A P OV ERT Y A SSESSME NT FO R T H E PH I LI PPI NE S Making Growth Work for the Poor A POV ERT Y A SSES SME NT FO R T H E PH I LI PPI NE S © 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 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this publication is subject to copyright. Because the World Bank Group encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: World Bank. 2018. Making growth work for the poor : a poverty assessment for the Philip- pines . Washington, D.C. : World Bank Group. https://hubs.worldbank.org/docs/imagebank/pages/docprofile.aspx?nodeid=29892868 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 individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such in- fringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Cover photo: Soraya Ututalum and ©World Bank. Further permission required for reuse. Cover design: Bianca Canoza Contents Foreword x Acknowledgements xii Acronyms and Abbrevations xiii EXECUTIVE SUMMARY 2 Growth and Poverty in the Philippines 2 Drivers of Poverty Reduction 4 Slower Progress Compared to Many Other East Asian Countries 5 Suggested Measures to Support Faster Poverty Reduction 9 CHAPTER ONE: Poverty Levels and Trends 17 Economic Growth and Challenges Over the Past Decade 18 Poverty and Inequality 22 Poverty in the Philippines: An International Comparison 26 Forces That Have Reduced Poverty 27 Reasons Why Poverty Has Not Declined as Fast as in Other East Asian Countries 29 CHAPTER TWO: Profile of Poverty and Inequality in Living Conditions 34 Characteristics of the Poor 35 Locations of the Poor 42 Non-Income Dimensions of Poverty 44 Education and Learning 44 Access to Health Services 44 Access and Quality of Basic Services 46 Sources of Household Income 49 Vulnerability to Disaster 50 Costs of Conflict 51 CHAPTER THREE: Labor Market Performance 55 Sector and Status of Employment of the Poor 56 Labor Market Status of Various Groups 58 Variation by Region 60 Variation by Area 61 Variation by Gender 61 Education and Labor Market Status 62 Minimum Wage in the Philippines 66 Returns to Education 67 CHAPTER FOUR: Interplay Between Income and Human Capital Accumulation 72 Disparities in Access and Quality of Education and Health Care 73 Education 73 Health Care 78 Vicious Cycle of Inequality of Income and Inequality of Education 82 Public Education Spending Trends 84 Efficiency and Effectiveness in Spending 84 Private Spending on Education and Disparities 88 iii Vicious Cycle of Inequality of Income and Inequality of Health Care 88 Expansion of Health Care Coverage 90 Public and Private Spending on Health 92 Vicious Cycle of Inequality from the Start of Life 93 Inequality in Child Malnutrition Outcomes 95 Costs of Child Malnutrition 96 CHAPTER FIVE: Role of Private and Public Transfers on Poverty and Inequality 101 Patterns and Distribution of Domestic and Foreign Remittances 102 Domestic Remittances 102 Foreign Remittances 105 Impact of Remittances on Poverty and Human Capital 107 Poverty Reduction 107 Other Impacts 107 Overview of Social Protection Programs in the Philippines 108 Impact of Pantawid Pamilya on Poverty Reduction 110 Impact of Pantawid Pamilya on Human Capital Building 115 CHAPTER SIX: Constraints on Poverty Reduction and Potential Policy Remedies 118 Constraints on Poverty Reduction 119 Potential Policy Remedies 120 References 129 ANNEX A: Intersectoral Labor Allocation 139 ANNEX B: Income Structure of Agriculture Households and Agriculture Sector Income and Employment Shares 141 ANNEX C: Natural Disaster Risk Effects on Investment 145 ANNEX D: The Poor Suffered Greater Loss of Well-Being for any Given Asset Loss 147 ANNEX E: Seasonality and Employment Dynamics 149 ANNEX F: Minimum Wage 153 ANNEX G: Returns to Education 154 ANNEX H: Pro-Poor Health Policies Introduced by the Government, 2005–2015 159 ANNEX I: Benefit Incidence Analysis 160 ANNEX J: Impact of Remittances 162 ANNEX K: Impact Evaluation Designs for the Pantawid Pamilya 167 ANNEX L: Impact of Conditional Cash Transfers on Remittances 169 iv Boxes Box 1.1. Poverty estimates using national and international poverty lines 22 Box 2.1. The poorest agriculture households 41 Box 2.2. Poverty among informal settlement families 43 Box 2.3. Poor informal settler families suffer from lack of adequate access to basic services 49 Box 2.4. The poor suffered greater loss of well-being for any given asset loss 52 Box 2.5. Vicious cycle of conflicts and poverty 53 Box 3.1. Minimum wage is high in the Philippines, but its coverage is low 66 Box 4.1. Determinants of fertility rate 83 Box 4.2. Public health spending and health care services 93 Box 4.3. Drivers of malnutrition 94 Box 4.4. High costs of childhood malnutrition in the Philippines 97 Box 5.1. Data sources for remittances 103 Box 5.2. The Pantawid Pamilyang Pilipino Program and Listahanan 109 Figures Figure 1.1. Economic growth in the Philippines 18 Figure 1.2. Regional growth disparities 20 Figure 1.3. Annual growth of household consumption per capita 20 Figure 1.4. Varying trends in per capita household income growth across the population 21 Figure 1.5. Poverty rate and number of the poor 23 Figure 1.6. Prosperity improvement in the Philippines compared with the East Asia and Pacific Region 24 Figure 1.7. Inequality of income 24 Figure 1.8. International comparison: share of wealth for the richest one percent 25 Figure 1.9. Growth and Poverty 25 Figure 1.10. Self-rated poverty and hunger in households 26 Figure 1.11. Decline of US$3.20/day poverty rate for each 1 percent increase in GDP per capita 27 Figure 1.12. Age-sex pyramid of household population, 2010 27 Figure 1.13. Contribution of income sources to poverty reduction, 2006–2015 28 Figure 1.14. Millions shifted out of agriculture 28 Figure 1.15. Greater earnings in manufacturing and services than in agriculture 29 Figure 2.1. Household size, poor versus non-poor (percentage of households) 35 Figure 2.2. Total fertility rate by income quintile 35 Figure 2.3. Average number of children, poor versus non-poor 36 Figure 2.4. Poverty rate by number of children 36 Figure 2.5. Poverty rate by age of household head 36 Figure 2.6. Poverty rate by gender of household head 37 Figure 2.7. Income sources by gender of household head 37 Figure 2.8. Educational attainment of household head 37 Figure 2.9. Poverty rate by educational attainment of household head 37 v Figure 2.10. Inequality of income 38 Figure 2.11. Highest educational attainment of the middle class versus average households 38 Figure 2.12. Employment status of household head 39 Figure 2.13. Poverty rate by class of work of household head 39 Figure 2.14. Share of households mainly relying on the specific income sources 40 Figure 2.15. Employment share by sector 40 Figure 2.16. Share of GDP per capita by main island group to total 42 Figure 2.17. Poverty rate by major island groups 42 Figure 2.18. Poverty share by major island groups 42 Figure 2.19. Poverty rate in urban and rural areas 43 Figure 2.20. Poverty share in urban areas 43 Figure 2.21. Poverty rate of high-conflict regions 43 Figure 2.22. Secondary enrollment and GNI per capita, 2015 45 Figure 2.23. Adult literacy and GNI per capita, 2015 45 Figure 2.24. PISA and TIMSS scores in East Asia and the Pacific 46 Figure 2.25. Health outcomes and services 47 Figure 2.26. Rates of stunting by gross national income per capita 48 Figure 2.27. Components of household income, 2015 50 Figure 3.1. Poverty rate by employment sector of household heads 56 Figure 3.2. Poverty rate by employment status of household heads 56 Figure 3.3. Poverty rate by employment sector of household heads 57 Figure 3.4. Poverty rate by employment status of household heads 57 Figure 3.5. Poverty shares by employment status of household heads in urban areas 57 Figure 3.6. Poverty shares by employment status of household heads in rural areas 57 Figure 3.7. Working-age population, labor force, and employment growth 59 Figure 3.8. GDP, employment, and real wage growth 59 Figure 3.9. Share of employment of the poor by sector 59 Figure 3.10. Share of employment of an average Filipino by sector 59 Figure 3.11. Changes in the composition of employment status over time 60 Figure 3.12. Labor market indicators, urban/rural 61 Figure 3.13. Daily wage, urban/rural 61 Figure 3.14. Labor market participation and employment ratios 61 Figure 3.15. Daily earnings, 2006–2015 61 Figure 3.16. Share of labor force with each grade completed by region 63 Figure 3.17. Wage growth by education level 63 Figure 3.18. Wage growth by wage quintile 63 Figure 3.19. Real daily wage in 2006–2012 by wage income groups 64 Figure 3.20. Real daily wage in 2012–2015 by wage income groups 64 Figure 3.21. Real daily wage in 2006–2012 by education groups 64 Figure 3.22. Real daily wage in 2012–2015 by education groups 64 Figure 3.23. Youth (20–29 years old) education level across income groups 65 Figure 3.24. Share of youth not in employment, education, or training 66 Figure 3.25. Rate of return for education by education level 68 Figure 3.26. Rate of return for another year of education 68 Figure 3.27. Rate of return for education by island group 68 vi Figure 3.28. Rate of return for education by additional years of schooling 69 Figure 3.29. Marginal effects of probability of wage employment with an additional year of schooling 69 Figure 3.30. Marginal effects of probability of wage employment with an additional year of schooling by education level 69 Figure 4.1. Government expenditure on education to GDP ratio, compared with other countries 73 Figure 4.2. Government expenditure on health to GDP ratio, compared with other countries 74 Figure 4.3. Net enrollment rate by level, 2006 and 2015 (all) 75 Figure 4.4. Net enrollment rate by level, 2006 and 2015 (poorest and richest quintiles) 75 Figure 4.5. Educational attainment rate among 22–24 year olds by income quintiles, 2006 and 2015 77 Figure 4.6. Educational attainment rate among 22–24 year olds by gender, 2006 and 2015 77 Figure 4.7. Reasons for not attending elementary school among 6–11 year olds in the poorest quintile, 2014 77 Figure 4.8. Reasons for not attending high school among 12–15 year olds in the poorest quintile, 2014 77 Figure 4.9. Immunization coverage of children, 2008 and 2013 79 Figure 4.10. Unmet needs among reproductive-age women, 2008 and 2013 80 Figure 4.11. Use of health facilities for antenatal care, by quintile and insurance status, 2008 and 2013 80 Figure 4.12. Quality of antenatal care, 2008 and 2013 81 Figure 4.13. Skilled birth attendance, 2008 and 2013 81 Figure 4.14. Infant mortality and under-five mortality rates 83 Figure 4.15. Total fertility rates 83 Figure 4.16. Adolescent fertility rates in EAP countries 83 Figure 4.17. Government basic education spending 84 Figure 4.18. Public spending and benefit incidence 86 Figure 4.19. National achievement test of primary and secondary education —mean percentage scores 87 Figure 4.20. Household monthly education expenditure by quintile of per capita total household expenditure in 2015 88 Figure 4.21. Per school-age child, tuition and non-tuition educational expenses by quintile of per capita total household expenditure in 2015 88 Figure 4.22. Health spending in the Philippines against international comparators, 2014 89 Figure 4.23. Household out-of-pocket spending on health as share of total health spending, 1995–2014 90 Figure 4.24. Household spending share for health by quintiles, 2009, 2012, and 2015 90 Figure 4.25. Health insurance coverage by quintile, 2008 and 2013 91 Figure 4.26. Impoverishing impact of health spending by quintile, 2009 and 2015 92 Figure 4.27. Malnutrition trends in the Philippines for children under five 93 Figure 4.28. Rates of malnutrition are highest in poor areas: Under-five stunting versus poverty rates in the Philippines by province 96 Figure 4.29. Rates of stunting for children under-five by urban/rural and gender, 2013 96 Figure 4.30. Rates of stunting for children under five, by income quintile, 2013 96 Figure 4.31. Rates of stunting for children under five by age group, 2013 96 Figure 4.32. Rates of return to investments to reduce stunting, by country 97 vii Figure 5.1. Foreign and domestic remittances 102 Figure 5.2. Incidence of domestic remittances by income quintile 103 Figure 5.3. Share of households with domestic remittances 103 Figure 5.4. Share of domestic remittance income to total household income 103 Figure 5.5. Share of households with domestic remittances by island group 104 Figure 5.6. Share of domestic remittance income to total household income by island group 104 Figure 5.7. Share of households with foreign remittances 105 Figure 5.8. Share of foreign remittance income as total household income 105 Figure 5.9. Share of households with foreign remittances by island group 106 Figure 5.10. Share of foreign remittance income in total household income by island group 106 Figure 5.11. Distribution of foreign remittances by income quintile 106 Figure 5.12. Annual foreign remittances by income quintile 106 Figure 5.13. Differences in consumption patterns by remittance-recipient status 108 Figure 5.14. Budget for social protection 110 Figure 5.15. Coverage of the poor 111 Figure 5.16. Distribution of program beneficiaries 111 Figure 5.17. Percentage of beneficiaries in the bottom 20 percent 111 Figure 5.18. Generosity per quintile 111 Figure 5.19. International comparison: generosity 112 Figure 5.20. Distribution of benefits, by income group 112 Figure 5.21. Progressivity of the Pantawid Pamilya 112 Figure 5.22 Distribution of beneficiaries, by poverty status 113 Figure 5.23 Distribution of benefits, by poverty status 113 Figure 5.24. Impact on national poverty rate 114 Figure 5.25. Impact on national income gap 114 Figure 5.26. Impact on national income inequality 114 Figure 5.27. Impact on poverty rate among beneficiaries 114 Figure 5.28. Impacts on school enrollment 115 Figure 5.29. Impacts on school attendance 115 Figure 5.30. Reported utilization of Pantawid cash grants 116 Figure A.1. Intersectoral labor allocation in selected East Asian countries 139 Figure A.2. Intersectoral labor reallocation in the Philippines 140 Figure B.1. Components of Agriculture Household Incomes 141 Figure B.2. Employment shares of agricultural household members by sub-sector activity, 2015 144 Figure D.1. Disaster losses in Manila from a once-every-25 year typhoon 147 Maps Map 1.1. GDP per capita by region 31 Map 1.2. Poverty rate by region 32 Map 3.1. Annual percentage reduction or increase in stunting rate, 1995–2015 70 Map 4.1. Secondary education enrollment by province 98 Map 4.2. Tertiary education enrollment by province 99 viii Tables Table 1.1. Top ten countries in East Asia by GDP growth and GDP per capita growth, 2006–2015 19 Table 1.2. Multiples of poverty line by national standards 23 Table 1.3. Poverty rates in selected East Asia countries 26 Table 2.1. Poverty rate of rural households with agriculture and remittances as main sources of income, 2015 40 Table 2.2. Poverty rate for regions prone to earthquakes 50 Table 2.3. Poverty incidence of high-conflict regions 53 Table 3.1. Employment and earnings status 58 Table 3.2. Labor force participation by region, average 2006–2015 60 Table 3.3. Employment, unemployment, and daily earnings, by educational attainment 62 Table 4.1. Public education spending and its share of total government expenditures, 2016 85 Table 4.2. Stunting rate in the Philippines for children under five by region, 2015 95 Table 5.1. Regions with highest and lowest coverage of domestic remittances, 2015 104 Table B.1. Components of household income 142 Table B.2. Breakdown of income shares by household 143 Table E.1. Employment Categories 149 Table E.2. Characteristics of the employed 150 Table E.3. Employment classification by sector 151 Table G.1. Returns to education to another year of schooling (OLS) 155 Table G.2. Returns to education by education level 156 Table G.3. Returns to education to another year of schooling (OLS) - island groups 156 Table G.4. Returns to education by educational level (OLS) – island groups 157 Table G.5. Marginal probability for wage employment by years of schooling (Probit) 158 Table G.6. Marginal probability for wage employment by education level (Probit) 158 Table J.1: Regression estimates for labor supply indicators (Adults 18 – 64 years old) 163 Table J.2: Regression estimates for school attendance (Children 5 – 17 years old) and child labor (Children 5 – 14 years old) 164 Table J.3: Regression estimates for Household Spending Patterns (as Share of Total Expenses) 165 Table L.1: Effects of CCT on remittances 170 ix Foreword To design the policies needed to help accomplish this goal, we need to understand not only the characteristics of the poor—who they are, where they are, and what their income sources are—but also the drivers of and impediments to poverty alleviation. We need to understand what efforts for poverty reduction have worked and what have not, and why. We need to understand which measures need to be strengthened and which need to be altered to put the country on a faster track to reducing poverty and becoming a middle- class society. By addressing these issues through in-depth analysis, the report Making Growth Work for the Poor: A Poverty Assessment for the Philippines, prepared by the World Bank, supports the poverty reduction efforts of the Philippine government. The report finds that increased wage income and the movement of workers out of agriculture, transfers from government social programs, and remittances from domestic and foreign sources were major forces in the poverty decline over the past decade. These gains were tempered by growth that was slower and had a less pro-poor pattern than in many other East Asian countries, as well as the high inequality of income and wealth and the adverse impacts of natural disasters and conflicts. The report emphasizes the importance of breaking the cycle of inequitable investment in human capital and lack of well-paying job opportunities that trap the poor in poverty, generation after generation. Children from poor households start life at a disadvantage. Malnourished and stunted, with poor access to quality health care, they are less likely to learn the skills they need and fulfill their potential. As adults, therefore, they earn low incomes and cannot afford to invest in their own children. They have little to meet their basic needs and nothing to save against emergencies. Frequent natural disasters buffet the poor, whose limited means to cope and disproportionate suffering push them deeper into poverty. Poverty is a threat to peace. In the parts of the country affected by conflict, where physical assets have been destroyed, families displaced, and human capital eroded, people are trapped in a cycle of conflict and poverty. In addition to the challenges of addressing poverty, the Philippines is hindered by the limited expansion of its middle class. In the East Asia region over 2002–2015, the share of population that is economically secure and middle class increased from just over one- fifth to nearly two-thirds, but the share in the Philippines increased from 37 percent to just 44 percent. The lack of well-paying jobs limited the gains for labor from structural transformation. Every year, 1 percent of the employment shifted out of agriculture, but x most of those workers end up in low-end services jobs. Such limited gains for labor could negatively affect the country’s long-term competitiveness. The report concludes that making the pattern of growth more inclusive and providing more well- paying jobs will be crucial to helping people achieve higher and more stable incomes. It claims that steps to accelerate poverty reduction include creating more well-paying jobs; improving productivity in all sectors, including agriculture; reducing income and wealth inequality through more investments in people and skills development, enhancing the ability of the poor to participate in growth; rebuilding conflict-affected areas; and better management of risks and protection of the vulnerable. We hope this report will stimulate debate on the implications of the poverty trends and profiles for the policy priorities and for accelerating progress on improving the lives of the Filipino people. of conflict and poverty. Mara K. Warwick World Bank Country Director Brunei, Malaysia, Philippines and Thailand xi Acknowledgements The report was undertaken with guidance from Mara K. Warwick (Country Director); Carolina Sanchez (Senior Director, Poverty and Equity Global Practice), Salman Zaidi (Practice Manager, Poverty and Equity Global Practice, East Asia), and Gabriel Demombynes (Program Leader). This report was prepared by a World Bank team led by Xubei Luo (Task Team Leader) with a core team including Sharon Faye Piza (Poverty and Labor Market Analysis); Pablo Acosta, Rashiel Velarde, and Angelo Santos (Remittances and Social Protection); Karima Saleh, Robert Oelrichs, Jewelwayne Salcedo Cain, and Ma. Vida A. Gomez (Health); Gabriel Demombynes (Nutrition); Takiko Igarashi, Raja Bentaouet