard to access, the could specify a date, an expenditure type, or a sum of program participants saved 81 percent more over savings, and the bank was committed to release the a year than a randomly selected group of bank cli- money only when such a goal was met. Nearly half ents who had been assigned to the traditional bank the clients opted for a date when expenses would be savings mechanism. incurred, such as Christmas, a birthday, or a gradu- ation. Others selected specific expenditures such as Sources: Ashraf, Karlan, and Yin 2006; Ashraf, Karlan, and Yin 2010. program for the poorest. Programs that introduced or scaled up, albeit with signifi- blur the line between these two systems— cant variations from country to country and and particularly between contributory and with gaps in coverage. Since the introduc- noncontributory schemes—appear to offer tion in 2008 of the Pantawid Pamilya pro- a way to extend coverage to the “missing gram in the Philippines—a geographically middle” of the distribution. Having started targeted conditional cash transfer program late to put in place a modern social protec- aimed at poor families with children or preg- tion system, countries in the region have an nant women—several cash transfers have opportunity to learn from international expe- been introduced, both conditional (such as rience, avoiding pitfalls such as the creation in Cambodia and Indonesia) and uncondi- of welfare traps and excessive labor taxation. tional. The largest such program is China’s Two main challenges are to avoid the frag- dibao or minimum living standard guaran- mentation of schemes for different types of tee, which was first introduced in urban areas workers, which limits mobility and results in and then in rural areas and currently covers uneven treatment, and to ensure that systems about 70 million people. The program tops are expanded in a fiscally sustainable man- up household incomes and provides several ner by expanding the revenue base (see also noncash benefits to beneficiaries, such as “Increasing the Redistributive Capacity of exemptions or reductions in school fees and the State”, below). subsidies on housing or utilities. Such programs are increasingly targeted based on needs or poverty, partly in order Strengthening social assistance to better respond to the changing nature of Although most countries in the region have poverty in those countries. In China, for placed limited emphasis on investing in social example, the dibao approach reflects a grow- protection, several social assistance and ing concern that although rural incomes were social insurance schemes have recently been growing, transitory poverty was on the rise A P o l i c y A genda f or I n c l u s ive G rowth   89 and its geographical dispersion was increas- Strengthening social insurance ing (Golan, Sicular, and Umapathi 2014). As far as social insurance is concerned, pen- Other countries, such as Fiji, are consolidat- sion systems exist in all countries, cover- ing their programs to improve targeting and ing at least part of the formal sector. Their effectiveness. Technology is also facilitat- focus therefore tends to be mostly urban. ing the targeting and administration of such Faced with the prospects of an aging popula- programs, for example by allowing the cre- tion, persistently high levels of informal sec- ation of large national social registries in the tor workers, and the rarity of the long-term Philippines (box 5.5) and in Indonesia (the employer-employee bond that characterized Unified Database). Smart card technology is the early experience of the newly industri- also offering new opportunities for program alized countries, developing East Asia and delivery. Pacific countries have resorted to a plural- Despite these positive developments, sev- ity of social pensions (noncontributory and eral lower-income or lower-capacity countries old-age transfers). Such pensions take dif- still lack a fundamental social assistance pro- fering forms: several Pacific Island countries gram (Cambodia, Lao PDR, Myanmar, and and Timor-Leste have introduced universal most of the Pacific Island countries). Thailand schemes; Thailand has introduced a pensions- and Vietnam, which lack such programs tested scheme available to all those without despite their income levels, remain anomalies. BOX 5.5  National household targeting system in the Philippines (Listahanan) The Philippines is one of the few large developing coun- preliminary results of targeting became available tries in the world with a complete, nationwide database (and to build trust in the system by making it very of the poor that is used by a large number of social transparent). programs to target interventions. This remarkable Since then, the Listahanan has become the back- instrument, known as the Listahanan, was constructed bone of the Philippines’s antipoverty programs; it is over six years starting in 2010. The government has now used to target 25 national programs, including formally adopted it as a targeting database, meaning the Pantawid Pamilya , the PhilHealth subsidized that all nationally implemented programs, including health insurance, and the social pension program. It the country’s core social assistance programs, use it to is also used by policy makers to plan disaster relief identify potential beneficiaries, provided that they meet assistance, by civil society to monitor government program-specific eligibility criteria. The data collection effectiveness, and by academics. The Listahanan effort to build the Listahanan took place in 2009–10, was updated in 2015 to cover 75 percent of the when data were collected on about 11 million house- population (15.8 million households), and a new holds (almost 60 percent of the Filipino households). set of proxy means tested models was developed. Of those, 5.2 million were identified as poor and Those revisions significantly reduced errors in mis- were targeted for government assistance. The process classifying poor and nonpoor households. Those involved three stages: methodological developments were accompa- •  The geographic identification of areas to be sur- nied by a communication effort to increase public veyed, on the basis of earlier poverty estimates awareness and understanding of the Listahanan •  The survey itself, which yielded information to and to broaden and deepen support for the instru- develop proxy means tested models to identify the ment within the government and among other poor in the absence of income information stakeholders. •  A community validation process to allow com- munities to challenge the list of poor once the Source: World Bank (forthcoming b). 90   RIDING THE WAVE formal sector pensions; and Malaysia and the complexities of health care provision. Wide- Philippines have introduced more tightly tar- ranging reforms rapidly achieved near-univer- geted schemes. China has brought 350 million sal health insurance coverage, accompanied additional working-age people, rural and by massive investments in the health insur- urban, into pension schemes in recent years. ance infrastructure and the supply of health Although these schemes vary significantly by services starting in the early 2000s,12 but costs age threshold and generosity, many (particu- have escalated and health care gains have larly China’s) have a high reliance on public slowed after a period of rapid improvement.13 subsidies. One of the design innovations of Further, total household out-of-pocket health the China project is that the elderly automati- expenditures, and the incidence of impover- cally benefit from programs if their household ishing and catastrophic health expenditures, members are making contributions, thereby have not appreciably decreased with the providing strong incentives to enroll in this introduction of health insurance (Zhang and voluntary scheme. Liu 2014). Similar innovations aimed at increas- These encouraging trends show that the ing coverage are seen in relation to health elements of a comprehensive social protec- risks, despite varying degrees of success in tion system can be put in place over time. introducing subsidized schemes and keeping Countries with high shares of poor and costs under control. In many of the lower- vulnerable populations can give priority income countries, large segments of the to putting in place or strengthening social population lack health insurance coverage assistance programs. Countries with higher because it is conditional on formal employ- shares of the economically secure and mid- ment status.11 Thailand provides a success- dle class can focus on expanding the cover- ful example of expanding health insurance age of social insurance. Recent experience coverage by relying on general revenues (box shows that the key to expanding social 5.6). Other countries, such as China, face insurance is to delink entitlement to social greater challenges in this area, because of the insurance from employment and from users’ BOX 5.6  Guarding against health risks: Thailand’s success In 2001, the Thai government introduced the Uni- to 100 percent of the Thai population. The insur- versal Coverage Health Scheme (UCS) for those who ance improved health care: infant mortality rates were not funded under the employer insurance pro- fell to 11 per 1,000, eliminating the infant mortal- grams (roughly three-quarters of the population). ity gap between rich and poor provinces, and access Under the scheme, all Thai citizens were eligible for to medical care increased for the poor (more than health care that was low cost (in the years that a min- for the nonpoor). But the program also closed the imal copayment was charged) or free. The UCS was gap in out-of-pocket health care payments between funded out of general revenues, rather than through classes. Before the scheme, 33 percent of the total employer contributions. Several cost-containment health expenditure was paid by the patient or fam- measures were simultaneously implemented, includ- ily, with payments highest among those near the pov- ing a single purchaser model; incentives for provid- erty line. Under the scheme, the full cost was covered ers to contain costs; and a quality control system for (minus a small copayment in some years, but collec- providers. Although these efforts contained costs, tion was irregular). Impoverishment caused by health overall health costs to the state nearly doubled dur- care expenses decreased from 2.7 percent in 2000 to ing the 10 years of implementation. 0.5 percent by 2009. Within a year after the UCS was implemented, health insurance coverage increased from 25 percent Source: Hanvoravongchai 2013. A P o l i c y A genda f or I n c l u s ive G rowth   91 direct contributions in order to increase cov- their borrowing when trying to cope with erage of informal sector workers, the poor, the impact of the storm. and the economically vulnerable. To achieve •  Market insurance (applied to either crops that expansion, it will be critical to increase or livestock) can mitigate risks through general revenue financing, for instance, pooling and diversification. Traditional through publicly financed contribution sub- insurance products often suffer from high sidies. Expanding the system in a fiscally transaction costs and might prove unaf- sustainable manner will require increasing fordable for poorer groups.14 Across East fiscal space by expanding the revenue base, Asia and Pacific there is significant scope whether taxes are earmarked—as in Japan, to expand insurance take-up, 15 includ- which funds half its basic pension from sales ing by strengthening regulatory frame- tax—or not. works. Examples of innovation include public-private partnerships such as the Increasing resilience to large Mongolian Livestock Insurance Pool, aggregate shocks which covers more than 10,000 herders. But increasing coverage of poorer groups Social protection measures can help house- remains a challenge: in the Mongolian holds deal with idiosyncratic shocks, but large example, only about 15 percent of herders aggregate shocks such as natural disasters call in the relevant areas are covered. Index- for different types of measures. First, ex ante based insurance helps to minimize the risk-mitigation strategies can reduce house- limitations of traditional contracts, but it holds’ exposure to certain risks. Investments might still require large subsidies to ensure in infrastructure can minimize the impact of that sufficient numbers enroll. In indexed- natural disasters, for example, and invest- based insurance, policies are linked to the ments in human capital can help households amount of rainfall over a given time, for diversify their sources of livelihood. Such instance, or commodity prices at a given measures are most effective, however, in the date. Thus, indexing can mitigate weather- case of well-defined, sufficiently large-scale related and price risks in agricultural pro- sources of risk. duction and promote agricultural invest- In addition, ex post measures can help ment and productivity. Similarly, policies people cope with shocks that materialize. can use clearly defined, exogenous disaster International evidence suggests that a pack- triggers to activate ex post income sup- age of interventions would be most effective, port and eventually to end disbursements especially for poorer households (Hallegatte (Acevedo 2015). and others 2017). Those interventions would •  Adaptive safety nets “can protect poor include the following: households from climate and other shocks •  Fostering financial inclusion helps people before they occur through predictable build up savings beyond physical assets transfers, building community assets, and that might be affected by the shock, as other programs that help them build resil- well as facilitates access to credit that ience; and can be scaled up to respond to allows people to cope with the conse- extreme events when they hit” (Acosta quences of the shock. The experience of 2016, 12) Cyclone Nargis in Myanmar in 2008, •  Targeted safety nets , in general, pro- however, illustrates the potential pitfalls vide support to households by increasing of increasing debt as a coping strategy for their coverage when a large-scale disaster those at the bottom of the distribution. In strikes, even if they are not designed with areas where farming was badly affected, such disasters in mind. Such benefits can be up to 50 percent of farmers lost their col- designed to be adaptive, so programs can be lateral, particularly land, after increasing scaled up when assistance is most needed. 