Kattan, Franco Russo, and Binh Thanh Vu (Education); Kevin Chua and Kevin Cruz (macro and fiscal); Frauke Jungbluth, Hanane Ahmed, and Felizardo Virtucio (Agriculture); Pia Peeters, Matthew Stephens, and Assad Baunto (Conflicts); Makiko Watanabe (Urban); Stephane Hallegatte, Artessa Saldivar-Sali, Lesley Jeanne Cordero, and Brian James Walsh (Disasters), Reno Dewina (International Comparison), Michael Dominic del Mundo and Yanan Li (Research Assistance); and Mildred Gonsalvez, Gia Mendoza, Corinne Bernaldez, Regina Calzado, Veronica De Leon, Joedie Perez (Administrative Support). Thanks to Soraya Ututalum for some of the photos and Bianca Canoza for the report layout. William Hurlbut and Caroline McEuen (Editing), The team would like to thank peer reviewers Professor Erwin Tiongson, Georgetown University, Dr. Ghazala Mansuri, Lead Economist, Poverty Global Practice, and Dr. Emmanuel Jimenez, Executive Director of International Initiative for Impact Evaluation. The team thanks Dr. Rosemarie G. Edillon, Deputy Director General, Philippines National Economic and Development Authority and Dr. Lisa Bersales, National Statistician, Philippines Statistics Authority for useful discussions, and the Philippines Statistics Authority for sharing the data for analysis. The team would also like to thank the members of the World Bank country team, especially Agata E. Pawlowska (Manager, Portfolio and Operations), Birgit Hansl, and Vickram Cuttaree (Program Leaders), Georgia A. Wallen (Country Program Coordinator), and Yolanda J. Azarcon (Senior Operations Officer) for comments and advice throughout the course of this work. Maribelle S. Zonaga (Senior Country Operations Officer), Maria Theresa G. Quinones (Senior Operations Officer) and Maria Loreto Padua (Senior Social Development Specialist) also provided useful comments on this report. xii Acronyms and Abbreviations ANC Antenatal care APIS Annual Poverty Indicators Survey ARMM Autonomous Region in Muslim Mindanao BIA Benefit incidence analysis BPO Business process outsourcing CCT Conditional cash transfer (program) CLHNS Cebu Longitudinal Health and Nutrition Survey CPI Consumer price index DHS Demographic and Health Survey DSWD Department of Social Welfare and Development EAP East Asia and Pacific ECCD Early childhood care and development ESC Education Service Contracting FDS Family Development Sessions FIES Family Income and Expenditure Survey GDP Gross domestic product GNI Gross national income GRDP Gross regional domestic product HFEP Health Facilities Enhancement Program ISF Informal Settler Families LFS Labor Force Survey LGU Local government units LIMC Lower-middle-income-class countries MDG Millennium Development Goals MNCHN Maternal, Neonatal, and Child Health and Nutrition NAT National achievement test NCR National Capital Region NEDA National Economic and Development Authority NEET Not in employment, education, or training NHA National Health Account xiii OCW Overseas Contract Workers OOP Out-of-pocket spending PER Public Expenditure Review PhilHealth Philippine Health Insurance Corporation PISA Programme for International Student Assessment PPAN Philippine Plan of Action for Nutrition PPP Purchasing power parity PSA Philippine Statistics Authority RCT Randomized control trial RDD Regression discontinuity design RHU Rural health units SWS Social Weather Stations TFR Total fertility rate TIMSS Trends in International Mathematics and Science Study UHNW Ultra-high net worth WDI World Development Indicators xiv 1 M A K I N G G ROW T H WO R K F O R T H E P O O R Executive Summary Growth and Poverty in the Philippines Robust growth in the Philippines over the past Despite the generally good economic performance, decade has helped to reduce the national poverty poverty remains high and the pace of poverty rate. Over 2006–2015, annual GDP growth averaged reduction has been slow compared with other 5.4 percent, up from 4.1 percent in 1996–2005 and 3.4 East Asian countries. Growth was slower and less percent in 1986–1995. With the population growing inclusive than in other high-performing countries at a relatively rapid 1.7 percent a year, this amounts in East Asia, and poverty reduction lagged (Table to about 3.8 percent in per capita GDP growth. The 1). Between 2006 and 2015, the Philippines poverty pace of poverty reduction has picked up recently. rate, as measured by the international poverty line The national poverty rate fell to 21.6 percent in 2015, (US$1.90/day), declined only 0.9 percentage points declining by an average of 1.2 percentage points per per year compared to 2–2.5 percent points in China, year over 2012–2015 compared with 0.6 percentage Indonesia, and Vietnam, and, as measured by the points per year over 2006–2015 (Figures 1 and 2). lower-middle-income-class poverty line (US$3.20/ Figure 1. National poverty rates and number of poor Figure 2. Poverty trends based on national and international poverty lines M A K I N G G ROW T H WO R K F O R T H E P O O R 25 30% 45% 40% 25% 20 35% 20% 30% Million Inidividuals 15 25% 15% 20% 10 10% 15% 5 10% 5% 5% 0 0% 0% 2006 2009 2012 2015 2006 2009 2012 2015 Number of Poor Poverty Rate National International Lower-middle-income-class Source: Staff estimates using various rounds of the Source: Staff estimates using various rounds of the Family Income and Expenditure Survey Family Income and Expenditure Survey 2 Table 1. Poverty rate in selected East Asian countries US$1.90/day US$3.20/day (international poverty line) (lower-middle-income-class poverty line) Decline per Decline per Country 2016a 2015b 2006a 2015b year year Thailand 0.7 0.0 0.1 6.2 1.1 0.7 China 18.8 1.9 2.4 43.5 20.2 3.3 Vietnam 19.5 2.8 2.1 51.3 11.6 5.0 Indonesia 27.5 7.5 2.2 65.6 34.0 3.5 Philippines 14.5 6.6 0.9 38.4 27.0 1.3 Source: Staff estimates. a. Data for Thailand are for 2006 and 2013, for China 2005 and 2012, for Vietnam 2006 and 2014. The Philippines uses income as the welfare measures, other countries use consumption. day), it declined only 1.3 percentage points per year, policies and changes to health insurance coverage compared with 3–5 percentage points for the same have resulted in increased use of health services, but three countries. In 2015, 22 million Filipinos—more household spending on health remains high and the than one-fifth of the country’s total population— quality of health services is uneven. Health outcomes still lived below the national poverty line. Even at for the poor have improved little, with high infant higher levels the Philippines lagged. The share of mortality rates, particularly among the poor, as well the population with per capita income above the as high fertility rates and high child malnutrition global middle-income line of US$15/day was only 9.2 rates that have not improved since the early 2000s. percent in 2015. This is small compared to Malaysia (65.7 percent), Thailand (35.4 percent), and China The Philippine government has formulated (19.4 percent). strategic plans focused on reducing poverty and improving the living conditions of its people to Other socioeconomic indicators show some meet these challenges. It has launched AmBisyon progress, but significant challenges remain. School 2040 (Philippines, NEDA 2016), a long-term vision enrollment has notably increased in recent years and to bring down poverty and improve the lives of is now at levels similar to those of other countries the poorest segments of the population, and the at the Philippines’ level of income. Recently, the Philippine Development Plan 2017–2022, the blueprint Philippines education cycle added universal and for the country’s development. Together, these M A K I N G G ROW T H WO R K F O R T H E P O O R mandatory kindergarten as well as two years of documents set ambitious goals and set out a plan for senior high school. Yet student learning lags behind the future with a central aim to convert the country that of many East Asian countries. The dropout rate into a prosperous middle-class society whose beyond elementary, particularly among the poor “people will live long and healthy lives, be smart and remained an important challenge. Access to clean innovative, and will live in a high-trust society.” To water and sanitation and electricity have improved achieve this vision, the government aims to reduce but remain unevenly distributed between the poor poverty to 13–15 percent by 2022, which will require and non-poor. Poor informal settler families, in annual poverty reduction of over 1 percentage point. particular, suffered from lack of access to basic services. At the same time, social safety nets were expanded to cover most of the poor, while the generosity of the cash grants were limited. Pro-poor 3 agricultural wages were the main contributor, This poverty assessment aims to inform these accounting for over 50 percent of the reduction in efforts through in-depth study of the varying poverty. While most of the poor in the Philippines impacts of growth on the living conditions of the continue to work in agriculture, data from the people of the Philippines. The report draws from a Labor Force Survey indicate that the share of the variety of sources—including the Family Income and population in primary production agriculture Expenditure Survey and the Labor Force Survey—to declined by nearly 1 percentage point each year, examine what has worked and what has not worked from 36 percent in 2006 to 28 percent in 2015. As in efforts to reduce poverty, as well as to identify even lower-end industry and services jobs paid more ongoing challenges. It identifies key elements that than agriculture jobs, those who shifted to non- affected the inclusiveness of growth, and provides agricultural jobs improved their circumstances. The policy suggestions to address the main constraints gradual movement of employment out of primary to accelerated poverty reduction and greater shared production agriculture and accompanying increase prosperity. in agricultural wages and for unskilled labor in recent years were therefore key drivers of poverty reduction. Drivers of Poverty Reduction Transfers from government social programs contributed about 25 percent of the reduction. The national conditional cash transfer program, Pantawid As illustrated in Figure 3, from a long-term Pamilya, expanded rapidly during this period, and perspective, the main forces powering the decline in became the primary government social assistance poverty between 2006 and 2015 were: program for the poor. It extends cash grants to 77 percent of poor households and contributes both • An increase in wage income and movement of to reducing poverty and to building human capital. employment out of agriculture; Estimates indicate that the program reduced the • Government transfers; and national poverty rate by up to 1.5 percentage points, • Remittances from domestic and foreign sources. lifting 1.5 million people out of poverty. This is consistent with global experience with the impact The increase in wage income and movement of of social safety net transfers on poverty. The World workers out of agriculture contributed about Bank report, The State of Social Safety Nets 2018, two-thirds of the poverty decline. Higher non- estimates that such transfers reduce the incidence Figure 3. Contribution of income sources to poverty reduction, 2006–2015 M A K I N G G ROW T H WO R K F O R T H E P O O R 2 Change in Poverty Incidence 0 -2 -4 -6 -8 Non-agriculture Government Domestic Agriculture wage Non-agriculture Foreign remittance Others Agriculture wage transfers remittance enterprise enterprise International poverty Lower-middle-income-class poverty Source: Estimates using various rounds of the Family Income and Expenditure Survey. International poverty is defined as household income per capita below US$1.90 a day (2011 PPP), and lower-middle-income-class poverty is defined as household income per capita below US$3.20 a day. 4 the non-poor, a high share comes from business. Its overall negative contribution to the aggregate poverty reduction observed in the Philippines during 2006–2015 may reflect the diverse nature of the work. Nevertheless, entrepreneurial income from agriculture-related activities is an opportunity to reduce rural poverty if efforts are made to address productivity constraints, access to finance, extension, and climate change. of international poverty by 36 percent. Moreover, even if the transfers do not lift beneficiaries above the international poverty line, they reduce the poverty gap by about 45 percent. Pantawid Pamilya also helped influence behavior change: it improved Slower Progress Compared school enrollment of older children, encouraged early childhood education, and increased the health- to Many Other East Asian seeking behaviors of beneficiaries. Countries Remittances from domestic and foreign sources contributed about 12 and 6 percent respectively. While the causes are complex, careful analysis Two-thirds of Filipinos (15 million households) of what has been holding the Philippines back receive domestic or foreign remittances. Foreign compared to many other East Asian countries points remittances are much higher in value; however, both to the pivotal role of three factors: transfer types have similar impacts on reducing the M A K I N G G ROW T H WO R K F O R T H E P O O R poverty rate (by up to 4 percentage points). This is • The lower pace and less pro-poor pattern because domestic remittances are more prevalent of growth than in many other East Asian among the poor, while foreign remittances, though countries; greater in value, are more common among the non- • High inequality of income and wealth; and poor. • The adverse impacts of natural disasters and conflict. By contrast, the contribution of entrepreneurial incomes to poverty reduction was a negative The lower pace and less pro-poor pattern of 15 percent. Entrepreneurial activities are economic growth: The annual growth rates of GDP varied: for poor rural households, a high share and GDP per capita in the Philippines—around 5.5 of entrepreneurial incomes typically come from percent and 3.6 percent respectively in 2006–2015— agriculture-related activities; for the urban poor, place the Philippines tenth in the East Asia Region, from self-employed, lower-end services; while for with annual growth rates 4–5 percentage points 5 lower than in China, the region’s top performer. and, worse, contributed to expatriation of highly Moreover, data from household surveys show even educated and skilled workers. In the long run, low slower growth in household income per capita wages for the highly educated and highly skilled are (on average, only 1.6 percent per year). These data likely to damage the economic competitiveness of indicate that less than 10 percent of the country’s the Philippines by making it even more attractive population has made it to the global middle class, for these groups to move abroad in search of more and more than 10 percent of Filipinos remain remunerative opportunities. vulnerable to falling into poverty. By contrast, elsewhere in the region, the economically secure For many other East Asian countries faster growth and middle class comprised nearly two-thirds of the in agricultural productivity has also been a key population in 2015, a significant increase from its driver of poverty reduction. Agriculture, which share of just over a fifth of the population in 2002. employs most poor people in the Philippines, has In the Philippines, the increase was very modest, experienced minimal growth in the past decade, from 37 percent to 44 percent, and its slower contributing to GDP growth by an average of 0.2 progress with poverty reduction is likely linked to percentage points (compared to 1.9 percentage the limited progress achieved in providing economic points for industry and 3.4 percentage points for mobility and growing the middle class. More services) over the period of 2006-15. In addition, rapid growth of the middle class can bring both compared with countries such as China, Indonesia, intrinsic and instrumental benefits—not only does Malaysia, and Thailand, the gains from structural it enjoy greater economic security and higher living transformation, or the shift from agriculture toward standards, but it can also provide instrumental non-agricultural activities, were limited, and growth benefits for others by being a key driver of growth, in labor productivity relied disproportionately an influential constituency for better governance, on within-sector productivity growth. As a and a major provider of income tax revenues needed result, the responsiveness of poverty reduction to to fund poverty reduction and risk-mitigating growth in these countries has been faster than in agendas, as well as investments for growth. the Philippines. In the past decade, the poverty rate in the Philippines, measured by the lower- High-performing East Asian countries built middle-income-class poverty line, declined by booming manufacturing sectors that provide only 0.3 percentage points for each percentage large numbers of labor-intensive jobs, absorbing point of growth of GDP per capita, compared with the surplus low-skilled labor from agriculture declines of 1.0 percentage point in Vietnam and 0.8 and providing them significantly higher wage percentage point in Indonesia. income. The Philippines has not fully developed a manufacturing base, which has placed it at a High inequality of income, wealth, and significant disadvantage. Workers moving out of opportunities: The Gini coefficient in the M A K I N G G ROW T H WO R K F O R T H E P O O R agriculture generally end up in low-end service jobs, Philippines has hovered around 45 percent over which limits the gains for labor from structural the past decade. The Credit Suisse Wealth Report transformation. Wages are a major source of income estimates that the top 1 percent owned more than for most households, so the 4 percent increase half of the nation’s wealth, the fourth highest after overall in real wages (for those reporting wages) the Russian Federation, Turkey, and Hong Kong, over the past decade indicated by the Labor Force SAR China. This high concentration of wealth Surveys helps explain the limited progress with may have contributed to strong vested interests poverty reduction compared to other countries in in the status quo by hindering acceptance of the the region. Furthermore, the even more limited reforms needed to prompt more inclusive growth real wage growth for the better educated, a mere and faster poverty reduction. Differences in the 2 percent for the college educated, probably quality of human resources, and in the incomes hindered the increase in the size of the middle class individuals and households can earn, drive a 6 large degree of the inequality of outcomes in Figure 4. Rates of stunting for children under 5, the Philippines. For instance, the effects of such by wealth quintile (2013) inequality show themselves from the earliest days 50% of life. Poorer children are at a disadvantage from the start. They have limited access to good-quality health care and early childhood education, which undermines their ability to succeed later in life. 40% Twenty percent of children under age five are malnourished and stunted (see Figure 4 for stunting by wealth quintile). Many poor people, including 30% in the younger generations, have limited education. In the labor force, for the poor households, only 31 percent have completed secondary education and 2 percent have benefitted from tertiary education, 20% compared to 59 percent and 15 percent of the non- poor, respectively. Workers with tertiary education earn nearly 4.5 times the wage of workers with no 10% education, and workers that have a high school education earn 1.8 times the wage of those with no education. The low level of education and skills means that the poor cannot compete for productive 0% jobs in the formal sectors, such as high-end services Poorest Second Middle Fourth Wealthiest or business-process outsourcing positions, which Severely Stunted Moderately Stunted require tertiary education. This constrains the total supply of skilled labor, which dampens the business Source: Food Nutrition and Research Institute (2015). environment for investors, perpetuating the cycle Note: The overall stunting rate is broken down into severe stunting (more than 3 standard of inequality of opportunity and inequality of deviations below the median) and moderate stunting (between 2 and 3 standard deviations below the median). outcomes. countries points to the extremely high human The adverse impacts of natural disasters and and economic costs of disasters and conflict, with conflict: Frequent natural disasters, including the poor suffering disproportionately. The same deadly typhoons that disproportionately strike level of asset loss affects poor and marginalized poor regions of the Philippines, and persistent people far more than their wealthier neighbors conflicts in parts of Mindanao continually push because their livelihoods depend on fewer assets vulnerable groups into poverty and jeopardize M A K I N G G ROW T H WO R K F O R T H E P O O R and their consumption is closer to subsistence long-term human capital development. Natural levels. Repeated and increasingly frequent natural disasters have caused an estimated US$23 billion disasters are undermining poverty reduction in in losses and damages in the Philippines since the Philippines. Where they occur, conflicts not 1990, making it one of the most disaster-prone only destroy physical assets, they also erode human countries in the world. Moreover, on average, capital through loss of life, injury, illness, denial more than a million Filipinos are impoverished of education and health services, and increased each year by natural disasters. Protracted conflict, malnutrition. This reduces the earning ability particularly in parts of Mindanao, has exacted and capabilities of the affected populations and a great toll on the local economy and trapped traps people in poverty. The confluence of weak people in poverty. Lack of security and justice and governance, conflict, and migration significantly economic stresses have intertwined to lock people affects the level and quality of initial human in poverty. A review of the experiences of other capital endowments of conflict-affected regions in Mindanao. 7 Closer examination of the profile of the of the poor live in rural areas, and the rural poverty poor clarifies the importance of each of rate is three times that of urban areas. Poverty rates these three factors, as well as how the are lowest in the National Capital Region (only 4 drivers of poverty work together to keep percent), and highest in the areas with risks due to the poor from achieving a better life. disasters and conflicts; two-fifths of the poor live in Mindanao; over 50 percent of the population in the Who are the poor? Poor Filipinos live in relatively Autonomous Region in Muslim Mindanao are poor. large households, with disproportionately low In addition, people living in informal settlements in educational attainment, headed by individuals who urban areas suffer high levels of multidimensional are self-employed or work in agriculture as laborers poverty; almost 40 percent of these informal settlers or smallholder producers. Households that rely on are in Manila. agriculture as the main source of income, such as farmers and fishermen, are the poorest. In 2015, What are their main income sources? Poor over 31 percent of the households with six or more households rely disproportionately on income members were poor, which was 10 percentage points from agriculture (including subsistence farming, higher than the national average. Over 25 percent agricultural wages, and agriculture-related of households headed by someone under 50 years self-employment), domestic remittances, and of age were poor in 2015, compared with less than government transfers. While the share of income 16 percent of those with heads over 50. Female- provided by agriculture has declined over time, headed households, which receive a large share of dependence on wages and salaries from agricultural income from remittances, are less likely to be poor. activities remains high—about two-thirds of Like most countries, the Philippines shows a strong enterprises are agriculture-related. Domestic negative correlation between poverty risk and the remittances and government transfers represent 7 level of education of the household head. High percent and 6 percent of total household income school education stands out as the key threshold— in poor households respectively, compared with 3 graduation reduces the risk of poverty to two-thirds percent and 1 percent in non-poor households. of the average. The drivers of poverty are mutually reinforcing. Where do they live? There are large regional The poor start life at a disadvantage. They are disparities across the Philippines—the regional hobbled by malnutrition, limited resources, poor GDP per capita in the National Capital Region is access to quality health care, and low education and five times that of Mindanao—and in some lagging skills, among other deficits. They earn low incomes, regions growth has slowed recently. Three-quarters M A K I N G G ROW T H WO R K F O R T H E P O O R 8 8 save little for the future, and are vulnerable to position to apply multiple policy tools to seize shocks. Lacking the requisite skills to take advantage opportunities, mitigate regional and global shocks, of job opportunities, the poor are generally limited and provide the basis for productive job creation to agricultural work or other low-paid jobs in rural and poverty reduction. areas; in urban settings, many end up in slums. Making the pattern of growth more inclusive and The vicious cycle of unequitable investment providing more well-paying jobs will help people in human capital and lack of well-paying job to achieve higher and more stable incomes. opportunities traps the poor in poverty generation The government can help end the vicious cycles after generation. The high concentration of income of unequal opportunity and outcomes that trap and wealth limit equality of opportunity and people in poverty, as well as set in place mutually impede equitable public service delivery, which reinforcing positive cycles that will create a growing is necessary for inclusive growth. In addition, middle class, well-integrated with other groups. frequent and severe natural disasters, as well as It can help improve service delivery for all and persistent conflicts in some parts of Mindanao, increase non-farm wage employment opportunities have limited the attractiveness of long-term through increased demand for manufacturing (foreign and domestic) investment, particularly goods and services. Finally, more progressive and investment in infrastructure, which has aggravated better-administered taxes can help finance needed regional disparities. The lack of land reforms and investments in both physical and human capital. unclear property rights have similarly discouraged investment in agriculture. With a low rate of Strong economic growth will be the basis for investment (20 percent of GDP), the economy is productive job creation and poverty reduction. In largely driven by consumption, which limits the the long run, productivity is the basis of everything. potential for more rapid structural transformation Addressing the key factors noted earlier—more well- and improvement in productivity. The low paying jobs; improved productivity in all sectors, increase in real wages, while it might be attractive especially in agriculture; ensuring people acquite the in the short run, is likely to negatively affect the skills they need; investing in health and nutirtion; competitiveness of the economy in the long run focusing poverty reduction efforts in Mindanao; and with the brightest leaving the country for better managing risks and protecting the vulnerable—can job opportunities. The high rate of emigration help achieve faster poverty reduction. from the Philippines, with 6 million Filipino migrants abroad, might indicate that this is already happening. Facilitate the creation of more well-paying jobs. M A K I N G G ROW T H WO R K F O R T H E P O O R A significant share of the poor have jobs with Suggested Measures to very low wages or are mired in involuntary underemployment. In the past decade, employment Support Faster Poverty grew at roughly the same rate as the working-age Reduction population, but a large portion of those jobs are poorly paid. Nearly 95 percent of the population in the labor force is employed. However, some 20 percent is underemployed, and to the extreme, some The Philippines has solid macroeconomic households earn as little as 50–100 pesos (US$1–2) fundamentals and its growth prospects remain a day. Many urban poor are trapped in low-wage, positive. With a healthy current account, strong low-productivity jobs in the informal service sector. international reserves, significant fiscal space, and Support for the creation of more well-paying jobs, low and stable inflation, the economy is in a strong 9 particularly semi-skilled jobs, for the majority of to find its specific niches in the services sector today’s labor force who have less than a high school and in regional and global value chains to education, can help reduce poverty and address capitalize on its growing services sector and inequality through higher wage incomes. enhance the productivity gains from structural transformation. • Improve the business environment to attract more investment. Underinvestment in • Strengthen backward and forward linkages human and physical capital has been a major to build on the comparative advantages of constraint to improved labor productivity skilled labor and create jobs for the unskilled. and has resulted in the low quality and high Linkages between the services sector and informality of jobs. Compared with most manufacturing and agriculture are critical high-performing countries in East Asia, to upgrading the domestic value. This would the Philippines investment-to-GDP ratio is include proficiency in English and good low. Investment in productive capacity, in information technology skills, as well as taking particular, has lagged in the manufacturing advantages of the time zone. In doing so, the sector. To attract more private investment, the Philippines could leverage strong performance business environment needs to be improved, in business-process outsourcing to expand particularly through addressing institutional other service-based sectors, such as tourism. constraints, strengthening competition in key This, in turn, could contribute to successful sectors, securing property rights, providing transformation by creating more productive risk management solutions and simplifying employment opportunities, including business regulations. To attract foreign and opportunities with skill requirements domestic investment, the government can play compatible with those of individuals from a key role by improving infrastructure and poor households. basic services delivery, as well as by providing targeted support to the self-employed or those working in small and medium-size enterprises, Improve productivity in all sectors, where large numbers of the poor are employed. especially agriculture. • Upgrade value chains to support strong The Philippines is a middle-income country whose and sustainable growth. Improve labor economy is becoming less dependent on agriculture productivity and moving up the value chain are for output and employment. Nevertheless, a proven basis for creating more well-paying agriculture remains important for poverty reduction jobs. The Philippines has gone from being an and employment as well as sustainable and equitable agricultural economy to a (low-end) service growth. Compared with many countries in the M A K I N G G ROW T H WO R K F O R T H E P O O R economy, without developing a manufacturing region, the sector performs below its potential sector. Labor productivity growth mainly stems for contributing to growth, employment, and from within-sector productivity growth. This is poverty reduction. Improvements in productivity, contrary to the development patterns of many diversification, and value-addition are crucial, as neighboring countries in East Asia, where well as progress in making agriculture more resilient booming manufacturing sectors created large to natural disasters and climate change. numbers of labor-intensive jobs, absorbing the surplus labor from agriculture. It is an ongoing • Enhance agricultural productivity. Over debate whether manufacturing can still deliver the past decade, productivity growth in the same productivity gains and well-paid the Philippines has lagged that of the best employment opportunities for the unskilled performers in East Asia, including China, workers as in the past. The Philippines needs Indonesia, and Vietnam. Agricultural 10 percent of GDP in Indonesia, 43 percent in Thailand, and 15 percent in the Philippines, which is higher than the agriculture share in productivity has been low and stagnant for 30 GDP). As agriculture’s share of GDP continues years. Farmers and fisherman remain among to fall and incomes and urbanization rise, the the poorest in the rural areas. Reasons for composition of agricultural output changes the persistent low productivity of agriculture as part of agricultural diversification. To include high input costs; small land sizes; reduce poverty in rural areas, support will insufficient ability to manage rainfall be needed for agricultural development variability and other natural hazards; lack and diversification through support for the of access to finance, applied research, and development of agribusiness, bringing in extension services; and limited connectivity various input providers and agro-processors, and links to market outlets. As evidenced in distributors, and retailers for value chain other middle-income countries in the region, development. structural transformation will attract workers out of agriculture as the manufacturing and service sectors expand. However, agriculture continues to be a large employer and Ensure that Filipinos acquire the skills absorption of surplus labor by manufacturing they need for the 21st century economy. and service sectors is not undertaken at a In recent years, the Philippines has made admirable M A K I N G G ROW T H WO R K F O R T H E P O O R fast pace, at least in the short run. Improving strides in education. Critical advances have been the income from agriculture will help address creation of both universal kindergarten and senior persistent poverty issues and contribute to high school education, with the first cohort of grade employment opportunities in rural areas. 12 students graduating in 2018. Key challenges now include making sure students in school are learning, • Support agribusiness and broader value reducing high dropout rates for the poor, and chain development. Within the structural developing socioemotional skills. transformation agenda, the role of agriculture is evolving, although slowly. • Boost learning in basic education overall The share of agribusiness in the GDP of and increase secondary enrollment and several countries in the region undergoing completion among the poor. To close gaps structural transformation is higher than that in education, two principle challenges of agriculture (agribusiness accounts for 33 11 remain. The first is that despite a high level a route to higher earnings for workers with of commitment by teachers and improved limited formal education. To take advantage learning environment, learning outcomes of this insight it will be necessary to develop are weak. The Philippines’ experience is teacher preparedness and training to actively similar to that of many countries around the foster these skills in all education and training, world that have boosted school completion including early childhood education, K–12 rates but still face quality challenges, which education, and tertiary education, as well as globally constitutes what the Bank’s 2018 regular and vocational training. World Development Report (World Bank 2017j) terms a “learning crisis.” The second challenge is that secondary enrollment is low Invest in health and nutrition. and dropout rate remain high among the poor beyond primary level. The returns to education Although the Philippines has implemented universal are high at college levels, but many among health coverage, it still has weakness in de facto the poor are not completing high school. health access and quality, rates of child malnutrition Improving education quality principally remain high, and the country has faced obstacles requires equipping teachers with the tools in implementing its reproductive health policies. A they need via effective training and materials. series of efforts in these areas are needed to boost Improvements of quality will help address the human capital and make possible a demographic second challenge, by attracting more students dividend. to stay in school. Other critical priorities are continuing efforts to improve budget • Boost health care quality and equity. The execution and the effective use of public Philippines has made great progress in education funds. Strengthening collection of expanding access to health care through learning outcome data including participation universal coverage via the Philippines Health to the international standardized students’ Insurance Program (PhilHealth). However, the assessments and use of the data to determine scope and quality of care available in public the direction of the ongoing basic education facilities remains limited and uneven. To break reform will be important. the cycle of poor health and poor income, public investment in health care needs to be • Develop socioemotional skills in addition improved to ensure easy access to basic good- to traditional technical skills and cognitive quality care and alleviate the burdens of out- skills. A recent World Bank report shows of-pocket payment. The top policy priority is the growing importance of socioemotional to expand the essential health benefits package skills for competitiveness in the global available to the poor. The next priority is to M A K I N G G ROW T H WO R K F O R T H E P O O R economy. A higher level of socioemotional develop a national strategy for quality of health skills is associated with greater probability care improvement. A third is to ensure that all of being employed and with higher daily of those who qualify for PhilHealth coverage earning. Therefore, worker competitiveness are enrolled and are aware that they are increasingly requires not only traditional insured. Limited and uneven access and quality technical and cognitive skills but also of health care contribute to the general health improved socioemotional skills. Moreover, challenges of the poor as well as to weaknesses such skills are associated with the greatest in reproductive health and nutrition as well as wage differential among workers with low general health challenges of the poor. educational levels. As a substitute for, instead of complement to, traditional technical and • Reduce child stunting. One in three children cognitive skills, socioemotional skills can offer in the Philippines under age 5 is stunted— 12 the principal marker of malnutrition—and allow informed parents to make their own stunting rates have been stagnant over a choices and achieve their desired family size. decade, even as other socioeconomic indicators A recent study estimated the economic gains have seen progress. Malnutrition in the womb from a full implementation of the RPRH law and during the first two years of life inhibits and suggested helping couples achieve the brain development, resulting in lower levels desired number of children can potentially of schooling, reduced cognitive function, have substantial economic benefits in terms of and lower earnings later in life. The returns more rapid economic growth. Critical aspects from investments to reduce malnutrition are of the law that need to be fully implemented extraordinarily high in the Philippines: each include expanding access to a wide range of peso invested results in a return of 44 pesos. modern contraceptives, especially for the poor, The Philippine Plan of Action for Nutrition as well as Comprehensive Sexuality Education provides a solid framework for tackling the to reduce teen pregnancies. challenge. The critical needs are to focus health interventions on the “first 1000 days” of a child’s life from conception through the first Focus poverty reduction efforts on two years of life, combined with multi-sector Mindanao. efforts involving education, social protection, agriculture, and water and sanitation. As the region is home to two-fifths of the poor, little progress on poverty is possible without inclusive • Fully implement the Responsible Parenthood growth in Mindanao. Five decades of violence and Reproductive Health (RPRH) Law. Filipino has depressed growth and poverty reduction. women in the poorest quintile have more than Conflict has affected over 60 percent of Mindanao’s five children on average and the fertility rate population. Over 50 percent of the population in has been steady in the past decades. One in ten ARMM lives below the poverty line. Economic girls age 15-19 is either pregnant or already a progress and poverty reduction in the Philippines mother. An increase in adolescent pregnancy will depend on the success of development in means higher maternal and infant mortality, as Mindanao. This will mean drawing on the region’s well as more school dropouts. At a macro level, untapped potential, linking lagging areas to growth the slow decline of fertility has robbed the centers, and strengthening peace-building efforts Philippines the opportunity for a “demographic in conflict-affected areas to break the cycles of dividend” of the sort that has been important insecurity. in economic development across East Asia. The total wanted fertility rate for the Philippines • Increase public investment in Mindanao. is 2.2 births per woman, 27 percent