92   RIDING THE WAVE BOX 5.7  (Almost) Ready for Winston On February 20, 2016, Tropical Cyclone Winston, 37 percent lost their entire dwelling, 74 percent sus- the most powerful storm on record in the Southern tained damage to their roofing, and 49 percent lost Hemisphere, cut a path of devastation across Fiji. their crops or harvest. Within four weeks, the majority The storm killed more than 40 people and caused of households had spent their entire additional social widespread destruction, which was particularly assistance, with 99 percent of expenditure on essential severe for the housing and agriculture sectors. Some items. Food and materials to repair damaged dwellings of the worst-hit areas were the largely rural provinces formed the two most important categories of expen- and outlying islands that had a high concentration of diture, followed by clothing and school supplies. Less households living below the poverty line. Total dam- than 1 percent of the assistance was spent on kava, age and losses were estimated at 22 percent of gross alcohol, or cigarettes, addressing a common con- domestic product (GDP) (US$957 million). cern that additional social assistance would be used In the aftermath, the government of Fiji drew on its unwisely for nonessential items. existing social program registries to provide cash trans- Three months after the cyclone, beneficiaries fers to affected citizens. Four weeks after Winston, under the Poverty Benefits Scheme were more likely the government transferred US$9.4 million to 43,524 to have recovered from the shocks they faced than beneficiaries covered under the three core social pro- comparable households that did not receive the addi- tection programs (the Poverty Benefits Scheme, the tional assistance. The recovery includes having recov- Care and Protection Scheme, and the Social Pension ered from sickness or injury; repaired their dwelling; Scheme). The benefit levels were set to the equivalent replenished their food stocks; remedied the damage of approximately three months of ordinary benefits, to to their agricultural land; repaired village or neigh- enable households to meet their immediate food needs borhood infrastructure; and resolved problems of without resorting to negative coping strategies like conflict, violence, or insecurity. Following these posi- selling productive assets. tive results, the government is expected to consider The initial results from a comprehensive impact ways to build a broader social registry that has the evaluation suggest that both poor and near-poor house- ability to target not just existing beneficiaries, but holds did suffer severe losses in the wake of Winston also the near poor and others excluded from current and that adaptive safety nets were effective in assist- programs. ing the poor to cope. Of all surveyed households in affected areas that received additional cash assistance, Sources: Doyle and others 2016; Erdman 2016; World Bank 2011. As the example of the relief effort in Fiji risk pools, and bonds transferring part of shows (box 5.7), having invested in creat- the risk to global capital markets or rein- ing social registries for existing targeted surance (Hallegatte and others 2017). programs meant that when disaster struck •  Early warning systems may also prove in 2016 following Cyclone Winston, the useful in reducing exposure to risks information base for delivering programs and enabling timely ex post assistance. to some of the most vulnerable groups was Providing information about possibly dis- readily available. ruptive weather events such as cyclones •  Disaster risk financing is also required to can enable households to take protective support the scaling up of social assistance measures, ranging from moving assets to and other reconstruction needs. Several higher or safer ground to evacuating. At measures offer protection against the fiscal the policy level, an early warning system implications of possible shocks, including that considers a broader set of shocks reserves, contingency credit lines, regional would rely on a predefined and frequently A P o l i c y A genda f or I n c l u s ive G rowth   93 monitored set of indicators across several capacity limitations on the administration domains (for example, macroeconomic, side limit the extent to which the system can weather-related, firm-sector, and house- be made more progressive (Park 2012). Given hold behaviors). Such indicators would these challenges, strengthening personal signal an impending crisis and predict who income taxation, increasing reliance on prop- would be affected and how they would erty taxes, and improving tax administration be affected. Trigger levels that categorize are priorities to mobilize additional revenues shock-related situations as normal, alarm- and increase tax fairness. ing, or crisis would be matched to national To strengthen personal income taxation, or local policies to manage the response governments may have to broaden the tax and alleviate the impacts (for example, base, review marginal tax rates, and eliminate evacuation plans). Indonesia, for example, differential treatment of labor income versus is developing a Crisis Monitoring and capital income (after allowing for corporate Response System, retrofitted to the data taxes).17 availability and country-specific behaviors Broadening the tax base is an option to that signal an impending economic crisis. consider in the many East Asia and Pacific countries characterized by high basic exemp- tion limits or high threshold levels at which Strengthening the institutions the top income rate is applied. In Fiji, income for inclusive growth tax is applied at an exemption threshold of nearly 1.7 times per capita income, and in Increasing the redistributive capacity Myanmar and Papua New Guinea the exemp- of the state tion is up to 1.2 times per capita income. Such To remove the constraints to inclusive growth high exemption thresholds clearly limit the experienced by different groups in the popu- coverage of the tax system. For example, in lation, governments need to implement a China, according to household survey data, tailored package of policies. Financing such 80 percent of workers are not liable to pay packages will prove expensive, so it will be personal income taxes (Lam and Wingender vital not only to mobilize more resources in a 2015). Similarly, the very high levels at which more progressive way (which in itself would top tax rates start to apply constrain revenue contribute to making growth more inclu- mobilization potential. Papua New Guinea sive), but also to adopt tools that can target stands out, with an applicability at 45 times resources more effectively. per capita income, whereas in Organisation for Economic Co-operation and Development Mobilizing resources through (OECD) countries, the highest threshold is 8 progressive taxation times per capita income. In China, the highest Developing East Asia and Pacific’s tax sys- marginal tax rate applies at about 35 times tems mobilize few resources. In 2010, the the national average wage, compared with region raised only around 15 percent of gross four times the average in OECD countries (de domestic product (GDP) in tax revenues, com- Mooij, Lam, and Wingender 2017). pared with 20 percent in Latin America and Reviewing marginal tax rates, especially 35 percent in advanced economies (Bastagli, for the top income bracket, is an option Coady, and Gupta 2015).16 Besides generally that raises concerns because of the expected low tax revenues, reliance on indirect taxes trade-off between equity and potential effi- is high, intake from personal income tax and ciency costs, for example, a reduction in property tax is limited, and tax administra- labor supply following an increase in the tion is poor. A reliance on indirect taxation tax rate. At an average of 32 percent, how- and low (effective) rates on higher incomes ever, top rates in the region are lower than in are making the system more regressive, and OECD countries. A wide range of estimates 94   RIDING THE WAVE exist for the magnitude of efficiency costs in Another option to mobilize resources OECD countries, commonly expressed as through progressive taxation is provided by the elasticity of taxable income.18 Estimates increasing reliance on property taxes. Such are even more limited in East Asia, so the an option, however, might be feasible only in optimal top marginal tax rate is difficult the medium term, because it requires build- to assess. A higher reform priority appears ing an effective administrative infrastructure. to be the level at which the top tax rate is But property taxes could significantly support applied, rather than the top tax rate itself. revenue mobilization, with expected progres- Simulations for China suggest that shorten- sive results and little inefficiency (unlike other ing personal income tax brackets could raise forms of wealth, property is both visible and revenues of around 2 percent of GDP (Lam immobile). In East Asia and Pacific as in other and Wingender 2015).19 Marginal tax rates developing regions, however, recurrent prop- for low-income workers are also an impor- erty taxes on immovable property are under- tant concern. The system of social security used, mobilizing a lower share of GDP than contributions in China creates a very high in OECD countries. Property tax revenues tax burden for the poorest workers because it are low in Indonesia at 0.35 percent of GDP, includes a minimum employee contribution. Mongolia at 0.15 percent, and Thailand at Eliminating the differential treatment of 0.20 percent (Norregaard 2015). By con- capital income versus personal income is a trast, in China, property taxes account for third option to increase the revenue base. 1.6 percent of GDP, which is comparable Capital incomes, including capital gains, are with the levels observed in OECD countries typically taxed at a lower rate than ordinary (de Mooij, Lam, and Wingender 2017).21 Best income, which has been identified as one practice suggests that existing property tax reason for the limited redistributive effect of regimes in the region could be shifted from taxation in developing countries (Bastagli, a focus on transaction-based taxation (which Coady, and Gupta 2015). Although a con- reduces spatial mobility) to recurrent (annual) solidated tax with a single tax base (includ- taxes on immovable property. That regime ing both capital and labor income) is often would provide a more stable income stream considered at the theoretical ideal, most (especially for local governments) and avoid advanced economies have moved toward a opportunities for collusion between buyer dual income tax, with capital incomes taxed and seller in understating the transaction separately (de Mooij, Lam, and Wingender (Norregaard 2015). Property taxes should 2017).20 Similarly, in Indonesia, dividends also adopt a progressive rate structure with are taxed at 10 percent and interest at special arrangements for poor, large families 20 percent, which is substantially lower than and asset-rich, income-poor households.22 the 30 percent top marginal tax rate that Finally, strengthening tax administration many dividend earners face for their salary can have substantial payoffs for efficiency incomes. Differences in treatment make it and equity, and it should be set as a prior- possible for business owners to move income ity (Bastagli, Coady, and Gupta 2015). 