lower Increasing public investment in Mindanao to M A K I N G G ROW T H WO R K F O R T H E P O O R than the actual fertility rate of 3.0 in 2013 boost development in areas where the bulk (recent DHS 2017 shows that actual fertility of the poor live would provide the basis for rate has declined to 2.7 births per woman). generating opportunities. As three-fifths of One important measure is to help households Mindanao’s production and employment is meet their need for contraception. A recent driven by agricultural value chains, investment study based on a natural experiment in Manila is particularly needed to support the agriculture shows that reducing access to contraception sector and improve connectivity. Complementary increases family size and decreases education efforts are needed to build human capital in attainment. Following through on the Mindanao and strengthen local governance. commitments of the 2012 RPRH Law will 13 • Support efforts to resolve conflict and targeted support to the poor and vulnerable to bring peace to Mindanao. Breaking the mitigate shocks, build up human capital, and cycle of insecurity and reducing the risk of provide an effective safety net for those times its recurrence requires a virtuous spiral of when it is needed, is crucial. Managing risks and restoring confidence in collective action protecting the vulnerable not only protects public between groups who have been in conflict investments in individuals and private assets, it also and transforming institutions to provide supports broader growth and capital accumulation a sustained level of security, justice, and through reducing repeated losses of physical jobs. This can be accomplished in three and human capital, and through increasing the steps: 1) Creating productive employment acceptable thresholds of natural risks for investors. opportunities, particularly for youth, who might otherwise be tempted to join extremists’ • Improve natural disaster risk management armed groups or organized crime; 2) Delivering systems. Poor people are more exposed to government programs and basic services more negative shocks—they are more likely to live effectively, which could anchor stabilization; in flood-prone areas in fragile housing, with and 3) Increasing programs to build human a large share of their meager income spent on capital by expanding coverage of basic staples—and are more vulnerable given their services, including health, education, and skills lack of capacity for prevention and limited development. Ultimately, enduring peace and ability to cope with and recover from shocks. development will hinge on the success of a Effective disaster prevention measures can political solution that addresses the causes yield high returns, especially when they are of violence—injustice, weak governance, correctly designed and implemented as part land dispossession, discrimination, and of a larger program of poverty reduction. sociocultural marginalization. Early warning systems, improved access to personal banking, insurance policies, and social assistance (such as cash transfers and public Manage risks and protect the works programs) can improve the capacity vulnerable. of individuals to cope with and recover from shocks and avoid well-being losses three-to-five Poor people are more vulnerable to negative shocks. times greater than their costs. Development They are more exposed to the risks through lack of post-disaster support systems, including of resources, more sensitive to the impacts due social safety nets, remittances, insurance, to an inability to cope with them, and lack the and other financial instruments can mitigate capacity needed to adapt to potential risks and the well-being losses of the poorest Filipinos therefore suffer repeated setbacks. Children from from natural disasters, even without directly M A K I N G G ROW T H WO R K F O R T H E P O O R poor families are particularly vulnerable not getting reducing asset losses. the needed education and health care. Providing 14 • Strengthen social protection systems. The Pantawid Pamilya conditional cash program keeps up with the evolving needs of transfer program has helped to provide poor poor beneficiaries, several improvements need households with much-needed financial to be considered. First, targeting efficiency augmentation to meet basic needs, and it has can be improved through regular updating provided an incentive to keep poor households’ of the roster of potential conditional cash children in school and healthy. It is important transfer beneficiaries in the Listahanan and to continue the cash assistance to poor cover by using the most updated database. Second, all poor households with children, and increase to strengthen the impact on building human the amount of transfers to sustain and enhance capital, it is important to move beyond access the gains, and to keep the convergence of to measure and monitor quality (that is, M A K I N G G ROW T H WO R K F O R T H E P O O R government efforts—in raising demand- monitor learning as well as school attendance, side pressures and supply-side responses—to and measure improved nutrition as well as maintain the program’s effectiveness in growth). achieving outcomes. To ensure that the 15 M A K I N G G ROW T H WO R K F O R T H E P O O R 16 C HAPTE R ON E Poverty Levels and Trends • Economic growth has been strong in the Philippines over the past decade. The national poverty rate declined 5 percentage points from 26.6 percent in 2006 to 21.6 percent in 2015. • The pace of poverty reduction was slower than many countries in the East Asia Region. In 2015, 22 million Filipinos, or more than one-fifth of the population, remained poor. Only a small share of the population has made it into the middle class and more than 10 percent remains vulnerable to falling back into poverty. Inequality of income and wealth remained high. • Recent improvement in inclusiveness and poverty reduction are reasons for hope. The poverty rate declined an average of 1.2 percentage points per year over 2012–2015, compared with 0.6 percentage points over 2006–2015. • Key drivers of poverty reduction include an increase in wage income, movement of workers out of primary production agriculture, government transfers, and remittances from domestic and foreign sources. Wages and salaries, entrepreneurial income, and transfers account for four-fifths of total household income. • The reasons poverty did not decline as quickly as in it did in many East Asian countries include: the slower pace and less pro-poor pattern of growth, high inequality of income and wealth, and disasters and conflict. M A K I N G G ROW T H WO R K F O R T H E P O O R The Philippine government formulated a strategic NEDA 2016),1 a long-term vision to enable the focus on reducing poverty and improving the living government to bring down poverty and improve conditions of its people. To guide this work, the the lives of the poorest segments of the population, government launched AmBisyon 2040 (Philippines, and the Philippine Development Plan 2017–2022 1 This represents the collective long-term vision and aspirations of the Filipino people for themselves and for the country for the next 25 years. The Vision states that “the Philippines will be a prosperous, predominantly middle-class society where no one is poor; our peoples will live long and healthy lives, be smart and innovative, and will live in a high-trust society.” 17 (Philippines, NEDA 2017), the blueprint for the country’s development. As a step forward in this Economic Growth and effort, the government aims to reduce poverty to Challenges Over the Past 13–15 percent by 2022, which will require annual poverty reduction of over 1 percentage point. Decade The main objective of this Poverty Assessment is to inform the efforts through a more in-depth The Philippines has experienced good economic understanding of the varying impacts of growth growth over the past decade (Figure 1.1A). Over on the living conditions of the people in the 2006–2015, the annual growth in GDP averaged 5.4 Philippines. The report draws from a variety percent, compared with 4.1 percent average growth of sources—including the Family Income and recorded in 1996–2005 and 3.4 percent in 1986–1995. Expenditure Survey and the Labor Force Survey— Underlying this strong performance was a stable to examine what has worked and what has not macroeconomic environment, achieved after a series worked in efforts to reduce poverty, as well as the of economic reforms that included liberalization current challenges. It identifies key elements that programs between 1986 and 1997 to improve affected the inclusiveness of growth and provides competitiveness, financial and corporate regulatory policy suggestions to address the main constraints reforms following the 1997 Asian Financial Crisis, to accelerated poverty reduction and greater shared and fiscal consolidation efforts in 2004–2007 (World prosperity. Bank 2013b). This chapter presents the achievements in economic The Philippine economy increasingly has been growth and poverty reduction over the past decade characterized by higher international reserves, and challenges ahead. It lays out the regional trends healthy current account surpluses, stable inflation, in output and employment, examines the poverty and declining debt ratios. While the GDP growth levels and trends, and provides a comparison with is impressive, the growth of GDP per capita, which other East Asian countries. is more relevant for changes in welfare, was slower, and related to the country’s rapid population growth. The growth rate in per capita terms was Figure 1.1. Economic growth in the Philippines A. Real GDP growth rates B. Real GDP per capita growth rates M A K I N G G ROW T H WO R K F O R T H E P O O R 8% 7% 7% 6% 6% 5% 5% 4% 4% 3% 3% 2% 2% 1% 1% 0% 0% -1% 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: Philippine Statistics Authority (PSA). Source: Philippine Statistics Authority (PSA). 18 Table 1.1. Top ten countries in East Asia by GDP growth and GDP per capita growth, 2006–2015 Average annual Average annual GDP growth per Country GDP growth Rank Country Rank capita (2006–2015) (2006–2015) China 9.60 1 China 9.05 1 Myanmar 9.21 2 Myanmar 8.40 2 Mongolia 8.43 3 Mongolia 6.73 3 Lao PDR 7.95 4 Lao PDR 6.14 4 Timor-Leste 7.02 5 Cambodia 5.30 5 Papua New Guinea 6.97 6 Vietnam 4.99 6 Cambodia 6.96 7 Papua New Guinea 4.57 7 Vietnam 6.12 8 Timor-Leste 4.49 8 Indonesia 5.63 9 Indonesia 4.27 9 Philippines 5.43 10 Philippines 3.80 10 Source: World Bank World Development Indicators (WDI) slower than the aggregate, again due to population recent years, however, growth performance in the growth (Figure 1.1B). The country’s per capita Philippines has been strong. In 2016, its growth rate income grew at an average of 3.6 percent annually, was among the highest in the East Asia Region. while the population grew at 1.8 percent over 2006– 2015. Since 2006, per capita income has continuously Performance is not uniform in the country, however, expanded except for the temporary contraction and there are wide disparities among regions (Map in 2009, due to the impact of the global financial 1.1).2 Of the four regional groups, the NCR and crisis. It has since rebounded, topping more than 4.0 Luzon vastly surpassed Visayas and Mindanao in percent growth in recent years. output production. The disparity between the regional groups is further accentuated, because the Nonetheless, compared with the East Asia and Pacific NCR has been growing faster than the rest of the Region, the gap between the Philippines and the country (Figure 1.2). The NCR is the wealthiest high-performing countries is significant. In 2015, its region in the Philippines, with its real per capita GDP per capita of US$2,635, in constant 2010 prices, income 16 times the Autonomous Region in Muslim M A K I N G G ROW T H WO R K F O R T H E P O O R was only 48 percent of the average for developing Mindanao (ARMM)). countries in the Region (US$5,507). Compared with other countries in East Asia, in the past decade, the Data from household surveys show that household average rates of GDP growth and GDP growth per income per capita growth was even lower for capita for the Philippines both rank tenth in the the Philippines than for other countries’ in the group, some 4-5 percentage points lower than the Region.3 On average, incomes were growing at only stellar performers such as China (Table 1.1). In the 1.6 percent each year over 2006–2015. The bottom 2 GDP per capita varies significantly within a region. In some cases, regional statistics may be driven by a few locations within the region. 3 GDP per capita and household income per capita come from two different sources. Much like the case in other countries, these two numbers are not expected to be the same. The gap between the two can come from multiple factors, including the distribution of the fruits of growth between capital and labor, as well as the possible missing coverage (often the very top end) of household surveys. 19 Figure 1.2. Regional growth disparities A. Per capita regional domestic product B. Cumulative growth of per capita (in 2000 prices) regional domestic product 250 135% 130% 200 Thousands Php constant 2000 125% 150 120% 100 115% 110% 50 105% 0 100% 2009 2010 2011 2012 2013 2014 2015 2009 2010 2011 2012 2013 2014 2015 NCR Luzon Visayas Mindanao NCR Luzon Visayas Mindanao Source: PSA, WB staff computation Source: PSA, WB staff computation 40 percent grew faster by over 1 percentage point was strong and inclusive, but growth in 2009–2012 compared with the overall population (2.9 percent was weak, which may be attributed to the global versus 1.6 percent). Household consumption per crisis, and benefited the rich more than the poor. capita growth was even lower, at 0.8 percent for the Between 2006 and 2009, mean income grew by 5 total population but 2.6 percent for the bottom 40 percent and all percentiles experienced positive percent4 (Figure 1.3) income growth; however, income growth was especially high (11 percent) for the bottom 20 The overall trends mask differences both in and percent of the distribution. Similarly, between 2012 between sub-periods (Figure 1.4). Household per and 2015, all percentiles experienced positive income capita income growth in 2006–2009 and 2012–2015 growth, with an overall increase in mean income of 6 Figure 1.3. Annual growth of household consumption per capita China (2008-2012) Malaysia (2007-2012) Cambodia (2008-2012) M A K I N G G ROW T H WO R K F O R T H E P O O R Thailand (2008-2013) Indonesia (2011-2014) Lao PDR (2007-2012) Vietnam (2010-2014) Philippines (2006-2015) 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% Bottom 40 Total population Source: WDI and staff estimates 4 Most of the household surveys in the region use consumption as the welfare measure, while the Philippines and Malaysia use income. The estimates for the rest of the countries in the East Asia Region almost mirror the GDP per capita growth. Note, however, that the bottom 40 percent of the population was growing much faster in Cambodia, Vietnam, and the Philippines 20 Figure 1.4. Varying trends in per capita household income growth across the population Source: Staff estimates using various rounds of FIES Note: A growth incidence curve (GIC) shows how the growth rate for a given quantile varies across quantiles ranked by income. A downward sloping GIC thus indicates that growth contributes to equalizing the distribution of income, and vice- versa for an upward sloping curve. percent, while the bottom 20 percent grew at about More than half of the available jobs were in the 16 percent. The most recent period had the broadest services sector, followed by agriculture and industry. impact on household income, as almost 80 percent In the recent decades, there was a gradual shift in of households experienced income growth of more jobs away from agriculture and into the services than 6 percent. In comparison, only the very rich sector, a structural change that started in the (top 5 percent) households reached this level in the 1970s. The share of agriculture in total employment 2009–2012 period, which suggests the vulnerability decreased from about 36 percent in 2006 to 28 of the economy to external shocks. percent in 2015, while the share of the services sector rose from about 50 percent to 56 percent. The strong economic growth contributed to However, unlike the experience in many other steady decline in the unemployment rate from 8.0 countries in the East Asia Region, where a large percent in 2006 to 6.9 percent in 2015. During this share of the unskilled labor moved from agriculture 10-year period, the labor force participation rate to manufacturing jobs, with consequent higher essentially did not change, with an average of 64.0 productivity and higher wages, in the Philippines percent of the working-age population employed or the share of total employment in manufacturing looking for jobs. The unemployment rate for men remained around 15 percent. The bulk of unskilled has consistently remained higher than the rate for employment moved to the informal services M A K I N G G ROW T H WO R K F O R T H E P O O R women, with men constituting nearly two-thirds subsector, which is typically low-wage and low- of the unemployed. Among age cohorts, young skill in nature. According to the Philippines workers between the ages 15 and 24 suffered the Development Report (World Bank 2013b), more highest unemployment, making up roughly half than three-quarters of the services sector is of the unemployed each year. As the age profile composed of low-pay or low-skill jobs, such as petty rose, the smaller the share of the total unemployed retail trade and public transportation. The low became. Despite the decline in unemployment, productivity in the services sector does not provide underemployment—those who worked but were the basis for large wage increases, which limits the willing and available to work “more adequately”— scope of poverty reduction and shared prosperity. has remained persistently high, in the range of 18–20 percent, since 2006. 21 Labor productivity in agriculture remained depressed over the past decade, and its growth was Box 1.1. Poverty estimates using the slowest of the three sectors. Between 2006 and national and international poverty lines 2015, agriculture productivity grew annually at 2.0 percent, compared with the services sector at National official poverty estimates in the Philip- 2.4 percent and the industry sector at 2.9 percent, pines are produced by the Philippine Statistics Authority (PSA). These are derived using income thereby widening the productivity gap between welfare aggregates evaluated against per capita agriculture and the other two sectors. Labor poverty lines that are set broadly following the productivity in the industry sector is roughly twice cost-of-basic-needs (CBN) approach. Using a that in the services sector and more than six times national reference food bundle based on expert that in the agriculture sector. opinion of what constitutes a nutritionally ade- quate bundle, province-specific bundles are set separately for urban and rural areas and reflect locally consumed commodities. These locally priced bundles that constitute food poverty lines Poverty and Inequality are scaled up by a constant food to non-food ratio to calculate total poverty lines. There are 163 poverty lines set that correspond to urban and rural areas of the 81 provinces in the country, The incidence of poverty has declined over the and these were benchmarked to 2009, when the past decade. Using the national poverty line, the current methodology was developed. poverty rate dropped 5 percentage points. Using the international poverty line of US$1.90 a day or the The methodologies of constructing the nation- al poverty line and the international extreme middle-income class poverty line of US$3.20-a-day poverty line differ in several aspects. Unlike the (in 2011 purchasing power parity), the poverty rate international poverty line, which is fixed and dropped 7.9 percentage points and 11.4 percentage updated by the consumer price index (CPI), the points respectively (Box 1.1).5 (Unless otherwise national poverty lines are updated by re-estimat- noted, this report uses the national poverty line to ing the food poverty lines at current prices. The measure the poverty rate and trends.) fixed food to non-food ratio used to derive the poverty lines assumes that the non-food price inflation is the same as food price inflation. This The reduction in poverty was minimal in the earlier explains the relatively slower reduction in poverty years of the decade, and a more rapid pace was using national poverty lines compared with the observed only recently (Figure 1.5A). In 2015, over international poverty lines. one-fifth of the population, or 22 million Filipinos, continued to live below the national poverty line (Map 1.2); of this number, roughly 8.2 million people the same trend of decline in the poverty rate is M A K I N G G ROW T H WO R K F O R T H E P O O R (8.1 percent of the population, as estimated from observed as with the national poverty line, though the Family Income and Expenditure Survey [FIES]), the magnitude and speed differ in the sub-periods did not have sufficient income to meet their basic 2006–2009 and 2012–2015, when a faster rate of food requirements, according to the national food decline is seen. The number of poor has declined poverty line.6 Using the international poverty line more rapidly in the recent years (Figure 1.5B) of US$1.90 a day and the poverty line for lower- middle-income-class countries (LMIC) of US$3.20 The Poverty rate declined over time—although a day, both in 2011 purchasing power parity (PPP), slowly in some periods—against various 5 Research shows that there has been a significant decline in multidimensional poverty over the past decade, although the magnitude of the decline in, and especially the dimensional contributions to, aggregate multidimensional poverty are quite sensitive to the alternatives considered. See Datt (2017). 6 Poverty rates vary significantly within a region. In some cases, regional statistics may be driven by a few locations within the region. 22 Figure 1.5. Poverty rate and number of the poor A. Poverty rate B. Number of the poor (in millions) 45% 35 32.8 32.6 38.4% 30.4 40% 30 27.4 35% 25 23.3 23.7 22.6 21.9 30% Million Indiividuals 26.6% 27.0% 25% 20 21.6% 20% 15 12.4 14.5% 15% 9.6 9.9 10 10% 6.7 6.59% 5 5% 0% 0 2006 2009 2012 2015 2006 2009 2012 2015 LMIC poverty line National poverty line International poverty line LMIC poverty line National poverty line International poverty line Source: Staff estimates using various rounds of FIES Source: Staff estimates using various rounds of FIES combinations of poverty lines (Table 1.2). In 2006, secure and are at risk of falling back into poverty a total of 3.8 percent of the population (3.2 million with a negative shock. individuals) were severely deprived (below 50 percent of the national poverty line); this declined Compared with many East Asian countries, the to 2.3 percent (2.4 million individuals) in 2015. The Philippines stands out for showing little dynamism relatively well-off (those living at least at twice the at both the low end (elimination of poverty) and level of the national poverty line), hovered around the high end (rise in economic security and global 40 percent in 2006–2012 but increased slightly, to middle class) (Figure 1.6A and B). In 2015, measured 41.6 percent, in 2015. However, in many cases, these by global standards, while only 6.6 percent of “well-off” households are far from economically the population was extremely poor (living below Table 1.2. Multiples of poverty line by national standards 2006 2009 2012 2015 Welfare level Population Share Population Share Population Share Population Share (million) (%) (million) (%) (million) (%) (million) (%) M A K I N G G ROW T H WO R K F O R T H E P O O R Severe deprivation 3.2 3.8 3.0 3.4 3.1 3.3 2.4 2.3 (< 0.5 poverty line) Quite poor 9.1 10.7 8.8 10.0 9.0 9.5 8.0 7.9 (0.5 – 0.75 poverty line) Poor 10.3 12.1 11.5 12.9 11.7 12.4 11.6 11.4 (0.75 – 1 poverty line) Vulnerable to poverty 9.6 11.2 10.0 11.3 10.6 11.3 12.14 11.9 (1 – 1.25 poverty line) Moderately well-off 18.9 22.2 20.8 23.5 21.6 23.0 25.3 24.9 (1.25 – 2 poverty line) Well-off 34.1 40.0 34.6 39.0 38.1 40.5 42.3 41.6 (> 2 poverty line) Overall 85.3 100.0 88.7 100.0 94.1 100.0 101.6 100.0 Source: Staff estimates using national poverty line and various rounds of FIES 23 Figure 1.6. Prosperity improvement in the Philippines compared with the East Asia and Pacific Region A. Population distribution by economic class in B. Population distribution by economic class in the Philippines, 2002–15 East Asia and Pacific, 2002–15 100% 100% Extreme poor (less than PPP $1.90 - a day) Extreme poor (less than PPP $1.90-a day) 90% 90% 80% 80% Moderate poor (PPP $1.90-$3.10 - a day) Moderate poor (PPP $1.90-$3.10 - a day) 70% 70% 60% 60% Vulnerable (PPP $3.10-$5.50 - a day) 50% Vulnerable (PPP $3.10-$5.50 - a day) 50% 40% 40% 30% 30% Economically secure (PPP $5.50-$15.00 - a day) Economically secure (PPP $5.50-$15.00 - a day) 20% 20% 10% 10% Global middle class (PPP $15.00 and higher - a day) Global middle class (PPP $15 and higher - a day) 0% 0% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: EAP Team for Statistical Development US$1.90 a day, 2011 purchasing power parity), 18.7 fifth of the population in 2002. In the Philippines, percent was moderately poor (between US$1.90 and the increase was very modest, a rise from 37 percent US$3.20 a day), and 30.8 percent of the population to 44 percent. There is still a long way to go for was economically vulnerable (between US$3.20 and the Philippines to achieve its goal of becoming a US$5.50 a day). Only 34.7 percent was economically middle-class society. secure (between US$5.50 and US$15 a day) and only 9.2 percent was in the global middle class Inequality in the Philippines is among the highest (above US$15 a day). Many East Asian countries, in the world. Inequality of income declined slightly particularly China and Vietnam, have fared better, during the periods 2006–09 and 2012–15, while it making significant improvements in the extent of slightly increased during 2009–2012, after the global economic vulnerability. In developing East Asia financial crisis. While the income Gini coefficient and the Pacific, the economically secure and middle declined from 47 percent in 2006 to 44 percent in 2015, class comprised nearly two-thirds of the region’s it is still higher than in the majority of developing population in 2015, up from its share of just over a countries in East Asia (Figure 1.7). The inequality Figure 1.7. Inequality of income A. Gini coefficient of the Philippines, 2006–2015 B. Gini coefficient of other countries in East Asia M A K I N G G ROW T H WO R K F O R T H E P O O R 60% 50% 45% 50% 40% 35% 30% 40% 25% Gini Ratio 20% 30% 15% 10% 20% 5% 0% Cambodia Lao Indonesia Malaysia China Thailand Vietnam Philippines 10% 0% 2006 2009 2012 2015 2012 2014 2012 2013 2013 2015 2012 2009 Source: Staff estimates using various rounds of FIES, measured by Source: PovCalnet and staff estimates, measured by household income per capita. household consumption per capita. 24 Figure 1.8. International comparison: share of wealth for the richest one percent 70% 60% 50% 40% 30% 20% 10% 0% Spain Sweden France Russia Japan Argentina Ireland Netherlands Norway Egypt Indonesia Taiwan Italy United Arab Emirates Saudi Arabia United States Brazil Hong Kong Czech republic South Africa Malaysia Greece Turkey Mexico Korea China Peru India Poland Singapore Chile United Kingdom Israel Belgium Australia Austria Switzerland Thailand Portugal Germany Denmark Canada Colombia Philippines Finland New Zealand Source: Credit Suisse Wealth Report, countries in East Asia marked in blue level in the Philippines is much higher if measured by middle-income-class poverty line. At the household wealth. According to the Credit Suisse Wealth Report, level, the improvement in income distribution the top 1 percent of the population owned more than accounted for some 40 percent and 50 percent of the half of the nation’s wealth (Figure 1.8). poverty reduction (Figure 1.9). During 2009–2012, poverty reduction was minimal, not only due to The role of growth and distribution in poverty the weak economic growth, but also the worsened reduction varied in the different sub-periods. distribution of income. During 2006–2009 and 2012–2015, growth was strong and inclusive. At the macro level, each percentage Families may perceive themselves as poor or moving point increase in GDP per capita resulted in 0.7 in and out of poverty differently from what the and 0.5 percentage points of decline in poverty rate, national poverty data show.7 The Social Weather respectively, against the US$3.20 per day lower- Stations (SWS), a private social research institution Figure 1.9. Growth and Poverty B. Contributions of household income growth and change in A. Responsiveness of poverty reduction to GDP per capita growth distribution to poverty reduction 1.0 100% % of total poverty reduction using the $3.20 Decline in $3.20/day poverty rate for each % 80% 0.8 M A K I N G G ROW T H WO R K F O R T H E P O O R 60% increase in GDP per capita 0.6 poverty line 40% 20% 0.4 0% 0.2 -20% -40% 0.0 2006-2009 2009-2012 2012-2015 2006-2015 2006-2009 2009-2012 2012-2015 2006-2015 Growth Change in Distribution Source: Staff estimates using the lower-middle-income-class poverty rate Source: Staff estimates using the lower-middle-income-class poverty rate and various rounds of FIES and various rounds of FIES Note: Residual component is minimal and are excluded in the figure. 7 Objective poverty and subjective poverty follow different methodologies of measurement. They are not meant to be comparable. 25 Figure 1.10. Self-rated poverty and hunger in households 60% 54.3% 50.5% Poverty in the Philippines: An 50% International Comparison 40% 30% From an international perspective, the rate of 20% 16.7% 13.3% poverty reduction in the Philippines has been slower than many East Asian countries over the past 10% decade (Table 1.3). The Philippines remains among 0% the countries with the highest poverty based on 2006 2009 2012 2015 SWS Poverty Rate SWS Hunger Rate both the US$1.90 a day and US$3.20 a day poverty lines in the region. The pace of extreme poverty Source: Social Weather Stations reduction in the Philippines averaged 0.9 percentage points per year between 2006 and 2015, less than in the Philippines, provides self-rated indicators half the 1.4 points per year decline in the developing of the status of poverty and hunger using regular world overall. The rate of reduction is also much surveys8 (Figure 1.10). Self-rated poverty estimates slower compared with other developing countries are consistently and significantly higher than the in East Asia using the lower-middle-income class national poverty estimates. Studies have shown that line of US$3.20 a day (Figure 1.11). The persistently while objective welfare measures affect perceptions, high level of inequality of income limited the underlying psychological factors inherent to the responsiveness of poverty reduction to growth in the respondent, such as demographic, socioeconomic Philippines. While this has improved in the most or health status, are also important (Pradhan recent years, it still lags behind countries such as and Ravallion 2000, Ravallion and Lokshin 2002, Vietnam and Indonesia. Angelilo 2014). Table 1.3. Poverty rates in selected East Asia countries US$1.90/day US$3.20/day Country Start year End year Start year End year China (2005–2012) 18.8 1.9 43.5 20.2 Indonesia (2006–2015) 27.5 7.5 65.6 34.0 M A K I N G G ROW T H WO R K F O R T H E P O O R Lao PDR (2007–2012) 18.3 15.3 65.2 57.8 Philippines (2006–2015) 14.5 6.6 38.4 27.0 Thailand (2006–2013) 0.7 0.0 6.2 1.1 Vietnam (2006–2014) 19.5 2.8 51.3 11.6 Source: Staff estimates based on international poverty lines 8 Calculated by SWS based on self-reported hunger in households. The survey questions on the family’s experience of hunger are directed to the household head. The survey question on hunger is as follows: “In the last 3 months, did it happen even once that your family experienced hunger and not have anything to eat?” Those who experienced hunger were further asked: “Only once, a few times, often or always?” SWS classify hunger into moderate and severe. Moderate hunger refers to those who experienced hunger only once or a few times in the last 3 months while Severe hunger refer to those who experienced hunger often or always. The hunger rate is made up of the moderate and severe hunger rates. Reference: SWS: Statistics for Advocacy (www.sws.org.ph) 26 The population of the Philippines is very young Figure 1.11. Decline of US$3.20/day poverty rate for (Figure 1.12). In 2010, half the population was each 1 percent increase in GDP per capita younger than 23.4 years. This is higher than the median age of 21.3 years recorded in 2000, but 1.2 significantly lower than in many other countries in East Asia, such as China and Thailand. Four in 1.0 10 of the household population are of school age 0.8 (5 to 24 years old). Regionally, ARMM had the highest percentage of school-age population at 49.1 0.6 percent of the household population; the NCR had 0.4 the lowest at 39.2 percent. For every 100 people in the working-age population, there were about 0.2 61 dependents (54 young dependents and 7 old 0.0 dependents). The demographic structure presents Thailand Lao PDR Philippines China Indonesia Vietnam 2006-13 2007-12 2006-15 2005-12 2006-15 2006-14 a clear opportunity for the country to reap the economic benefits of demographic dividends in the Source: Staff estimates based on lower-middle-income-class poverty lines coming years if it can manage to improve the skills of the labor force and create productive jobs. At the Figure 1.12. Age-sex pyramid of household population, 2010 same time, it raises significant challenges for more Total population = 92.1 million inclusive growth should the high fertility rate persist for women in the poorer households. Age group 85 and Over 80 - 84 75 - 79 70 - 74 Male Female Forces That Have Reduced 65 - 69 M A K I N G G ROW T H WO R K F O R T H E P O O R 60 - 64 55 - 59 Poverty 50 - 54 45 - 49 40 - 44 35 - 39 30 - 34 25 - 29 Empirical analysis shows that the observed decline 20 - 24 15 - 19 in poverty over 2006–2015 is attributable mainly 10 - 14 to an increase in wage income and movement 5-9 Under 5 of employment out of agriculture,9 government -6% -4% -2% 0% 2% 4% 6% transfers, and remittances from domestic and Source: Philippine Statistics Authority foreign sources (Figure 1.13). 9 Employment in agriculture focused on workers who were engaged only in activities related to primary production. 27 Figure 1.13. Contribution of income sources to poverty reduction, 2006–2015 2 0 Change in Poverty Incidence -2 -4 -6 -8 Non-agriculture Government Domestic remittance Agriculture wage Non-agriculture Foreign remittance Others Agriculture wage transfers enterprise enterprise International poverty Lower-middle-income-class poverty Source: Estimates using international poverty lines and various rounds of FIES. Extreme poverty is defined as household income per capita below US$1.9 a day (2011 PPP); and lower middle-income-class poverty is defined as household income per capita below US$3.2 a day. Increase in wage income and movement out of percent in 2000, 12 percent in 2010, and 10 percent primary production agriculture. The increase in in 2016), and a decline in the share of agriculture wage income and movement of workers out of in employment. While most of the poor continued agriculture contributed about two-thirds of the to work in agriculture, this share of the population poverty decline.10 The major contribution is from the gradually declined by nearly 1 percentage point increase of non-agricultural wages, which accounted each year, from 36 percent in 2006 to 28 percent for over 50 percent of the reduction in poverty (out in 2015 (Figure 1.14). Even lower-end industry and of a total of two-thirds for non-agricultural and services jobs paid more than agriculture jobs (Figure agricultural income combined). The Philippines 1.15). Agricultural wage incomes account for about has been experiencing a decline in the share of one-eighth of the reduction of extreme poverty for agriculture to GDP over time (22 percent in 1990, 14 2006–2015. The gradual movement of employment Figure 1.14. Millions shifted out of agriculture 2015 M A K I N G G ROW T H WO R K F O R T H E P O O R 2012 2009 2006 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Agriculture Manufacturing Other Industry Trade Finance, Real Estate, Renting & Business Activities Transportation, Storage & Comminication Public administration & Defense Accomodation, Food & Recreation Services Other Services Source: LFS 2006 and 2015 10 In this report, the January rounds of the Labor Force Survey data are used which are merged with the corresponding Family Income and Expenditure Survey data in relevant analysis. 28 Figure 1.15. Greater earnings in manufacturing and services than in agriculture Monthly wage, 2015 25,000 20,000 15,000 10,000 5,000 0 Laborers & Farmers, Forestry Service Workers Trades & Related Plant & Machine Clerks Technicians and Special Professionals Gov't. officials, Unskilled Workers Workers and and Shop and Workers Operators and Associate Occupations Corporate execs, Fishermen Market Sale Assemblers Profesionals Managers Source: LFS 2015, vertical axis is 2015 Philippine pesos out of agriculture, as well as the accompanying Remittances. Remittances from domestic and increase in agricultural wages and for unskilled foreign sources contributed about 12 and 6 percent labor in recent years, are among the key factors of of the poverty reduction, respectively. Two-thirds of poverty reduction. Filipinos, or 15 million households, receive domestic and foreign remittances. Foreign remittances are Government transfers. Transfers from government much higher in value. Both transfer types have social programs contributed about 25 percent of similar impacts on reducing the poverty rate (by the reduction in poverty. The national conditional up to 4 percentage points), and domestic transfers cash transfer program, Pantawid Pamilya, expanded have higher impact on the poverty gap. Domestic rapidly during this period, and became the primary remittances are more prevalent among the poor, government social assistance program for the while foreign remittances, though greater in value, poor. It extends cash grants to 77 percent of poor are more common among the non-poor. Domestic households and contributes both to reducing remittances reduce inequality, while foreign poverty and to building human capital. It is remittances increase it. estimated that the program reduced the national poverty rate by up to 1.5 percentage points (lifting The contribution to poverty reduction of 1.5 million people out of poverty) in 2015. This is entrepreneurial incomes was a negative 15 percent. consistent with the global experience with the Entrepreneurial activities are varied: for poor rural M A K I N G G ROW T H WO R K F O R T H E P O O R impact of social safety net transfers on poverty. households, a high share of the entrepreneurial The World Bank report, The State of Social Safety incomes come from activities related to primary Nets 2018, estimates that such transfers reduce production in agriculture; for the urban poor, from the incidence of international poverty by 36 self-employed, lower-end services; while for the non- percent. Moreover, even if the transfers do not lift poor, a high share might come from business. Its beneficiaries above the international poverty line, overall negative contribution to poverty reduction they reduce the poverty gap by about 45 percent. may reflect the diverse nature of the work, including Pantawid Pamilya also helped influence behavior the declining importance of entrepreneurial change. It improved school enrollment of older agriculture activities for the poor. children, encouraged early childhood education, and increased the health-seeking behaviors of beneficiaries. 