23 into categories taxed at lower rates. That Concentrating enforcement efforts on large tax treatment does not recognize the real- taxpayers—for example, by establishing ity of compensation packages such as stock large-taxpayer offices—is likely to result options or equity stakes, which blur the line in the biggest payoffs for both revenue and between capital and labor incomes. In China, equity. Improving compliance, including after accounting for taxes on corporate prof- through increased withholding of income its, a 15 percent tax on dividends would taxes on capital and use of third-party report- be broadly consistent with a 35 percent ing, could remedy one of the most striking top marginal tax rate on labor income (de features of tax systems in the region, namely, Mooij, Lam, and Wingender 2017). the extremely low share of the population A P o l i c y A genda f or I n c l u s ive G rowth   95 that pays taxes (as proxied by the share power. Regulatory agencies could play a vital of registered taxpayers). In a sample of role in ensuring that East Asia’s large firms eight countries, an estimated 14 percent of do not end up stifling competition. In par- households are registered taxpayers, blunting ticular, family firms often consolidate their the effectiveness of income taxes in mobiliz- holdings through pyramidal holding struc- ing and redistributing resources. Although tures, which guarantee a disproportionate many countries withhold taxes for capital share of votes relative to the owners’ capital incomes (Thailand is a notable exception), (Kandel and others 2015). Another example which significantly improves compliance, is the use of dual-class shares, which sepa- gaps remain with capital gains. That result rate control rights and cash flow rights and contrasts with labor incomes, which are which have recently become popular with largely subject to withholding and thus face tech companies, including some in East a higher effective tax burden. In Indonesia, Asia and Pacific. Although many of these capital gains (for example, on stocks or real practices are legal, they stifle competition. estate) are subject to the standard personal Countries such as Israel and the United States income tax, but there is no withholding, so have fought these pyramidal structures; compliance is limited (World Bank 2016b). Israel, for example, has banned companies As a result, only 5 percent of personal income with more than two levels of subsidiaries tax revenues in Indonesia are from capital (Bebchuk 2012). Measures are also needed incomes, with the rest coming from withhold- to strengthen asset disclosure and corporate ings on salaries. Finally, because compliance transparency, and to reduce the opacity of is weaker for taxes that are bureaucratically ownership structures, which are particularly more cumbersome, are harder to understand, prevalent in family-controlled firms. or are easier for taxpayers to bribe their way out of (Usui 2011), efforts to simplify the tax Coping with aging system and increase transparency are also likely to pay off. Populations in developing East Asia and Pacific are aging more rapidly than any region in history, a trend driven by sharp Leveling the playing field for different declines in fertility and steady increases in economic actors to compete fairly life expectancy. In many countries, aging is Perceptions of growing inequalities and occurring at low-income levels—people are unequal opportunities, along with high or “getting old before getting rich” (World Bank growing concentrations of income, wealth, 2016c, 9). Significant differences can be seen and economic power, are all signals that across countries. Populations in some middle- governments could do more to allow fairer income countries (including China, Indonesia, competition among different actors. Priorities Malaysia, Mongolia, Thailand, and Vietnam), in this area include boosting competition, which are already relatively old, are aging rap- reducing opportunities for revenue capture idly. In poorer countries (including Cambodia, in the more regulated sectors of the economy, Lao PDR, Myanmar, Papua New Guinea, tackling nepotism and corruption, and more and the Pacific Island countries), populations generally strengthening and enforcing the are relatively young and still have decades of rule of law. potential demographic dividends to realize. Given the historical role that large com- In all countries, however, aging raises policy panies, especially family-owned companies, challenges and poses economic and social have played in the development of the region, risks, with the speed of change heightening policies need to increase scrutiny of their role the urgency of policy action. and of the potential for mutually reinforc- When the labor forces of a country age, ing links with the concentration of political economic growth tends to slow. More 96   RIDING THE WAVE immediately, countries face the fiscal chal- are already following this path, such as lenges of providing sustainable health, pen- Korea and Singapore, which have dedi- sion, and long-term care systems. Relatedly, cated programs for training older workers. poverty rates are higher among the elderly, However, use of these programs remains not only because health and pension insur- limited. ance is insufficient, but also because tradi- •  Adaptation of the health sector . The tional family- and community-based support health challenges of aging are seen in networks are being undermined by declining the increased incidence of noncommu- fertility. Rapid urbanization and migration nicable diseases such as cardiovascular are also leading to the dispersion of extended diseases, cancer, chronic respiratory dis- families.24 That said, the region starts from eases, and diabetes, which reinforces the a favorable position. People already have need to transform health delivery sys- long working lives. Entitlements to pen- tems. Priorities include strengthening sions, health care, and long-term care are still primary care, reducing overprovision of modest, so this is an opportunity to put in services, improving coordination among place systems now that can be sustainable in providers, and instituting transparent and the future. Household savings in the region ­ evidence-based processes for prioritizing are already high, and people tend to save new technologies and drugs. The aim is until later in life. Finally, people in devel- to manage the treatment of older patients oping East Asia and Pacific have in recent with chronic conditions affordably and at decades seen a steady increase in the number the right levels of the system. Such reforms of years lived in good health. would be important without aging, but the In addition to putting in place compre- aging population makes them even more hensive social protection systems, older or important. Aging is also creating a rap- rapidly aging countries need to implement idly growing demand for and market for several complementary policy reforms to elder- and long-term care, as traditional adapt to changing demographics (World family networks become increasingly Bank 2016c): stretched. Innovation is urgently needed here, including with respect to home- and •  Fertility and childcare policies. Although community-based care (“aging in place”), several countries, including Mongolia and which can provide more positive experi- Vietnam, have introduced direct monetary ences for older people at a lower cost than incentives to support fertility, international residential care. Although the state may evidence suggests that those measures are provide financing for elder- and long-term not very effective or cost-effective. Other care (with copayments for all but the poor measures have more potential to change and the disabled), the private sector will the trade-offs between women’s fertility likely play an important role in providing and labor market decisions, such as longer such care. maternity and paternity leave, which in the •  Extension of productive working lives . region are typically short and available only Many policies extend productive working to the minority of workers in the formal lives by eliminating incentives for older sector. Subsidized childcare has the same workers in the formal sector to retire potential, whether delivered as subsidies to too early or by simplifying the return to households, as in the Republic of Korea, or work for women who have had children. to employers, as in Malaysia. Longer productive lives can also be sup- •  Upgrading of education and skills systems. ported by specific reforms—for example, Countries need to support lifelong learn- by targeting labor market institutions ing to upgrade workers’ skills throughout that discourage employment of older the life cycle. Some countries in the region workers, such as seniority wage systems, A P o l i c y A genda f or I n c l u s ive G rowth   97 and by using incentives to adjust work- the institutional foundations for efficient places to accommodate older workers urban expansion. Such foresight will enable (for instance, through greater use of flex- countries to take advantage of urbaniza- ible and part-time work arrangements). tion’s potential to boost productivity, reduce •  Opening up of aging labor markets to service delivery costs (through economies of greater inflows of young immigrants . density in serving more compact locations), Opening labor markets will also help stave and, more generally, deliver inclusive growth off countries’ declining working-age popu- (Baker and Gadgil 2017). Urbanization inter- lations. Singapore and Hong Kong SAR, sects with a variety of institutional arrange- China, offer examples of the economic ments at the local level, from centralized to benefits of more open immigration benefits. decentralized administrations, often varying by sector as well as by country. Ensuring that Countries with younger populations, this evolving landscape becomes an effec- meanwhile, need to prepare now for rapid tive environment for the delivery of key ser- aging in the future. Such preparation requires vices and contributes to inclusive growth will setting up pension and health systems that require a set of coordinated interventions can achieve significant coverage and are also along two dimensions. sustainable. Countries also can maximize First, it is vital to support urban develop- the demographic dividend from their still- ment and efficient land use along the entire expanding labor forces through investments rural-urban spectrum. The focus of spatial that raise productivity and maximize youth policies should not be to create spatial equal- employment. ity; rather, the focus should be on promoting growth and long-run convergence in living standards, building on the specific charac- Institutions for efficient urbanization teristics and economic strengths of different The structural transformation in East Asia environments. Cities along the rural-urban and Pacific is changing the distribution of the spectrum can play different roles and func- population along the rural-urban spectrum. In tions. Primary cities can lead in innovation countries with low urbanization rates, urban and technology. Secondary cities can provide expansion is still dominated by the largest the manufacturing backbone of the country. cities; in more urban countries, urban expan- And smaller cities can act as the interface sion is happening rapidly in cities of all sizes. between urban and rural areas. Spatially In China and Vietnam, however, inefficient aware policies can promote such development constraints on land allocation are suppress- patterns by doing the following: ing potential productivity gains from urban agglomeration, both in large metropolitan •  Encouraging efficient urban agglom- areas and in secondary cities. Further, impres- eration, which requires developing fluid, sive reductions in monetary poverty have transparent land markets, including clari- not always been matched by improvements fying property rights, strengthening title in urban infrastructure, including access to registration, and creating institutions that water, sanitation, and good-quality housing. will allow for transparent, market-based That lagging investment harms welfare. It can land valuation. It also requires effective exclude the poorest, especially new rural-to- urban planning, with master plans linked urban migrants, and it can lead to high levels to the budget process. of nonmonetary poverty among those who •  Providing incentives to cluster service and have reached economic security, and to a knowledge-based sectors in large cities lesser extent among the middle class. and relocating land-intensive industry to For countries where most of the urbaniza- smaller cities and towns. For such a spa- tion has yet to happen, it is essential to lay tial reallocation of economic activities to 98   RIDING THE WAVE occur, red tape needs to be reduced and (box 5.8). Many East Asian countries have changes in land-use allocations need to be strongly centralized systems for service deliv- facilitated. ery. However, the delivery of core urban •  Investing in connections between small services (including public transport, road, cities/market towns and the rural areas water supply and sewerage, solid waste man- they serve both as markets and as centers agement, and drainage) is often better man- for the delivery of services. aged at the local level. Cities can be given more responsibility for service delivery, pro- Within cities, spatially aware policies vided that local authorities have the neces- can foster efficient land use through transit-­ sary resources, including technical capacity. oriented development (TOD) or smart In addition, land-use policies must reserve growth that prioritizes new developments rights of way, so that basic services and con- along established public transport routes. nective infrastructure can be extended as cit- Such policies would also favor more inclu- ies expand. Finally, policies must consider sive urbanization by ensuring a better spatial the needs of and constraints faced by rural- match between jobs, public transport, afford- to-urban migrants to prevent new forms of able housing, service provision, and urban exclusion as they continue to be the engine of amenities such as recreational areas. In addi- urbanization. tion, efforts to