29 for labor have been limited. Compared with other Reasons Why Poverty Has Not East Asian countries, including China, Indonesia, Declined as Fast as in Other Malaysia, and Thailand, the reallocation of agricultural labor toward sectors with higher or East Asian Countries faster productivity, such as manufacturing, was more limited.11 For wage earners, real wage gains were limited in the past decade (only 4 percent Poverty declined only marginally, particularly over increase in 2006–2015). A large share of the unskilled 2006–2012, despite good economic growth. The main agricultural workers ended up in the low-end reasons poverty in the Philippines did not decline services sector, which limited the productivity gains as fast as in other East Asia and Pacific Region from the structural transformation. countries include: lower pace and less pro-poor pattern of growth, high inequality of income and High inequality of income and wealth. The wealth, and disasters and conflict. Philippines has one of the highest levels of inequality in the world. Measured by household Lower pace of growth in household income per income, the Gini coefficient hovers around 44–47 capita. The annual growth rate of GDP in the percent, with declining trends only in recent years, Philippines of around 5.5 percent in the past decade but it is higher than many neighboring countries translates into a 3.6 percent growth in per capita in the Region. The top 1 percent owned more than terms for the high population growth rate of 1.7 half of the nation’s wealth, according to the Credit percent. Data from household surveys show that Suisse Wealth Report. Differences in the quality growth in household income per capita was also of human capital (from an unfair start in life), as lower for the Philippines than for other countries in well as difference in the incomes individuals and the Region. On average, these incomes were growing households can earn, drive a large degree of the at only 1.6 percent each year during 2006–2015. inequality of outcomes in the Philippines. The This lagged the stellar-performing countries in the high concentration of wealth could result in strong East Asia and Pacific Region. While the bottom 40 vested interests in maintaining the status quo, which percent experienced growth of 2.9 percent each year, could hinder the reforms needed to facilitate more the low growth rate of household income limited the inclusive growth and poverty reduction. pace of poverty reduction. Despite the good economic growth, only a small share of the population has made Natural disasters and conflicts. Regions affected it to the middle class, and more than 10 percent of by natural disasters and conflicts have lower living Filipinos remained vulnerable to falling into poverty. standards and higher poverty rates than more placid areas.12 Changes in weather patterns are shifting the M A K I N G G ROW T H WO R K F O R T H E P O O R Less pro-poor pattern of economic growth with path of seasonal natural disasters, and El Niño may limited gains from structural transformation for intensify, with particularly harsh consequences for labor. Agriculture, which employs most poor people the poor. The poorest regions of the country, where in the Philippines, has experienced minimal growth agriculture accounts for a disproportionate share of in the past decade, contributing to GDP growth by income and the capacity to manage risk is particularly an average of 0.2 percentage points (compared to 1.9 weak, face increased vulnerability to shocks. percentage points for industry and 3.4 percentage Protracted conflict, particularly in parts of Mindanao, points for services) over the period of 2006–15. The has exacted a great toll on the local economy and efficiency of productive resource allocation has trapped people in poverty. been low and gains from structural transformation 11 See Annex A for more details on the comparison of intersectoral allocation and productivity growth between the Philippines and other East Asian countries. 12 See more details in World Bank (2017d). 30 31 M A K I N G G ROW T H WO R K F O R T H E P O O R Map 1.1. GDP per capita by region M A K I N G G ROW T H WO R K F O R T H E P O O R 32 Map 1.2. Poverty rate by region 33 M A K I N G G ROW T H WO R K F O R T H E P O O R C HA PTE R TWO Profile of Poverty and Inequality in Living Conditions • Poor households are large, have low educational attainment, rely on self- employment or agricultural work. Farmers, fisherfolk, and other agricultural workers are often the poorest. • Three-quarters of the poor live in rural areas, and the rural poverty rate is three times that of urban areas. Poverty rates are highest in the high disaster-risk and conflict-affected areas. Two-fifths of the poor live in Mindanao, and over 50 percent of the population in ARMM are poor. • Education access in the Philippines are slightly better than those of other countries at its level of income, but health services access and health outcomes lag. • Access to basic services and ownership of communication and mobility assets significantly improved over the past decade, but the poor still fared worse than the non-poor. Informal settlements are the most visible manifestation of multidimensional poverty in urban areas. • Natural disasters impose extremely high economic and human costs in the Philippines, and the poor are the most exposed to risk and the least able to cope. The same level of asset loss affects poor and marginalized people far more than wealthier people because their livelihoods depend on fewer assets, and their M A K I N G G ROW T H WO R K F O R T H E P O O R consumption is closer to subsistence levels. • Protracted conflict, particularly in parts of Mindanao, has exacted a great toll on the local economy and trapped people in poverty. This chapter examines the characteristics of the poor also takes stock of the non-income dimensions in the Philippines and their distribution by rural of poverty, including access to basic services and and urban areas and among regions of the country, ownership of durable assets, among the poor and taking into consideration factors that contribute non-poor. to poverty, such as disaster risks and conflicts. It 34 Characteristics of the Poor Figure 2.1. Household size, poor versus non-poor (percentage of households) 70% Households of larger size, with higher dependency 60% 58.3% ratios, headed by younger males, with lower 50% education, and employed in agriculture are more likely to be poor. (Poverty throughout this chapter is 40% based on the national poverty line.) 30% 26.7% 20.6% 19.1% 18.2% 20% 17.1% 12.7% Poverty rate increases monotonically with household 10% 11.7% 5.7% 6.0% size. Households with six members or more are 0.7% 3.3% 0% much more likely to be poor. In 2015, over 31 percent 1 Person 2 Persons 3 Persons 4 Persons 5 Persons 6+ Persons of the households with six members or more were Poor Non-Poor poor, 10 percentage points higher than the national Source: Staff estimates using FIES 2015 average. Among all the poor households, 58 percent have six members or more; in contrast, only 27 Figure 2.2. Total fertility rate by income quintile percent of all the non-poor households have six or more members (Figure 2.1). 6 5.2 5 Women in poor households have more children. With a total fertility rate of 3 (recent DHS 2017 4 3.7 showed that TFR has declined to 2.7 births per 3 3.1 3 woman), the Philippines has high population 2.4 growth—1.7 percent per year—more than double the 2 1.7 average for the East Asia Region (0.7 percent), and nearly 50 percent higher than the world average (1.2 1 percent). While total fertility has been declining, 0 it remained high for the poor. A typical Filipina All Poorest 2 3 4 Richest woman in the poorest 20 percent of the population Source: PSA 2013 has 5.2 children in her lifetime, compared with 1.7 children for those in the richest 20 percent (Figure rate of 21.6 percent. Poor households have, on 2.2). According to the Department of Health, over average, nearly four children (1.5 young children M A K I N G G ROW T H WO R K F O R T H E P O O R 80 percent of married women wanted to either under 5 years old, and 2.4 young school-age space their births or limit childbearing (PSA 2013). children between 5 and 17 years old), while non- poor households have only 2.4 children (1.3 young Households with a high child dependency ratio are children under 5 years old, and 1.1 young school- more likely to be poor (Figure 2.3 and Figure 2.4). age children between 5 and 17 years old). There is a In 2015, nearly 40 percent of the households with strong association between household size and child three or more young children (below 17 years of dependency ratio. age) are poor, compared with the national poverty 35 Figure 2.3. Average number of children, Figure 2.4. Poverty rate by number of children poor versus non-poor 42% 39.4% 3.0 37% 2.5 2.4 32% 2.0 26% National Level 1.5 21% 1.5 1.3 15.2% 1.1 16% 1.0 11% 8.6% 5.6% 0.5 5% 0% 0.0 None 1 dependent 2 dependents 3+ dependents <5 5 to 17 Number of Dependents (<17 y.o.) Poor Non-Poor Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 The risk of poverty declines with the age of Figure 2.5. Poverty rate by age of household head M A K I N G G ROW T H WO R K F O R T H E P O O R household heads (Figure 2.5). Poor households are 30% more likely to be headed by younger adults. Nearly 27.7% 25.5% 60 percent of poor households have heads younger 25% National Level than 50 years old, compared with only 44 percent of 20% non-poor households. Over 25 percent of households 16.5% 15.3% with heads aged below 50 were poor in 2015, 15% compared fewer than 16 percent of households with 10% heads over 50. 5% Female-headed households are less likely to be 0% poor (Figure 2.6). The low poverty risks of female- < 30 31 to 49 50 to 64 65+ headed households are related to the high share Household Head Age Source: Staff estimates using FIES 2015 36 Figure 2.6. Poverty rate by gender of household head Figure 2.7. Income sources by gender of household head 25% 50% 43.8% 45% National Level 40% 36.5% 20% 23.4% 35% 30% 15% 24.1% 25% 19.6% 20.7% 20% 16.4% 14.1% 14.8% 10% 15% 10% 6.7% 6.7% 4.2% 4.8% 5% 1.8% 5% 0% Agri Wages Non-Agri Wages Entrepreneural Domestic Foreign Others Activity Remittances Remittances 0% Male Headed Female Headed Male Female Source: Staff estimates using FIES 2015 Source: Staff estimates based on FIES 2015 Note: In percent of foreign remittances in their household income and Figure 2.9). High school education stands out as (Figure 2.7). Female-headed households have a the key threshold. Households headed by individuals lower share of income from agriculture wages, non- who have not graduated from high school have a agricultural wages, and entrepreneurial activity, poverty risk that is higher than the average. Nearly but a significantly higher share from remittances, 60 percent of households headed by individuals both domestic and foreign. On average, foreign with no education and 40 percent of those headed remittances represent 20 percent of total household by an individual who did not complete primary income, and domestic remittances represent education are poor. High school graduation reduces 7 percent for female-headed households. This the risk of poverty to two-thirds of the average compares with 7 percent and 5 percent for male- rate for the population. Among households headed headed households. by individuals with complete tertiary education, the risk of falling into poverty is minimal. Eighty Like most countries, the Philippines shows a strong percent of the poor live in households headed by negative correlation between poverty risk and the individuals with less than a high school education. level of education of the household head (Figure 2.8 Figure 2.8. Educational attainment of household head Figure 2.9. Poverty rate by educational attainment of household head M A K I N G G ROW T H WO R K F O R T H E P O O R 40% 37.7% 60% 57.6% 35% 50% 30% 26.8% 40.1% 25% 22.6% 40% 20% 17.0% 17.7% 15.7% 15.1% 15% 13.6% 30% 26.7% 11.1% 25.0% 9.6% National level 10% 6.4% 20% 5% 2.0% 2.6% 13.3% 1.4% 0.0% 0.7% 0% 10% 7.3% 3.2% 0.0% 0% Poor Non-Poor Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 37 Figure 2.10. Highest educational attainment of the population A. By age group, 2015 B. Poor adults versus non-poor adults 65 and above 6.1% 60.2% 20.0% 13.7% 2006 1.5% 23.2% 40.5% 34.9% Non-poor 50-64 2.3% 40.1% 33.2% 24.4% 2015 1.1% 19.1% 46.3% 33.5% 25-49 1.2% 23.8% 42.0% 33.0% 2006 4.9% 50.4% 38.2% 6.6% Poor 15-24 0.8% 20.8% 49.9% 28.5% 2015 4.1% 41.4% 45.5% 9.0% No grade completed Elementary High school College No grade completed Elementary High school College Source: Staff estimates based on LFS 2015 Source: Staff estimates based on LFS 2006 and 2015 Education of the entire population, including the middle-class households (above US$15 a day), some poor and the non-poor, improved over time, but low 70 percent of which are headed by someone with a levels of education remain a common characteristic post–high school education (including vocational of the poor (Figure 2.10). The younger cohorts of programs and college) compared with less than 25 workers are more educated than the older cohorts. percent for average households (Figure 2.11). In 2015, over 75 percent of the population in the 15– 24 age group had completed secondary education or In the Philippines, having a decent-paying job, not above, compared with about 35 percent of the oldest just having a job, is the key factor that conditions in the 65 and above cohort. However, almost half of living standards (Figure 2.12). Nearly 90 percent poor adults have no more than primary schooling; of the poor live in households whose heads are less than 10 percent have any postsecondary employed, compared with 80 percent of the non- education. This sharply contrasts with the education poor. But the type of work matters. Nearly 30 levels of the non-poor. For the high end, among percent of the households headed by individuals Figure 2.11. Highest educational attainment of the middle class versus average households M A K I N G G ROW T H WO R K F O R T H E P O O R 60% 51.6% 50% 40% 30% 24.6% 20.1% 20% 18.4% 17.8% 13.9% 14.1% 12.0% 10% 8.3% 5.2% 4.2% 2.3% 3.4% 3.5% 0.2% 0.5% 0% No Schooling Unfinished Elementary Graduate Unfinished Highschool Graduate Vocational / Unfinished College College Graduate Elementary highschool Programs Middle Class Average Household Source: Staff estimates using FIES 2015 38 Figure 2.12. Employment status of household head Figure 2.13. Poverty rate by class of work of household head 100% 35% 89% 30% 90% 30% 79% 80% 25% 22% 23% 70% 20% 60% 15% 14% 50% 10% 11% 10% 40% 5% 30% 5% 21% 20% 0% 11% 10% 0% Employed Not Employed Poor Non-Poor Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 with self-employment (without employees) are poor, of the households headed by individuals who are a significantly higher percentage than the national unemployed are poor. average of 21.6 percent (Figure 2.13). Households M A K I N G G ROW T H WO R K F O R T H E P O O R headed by individuals holding government jobs Over 30 percent of poor households reported or working in their own business with pay are their most important source of income to be much less likely to be poor. Households headed by entrepreneurial activities (Figure 2.14).13 In contrast, individuals working without pay are also less likely over 50 percent of non-poor households reported to be poor, which might be related to migration their most important sources of income as salary / remittances received from other members of from non-agricultural activities. Eleven percent the household. It is also probable that because of the non-poor households reported that their of the strong role of remittances, only 12 percent most important income source was assistance from 13 For agricultural households, a large share of the entrepreneurial activities is agriculture-related. See Annex B for details of the income sources in agriculture households and the agriculture subsection on income and employment shares. 39 Figure 2.14. Share of households mainly relying Figure 2.15. Employment share by sector on the specific income sources 19% Salary from Agri Activity 4% Poor 58% 12% 11% 6% 13% 30% Salary from Non-Agri Activity 53% 31% Enterpreneurial Activities 20% 3% Assistance from Abroad Non-poor 20% 17% 21% 9% 33% 11% 14% Assistance from Domestic 5% 0% Pension and Benefits 3% All 29% 16% 19% 8% 29% 3% Other Sources 3% Poor Non-Poor Agriculture Industry Trade Transpo & Comm Other services Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 abroad, compared with only 3 percent of the poor households that receive the highest share of their households. In the meantime, 14 percent of the poor income from agriculture, 48.5 percent are poor, far households reported that their most important more than the 29.8 percent for rural households income source was assistance from domestic in general. In addition, a substantial share of sources, compared with 5 percent of the non-poor agricultural households are vulnerable to falling into households. Poor households received a negligible poverty—14 percent of the population that relies on amount in pensions and benefits. agriculture lived between 100 and 125 percent of the poverty line. Farmhands and laborers, corn farmers, Nearly 60 percent of the poor work in agriculture, coconut farmers, and fisherfolk are among the very twice the national average, and three times the poorest (Box 2.1). ratio of the non-poor (Figure 2.15). Overall, the second-largest share of employment is in trade. Poor agricultural households also are typically more Manufacturing and other industries account for vulnerable to shocks. First, agricultural production about 16 percent of the total employment, but only is more exposed to natural disasters, and thus 12 percent of the poor are employed in industries more vulnerable than other sectors. Households compared with 17 percent of the non-poor. depending largely on agriculture for their income face substantial risks of falling back into poverty. A Households for which agriculture is the main natural disaster tends to affect the cropping cycle, M A K I N G G ROW T H WO R K F O R T H E P O O R source of income are the poorest (Table 2.1). Among with farmers potentially less able to plant crops, Table 2.1. Poverty rate of rural households with agriculture and remittances as main sources of income, 2015 Number of poor Type of household Population (million) Poverty rate (percent) (million) All households 101.6 21.6 21.9 Of which, rural households 58.0 29.8 17.3 Of which, households with agriculture as main sources of 18.4 48.5 8.9 income Source: Staff estimates using FIES 2015 40 Box 2.1. The poorest agriculture households Over the past decade, the top five occupations of the household head of the population at the bottom decile of per capita household income remained corn farmers, farmhands and laborers, coconut farmers, inland and coastal water fisherfolk, and rice farmers. The first four groups have the poverty rates twice or three times the national average. Rice farmers represent a high share of the poorest decile (16 percent in 2006 and 10 percent in 2015), while their incidence of poverty is lower than that of the other four occupations. The average house- hold per capita incomes in the households headed by farmhands and laborers, corn farmers, and coconut farmers are the lowest, accounting for only 55–60 percent of the average rural household income per capita. The average income per capita of households headed by rice farmers was higher than the other four types of households, with an income level close to the rural average. Much like the trend of changes of national poverty, the poverty rate of these occupations declined. As the labor force gradually moved out of agriculture, the share of the population in each of these occupations declined over time. 2016 2015 Top 5 occupations Share at Poverty rate Top 5 occupations Share at Poverty rate of household heads bottom 10% of of household heads bottom 10% of for population at the income distri- for population at the income distri- bottom 10% of income bution bottom 10% of income bution distribution distribution Corn Farmers 17.7 65.2% Farmhands and laborers 21.4 48.8% Farmhands and laborers 17.0 56.4% Corn farmers 13.5 61.0% Rice farmers 16.1 39.3% Rice farmers 10.1 31.1% Coconut farmers 7.5 52.7% Coconut farmers 5.2 45.6% Inland and coastal waters 6.3 51.8% Inland and coastal waters 5.2 39.9% fisherfolk fisherfolk thus perpetuating the period of income uncertainty. (roughly ₱19,000–22,000 per household per year in Second, a large share of the poor agriculture 2015). But the expenditure per capita of the richest households are smallholder farmers and are often quintile is 4.5 times that of the poorest quintile net food buyers. In the late 2000s, when the price (₱488,000 versus ₱107,000). While rice accounts for M A K I N G G ROW T H WO R K F O R T H E P O O R of rice increased sharply during the global food only 5 percent of total household expenditure for crisis, millions of poor Filipinos, many of them poor the richest quintile, it is 20 percent for the poorest farmers and net food purchasers, suffered. quintile. In particular, households headed by farmers are net rice purchasers. Because 19 percent of their Poor households spend some 70 percent of their household expenditure was used for rice purchase in income on food in the Philippines, with rice as the 2015. The poorest households are more vulnerable to main staple and the greatest single expenditure. rice price changes (see World Bank 2017d, box I.C.4). The rice price in the Philippines is twice that in Thailand and Vietnam, and considerably higher than the world price. Regardless of the level of household income, expenditure on rice is similar 41 Figure 2.16. Share of GDP per capita by Locations of the Poor 350% main island group to total 300% 250% Three-quarters of the poor live in rural areas. Poverty rates are lowest in the NCR and highest in 200% Eastern Visayas and Mindanao. Two out of five poor 150% people in the Philippines live in Mindanao, and over 100% 50 percent of the population in ARMM is poor. 50% Regional disparities are wide, and some lagging 0% NCR Luzon Visayas Mindanao regions are growing even more slowly in recent 2006* 2009 2012 2015 years than they did in the past. The development Source: Philippine Statistics Authority gaps between NCR and the rest of the country, * 2006 data at constant 1985 prices; 2009, 2012, 2015 data are at constant 2000 particularly the areas outside Luzon, widened over prices. The numbers are not directly comparable time. Regional GDP per capita in NCR increased to about three times the national average in 2015, while Figure 2.17. Poverty rate by major island groups those of Visayas and Mindanao were 64 percent and 45% 60 percent, respectively (Figure 2.16). 