incorporate risk reduction into In creating more efficient institutions land-use and infrastructure planning could for urbanization, metropolitan areas face a boost growth and ultimately save lives. conundrum. Their governance structures are A second priority is strengthening urban often fragmented, which hampers integrated governance to improve the delivery of ser- urban planning and investment and results vices and provision of urban infrastructure in a failure to realize economies of scale or BOX 5.8  Institutions for effective urban governance: lessons from Malaysia Many problems related to urban planning, develop- accountability. Such services could include intraur- ment, and service delivery may at first glance appear ban highways and federal intraurban roads, urban to be technical or financial. As a recent study in public transport including buses and rail, drainage Malaysia found, however, the underlying causes tend and flood mitigation, solid waste management and to be institutional, and they can significantly ham- disposal, and emergency services. This priority also per cities’ competitiveness. Among the key challenges involves strengthening metropolitan governance for identified were the centralization or federalization of some functions and services, particularly planning, urban service delivery, constraints in urban and spa- and investing in city performance indicators, which tial planning, and financial and technical constraints are essential for tracking improvements in urban ser- facing local authorities. In the case of Malaysia, the vice delivery. following are recommended priorities for strengthen- The second priority is to increase the capacities ing institutions for effective urban governance. of local authorities in specific financial and techni- The first priority is to deliver selected urban ser- cal areas and ensure adequate local agency staffing. vices by shifting more management and decision Areas such as property assessments are weak and making to the local level and to enhance the system would benefit from updated systems. The system of of local performance indicators. Creating indicators fiscal transfers also needs to be revised to be more involves identifying services to be delivered locally transparent, predictable, and formula based. using criteria such as economies of scale, external- ity effects, equity, and local responsiveness and Source: World Bank 2015a. A P o l i c y A genda f or I n c l u s ive G rowth   99 agglomeration. Those structures also lead investments in basic services and infrastruc- to unequal coverage and quality of services ture. The central government provided both and infrastructure. Such inefficiencies and earmarked and non-earmarked transfers. The inequalities are likely to be deepened by fur- earmarked transfers were allocated mostly to ther decentralization in service delivery. The investments in basic services (including rural inequalities may particularly affect vulner- roads) and social institutions (such as pub- able groups such as internal migrants, who lic utilities, medical facilities, and schools) may face limitations in choosing to settle in a and used cost-sharing agreements with local particular jurisdiction. Regional government governments. authorities and other mechanisms (voluntary Greater coordination is needed between or mandated by law) can help coordinate ser- different levels of government and different vice provision across administrative bound- jurisdictions, especially in the case of mega- aries and improve the efficiency and equity cities. Uncoordinated actions by jurisdic- of service delivery, while continuing to be tions in a larger metropolitan area can lead responsive to local needs. to suboptimal investment in public goods, Countries should develop effective sys- depletion of local resources (the “tragedy of tems of property tax assessment and collec- the ­commons”), and a race to the bottom in tion as quickly as possible. Such systems, taxes, hindering local finances. In large cities, which are mostly lacking, can help mobilize intensive development and redevelopment of the resources required to pay for local ser- intracity areas can also help reduce the need vices, foster accountability for the delivery of for further expansion and the resulting exten- such services, and encourage the development sion of service networks. of land. Property taxation reforms are slow, Throughout this quest to reform institu- incremental processes. Their technical cred- tions for more efficient urbanization, govern- ibility and political feasibility hinges on good ments must adopt people-centered policies. land records and a valuation infrastructure This approach requires being aware of how (including standards, professional monitor- individuals use services, their expectations ing, access to market data, and automated for access and quality, and whether they fully valuation systems). To broaden the use of appreciate the economic and health benefits property taxation, therefore, governments of services (such as sanitation). An example must implement measures aimed at strength- of failing to incorporate the user perspective ening investments in urban areas, includ- is that major roads are often designed in ways ing improvements in land registration and that do not respect (or they even disrupt) the increased transparency in the public informa- nature of the settlements they cross, with del- tion on parcel definition and land values. eterious consequences for road safety. Fiscal transfers are likely to remain the key source of financing for urban administra- tions over the medium term. Such transfers Notes should be more transparent and predictable, 1. Ger is the name of Mongolian traditional and ideally should be based on a formula that tents, which in urban areas are used in at least partly reflects local needs. Incentive peripheral and underserved areas (that is, payments could also be introduced, based slums). on solid performance indicators for the local 2. See World Bank 2012, 2015b chapter 2.B, “Realizing the Economic Potential of Women,” level. The 1970 New Economic and Social for an analysis of these issues in the case of Development Plan in Japan is an example Malaysia. of equitable funding for public services. It 3. O n e - q u a r t e r o f m o d e r a t e p o o r a n d has been widely credited with first reducing economically vulnerable women do not have disparities in living standards and then con- access to potable water, and depending on tributing to convergence in incomes between the country, these shares rise significantly for leading and lagging areas in Japan, through the extreme poor. 100   RIDING THE WAVE 4. In Lao PDR, 21 percent of rural children initiated in 2010 and subsequently expanded, have never been to school (as of 2010), with a rapid increase in debit-card issuance. and the rural-urban gap in attainment of In Vietnam, account penetration rose from lower secondary education increased from 21 percent in 2011 to 31 percent in 2014, 29 percentage points to 42 percentage points and the share of households that save with between 2000 and 2011. In Cambodia, a formal financial institution rose from Lao PDR, and the Solomon Islands, less than 8 percent to 15 percent. 70 percent of children who enter primary 8. For instance, significant rural-urban school actually complete it. In the Philippines, disparities persist with respect to the number among adolescents who reached the end of of bank branches, ATMs, and point-of-sale primary school, 94 percent of those from the terminals. In the Philippines, 36 percent of richest households, but only 69 percent of municipalities do not have a banking office, those from the poorest households, continued and 12 percent lack access to all financial to lower secondary school (as of 2008). In service providers (including pawnshops, Vietnam, ethnic minority students drop out in remittance agents, money changers, and higher numbers at earlier ages than students mobile banking agents). These mainly remote from the Kinh majority, and are 10 times less provinces have a high concentration of poor likely to attend university. In Timor-Leste, and marginalized farmers and fishermen and the net secondary school enrollment rate are often vulnerable to natural and other was only 38 percent (as of 2012). In China, disasters. the gross tertiary enrollment rate reached 9. For instance, 68 percent of adults in Lao PDR 40 percent in 2015, but among rural and thought they needed more information about ethnic minority students, the rate is far lower managing money (Finmark Trust 2015). and has increased slowly. 10. In Indonesia, 48 percent of firms consider The uneven quality of education is access to finance the top business environment noticeable in achievement disparities, which constraint. Only about one-half of small and start in the early grades. For example, early- medium enterprises (SMEs) hold savings or grade reading assessments indicate that in checking accounts, and only 17 percent have Cambodia and Lao PDR, 13 percent of a bank loan or line of credit (Indonesia— students in grade 3 were unable to read a Enterprise Survey 2015, World Bank single word. See World Bank (forthcoming a). database, http://microdata.worldbank.org​ 5. As shown by the Vanuatu Literacy Education /index.php/catalog/2665). In China, the Programme’s Effective Literacy and Numeracy majority of SMEs consider access to finance Practices Database (UNESCO 2009). to be the biggest obstacle to their operations 6. In 2015, 75,000 university students were and growth. enrolled through these special plans. 11. Health insurance covers 85 to 95 percent of Also, to address regional inequalities, the the population in China, Mongolia, and the Collaborative Admission Plan for Supporting Philippines and 40 to 60 percent in Indonesia the Central and Western Regions arranges and Vietnam. Coverage is much lower in dedicated college matriculation quotas for poorer countries such as Cambodia, Lao universities in the eastern region to admit PDR, and Papua New Guinea, although students from provinces of the central and efforts have been made to remove financial western regions. This plan admitted 900,000 barriers for targeted subgroups, such as the students during 2011–15, and the gap in poor, and specific services, such as maternal college admission rates between the national and child health. average and the worst-performing provinces 12. Social health insurance was first introduced fell from 15 percentage points in 2010 to to cover formal sector workers in 1998. It under 5 percentage points in 2015. expanded to the rural population in 2003 and 7. In China, account penetration among adults to urban informal sector workers, the poor, rose from 64 percent to 79 percent between children, and the elderly in 2007. As a result, 2011 and 2014, as 188 million people became the social health insurance system covered first-time bank account holders. Bankcard- more than 95 percent of citizens in both based rural cash withdrawal services were urban and rural areas by 2011. The depth A P o l i c y A genda f or I n c l u s ive G rowth   101 of insurance coverage has expanded 21. However, in China, property taxes are more gradually. primarily transaction based, whereas 13. For instance, perverse provider incentives recurrent taxes are more important in other create a strong bias toward excessive countries. treatment, including medically unnecessary 22. For example, asset-rich, income-poor high-technology services, irrational drug households include pensioners who live in prescriptions (including, in particular, large houses, but on relatively low incomes excessive use of antibiotics), and lengthy (pension). They could defer their tax liability hospital stays (see China Joint Study until they die or sell their house—effectively Partnership 2016). this amounts to the tax being paid in the form 14. In China, the penetration of agricultural of an equity stake in the house. insurance remains extremely low despite 23. Note also that some East Asia and Pacific significant fiscal subsidies, thus reflecting countries have quite low administration costs. the high transaction costs associated with the For example, Indonesia has the seventh- fragmented smallholder farming system. lowest enforcement costs per unit of revenue, 15. In China, only 7 percent of farming slightly above Norway and below the United households and 6 percent of those in the States (Claus, Martinez-Vazquez, and Vulovic poorest quintile hold any type of non–life 2014). insurance product. In Lao PDR, 77 percent 24. For instance, see Cai and others (2012) for an of adults do not hold any kind of insurance analysis of poverty among the elderly in rural product. China. 16. Across the region, tax revenues ranged from a low of 5 percent in the Federated States of Micronesia to 27 percent in Timor-Leste in References 2012. Acevedo, M. C. 2015. “The Effectiveness of 17. A recent reform proposal for China includes Ex-Ante Risk Management Strategies in Latin a comprehensive tax on capital income with America and the Caribbean.” Unpublished a rate similar to the top marginal tax rate on paper, World Bank, Washington, DC. labor income after allowing for corporate Acosta, P. 2016. What Makes Social Protection taxes (de Mooij, Lam, and Wingender 2017). Systems Adaptive? https://unfccc.int/files​ 18. Cross-country evidence from OECD countries /­a daptation/application/pdf/presentation_3​ shows that cutting top rates did not result _pablo_acosta_-_ftc_standing_committee.pdf. in higher growth, which suggests that the Ashraf, N., D. Karlan, and W. Yin. 2006. elasticity of labor supply in relation to income “Tying Odysseus to the Mast: Evidence tax cannot be very high (Piketty, Saez, and from a Commitment Savings Product in the Stantcheva 2014). Replicating this analysis Philippines.” Quarterly Journal of Economics for East Asia and Pacific also finds no 121 (2): 673–97. relationship between growth and changes in ———. 2010. “Female Empowerment: Impact top marginal income tax rates, although this of a Commitment Savings Product in the result is difficult to interpret, because under Philippines.” World Development 38 (3): the current schedules, the top tax rate applies 333–44. to a very small number of taxpayers. Baker, J., and G. Gadgil, eds. 2017. East Asia and 19. In the simulated schedule, more than half Pacific Cities: Expanding Opportunities for the of workers would face a marginal tax rate Urban Poor. Washington, DC: World Bank. of less than 5 percent, with three-quarters of Bastagli, F., D. Coady, and S. Gupta. 2015. workers paying an average tax rate of less “Fiscal Redistribution in Developing than 20 percent. Under the new brackets, Countries: Overview of Policy Issues and less than 10 percent of taxpayers face the top Options.” In  Inequality and Fiscal Policy , marginal tax rate of 45 percent, compared edited by B. Clements, R. de Mooij, S. Gupta, with 0.1 percent of taxpayers at the moment. and M. Keen, 57–76. Washington, DC: 20. It has also been argued that a dual income tax International Monetary Fund. might be easier to administer in low-capacity Bebchuk, L. 2012. Corporate Pyramids in the settings (Bird and Zolt 2011). Israeli Economy: Problems and Policies . 102   RIDING THE WAVE Report prepared for the Committee on in Soft Infrastructure, edited by W. R. Lam, Increasing Competitiveness in the Economy, M. Rodlauer, and A Schipke. Washington, DC: May 2012. http://mof.gov.il/Committees​ International Monetary Fund. /­C ompetitivenessCommittee/FinalReport​ Doyle, J., O. Ivaschenko, J. Kim, and J. Sibley. _ExpertOpinion2.pdf. 2016. Does Manna from Heaven Help? The Belfield, C. R., M. Nores, S. Barnett, and Role of Cash Transfers in Disaster Recovery: L. Schweinhart. 2006. “The High/Scope Perry Lessons from Tropical Cyclone Winston in Fiji. Preschool Program Cost–Benefit Analysis Washington, DC: World Bank. Using Data from the Age-40 Followup.” Erdman, J. 2016. “Tropical Cyclone Winston Journal of Human Resources 41 (1): 162–90. Makes Category 5 Landfall; Strongest on Bird, R., and E. Zolt. 2011. “Dual Income Record in Fiji.” Weather Channel. February 22, Taxation: A Promising Path to Tax Reform for 2016. https://weather.com/storms/hurricane​ Developing Countries.” World Development /­news​/­tropical-cyclone-winston-fiji-strongest​ 39 (10): 1691–703. -landfall. Brinkman, S., A. Hasan, H. Jung, A. Kinnell, and Finmark Trust. 2015. “FinScope Consumer M. P. Pradhan. 2015. “The Impact of Survey 2014, Lao PDR.” Latest News Expanding Access to Early Childhood Services (blog), July 13. https://www.finmark.org.za​ in Rural Indonesia: Evidence from Two Cohorts /­finscope-consumer-survey_laos-2014/. of Children.” Policy Research Working Paper Golan, J., T. Sicular, and N. Umapathi. 2014. Any 7372, World Bank, Washington, DC. Guarantees? China’s Rural Minimum Living Cai, F., J. Files, P. O’Keefe, and D. Wang. 2012. Standard Guarantee Program. Washington, The Elderly and Old Age Support in Rural DC: World Bank. China: Challenges and Prospects. Washington, Goldin, C. 2006. “The Quiet Revolution DC: World Bank. That Transformed Women’s Employment, China Joint Study Partnership. 2016. Deepening Education, and Family.” American Economic Health Reform in China: Building High- Review: Papers and Proceedings 96 (2): Quality and Value-Based Service Delivery. 1–21. Policy summary. http://documents.worldbank. Hallegatte, S., A. Vogt-Schilb, M. Bangalore, and o r g / c u r a t e d / e n / 8 0 0 9 1 1 4 6 9 1 5 9 4 3 3 3 0 7​ J. Rozenberg. 2017. Unbreakable: Building the /­Deepening-health-reform-in-China-building​ Resilience of the Poor in the Face of Natural -high-quality-and-value-based-service-delivery​ Disasters. Washington, DC: World Bank. doi: -policy-summary. Washington, DC: World doi:10.1596/978-1-4648-1003-9. Bank Group. Hanvoravongchai, P. 2013. “Health Financing Claus, I., J. Martinez-Vazquez, and V. Vulovic. Reform in Thailand: Toward Universal 2014. “Government Fiscal Policies and Coverage under Fiscal Constraints.” Universal Redistribution in Asian Countries.” In Health Coverage Studies Series 20. World Inequality in Asia and the Pacific: Trends, Bank: Washington, DC. Drivers, and Policy Implications, edited by Hasan, A., M. Hyson, and M. C. Chang, C. Rhee, J. Zhuang, and R. Kanbur. Manila: eds. 2013. Early Childhood Education Asian Development Bank. and Development in Poor Villages of de los Reyes, V. R., R. Librojo, and J. L. Capacio. Indonesia: Strong Foundations, Later 2016. Offering Guideposts for Reform Efforts: Success. Washington, DC: World Bank. http:// Lessons from the Agrarian Reform Program www-wds.worldbank​ . org/external/default​ in the Philippines. Manila: Department of /­WDSContentServer​/­WDSP/IB/2013/06/17/00 Agrarian Reform. 0442464_20130617121704/Rendered/PDF/7 Demirguc-Kunt, A., L. Klapper, D. Singer, and 84840PUB0EPI0000PUBDATE06011020130​ P. Van Oudheusden. 2015. The Global .pdf. Findex Database 2014: Measuring Financial Haque, T. and T. Packard. 2014. “Well-being from Inclusion around the World. Washington, DC: Work in the Pacific Island Countries.” World World Bank. Bank East Asia and Pacific Regional Report. de Mooij, R., W. R. Lam, and P. Wingender. Washington, DC: World Bank. 2 0 1 7 . “ M o d e r n i z i n g t h e Ta x P o l i c y Heckman, J. J., S. H. Moon, R. Pinto, P. A. Savelyev, Regime.” In Modernizing China: Investing and A. Yavitz. 2010. “The Rate of Return to A P o l i c y A genda f or I n c l u s ive G rowth   103 the High/Scope Perry Preschool Program.” Tale of Three Elasticities.” American Economic Journal of Public Economics 94 (1): 114–28. Journal: Economic Policy 6 (1): 230–71. Hou, J., and X. Huo. 2016. “Property Rights Tan, S., N. Heerink, and F. Qu. 2006. “Land Insecurity, Land Transaction Restrictions, Fragmention and Its Driving Forces in China.” and Welfare Loss: Empirical Evidence from Land Use Policy 23: 272–85. Specialized Farmers in China.” Paper prepared UNESCO (United Nations Educational, Scientific, for the 17th Annual World Bank Conference and Cultural Organization). 2009. Vanuatu on Land and Poverty, March 14–18, World Literacy Education Programme’s Effective Bank, Washington, DC. Literacy and Numeracy Practices Database, Ibarraran, P., and D. Rosas. 2009. “Evaluating accessed June 27, 2017. http://www.unesco​ the Impact of Job Training Programmes in .org/uil/litbase/?menu=8&programme=34. Latin America: Evidence from IDB Funded Usui, N. 2011. “Tax Reforms toward Fiscal Operations.” Journal of Development Consolidation: Policy Options for the Effectiveness 1 (2): 195–216. Government of the Philippines.” Policy note, Indonesia—Enterprise Survey 2015, World Bank Asian Development Bank, Manila. database, http://microdata.worldbank.org​ Ve z z a , E . , B . G a r c í a , G . C r u c e s , a n d /­index.php/catalog/2665. J. Amendolaggine. 2014. “Programa Juventud y Jung, H., and A. Hasan. 2014. “The Impact Empleo, Ministerio de Trabajo, República of Early Childhood Education on Early Dominicana: Informe de Evaluación de Impacto Achievement Gaps: Evidence from the Cohortes 2008–2009.” Unpublished paper, Indonesia Early Childhood Education and World Bank, Washington, DC. Development (ECED) Project.” Policy We l l s - D a n g , A . 2 0 1 2 . E t h n i c M i n o r i t y Research Working Paper 6794, World Bank, Development in Vietnam: What Leads to Washington, DC. Success? Washington, DC: World Bank. Kandel, E., K. Kosenko, R. Morck, and Y. Yafeh. World Bank. 2011. Assessment of the Social 2015. “The Great Pyramids of America: Protection System in Fiji and Recommendations A Revised History of US Business Groups, for Policy Changes . Washington, DC: Corporate Ownership and Regulation, World Bank. 1930–1950.” Unpublished paper, May. ———. 2012. Malaysia Economic Monitor 2012: Lam, W. R., and P. Wingender. 2015. China: How Unlocking Women’s Potential. Washington, Can Revenue Reforms Contribute to Inclusive DC: World Bank. and Sustainable Growth. Washington, DC: ———. 2014a. East Asia Pacific at Work: International Monetary Fund. Employment, Enterprise, and Well-being . Naidoo, D., T. Packard, and I. Auwalin. 2015. Washington, DC: World Bank. Mobility, Scarring and Job Quality in ———. 2014b. Global Financial Development Indonesia’s Labor Market. Washington, DC: Report 2014: Financial Inclusion. Washington, World Bank. DC: World Bank. Narayan, D., L. Pritchett, and S. Kapoor. 2009. ———. 2015a. Achieving a System of Competitive Moving Out of Poverty. Vol. 2 of Success Cities in Malaysia. Washington, DC: World from the Bottom Up. New York: Palgrave Bank. Macmillan. ———. 2015b. Malaysia Economic Monitor 2015: Norregaard, J. 2015. “Taxing Immovable Property: Transforming Urban Transport. Washington, Revenue Potential and Implementation DC: World Bank. Challenges.” In Inequality and Fiscal Policy, ———. 2016a. East Asia and Pacific Economic edited by B. J. Clements, R. A. de Mooij, Update, October 2016: Reducing and S. Gupta. Washington, DC: International Vulnerabilities. Washington, DC: World Bank. Monetary Fund. ———. 2016b. Indonesia’s Rising Divide. Park, C.-Y. 2012. “Taxes, Social Transfers, and Washington, DC: World Bank. Inequality in Asia.” Accessed June 27, 2017. ———. 2016c. Live Long and Prosper: Aging in https://www.imf.org/external/np/seminars​ East Asia and Pacific. Washington, DC: World /­eng/2012/asiatax/pdf/park.pdf. Bank. doi:10.1596/978-1-4648-0469-4. Piketty, T., E. Saez, and S. Stantcheva. 2014. “The ———. 2017a. Pacific Possible: Long-term Optimal Taxation of Top Labor Incomes: A Economic Opportunities and Challenges for 104   RIDING THE WAVE the Pacific Island Countries. Washington, DC: World Bank and World Health Organization. World Bank. 2014. Global Civil Registration and Vital ———. 2017b. Financing Pacific Governments Statistics—Scaling Up Investment Plan for Pacific Development. Washington, DC: 2015–2024. Washington, DC: World Bank. World Bank. Yan, X. 2013. Land Tenure Arrangements, ———. Forthcoming a.“Growing Smarter: Factor Market Development, and Agricultural Learning and Growth in East Asia and the Production in China: Evidence from Henan Pacific.” Washington, DC: World Bank Province. Weikersheim, Germany: Margraf. ———. Forthcoming b. Breaking the Intergenerational Zhang, L., and N. Liu. 2014. “Health Reform Transmission of Poverty: The Philippines’ and Out-of-Pocket Payments: Lessons from Conditional Cash Transfer Program. Washington, China.” Health Policy and Planning 29 (2): DC: World Bank. 217–26. Statistical Appendix: Poverty data availability for A developing East Asia and Pacific The East Asia and Pacific poverty reposi- survey design, questionnaire, fieldwork, and tory hosts 75 household surveys from 19 overall methodology in collecting and pro- countries (table A.1). These data are used cessing information. for monitoring poverty and shared prosper- Coverage of the surveys increased from ity and, more broadly, can be used to iden- 59 percent of villages in 1993/94 to 100 percent tify economic class. In addition, 20-point in 2004. grouped data for China are available for After the 2004 survey, Living Standards the following years: 2002, 2005, 2008, and Measurement Study surveys were collected 2010–13. This report relies on estimates every year from 2007 through 2013. The based on the data available and approved sample sizes for the 2004 and 2009 surveys for use by the respective National Statistical were close to 12,000 households; in other Offices as of June 2017. years, the sample sizes were about 3,600 The frequency with which household sur- households. veys are conducted varies across countries. All of the surveys were collected over the National representative household surveys are entire year and are representative of the entire conducted annually in Cambodia, Indonesia, country. and Thailand; every two years in Vietnam; and every three years in the Philippines. For other countries, the survey gap can be China five years or more. China’s population reached 1.37 billion in 2015. Between 2002 and 2015, China expe- Country-specific notes rienced low demographic growth of slightly above 0.5 percent per year, less than the average Cambodia in the East Asia and Pacific region (0.7 percent). The Cambodia Socio-Economic Survey has Given China’s size, estimates of poverty and been conducted since 1993/94. It collects the number of poor in China greatly influence information on households and individual regional and global aggregates. household members. Estimates of poverty are typically based on In 2004, the survey adopted a more analysis of household-level data from large- standardized methodology, including scale nationally representative surveys. 