40% 40.4% 36.2% 34.5% 35% Poverty rates also are very low in the NCR and 30% 28.2% high in Mindanao, and they have remained largely 25% 21.7% unchanged over time (Figure 2.17 and Figure 2.18). 20% Nearly two-fifths of the poor live in Mindanao. The 15.7% 15% decline in the poverty rate at the national level in 10% 2012–2015 came primarily from regions outside of 4.7% 3.9% 5% NCR.14 0% 2006 2009 2012 2015 Poverty is mainly a rural phenomenon (Figure 2.19 NCR Luzon Visayas Mindanao and Figure 2.20). Nearly 80 percent of the poor live Source: Staff estimates using various FIES rounds in rural areas. The rural poverty rate and the urban poverty rate both declined over time, particularly Figure 2.18. Poverty share by major island groups after 2012, but the rural poverty rate remained 40% around three times the urban rate.15 35% M A K I N G G ROW T H WO R K F O R T H E P O O R 30% While the urban poverty rate is lower than the rural 25% poverty rate, disparities in living conditions are most 20% evident in urban areas. Shelter inequalities depict significant polarization in the distribution of wealth 15% and resources in cities. Informal settlements are 10% the most visible manifestation of multidimensional 5% poverty in the urban areas of the Philippines (Box 0% NCR Luzon Visayas Mindanao 2.2). In most cities, informal settlement communities 2006 2015 with no security of tenure and inadequate access to Source: Staff estimates using various FIES rounds 14 Large migration to the NCR might be one of the reasons for the limited reduction in poverty there. 15 FIES (2003, 2006, 2009, 2012, 2015). 42 Figure 2.19. Poverty rate in urban and rural areas Figure 2.20. Poverty share in urban areas 40% 37.1% 25% 36.6% 35.2% 21.7% 35% 20.8% 20.8% 21.1% 29.8% 20% 30% 25% 15% 20% 15% 12.8% 12.7% 12.5% 10.6% 10% 10% 5% 5% 0% 2006 2009 2012 2015 0% Urban Rural 2006 2009 2012 2015 Source: Staff estimates using various FIES rounds Source: Staff estimates using various FIES rounds basic services coexist with exclusive, fully serviced, and gated communities (World Bank 2017c). This Box 2.2. Poverty among informal settlement families phenomenon is most pronounced in Metro Manila, which houses almost 40 percent of the total informal The Philippines has more than 1.5 million infor- settlements in the country. Many informal settlers mal settler families (ISFs), nearly 600,000 (or 40 live in chronic urban poverty, confronted daily by percent) of whom live in Metro Manila, according physical, economic, social, legal, and environmental to the estimates of a recently developed National risks.16 According to a recent survey covering 21 Informal Settlements Upgrading Strategy. Three million people in Manila, or one out of every four communities in Metro Manila, households may Metro Manila residents, rely on informal housing. earn as little as ₱50 to ₱100 per day (World Bank As with other developing countries, the pervasive- forthcoming a). ness of informal settlements in the country can be traced to low income, unrealistic and inade- The poverty rate is persistently higher in high- quate urban planning, lack of serviced land and conflict regions. (Box 2.2) It is greater in the affordable social housing, and a dysfunctional legal system.a conflict-affected areas of Mindanao and Eastern Visayas than elsewhere. Over 50 percent of the The deprivations that accompany informal population in ARMM is poor (Figure 2.21). housing include increased vulnerability to natural disasters, inadequate access to infrastructure and Figure 2.21. Poverty rate of high-conflict regions services, and a lack of physical safety and tenure security, all of which exacerbate and deepen the 60% experience of urban poverty. A majority of the 50% M A K I N G G ROW T H WO R K F O R T H E P O O R people are forced to remain in informal settle- ments for decades given the lack of affordable 40% housing options. Informal settlements are the 30% most visible manifestation of multidimensional poverty in the urban areas of the Philippines. 20% Slum households’ experience of informality is not 10% limited to the spaces they inhabit, but extends 0% to their jobs, modes of transport, and access to 2006 2009 2012 2015 basic services. ARMM Central Mindanao Northern Mindanao Zamboanga Peninsula Southern Mindanao a ICF International and HUDCC 2014. Note: The box is drawn from the World Bank (2017c) Source: FIES various rounds 16 The information of the informal settlers may not be fully representative in the official household surveys, because some ISF might not be covered. 43 Non-Income Dimensions the Philippines data used for this comparison predates many recent education reforms.17 The of Poverty Philippines is participating in the 2018 Programme for International Student Assessment (2018), which will allow for an updated analysis of the Poverty is more than a state of financial deprivation. performance of the education system relative to It encompasses a range of socioeconomic factors other countries. that collectively tend to lock the poor into their condition. These non-income dimensions include access to education and health care services as Access to Health Services well as to basic services such as clean water and sanitation. This section examines these aspects of Health outcomes and health care services access poverty in the Philippines. are more worrisome. The Philippines did not fare well in many health outcome indicators compared with countries with similar income (Figure 2.25). Education and Learning The Philippines had a life expectancy at birth of about 68 years in 2015. The country had not met Rates of adult literacy and school enrollment in the Millennium Development Goal targets 4 and 5, Philippines are similar to those of countries with the related to maternal and child health, by 2015. same level of income. While the primary school net Its under-5 child mortality was at 28 per 1,000 enrollment rate (96 percent) is as high as in many live births in 2015. Vietnam performs better, other East Asian countries, net secondary school while Indonesia and the Philippines have similar enrollment (Figure 2.22) in the Philippines (67.4 outcomes, but are slightly worse-off than other percent) ranked behind Malaysia (68.5 percent), countries of similar income. Children from poor Indonesia (75 percent), and Thailand (82.6 percent) households suffer from an unequal start in life in 2015. The adult literacy rate for the Philippines compared with those from rich households. M A K I N G G ROW T H WO R K F O R T H E P O O R (Figure 2.23) does not differ substantially from most East Asian counties of the similar income level. Childhood malnutrition is pronounced in the Philippines. One in three children under age five In terms of learning, the performance of the is stunted—the principal marker of malnutrition. education system in the Philippines is far below the Based on the worldwide data, a country at the regional average (Figure 2.24). The difference in test Philippines’ level of income would be expected scores between the Philippines and high-performing to have a stunting rate of 20 percent, rather than countries like Vietnam is equivalent to three years’ its actual level of 33 percent. The stunting rate worth of learning. An important caveat to this in the Philippines is substantially above that finding is that because the Philippines has not been of wealthier countries in the region, including part of a recent international learning assessment, Thailand, Malaysia, and China, but also above that 17 The most recent international learning assessment data for the Philippines is from the 2003 Trends in International Math and Science Study. 44 Figure 2.22. Secondary enrollment and GNI per capita, 2015 Figure 2.23. Adult literacy and GNI per capita, 2015 M A K I N G G ROW T H WO R K F O R T H E P O O R of Vietnam, which has a much lower per capita gross The Philippines has also lagged other countries national income (GNI) than the Philippines (Figure around the world in its long-term reduction in 2.26). Both the Philippines and Indonesia have much stunting. The prevalence of stunting in the country higher levels of stunting than would be expected for has been roughly flat since the early 2000s. Taking their GNI. Filipino children from poor households into account the earlier decline, the average suffer particularly high rates of malnutrition. annualized rate of decline over the period 1995–2015 45 Figure 2.24. PISA and TIMSS scores in East Asia and the Pacific Source: OECD; TIMSS; cited from Growing Smarter: Learning and Growth in East Asia and the Pacific, forthcoming. was 1 percent. Many countries around the world have of non-poor households (85 percent in 2006 had much faster rates of decline—notable examples and 94 percent in 2015, respectively). include Brazil, China, the Islamic Republic of Iran, and Vietnam. Map 3.1 shows rates of change for • Access to clean water also remains a problem countries around the world. for poor households. Only 31 percent had access to clean water in 2015, compared to 23 percent in 2006. This is in sharp contrast with Access and Quality of Basic Services the non-poor households: 54 percent of the non-poor households had access to clean water Poor households often have limited access to in 2006 and 61 percent in 2015. M A K I N G G ROW T H WO R K F O R T H E P O O R sanitation, clean water, electricity, and household assets. Access to all these improved over 2006–2015 • Poor households’ access to electricity improved among poor households. However, significant significantly, from 55 percent in 2006 to disparity between non-poor and poor households 75 percent in 2015. However, there is still a persisted. The proportion of poor households with considerable difference when compared to the access to basic services remained significantly below proportion of non-poor households that have that for non-poor households. quasi-universal access to electricity (92 percent in 2006 and 95 percent in 2015). • The access of poor households to sanitation— flush or water-sealed toilets—has significantly • The proportion of poor households with access improved, from 49 percent in 2006 to 71 to at least one communication asset, including percent in 2015, but it remains well below that mobile and landline phones, television sets, 46 Figure 2.25. Health outcomes and services A. Life expectancy at birth, 2015 B. Under-5 child mortality rates, 2015 M A K I N G G ROW T H WO R K F O R T H E P O O R C. Maternal mortality rates, 2015 47 Figure 2.26. Rates of stunting by gross national income per capita Source: World Bank (2016) Note: This shows a scatter plot of stunting rates versus gross national income per capita for countries around the world, with countries in East Asia and the Pacific highlighted. The solid line shows the general trend across countries. and personal computers, increased from 65 significantly lower than the numbers for the percent in 2006 to 85 percent in 2015. The non- non-poor households, at 59 percent in 2006 poor have near universal access, 94 percent in and 62 percent in 2015, respectively. 2006 and 98 percent in 2015. The 20 percentage point increase for the poor from 2006 to 2015 A high percentage of individuals in informal urban reduced the disparity between the poor and settlements have limited access to basic services. non-poor households. About 3 percent live on lots without the consent of the owners.18 People living in informal settlements • The access to at least one mobility asset (car, have lower access to basic services in multiple areas. motorcycle, or motorboat) by poor households The difference between the ISF in Metro Manila and is low, despite a significant improvement from the overall population in Metro Manila is sharp in 3 percent in 2006 to 13 percent in 2015. Similar several areas. Less than 60 percent of Metro Manila M A K I N G G ROW T H WO R K F O R T H E P O O R improvement, from 23 percent to 37 percent, is ISFs have access to durable assets, while 80 percent witnessed for the non-poor. The gaps in access of the Metro Manila population has access to at between the poor and non-poor remained least one. Only 55 percent of the ISF have access to wide. good housing materials, while nearly 80 percent of the population does. And only 40 percent of the • The share of poor households with at least ISFs have access to clean water, while more than 85 one durable asset, such as a refrigerator, percent of the population has access to clean water. stove with a gas range, washing machine, or air conditioner increased slightly, from 7 percent in 2006 to 12 percent in 2015. This is 18 Living on lots without the consent of the owners is used here as a proxy for informal settlement. The share of population living in lots without owner consent is lower than the reported share of the population living in informal settlements, drawing from the slum survey in Metro Manila. Drawing from FIES 2015, about 6 percent of the population in Metro Manila live in lots without the consent of owners. Possible reasons include the differences in sample frame and in definition. See World Bank (2017e) for details. 48 Sources of Household Income Box 2.3. Poor informal settler families suffer from lack of adequate access to basic services Salaries and wages, entrepreneurial incomes, and Informal settlements are a manifestation of transfers and remittances are the most important poverty and inequality in urban areas. A survey sources of household income in the Philippines. In of 3,000 ISFs in Metro Manila found that almost 2015, these sources combined accounted for four- half of them obtain water through vendors, fifths of household income. Over half of this was paying 9–13 times more for the delivery of water than households living in adjacent, fully from salaries and wages, while enterprise income serviced neighborhoods.a The major constraint provided about a quarter (Figure 2.27). that prohibits ISFs from accessing potable water is the connection fee. New connection fees are Among poor households, salaries and wages account equivalent of US$97 and US$176 in Manila West for the biggest share of income, though less than for and Manila East, respectively.b Additionally, there non-poor households, made up in part by a larger are administrative requirements such as proof of land title, which prevent many households share in enterprise income. Disaggregating these from connecting individually. While 93 percent components further reveals that agriculture is still a of the urban poor sampled report having access significant source of income for poor households.19 to water-sealed septic tanks,c many of them are Over one-third of salaries and wages come from improperly designed and hardly maintained, agricultural activities and about two-thirds of allowing human waste to pollute the water. The enterprises are agriculture-related. Interestingly, the Philippines has a very high electricity rate—the fifth most expensive in the world, averaging at share of transfers and remittances is the same for US$0.24 per kilowatt-hour in 2012.d The high elec- both poor households and non-poor households; the tricity rate forces many ISFs to resort to shared difference is the source. Remittances and transfers connection or “jumping,” illegal connection to to poor households come from domestic sources, neighbors or public electricity. Access to educa- while those for non-poor households are mainly tion is also limited due to financial constraints. from foreign sources. In the Philippines, public education is provided for free from kindergarten to 12th grade. How- ever, surveyed ISFs responded that expenses for The share of pensions and retirement benefits is textbooks, school supplies, uniform, lunches, minimal. This indicates the lack of financial security and transportation costs are often a burden they for the elderly. Imputed rent of owner-occupied cannot afford.e dwellings and other minor sources20 complete the Source: The box is drawn from World Bank (2017c). composition of household incomes. Among poor a. Ballesteros 2010. households, a greater part of the residual is related b. ADB 2014. to agricultural production, particularly subsistence c. World Bank 2016d. M A K I N G G ROW T H WO R K F O R T H E P O O R d. International Energy Consultants 2012. farming. e. World Bank 2016d. The sources of income have changed structurally over the past decade. Dependence on agricultural incomes the share of wages and salaries from non-agricultural has declined, particularly among poor households. employment increased by about 4 percentage points. The share of incomes from agricultural enterprises Despite that, agriculture still accounts for two-fifths has dropped by 10 percentage points. Subsequently, of the incomes of the poor. 19 Due to data limitation, agriculture refers to only primary agricultural production and agricultural employment refers to farm laborers. Any selling of agricultural produce in the market was captured under business in the survey questionnaire 20 Includes net share of agricultural production, subsistence farming, dividends from investments, interest from bank deposits and loans extended to other households, rental income from lands and properties not used for agriculture, and gifts received. 49 Figure 2.27. Components of household income, 2015 Poor households Non-poor households Pensions and Others Pensions and retirement Others 4.9% retirement benefits benefits 8.5% 3.9% 0.5% Rental value of owner- Rental value of owner- occupied dwelling occupied dwelling Salaries and 7.1% 8.9% wages 41.6% Salaries and Transfers and Transfers and wages remittances remittances 47.7% 14.7% 14.7% Entrepreneurial incomes Entrepreneurial incomes 19.9% 27.6% Source: Estimates using FIES 2015 The share of government transfers in the incomes Vulnerability to Disaster of poor households has increased in recent years, from 0.1 percent of in 2006 to 6 percent in 2015. This reflects the government’s effort to improve social As in many other countries, the poor and vulnerable programs. Most notable among those programs is the suffer the most due to their higher exposure to disaster Pantawid Pamilyang Pilipino Program, a CCT program risks (including living in the wrong locations and that accounted for three-quarters of the government greater reliance on agriculture) and more limited transfers21 received by poor households in 2015. Such capacity to cope (due to lower savings to serve as a programs partly offset the lack of safety nets and buffer against difficult times). The disaster-prone income security, particularly among poorer households. regions have a higher poverty rate (Table 2.2). Table 2.2. Poverty rate for regions prone to earthquakes M A K I N G G ROW T H WO R K F O R T H E P O O R Regions with degree IX-XII in the Mercalli Poverty Rate Scale for Earthquake Intensity 2006 2009 2012 2015 Bicol Region 44.2 44.2 41.1 36.0 Eastern Visayas 41.5 42.6 45.2 38.7 Western Visayas 29.1 30.8 29.1 22.4 Caraga 49.3 54.4 40.3 39.1 Southern Mindanao 30.6 31.4 30.8 22.0 Source: FIES and staff estimates 21 Other government transfers include cash assistance or relief during calamities and programs initiated by local governments, such as scholarships and benefits for the elderly. 50 Natural disasters impose extremely high economic and human costs in the Philippines.22 The Philippines is located on the “Pacific Ring of Fire,” affected buildings, infrastructure, equipment, and a line of volcanic and seismic activity that runs agriculture, as made clear by the estimate that, along the edge of the Pacific Ocean. An average on average, upwards of a million Filipinos are of 20 tropical cyclones hit the country every year, impoverished each year by natural disasters. of which 5 to 7 are destructive (Bowen 2016). The increase in temperature due to climate change is The same level of asset loss affects poor and projected to lead to more intense tropical storms marginalized people far more than wealthier people and typhoons. The Philippines currently has the because their livelihoods depend on fewer assets, second-highest level of disaster risk in the world and their consumption is closer to subsistence and is the eighth-most vulnerable country to the levels. The poor and vulnerable cannot rely on effects of climate change (United Nations University savings to cope with the impacts of losses, placing and Alliance Development Works 2014). Manila their health and education at greater risk and is the fourth-most exposed city in the world to potentially requiring more time to recover and natural disasters. Approximately 74 percent of the reconstruct. A recent World Bank report applies the country’s population and 60 percent of its land socioeconomic resilience methodology and finds area are susceptible to multiple natural hazards annual “well-being” costs (or impact on quality of (GFDRR 2014). Natural disasters have caused life) to make explicit the impacts of natural disasters an estimated US$23 billion in damages in the on consumption and to account for the more severe Philippines since 1990, making it one of the most impact of asset loss and consumption loss on well- disaster-prone countries in the world (World Bank being of the poor and to identify the socioeconomic 2017d). The repeated and increasingly frequent capacity of different regions (Box 2.4). natural disasters are undermining poverty reduction M A K I N G G ROW T H WO R K F O R T H E P O O R in the Philippines. Typhoon Yolanda, one the strongest typhoons ever recorded, affected some of the country’s poorest regions and resulted in nearly 6,300 casualties, 4.1 million people displaced, and Costs of Conflict pushed up to an additional one million people into poverty (Philippines, NEDA 2013, p. 3; UNISDR 2015, p. 49). Over 10 percent of Filipinos lived just The Philippine archipelago is home to some of the above the national poverty line in 2015. Shocks, world’s longest-running subnational armed conflicts. such as natural disasters, can push and even trap Protracted conflict, particularly in some areas of them in poverty. However, the costs of natural Mindanao, has exacted a great toll on the local disasters go well beyond the replacement costs of economy and trapped people in poverty. Security, 22 See Annex C for more discussion on the disasters and investment. 