105 106   RIDING THE WAVE TABLE A.1  Available household surveys for countries in developing East Asia and Pacific Country Survey code name 1996 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 CHN* X X X X X X X FJI HIES X X FSM HIES X X IDN SUSENAS X X X X X X X X X X X X X X X X KIR HIES X LAO LECS X X X MNG HSES X X X X MYS HIS X X X X X PHL FIES X X X X X X PNG HIES X X SLB HIES X X THA SES X X X X X X X X X X X X TLS TLSS/TLSLS X X TON HIES X X TUV HIES X VNM VHLSS X X X X X X X VUT HIES X WSM HIES X Note: * China is group data. From 2013, the survey name is Integrated Urhan and Rural Household Survey; and prior to that there are two different surveys: Urban Household Survey and Rural Household Survey. HIES = Household Income and Expenditure Surveys; SUSENAS = National Socio-Economic Survey (Indonesia); LECS = Lao Expenditure and Consumption Surveys; HIS = Household Income Survey; HSES = Household Socio-Economic Surveys; FIES = Family Income and Expenditure Survey; SES = Socio-Economic Survey; TLSS/TLSLS = Timor-Leste Living Standards Survey/Timor-Leste Survey of Living Standards; VHLSS = Vietnam Household Living Standards Surveys. S T A T I S T I C A L A P P E N D I X A    107 China’s National Bureau of Statistics has module data. Module data contain more never given the World Bank access to the detailed information. household-level data from its nationally rep- In 2005–10, SUSENAS was conducted bian- resentative large-scale household surveys. nually—every February and July (2005–06) The World Bank’s poverty estimates for and then every March and July (2007–10). China are based on grouped data, which is From 2011 to 2014, SUSENAS was information on mean per capita consumption ­ c onducted quarterly: in March, June, levels for groups of the population equally September, and December. In 2011, the sur- spaced across the distribution. The 20-point vey underwent major changes in the sampling grouped data are available for urban and methodology, level of representativeness, and rural areas separately. weighting because of the availability of 2010 The grouped data have been made avail- census data. able to the World Bank with a lag of three Beginning with 2015, the surveys have years. For example, the grouped data from been conducted twice a year, in March and 2013 household surveys were made available September. In 2015, BPS instituted several in the first quarter of 2016. changes: (a) reducing the number of food items Grouped data for China are available every in the consumption module from 255 to 100; three years from early 2000s (2002, 2005, (b) changing the sample frame to stratify pri- 2008) and each year since 2010 (2010, 2011, mary sampling units by wealth levels instead of 2012, and 2013). A survey break occurred in by education levels; (c) dropping the quarterly 2013, so these estimates are not comparable approach and returning to biannual sampling with the previous ones. (March and September); and (d) increasing the sample to 300,000 households. Overall, SUSENAS uses a two-stage Fiji sampling method: selecting census blocks and ­ The Fiji Island Bureau of Statistics (FIBOS) selecting households within each designated ­ conducted a nationwide Household Income census block. The Probability Proportional and Expenditure Surveys (HIES) in 2002/03, to Size (PPS) method is used when select- 2008/09, and 2013/14. census blocks from the Master Sampling ing ­ The survey is nationally representative for Frame (MSF), which can be obtained from urban and rural areas, divisions, and strata, the population census. and the sample was drawn from the 1996 SUSENAS is conducted in all areas in and 2007 population censuses according Indonesia, with a variety of sample sizes. to a two-stage, stratified, random sampling For measuring poverty, the February design. Data collection is continuous over a (2002–07) or March (2008–15) rounds are 12-month period. used. The sample size is varied each year. The survey sample size was 5,245 house- In 2000 and 2001, the sample size was holds in 2002/03; 3,573 in 2008/09; and about 200,000 households; about 10,000 6,020 in 2013/14. households from 2002 to 2006; between 66,000 and 72,000 from 2007 to 2014; and 286,000 in 2015. Indonesia Indonesia’s National Socio-Economic Lao People’s Democratic Republic Survey (SUSENAS) is carried out regularly by the Central Bureau of Statistics (BPS). The Lao Statistics Bureau has conducted the The first SUSENAS was conducted in 1963. Lao Expenditure and Consumption Surveys Since 1992, SUSENAS has been collecting (LECS) at five-year intervals since 1992/93. two different types of data: core data and The fifth and most recent round (LECS 5) 108   RIDING THE WAVE was implemented between April 2012 and All poverty estimates from the 1996 survey March 2013. took account of the clustered, stratified, and LECS 5 is a nationally representative sur- weighted nature of the sample. vey designed to generate representative pov- The 2009/10 Household Income and erty estimates at national and provincial Expenditure Survey (HIES) conducted by levels. The sizes of survey samples were 8,092 the National Statistical Office of Papua (2002); 8,296 (2007); and 8,226 (2012). All New Guinea ran from June 2009 until surveys are comparable. January 2011. One-third of interviews were in 2009, two-thirds in 2010, and less than 2 percent in January 2011. Malaysia The total sample of the 2009/10 HIES The Household Income Survey (HIS) is was 3,600 households from 10 strata, which ­ carried out by the Departments of Statistics, covered rural and urban sectors in the four Malaysia. Since 1973, the HIS has been regions (Southern, Highlands, Momase, and carried out twice in five years, that is, two ­ Islands), along with the metro area. surveys within each Malaysia Development The 1996 and 2009/10 surveys are not Plan period. comparable. The main objectives of the HIS are to measure the economic well-being of the pop- The Philippines ulation; to collect information on income dis- tribution patterns of households, classified by The Family Income and Expenditure Survey various socioeconomic characteristics; and to (FIES) is a nationwide survey of households provide the base data for the calculation of undertaken every three years by the National the poverty line income. Statistics Office (NSO). It is the main source of data on family income and expenditure. The FIES is designed to provide income Mongolia and expenditure data that are nationally and The National Statistical Office of Mongolia regionally representative. conducted Household Socio-Economic Surveys The survey used four replicates of the 2003 (HSES) in 2010, 2011, 2012, and 2014. Master Sample (MS) created for household All surveys are nationally representa­ surveys on the basis of the 2000 Census of tive and cover the whole of Mongolia. They Population and Housing. aim to evalu­ate and monitor the income and The 2003 MS was designed to produce the expenditure of households, update the basket sample size needed for large surveys like the and weights for the consumer price index, and FIES. To facilitate subsampling, the 2003 MS provide inputs to the national accounts. The was designed to readily produce four replicate HSES covers a 12-month period for analysis. samples from the full set of sampled primary All available surveys (2010, 2011, 2012, sampling units. and 2014) are comparable. The FIES enumeration is conducted twice. The first visit is done in July, covering the first semester (January to June) as the r ­eference Papua New Guinea period; the second visit is made in January of The first Papua New Guinea Household the succeeding year, covering the second semes- Survey (PNGHS) was undertaken in 1996, ter of the reference year (July to December). with a total sample of 1,144 households. The The same set of questions is asked in both visits. fieldwork for PNGHS was staggered over the The sample size for each FIES is about full calendar year to allow for any seasonality. 40,000 households: 39,615 (2000); 42,094 The samples were drawn from enumera- (2003); 38,483 (2006); 38,400 (2009); tion areas of the 1990 census, using stratified 40,171 (2012); and 41,544 (2015). sampling (15 strata). All FIES results are comparable. S T A T I S T I C A L A P P E N D I X A    109 Thailand Vietnam The National Statistical Office (NSO) con- Vietnam carried out Living Standards Surveys ducted its first survey on household income (VLSS) in 1992/93 and 1997/98, with exten- and expenditure, the Household Expenditure sive technical support from international part- Survey, in 1957. In 1968/69, the title was ners. Then the government financed a series of changed to Socio-Economic Survey (SES), and Vietnam Household Living Standards Surveys that survey was conducted every five years (VHLSS) every other year (2002, 2004, 2006, until 1986. 2008) using an approach that was similar to In 1986, NSO was requested to conduct the earlier VLSS. The main objective was to the SES every two years, to keep up with the maintain comparability over time. country’s economic circumstances and rapid Strict comparability came at too high a expansion, as well as with data needs for cost, however. The 2010 VHLSS used a new ­ policy planning and the formulation of pov- master sample based on the 2009 Housing erty reduction strategies. and Population Census, including a new set After the economic crisis in 1997, NSO of communes and enumeration areas. The conducted special rounds of the SES in 1999 VHLSS household questionnaire was sub- and 2001 to measure the impact of the crisis stantially revised (including revisions to the on the Thai population. core consumption module) and shortened. In Since 2006, the SES has been conducted addition, an updated methodology was used annually. In even-numbered years, income, to construct a more comprehensive consump- expenditure, and debt are surveyed, and in tion (welfare) aggregate. As a result, the 2010 odd-numbered years, only expenditure is VHLSS and related welfare aggregates repre- surveyed. sent a break with the 2002–08 VHLSS series. Timor-Leste Pacific Island Countries The first national Timor-Leste Living Among Pacific Island countries, survey data Standards Survey (TLSS) was undertaken in are available from Kiribati, the Federated 2001 from August to November. The 2001 States of Micronesia, Samoa, the Solomon TLSS had a modest, though nationally repre- Islands, Tonga, Tuvalu, and Vanuatu. sentative, sample of 1,800 households cover- The Federated States of Micronesia, the ing 1 percent of the population. Solomon Islands, and Tonga have two data The second national Timor-Leste Survey points, and the remaining countries have a of Living Standards (TLSLS) took place from single data point. The average gap between January 2007 to January 2008, with a sample surveys is about eight years. of 4,500 households. The two surveys in each country—the The 2001 and 2007 surveys are comparable. Federated States of Micronesia, the Solomon The latest survey available is the TLSLS Islands, and Tonga—are not comparable. 2014, but the data had not been cleared for public use at the time of writing. Statistical Appendix: Methodology for the estimation B of internationally comparable poverty estimates for East Asia This appendix describes briefly the main spatially deflated terms, and occasionally in element of the methodology adopted to mea- ­ both forms. Methodologies vary by country. sure poverty, namely, the use of welfare aggre- China, Indonesia, and Mongolia, for example, gates for the analysis and for the construction release data that are already spatially adjusted, of internationally comparable annual data whereas other countries do not. The more series. diverse and vast a country, the more the spa- With the exception of the application of tial adjustments will affect poverty estimates. different thresholds, the same methodological The East Asia and Pacific Poverty estimates, considerations apply to the measurement of available from 2002 onward,1 adopt as spatial economic classes. deflators the ratio of location-specific poverty lines (urban/rural, regional, provincial) to the national poverty line. This deflation is applied Constructing internationally to most East Asia and Pacific countries’ nomi- comparable welfare aggregates nal welfare aggregate, with the exception of Either consumption or income can be used Thailand’s and the Solomon Islands’ aggre- as a measure of welfare. Most countries in gates. Thailand ­ produces ­ household-level developing East Asia and Pacific report their poverty lines that are adjusted for demograph- official poverty measures on the basis of con- ics and economies of scale. In the Solomon sumption. Only Malaysia and the Philippines poverty line in the rural areas Islands, the food ­ use income. is underestimated compared with urban areas. For monitoring poverty, the World Bank Using a comparison between the two lines to mostly employs the official welfare aggregates construct a spatial deflator would significantly used by countries for calculating the national underestimate rural poverty incidence. poverty rate. Official sources provide the con- In addition to spatial deflation, adjust- sumption and income aggregates in either ments are also sometimes performed over nominal (non–spatially deflated) terms or time, even within a given survey period.2 111 112   RIDING THE WAVE Producing regional poverty reduction in poverty rates associated with a estimates percentage change in mean per capita income. It is estimated, as follows: The welfare indicator obtained after the spatial and temporal adjustment previously •  For the interpolations, the GEP is estimated described is expressed as per capita household between the two closest available surveys consumption and income in annual terms. before and after the missing year (the inter- Those are then translated into purchasing polation period). The GEP is then applied to power parity (PPP) terms to reflect purchas- the poverty estimate obtained from the last ing power differences across countries and household survey before the missing year, compared with the international poverty lines using known changes in GDP per capita. (or the relevant class thresholds). •  For the extrapolations, two methods are The international poverty lines—originally used. In the first method, applied to coun- proposed by the World Bank in 1990 to facili- tries for which at least two surveys are tate poverty comparisons across countries—in available, the GEP estimated from two 2011 PPP stand at US$1.90 a day for extreme previous surveys and annualized is applied poverty. The East Asia and Pacific region of to the latest available survey, using known the World Bank has traditionally adopted changes in GDP per capita. If the pro- a higher poverty line to capture moderate jections relate to the future rather than poverty. Such a line is currently estimated at to known GDP per capita changes, the US$3.10. These lines are the basis for poverty second method uses projected GDP per measurement across the region. capita. In the second method, a neutral Regional poverty rates are obtained as shift in the distribution is assumed. This population-weighted averages of country- method is used in countries for which only specific poverty rates using the international one data point is available and in China. poverty lines at 2011 PPP prices. In the case of China, the extrapolation applies the distribution obtained from 20 grouped data points. Constructing the internationally For the projections used in this report, GDP comparable annual series per capita for the years 2016 through 2018 is of poverty headcounts and obtained from World Bank simulation mod- number of poor people living els; for the years from 2019 through 2021, in poverty it is obtained from the World Economic The annual series of poverty estimates pre- Outlook of the International Monetary Fund sented in this report reflect a combination of (IMF). Longer-term projections through 2030 make the assumption that GDP growth from •  Actual-year data for which the estimates 2021 is constant. are calculated on the basis of available Finally, the ratio of the number of poor survey data (table A.1) and to the total population in the region is used •  Estimates for other years based on inter- to obtain regional poverty numbers from the polation (that is, constructing estimates headcounts calculated based on the methodol- for years between those for which house- ogy just described. Total population estimates hold survey data are available to directly and projections are derived from the World estimate poverty) and extrapolation (that Bank’s Health Nutrition and Population is, projecting beyond the most recent year Statistics database. Data are received from for which survey data are available). the World Bank’s Development Data Group, The interpolated and extrapolated estimates which updates and publishes them twice a are obtained on the basis of available house- year, in April and July. hold survey estimates and the growth elas- A note of caution: the regional estimates ticity of poverty (GEP) to gross domestic presented in this report may not compare with product (GDP). The GEP is the percentage those published in PovcalNet (http://iresearch​ S T A T I S T I C A L A P P E N D I X B    113 .worldbank.org/PovcalNet/povOnDemand​ 2. Nominal in the context of this note refers .aspx) because of methodological differences to non–spatially adjusted. In addition, relating to spatial cost-of-living differences the concept of over time refers to the and the ways estimates are interpolated. months during which a particular survey is undertaken, not the years in the macro-GDP sense of real GDP. Some national statistics Notes offices not only make a spatial adjustment 1. Longer time series presented in this report but also make a temporal adjustment of are provided by PovcalNet, http://iresearch​ household incomes and consumption to a .worldbank.org/PovcalNet/povOnDemand​ particular month of the year (or to a yearly .aspx. average) for a given survey. Statistical Appendix: Supplementary tables C TABLE C.1  Growth rates in developing East Asia and Pacific, 1960–2015 1960–69 (%) 1970–79 (%) 1980–89 (%) 1990–99 (%) 2000–07 (%) 2008–15 (%) East Asia and Pacific 8.9 5.0 5.3 3.7 4.6 4.0 Japan 10.4 4.1 4.4 1.5 1.5 0.2 NIEs 7.8 10.3 8.3 6.3 5.3 3.1 Hong Kong SAR, China 4.5 9.0 7.4 3.6 5.3 2.6 Korea, Rep. 8.3 10.5 8.6 6.7 5.4 3.1 Singapore 8.9 9.5 7.8 7.3 6.5 4.4 Taiwan, China 9.9 10.9 8.5 6.6 4.9 2.8 Developing EAP 3.9 7.3 7.7 8.2 9.1 7.8 Cambodia … … … 7.2 9.6 6.1 China 3.0 7.4 9.8 10.0 10.5 8.6 Indonesia 3.7 7.8 6.4 4.8 5.1 5.6 Lao PDR … … 4.1 6.3 6.7 7.9 Malaysia 6.5 7.7 5.9 7.2 5.6 4.6 Mongolia … … 6.0 −0.3 6.6 8.2 Myanmar 3.0 4.4 1.9 6.1 12.9 7.9 Papua New Guinea 6.4 3.9 1.4 4.3 2.2 7.7 Philippines 5.1 5.8 2.0 2.8 4.9 5.3 Thailand 7.8 7.5 7.3 5.2 5.3 2.9 Vietnam … … 4.5 7.4 6.9 5.9 ECA … … 2.2 −0.4 6.1 2.8 LAC 6.1 6.9 2.5 2.7 3.5 2.9 MNA 9.5 5.5 1.5 4.2 4.8 2.3 SAR 4.1 3.0 5.6 5.4 6.7 6.5 SSA 4.6 4.0 1.7 1.9 5.1 4.1 Source: World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world-development-indicators; IMF, World Economic Outlook Database, https://www.imf.org/external/pubs/ft/weo/2017/01/weodata/index.aspx. Note: Values for regional aggregates denote GDP-weighted means. EAP = East Asia and Pacific; ECA = Europe and Central Asia; GDP = gross domestic product; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; NIE = newly industrialized economy; SAR = South Asia; SSA = Sub-Saharan Africa; . . . = missing. 115 116   RIDING THE WAVE TABLE C.2  Tax revenue in developing East Asia and Pacific, percentage of GDP, 2013 Taxes on income, Taxes on Taxes on Taxes on profits, payroll goods international GDP per and and Taxes on and trade and Other capita capital workforce property services transactions taxes (US$, 2011 Taxes gains (%) (%) (%) (%) (%) (%) PPP) East Asia and Pacific 15.7 5.2 0.2 0.0 7.0 3.1 0.3 11,375 East Asia 14.2 5.6 0.0 0.0 7.5 1.6 0.3 11,130 Cambodia 13.6 2.9 0.0 0.0 7.9 2.8 0.0 2,955 China 10.6 2.9 0.0 0.0 7.2 0.5 0.0 11,805 Indonesia 13.0 5.6 0.0 0.0 5.4 0.5 1.5 9,675 Lao PDR 16.1 3.0 0.0 0.1 11.1 1.9 0.0 4,800 Malaysia 14.8 10.5 0.0 0.0 3.0 0.4 0.9 23,419 Mongolia 17.2 3.8 0.0 0.0 11.2 2.2 0.1 10,757 Myanmar 7.1 … … … … … … … Philippines 13.6 6.4 0.0 0.0 4.5 2.7 0.0 6,365 Thailand 16.9 7.4 0.0 0.0 8.5 0.9 0.1 14,943 Timor-Leste 1.6 0.6 0.0 0.0 0.8 0.2 0.0 2,039 Vietnam 19.1 7.7 0.0 0.0 8.8 2.2 0.3 5,125 Pacific Island countries 18.7 5.8 0.3 0.1 7.3 5.0 0.3 4,997 Fiji 24.6 6.0 0.0 0.0 13.1 5.5 0.0 7,897 Kiribati 14.9 3.8 2.5 0.0 0.0 8.6 0.0 1,694 Marshall Islands 17.3 6.8 0.0 0.2 6.0 4.2 0.1 3,671 Micronesia, Fed. Sts. 5.2 2.5 0.0 0.0 1.5 1.2 0.1 3,300 Palau 18.9 3.3 0.5 0.0 7.9 4.6 2.7 13,152 Samoa 23.3 5.6 0.0 0.2 14.7 2.9 0.0 5,499 Solomon Islands 29.5 9.8 0.0 0.3 7.5 11.9 0.0 2,043 Tonga 17.0 3.1 0.0 0.0 12.0 1.9 0.0 4,888 Tuvalu 18.9 11.3 0.0 0.0 3.0 4.5 0.0 3,528 Vanuatu 17.2 … … … … … … 2,890 ECA 18.3 4.4 0.0 0.0 12.0 1.6 0.2 13,501 LAC 17.5 5.4 0.0 0.2 9.2 2.3 0.3 13,731 MNA 14.2 4.3 0.0 0.6 7.8 1.1 0.4 11,574 SAR 12.9 4.4 0.0 0.0 5.8 2.3 0.4 5,067 SSA 21.7 7.9 0.1 0.1 7.1 6.2 0.3 3,324 Source: International Monetary Fund, Government Finance Statistics Yearbook and data files, http://data.imf.org/?sk=E86E9088-3830-4CA3-B240-1B0EC5E15221. Note: Data refer to 2013, except for China (2011) and Indonesia (2007). ECA = Europe and Central Asia; GDP = gross domestic product; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; PPP = purchasing power parity; SAR = South Asia; SSA = Sub-Saharan Africa; . . . = missing. S T A T I S T I C A L A P P E N D I X C   117 TABLE C.3  Access to basic services in developing East Asia and Pacific, 2015 Primary Secondary Access to Access to school school sanitationb improved Access to Access to GDP per enrollmenta enrollmenta (%) waterc (%) telecomd Internete capitaf East Asia and Pacific 100.5 83.1 74.8 93.6 13.4 44.9 12,308 Cambodia 97.4 38.2 42.4 75.5 1.6 19.0 3,278 China 103.9 94.3 76.5 95.5 16.5 50.3 13,572 Indonesia 89.7 75.0 60.8 87.4 8.8 22.0 10,385 Lao PDR 95.1 50.8 70.9 75.7 13.7 18.2 5,345 Malaysia 94.6 69.1 96.0 98.2 14.3 71.1 25,312 Mongolia 94.9 86.3 59.7 64.4 8.7 21.4 11,478 Myanmar 94.5 48.3 79.6 80.6 1.0 21.8 4,931 Philippines 96.0 67.4 73.9 91.8 3.0 40.7 6,938 Thailand 92.4 79.6 93.0 97.8 7.9 39.3 15,347 Vietnam 98.0 … 78.0 97.6 6.3 52.7 5,667 Fiji 95.1 83.4 91.1 95.7 8.1 46.3 8,756 Kiribati 96.4 69.1 39.7 66.9 1.4 13.0 1,873 Micronesia, Fed. Sts. 83.0 … 57.1 89.0 6.5 31.5 3,284 Palau 99.0 95.2 100.0 95.3 33.8 27.0 14,386 Papua New Guinea 86.0 … 18.9 40.0 2.0 7.9 2,723 Samoa 96.1 79.5 91.5 99.0 5.6 25.4 5,574 Solomon Islands 80.7 42.2 29.8 80.8 1.3 10.0 2,067 Timor-Leste 96.6 51.8 40.6 71.9 0.2 13.4 2,253 Tonga 87.2 75.4 91.0 99.6 12.4 45.0 5,198 Tuvalu 84.6 69.8 83.3 97.7 20.2 42.7 3,687 Vanuatu 98.9 51.6 57.9 94.5 1.8 22.4 2,806 ECA 92.4 86.7 94.1 94.8 21.4 42.9 13,224 LAC 91.8 73.7 81.0 93.8 17.8 42.6 13,938 MNA 95.6 71.5 88.0 89.7 16.1 31.2 11,728 SAR 89.6 51.4 29.6 64.6 1.3 14.7 3,259 SSA 77.1 33.2 39.8 91.2 2.6 11.6 4,635 World 89.1 65.5 63.6 89.3 16.7 35.6 13,682 Sources: World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world-development-indicators; national statistical authorities. Note: Data for EAP refer to 2015, with the following exceptions. Primary school enrollment refers to 2014, except Fiji (2013), Malaysia (2012), Papua New Guinea (2012), the Philippines (2013), Solomon Islands (2007), Vanuatu (2005), and Vietnam (2013). Secondary school enrollment refers to 2014, except Cambodia (2008), Fiji (2012), Kiribati (2005), Palau (2013), Solomon Islands (2012), the Philippines (2013), Thailand (2012), Vanuatu (2010), and South Asia (2008). Access to sanitation: Tuvalu (2013). Access to improved water: Palau (2011). Access to Internet: Palau (2004). GDP per capita: Papua New Guinea (2014). Data for all other regions and the world refer to 2012. Regional aggregates denote population- weighted means. ECA = Europe and Central Asia; GDP = gross domestic product; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa; . . . = missing. a. Net enrollment, except gross enrollment for China. b. Access to improved sanitation facilities refers to the percentage of the population using improved sanitation facilities. These include flush/pour flush (to piped sewer system, septic tank, pit latrine), ventilated improved pit (VIP) latrine, pit latrine with slab, and composting toilet. c. Access to improved water is defined as the percentage of the population using improved drinking water sources. These include piped water on premises (piped household water connection located inside the user’s dwelling, plot, or yard), and other improved drinking water sources (public taps or standpipes, tube wells or boreholes, protected dug wells, protected springs, and rainwater collection). d. Access to telecom refers to the sum of active analog fixed telephone lines, voice-over-IP (VoIP) subscriptions, fixed wireless local loop (WLL) subscriptions, ISDN voice-channel equivalents, and fixed public payphones, per 100 people. e. Access to Internet is defined as the number of individuals who have used the Internet (from any location) in the last 12 months per 100 people. Internet can be used via a computer, mobile phone, personal digital assistant, games machine, digital TV, and so on. f. GDP per capita, purchasing power parity (PPP) in constant 2011 US$. 118   RIDING THE WAVE TABLE C.4  Access to sanitation and improved water in developing East Asia and Pacific, rural versus urban, 2012 Access to sanitationa (%) Access to improved waterb (%) Rural Urban Rural Urban East Asia and Pacific 57.8 75.9 84.5 97.3 Cambodia 26.4 79.9 62.9 91.9 China 60.9 84.4 88.5 97.5 Indonesia 45.5 71.4 77.2 93.6 Lao PDR 50.5 90.4 64.9 83.7 Malaysia 95.4 96.1 92.4 100.0 Mongolia 39.2 66.1 53.8 68.0 Myanmar 77.1 84.3 74.4 92.7 Philippines 67.8 76.8 88.7 93.4 Thailand 96.1 89.9 96.8 97.5 Vietnam 64.8 90.8 91.9 98.0 Fiji 88.4 93.4 91.2 99.5 Kiribati 30.6 51.2 50.6 87.4 Marshall Islands 55.5 84.2 97.5 93.4 Micronesia, Fed. Sts. 49.0 85.1 87.4 94.8 Palau 100.0 100.0 86.0 97.0 Papua New Guinea 13.3 56.4 32.8 88.0 Samoa 91.1 93.3 98.8 97.4 Solomon Islands 15.0 81.4 77.2 93.2 Timor-Leste 26.8 69.0 60.5 95.2 Tonga 89.4 97.6 99.5 99.2 Tuvalu 80.2 86.3 97.0 98.3 Vanuatu 55.4 65.1 88.4 98.3 ECA 90.9 96.8 88.9 98.6 LAC 62.3 86.1 82.3 97.1 MNA 80.2 93.5 82.7 94.6 SAR 22.6 40.7 52.6 85.8 SSA 29.9 61.1 89.2 95.4 World 46.5 79.3 81.5 96.5 Source: World Bank World Development Indicators (database), https://data.worldbank.org/data-catalog/world-development-indicators. Note: All data refer to 2012, except for access to improved water in rural Palau (2011). Regional aggregates denote population-weighted means. ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MNA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. a. Access to sanitation is defined as the percentage of the population with access to improved sanitation facilities, including flush/pour flush (to piped sewer system, septic tank, pit latrine), ventilated improved pit latrine, pit latrine with slab, and composting toilet. b. Access to improved water is defined as the percentage of the population with access to an improved drinking water source, including piped water and other improved drinking water sources (public taps, standpipes, tube wells, boreholes, protected dug wells, protected springs, and rainwater collection). S T A T I S T I C A L A P P E N D I X C   119 TABLE C.5  Percentage of self-reported membership in different organizations, by economic security status PHL THA MYS CHN Organization NS S NS S NS S NS S Religious 60 55 43 25 37 24 7 4 Sporting 34 24 34 23 23 13 11 9 Arts, music, education 24 21 32 17 20 9 9 7 Labor union 21 14 18 11 20 14 7 8 Political party 22 16 17 10 24 15 7 9 Environmental 29 21 24 14 16 7 3 2 Professional association 18 19 22 14 13 6 3 2 Humanitarian/charity 30 24 22 16 17 10 3 1 Consumer 20 13 22 12 15 9 4 1 Self-help or mutual aid 18 13 61 53 16 8 4 3 Other 20 20 16 10 9 5 1 1 Source: Wai Poi and others 2016 using World Values Survey data from Wave 6, 2010–14. Note: Economic security is defined here on the basis of answers to the World Values Survey rather than on economic welfare as in the rest of the report. More specifically, the secure are those who never went without enough food in the past year, never or rarely went without medical treatment they needed in the past year, and did not deplete savings (“saved money” or “just got by”) in the past year. For Thailand, only two of the criteria have to be met. This adjustment was made to ensure that the definition of security is roughly consistent with the economic definition obtained by household survey data. CHN = China; MYS = Malaysia; NS = nonsecure; PHL = Philippines; S = secure; THA = Thailand. TABLE C.6  Percentage of self-reported lack of confidence in public institutions, by economic security status PHL THA MYS CHN Institution NS S NS S NS S NS S Churches 6 5 19 27 12 8 72 76 Army 28 36 21 32 20 20 10 6 Press 33 33 44 59 35 31 34 28 Television 26 24 39 50 32 29 30 26 Labor unions 48 57 50 56 28 21 37 33 Police 31 42 46 54 26 26 31 23 Courts 34 38 20 25 21 20 24 17 Government 41 50 41 48 23 25 10 7 Political parties 53 55 59 63 41 37 20 12 Parliament 42 43 46 55 37 30 18 10 Civil service 34 39 34 41 23 22 26 20 Universities 22 18 9 14 20 15 13 9 Major corporations 36 35 32 40 33 24 36 31 Banks 23 26 14 16 23 18 16 11 Sources: Wai Poi and others 2016 using World Values Survey data from Wave 6, 2010–14. Note: Economic security is defined here on the basis of answers to the World Values Survey rather than on economic welfare as in the rest of the report. More specifically, the secure are those who never went without enough food in the past year, never or rarely went without medical treatment they needed in the past year, and did not deplete savings (“saved money” or “just got by”) in the past year. For Thailand, only two of the criteria have to be met. This adjustment was made to ensure that the definition of security is roughly consistent with the economic definition obtained by household survey data. Respondents were asked how much confidence they had in each institution. They could choose “a great deal,” “quite a lot,” “not a lot,” or “not at all.” Lack of confidence as measured in the table reflects the last two answers combined. CHN = China; MYS = Malaysia; NS = nonsecure; PHL = Philippines; S = secure; THA = Thailand. 120   RIDING THE WAVE TABLE C.7  Indicators of financial inclusion, percentage of population age 15 and above, 2014 Account in formal institution Formal savings Formal borrowing Country All Female Rural Poor LE All Female Rural Poor LE All Female Rural Poor LE Cambodia 12.6 10.7 11.4 8.8 10.5 3.6 2.6 3.0 2.7 3.1 27.7 29.4 28.5 25.8 29.3 China 78.9 76.4 74.3 72.0 72.8 41.1 41.2 37.5 30.5 37.1 9.6 8.7 7.5 5.9 7.1 Indonesia 35.9 37.2 28.5 21.9 15.8 26.6 26.8 21.3 13.8 11.0 13.1 11.2 11.4 11.3 9.5 Malaysia 80.7 78.1 73.7 75.6 58.6 33.8 32.3 32.6 25.6 22.3 19.5 16.6 17.0 15.2 9.5 Philippines 28.1 33.9 24.6 14.9 15.2 14.8 15.9 11.2 4.9 6.2 11.8 13.6 12.2 8.2 7.7 Thailand 78.1 75.4 78.2 72.0 73.2 40.6 37.9 35.8 31.3 38.1 15.4 13.7 16.7 16.2 14.7 Vietnam 30.9 31.9 27.0 18.7 15.3 14.6 13.7 11.9 9.1 9.0 18.4 21.3 20.7 19.9 24.8 Source: Global Findex 2014 database, http://www.worldbank.org/en/programs/globalfindex. Note: All = all adults; Female = females; Rural = rural population; Poor = lowest 40 percent of income distribution; LE = least educated—primary only. Reference Wai-Poi, M., R. Purnamasari, T. Indrakesuma, I. Uochi, and L. Wijaya. 2016. “East Asia’s Rising Middle Classes.” Background paper for Riding the Wave: An East Asian Miracle for the 21st Century . Washington, DC: World Bank. Statistical Appendix: A profile of economic class in D developing East Asia and Pacific 121 122   TABLE D.1  Key socioeconomic characteristics of the various economic classes Economically Extreme poor Moderate poor vulnerable Economically secure Middle class Rich RIDING THE WAVE Population and Less than 3 percent of 10 percent of about one quarter of about half of about 17 percent of less than half a consumption shares population population the population population population percentage point of population Location Predominantly rural—31 Predominantly rural— Mostly rural—39 Slightly more urban Predominantly urban— Predominantly urban— percent living in urban areas 35 percent living in percent living in urban than rural (51 percent) 70 percent living in 78 percent living in urban areas areas and about the same as urban areas urban areas the regional average Transport Almost no access to private Limited access to Less likely to have More likely to have Car ownership The only class for modes of transport, such as private modes of access to private modes access to private (28 percent) much which car ownership motorcycles or cars transport of transport modes of transport lower than motorcycle (70 percent) is higher than lower economic ownership (67 percent) than motorcycle classes; motorcycle ownership (48 percent) ownership (72 percent) more common than car ownership (9 percent) Education Education level 4 years: Education level 5.2 Education level 6 years: Above-average human Education level 10 Education level among 3- to 4-year-olds, years: among 3- to among 3- to 4-year- capital; education level years; among 3- to 13 years—almost twice 14 percent are in school 4-year-olds, 18 percent olds, 24 percent are 8 years: among 3- to 4-year-olds, 55 percent the average for the (about half the EAP are in school; after in school (lower than 4-year-olds, 43 percent are in school; for region; in contrast to average); after enrollment enrollment gap the regional average are in school; for ages 7–12, almost all most other classes, gap closes for 7- to 12-year- closes, it reemerges of 28 percent); for ages ages 7–12, 99 percent children (97 percent) enrollment rates are olds, it reemerges by by adolescence—for 7–12, 98 percent (over are in school; and in are enrolled; and in relatively constant adolescence—for those those ages 13–15, the regional average) adolescence (ages adolescence (ages over the life cycle; ages 13–15, 75 percent 83.6 percent are in are in school; and in 13–15), 94 percent 13–15), enrollment rates among 3- to 4-year- are in school (compared school. adolescence (ages of the economically are virtually constant at olds, 86 percent are with the EAP average of 13–15), in contrast secure are in school. 97 percent, in contrast enrolled; for ages 7–12, 88 percent). to the extreme and to the lower classes. almost all are enrolled moderate poor, (99 percent); and in children of this group adolescence (ages are no longer behind 13–15), enrollment rates the average, with are almost constant 90 percent in school. (98 percent). table continues on next page TABLE D.1  Key socioeconomic characteristics of the various economic classes (continued) Economically Extreme poor Moderate poor vulnerable Economically secure Middle class Rich Access to basic 33 percent lack access to 24 percent of the 20 percent of the 16 percent lack 6 percent lack access to 4 percent lack access to services clean water; 36 percent moderate poor lack economically access to clean water; clean water; 0.9 percent clean water; 0.3 percent lack access to improved access to clean water; vulnerable lack access 6 percent lack access to lack access to improved lack access to improved sanitation (compared with 27 percent lack access to clean water; 19 improved sanitation. sanitation. sanitation. 17 and 15 percent for EAP, to improved sanitation. percent lack access to 24 percent lack access 10 percent lack access 5 percent do not respectively). improved sanitation. 54 percent lack access to proper housing. to proper housing. have access to proper 65 percent lack access to to proper housing. 42 percent do not housing. 96 percent have access 98 percent have access proper housing (33 percent have access to proper 87 percent of the to electricity. to electricity. 99 percent have access EAP average). housing. moderate poor have to electricity. 78 percent have access access to electricity. 91 percent have access to electricity (91 regional to electricity. average). Employment For youth ages 15–24, For youth ages 15–24, For youth ages 15–24, For youth ages 15–24, For youth ages 15–24, For youth ages 15–24, 32 percent are neither 29 percent are neither 22 percent are neither 15 percent are neither 9 percent are neither 10 percent are neither in school nor gainfully in school nor gainfully in school nor gainfully in school nor gainfully in school nor gainfully in school nor gainfully employed (compared employed. The employed. The employed. The employed. The employed. The with the EAP average of unemployment rate unemployment rate (for unemployment rate (for unemployment rate (for unemployment rate 19 percent). (for those ages 15–64 those ages 15–64 who those ages 15–64 who those ages 15–64 who (for those ages 15–64 who are no longer in are no longer in school) are no longer in school) are no longer in school) who are no longer in school) is 4 percent is 3 percent. is 4 percent. is 4 percent. school) is 3 percent. (4 percent of EAP on Self-employment Self-employment Self-employment rates Self-employment rates average). rates among the rates among the are about one-third. equal about 15 percent. extreme poor, the economically secure moderate poor, and are equal to about one- the economically third. vulnerable are about 50–60 percent. Source: World Bank, East Asia and Pacific Team for Statistical Development. Note: Data on the rich refer to the group living on more than US 50 2011 PPP a day. Given the overall small coverage of this group in household surveys, these data are presented for comparative purposes only. EAP = East Asia and Pacific. S T A T I S T I C A L A P P E N D I X D    123 ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage 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, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (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), processed chlorine–free (PCF), or enhanced elemental chlo- rine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. D eveloping East Asia and Pacific has led the way in showing how rapid and broadly shared growth can lift millions out of poverty. As this book shows, the region has achieved even more: the wave of prosperity across the region since the 1980s has lifted three out of five of its citizens into economic security, where their risk of falling into poverty is minimal. And a solid middle class has emerged in most countries. These successes, however, do not guarantee that inclusive growth— growth that reduces poverty and delivers upward mobility and economic security for all—is ensured. The region has become more diverse, with progress varying across countries and extreme poverty increasingly concentrated among specific groups. Roughly one-fifth of the region’s population remains at risk of falling into poverty, and prospects for upward mobility are increasingly elusive across the income distribution, reflecting a growing concentration of income and wealth and limited access to basic social services. Challenges old and new, including rapid aging and less certain growth prospects, are also increasing the premium on economic security for all. Riding the Wave is about how countries across the region can effectively confront these challenges and achieve inclusive growth.  iding the Wave provides a comprehensive, nuanced, and accessible “R analysis of East Asia’s remarkable success in lifting hundreds of millions of its citizens out of poverty. Looking forward, the report develops an eclectic framework, demonstrating how factors such as economic security, structural change, urbanization, and demographics will shape future trends. The authors are to be congratulated for this must-read analysis, which will be of great value to policy makers, scholars, and anybody interested in international development challenges.” —Hal Hill, H. W. Arndt Professor Emeritus of Southeast Asian Economies, College of Asia and the Pacific, Australian National University The report offers a novel analytical approach that enhances our “ understanding of the region’s dynamics of economic inequality, social mobility, and security. The discussion on implied policy measures is greatly beneficial to both academics and policy makers. A must-read indeed for anyone interested in the subject.” Dr. Somchai Jitsuchon, Research Director, Inclusive Development, — Thailand Development Research Institute, Bangkok ISBN 978-1-4648-1145-6 SKU 211145