51 Box 2.4. The poor suffered greater loss of well-being for any given asset loss A socioeconomic resilience assessment conducted by the government found that the Philippines suffers asset losses of around Php182 billion, and well-being losses (or impact on quality of life) of around ₱208 billion each year due to natural disasters. However, while the asset losses of the poorest Filipinos account for only 7 percent of total asset losses (₱12.2 billion per year), they suffer 27 percent of the total well-being losses (₱56 billion per year). The same peso value of asset losses has a greater impact on the Disaster losses in Manila from a once-every-25 year typhoon well-being of the poor than of the 18,000 non-poor. For example, a once- every-25-years typhoon in Manila 16,000 causes ₱2,700 in asset losses per Poorest 14,000 Disaster losses (Php per capita) capita for the poorest quintile, while quintile the wealthiest quintile loses assets 12,000 Q2 worth ₱16,600 per capita. However, 10,000 these losses affect the poorest and Q3 wealthiest residents of the capital 8,000 very differently: equivalent well- Q4 6,000 being losses are nearly four times higher than asset losses (₱10,200 4,000 Richest per capita) for the poorest quintile, quintile while the wealthiest quintile 2,000 experiences well-being losses 0 of roughly a third of asset losses Asset loss Well-being loss (₱4,600 per capita). Socioeconomic capacity, defined as the ratio of asset–to–well-being losses, measures the capacity of individuals to minimize the effects of natural disasters on their well-being. For example, a population with socioeconomic capacity twice as large as another will experience half the well-being losses for the same asset losses. The metric is defined for each province in the Philippines and varies widely across regions. Due to factors that condition the resilience of a region, such as quality of housing and infrastructure, financial inclusion, social protection, diversification, early warning systems, and remittances, regions in eastern Visayas and Mindanao are characterized by lower socioeconomic capacity. Despite their relative ability to cope with disasters, well-being losses in Luzon and the Eastern Visayas are high due to the elevated exposure of those regions to typhoons and earthquakes. M A K I N G G ROW T H WO R K F O R T H E P O O R Source: World Bank Group (2017f); also see Annex D. 52 Box 2.5. Vicious cycle of conflicts and poverty Conflicts not only destroy physical assets, they erode human capital through death, injuries, and illnesses, denial of education and health services, as well as malnutrition, reducing the earning ability and capabilities of affected household members.a Communities with high conflict intensity, such as in ARMM, have low and worsening human capital indicators compared with areas with low intensities of conflict (such as Davao) or in peaceful communities. On the Human Development Index, the provinces in ARMM languished at the bottom in 2012, while Davao provinces are close to the national average of 0.644. The low levels of human capital endowments in certain communities is both a result of neglect in the provi- sion of social services and the deterioration of available human resources. These are borne out by three fac- tors: local governance failure where corruption and weak governance limited the provision of basic services, violent conflicts that further disrupt the provision of basic services, and emigration of the most skilled. As a result, the areas of high conflict intensity, such as ARMM, have been trapped in a vicious cycle of conflict and poverty, with low physical and human capital investment due to low degree of predictability, low value– added products, low technology, small firm size, large informal sector, and prevalence of “shadow econo- mies.” a For more details, see World Bank 2017h. justice, and economic stresses are closely linked (Box Poverty is higher in the conflict-affected areas of 2.5). Addressing the causes of conflict and providing Mindanao and Eastern Visayas. Two out of five poor jobs and economic opportunities are crucial for people in the Philippines live in Mindanao, and over resolving the root causes of conflict and violence. 50 percent of the population in ARMM is poor. Poverty incidence is persistently higher in high- conflict regions (Table 2.3). Table 2.3. Poverty incidence of high-conflict regions 2006 2009 2012 2015 M A K I N G G ROW T H WO R K F O R T H E P O O R IX - Zamboanga Peninsula 45.0 45.8 40.1 33.9 X - Northern Mindanao 39.0 40.1 39.5 36.6 XI - Southern Mindanao 30.6 31.4 30.8 22.0 XII - Central Mindanao 37.9 38.3 44.8 37.3 ARMM 47.1 47.5 55.8 53.7 Source: FIES varies rounds 53 M A K I N G G ROW T H WO R K F O R T H E P O O R 54 C HAPTE R THRE E Labor Market Performance • The key challenge in the Philippines labor market lies in the quality of jobs: most of the poor are working poor. They have been deprived of the opportunity to benefit from growth, not because of unemployment but because of the low pay level of the available jobs or underemployment. • Employment gradually shifted out of primary production agriculture. Unlike many countries in East Asia, where labor-intensive manufacturing absorbed most of the surplus agricultural labor, in the Philippines, they have moved into less well-paying services. Real wages, particularly for workers in the private sector, increased only marginally in the past decade, which limited the gains for labor from structural transformation and could negatively affect the Philippines’ long-term competitiveness. • The population has become more educated over time, and the younger cohort is more likely to be employed in jobs with better pay compared with the older generations. However, the poor, including younger workers from poor households, remained less educated and more likely to be consigned to low-paid jobs. • There is a mismatch between skills supply and demand, particularly for workers in skill-intensive occupations. Workers with higher levels of educational attainment report longer delays in finding employment and are more likely to be unemployed. • The labor market segmentation between urban and rural as well as between NCR M A K I N G G ROW T H WO R K F O R T H E P O O R and the rest of the country persisted. The less educated, women, and youth face the greatest challenges in finding employment with a decent wage. Women in the labor force are more educated than men, but they earn less than men at every level of education. • More education is strongly associated with wage employment and higher earnings. Few Filipinos who have not completed secondary education hold well-paying jobs. 55 This chapter assesses recent labor market average of 21.6 percent. Households headed by performance in the Philippines, focusing on the individuals working in agriculture or self-employed quality of jobs, structural transformation, and have the highest poverty rate. This highlights the disparities between regions. Using the Labor Force challenge of job quality—having a job is not a ticket Surveys (LFS) for 2006, 2009, 2012, and 2015, it out of poverty; many poor families are headed by reviews key labor market indicators, including active the working poor and those working in agriculture. population, labor force participation, employment, Households headed by the unemployed, a large share unemployment, and underemployment, and wage of which are migrant households, or own family income based on region and population groups, businesses, had the lowest poverty rate. such as educational attainment, gender, age, and other characteristics, and examines the returns to The nature of employment tells the story with different levels of education. vivid numbers. Three out of four households are headed by someone who is employed (Figure 3.3). The poverty share of households headed by the wage-employed is similar to their population share Sector and Status of (roughly 45 percent), but those headed by the self-employed are more likely to be poor—while 31 Employment of the Poor percent of the population lives in households headed by the self-employed, they represent 43 percent of the poor (Figure 3.4). While nearly 18 percent of Poor-quality jobs (or “in-work poverty”), rather the population is in households headed by someone than unemployment, is the key challenge in the who is not employed (including those not seeking Philippines.23 Poverty (based on the national poverty employment), they represent only 9 percent of line) is closely associated with the employment the poor. Similarly, 6 percent of the population is sector and activity status of the household head, not in households headed by someone with a family whether the household head is employed (Figure business, but they represent only 4 percent of the 3.1 and Figure 3.2). The poverty rate of households poor. headed by the employed is similar to the national Figure 3.1. Poverty rate by employment sector of Figure 3.2. Poverty rate by employment status of household heads household heads 45% 45% 40% 42% 40% M A K I N G G ROW T H WO R K F O R T H E P O O R 35% 35% 30% 30% 25% 25% 30% Overall Poverty Overall Poverty Rate 20% Rate 20% 19% 15% 21% 15% 10% 10% 12% 13% 12% 5% 5% 0% 0% Not Employed Employee Self-Employed With Family Agriculture Industry Services Business Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 23 See more discussion in World Bank (2016f). 56 Figure 3.3. Poverty rate by employment sector of Figure 3.4. Poverty rate by employment status of household heads household heads 6.4 % 3.9 % 31.1% 43.1% Population Shares Poverty Shares With Family Business With Family Business Self-Employed Self-Employed Employee Employee 44.8% Not Employed Not Employed 43.3% 17.7% 9.6 % National National Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 In both urban and rural areas, the working poor true in rural areas (40 percent versus 47 percent). account for 85 percent of the poor. In urban areas, This is consistent with the observations of a higher a higher share of the poor live in households headed share of wage employment in urban areas and self- by the wage-employed (54 percent), compared with employment in rural areas. Households headed by those headed by the self-employed (30 percent).24 those not employed are a relatively small group This is consistent with the finding of a recent that represents less than 10 percent of the poor; and case study of extreme poverty in the Philippines, households headed by those with a family business which suggests that the primary constraint facing account for only 4 percent (Figure 3.5 and Figure poor households in urban areas is the low level of 3.6). wages paid to unskilled workers.25 The opposite is Figure 3.5. Poverty shares by employment status of Figure 3.6. Poverty shares by employment status of household heads in urban areas household heads in rural areas 2.1 % 4.4 % 30.1% 46.6 % M A K I N G G ROW T H WO R K F O R T H E P O O R Poverty Shares Poverty Shares With Family Business With Family Business Self-Employed Self-Employed Employee Employee 54.1% Not Employed Not Employed 40.4 % 13.7% 8.5 % Urban Rural Source: Staff estimates using FIES 2015 Source: Staff estimates using FIES 2015 24 The self-employed do not have wages reported in the LFS, so do those employed in family owned business without pay. 25 See more details in World Bank (forthcoming a). In the report, respondents in poor urban communities report that wages for retail, household, or construction labor are often very low. Households may earn as little as PHP 50 to PHP 100 per day. 57 Labor Market Status net positive new job creation, employment growth was at par with the working-age population growth of Various Groups (both at about 20 percent over the period), and even slightly faster than the labor force growth (about 16 percent for the same period), resulting in a decline An individual’s labor market status has an in the unemployment ratio (Figure 3.7). However, a important influence on his or her poverty risk. large share of the employment created is with low Overall, the labor market in the Philippines is wages. The rapid expansion of employment also characterized by a high employment rate, a low might have exerted negative pressure on wages. unemployment rate, a high underemployment rate, and a limited increase in the real wage level (Table Millions shifted out of primary production 3.1). The average employment and earning status agriculture in the past decade (Figure 3.9 and Figure has changed little over the past decade. The ratio 3.10).27 Unlike in many neighboring East Asian of working-age population to total population was countries where surplus agricultural labor moves about 66 percent over the past decade. Labor force to labor-intensive manufacturing, the majority of participation declined slightly, from 63 percent in the workers in the Philippines who moved out of 2006 to 61 percent in 2015. The ratio of employment agriculture went to services. The share of the total to working-age population is nearly 60 percent labor force working in agriculture declined from and has changed little over time. Unemployment 36 percent in 2006 to 28 percent in 2015, the share declined from nearly 8 percent in 2006 to 6 percent working in industry increased from 15 percent to in 2015, but this masks the challenges associated 17 percent only, and the share working in services with low-quality jobs. Over the past decade, the increased from 50 percent to 56 percent. The share underemployment rate hovered around 20 to 22 of the poor working in agriculture declined from percent.26 67 percent in 2006 to 58 percent in 2015, the share working in industry increased from 10 percent to Over 2006–2015, labor demand caught up with the 13 percent only, and the share working in services fast-growing labor supply, but the quality of jobs increased from 23 percent to 29 percent. created has been lagging behind. With consistent 3.8 While GDP increased by about 60 percent, Table 3.1. Employment and earnings status Working-age Employment Labor force Under- M A K I N G G ROW T H WO R K F O R T H E P O O R population to working-age Unemployment Daily wage Year participation employment (>=15)/total population rate (2006 pesos) rate rate population ratio 2006 65% 63% 58% 8% 22% 259 2009 67% 63% 58% 7% 20% 263 2012 67% 62% 58% 7% 21% 263 2015 67% 61% 58% 6% 20% 269 Growth, 2006–15 1% -2% -1% -2% -2% 4% Source: Staff estimates using various rounds of LFS 26 Underemployment is defined as “persons in underemployment are all those who worked or had a job during the reference week but were willing and available to work “more adequately”; following the International Labor Organization definition. 27 See Annex E for discussions on employment seasonality and employment dynamics. 58 Figure 3.7. Working-age population, labor force, Figure 3.8. GDP, employment, and real wage growth and employment growth 125% 200% 175% 120% 150% 115% 125% 110% 100% 75% 105% 50% 100% 25% 95% 0% 2006 2009 2012 2015 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Working age population Employment Labor force GDP (2006=100) Employment Real wages (2006=100) Source: Staff estimates using LFS, various rounds Source: Staff estimates using LFS, various rounds employment increased nearly 20 percent, but real in private establishments increased, and that of wages were stagnant, with only a 4 percent increase wage employment in private households, self- in real terms over the period 2006–2015 (Figure employment, and unpaid work declined (Figure 3.8).28 The minimal increase in wage suggests the 3.11). However, the increase in wages for workers in limited gain for labor in the process of the structural private establishments was only some 1.5 percent, shift of employment. while the increase for government employees was 11 percent, and that for private household workers There was a positive shift to employment with was 9 percent. Increases in wages among workers in higher earning potential, but real wage increase was government and in private households are related to minimal. In 2006–2015, the share of employment key wage legislation that led to more rapid growth Figure 3.9. Share of employment of the poor by sector Figure 3.10. Share of employment of an average Filipino by sector 2015 58% 13% 29% 2015 28% 17% 56% 2012 62% 11% 27% 2012 31% 15% 54% 64% 11% 26% M A K I N G G ROW T H WO R K F O R T H E P O O R 2009 2009 33% 15% 52% 2006 67% 10% 23% 2006 35% 15% 50% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% % of employment in each sector each year % of employment for each sector each year Agriculture Industry/manufacture Service Agriculture Industry/manufacture Service Source: Staff estimates using LFS, various rounds Source: Staff estimates using LFS, various rounds 28 The information collected from the LFS is from a sole informant for each household. The respondent is either the household head or the spouse or, in their absence, any responsible adult member of the household. Second-hand accounts of sensitive information such as wage and salary may be underestimated (or overestimated). Due to the data limitation, the analysis of real wage covers the workers who reported positive wage only. The earning of those self-employed and work without paid are not included in the statistics. The results related to wage need to be interpreted with caution. 59 Figure 3.11. Changes in the composition of employment status over time 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2006 2009 2012 2015 Private Household Private Establishment Gov't/Gov't Corporation Self Employed Employer With Pay (Family Owned Business) Without Pay (Family Owned Business) Source: Source: LFS, various rounds. in public sector wages.29 The minimal increase in real a low employment rate, a high unemployment rate, wages in the private sector, where the largest share a low underemployment rate, and a high daily wage of workers is employed, might suggest the limited (Table 3.2). Mindanao and Visayas are at the other gains for labor from structural transformation. end of the spectrum. They are characterized by a low active population rate, a high participation rate, a high employment rate, a low unemployment rate, a Variation by Region high underemployment rate, and a low daily wage. These are all consistent with the wide variation in The labor market environment in NCR is very labor supply and demand, and thus the quality of different from other regions. It is characterized by a jobs, across regions. high active population rate, a low participation rate, Table 3.2. Labor force participation by region, average 2006–2015 Working-age Employment Labor force population to working-age Unemploy- Underem- Daily wage Region participation (>=15)/total population ment rate ployment rate (2006 pesos) rate population ratio M A K I N G G ROW T H WO R K F O R T H E P O O R NCR 70% 61% 55% 10% 12% 402 Luzon (without NCR) 67% 62% 57% 7% 20% 259 Visayas 66% 63% 59% 6% 23% 208 Mindanao 64% 63% 60% 5% 24% 208 Philippines 66% 62% 58% 7% 21% 263 Source: Staff estimates using various rounds of LFS 29 Two laws on the public sector salary standardization were implemented in the last 10 years. The first was Joint Resolution No. 4 by the Fourteenth Congress that authorized President Gloria Macapagal-Arroyo to modify the compensation package for government, military, and uniformed personnel. The revised compensation took effect a year after it was signed in July 2008 for employees in national government offices, and after eighteen months for employees in local government. The salary increase was implemented in equal tranches over four years. Another round of salary standardization through Executive Order No. 201 took effect in July 2016. This legislation ensured comparability of wages in government, particularly management-level positions, with prevailing rates in the private sector. This new adjustment in wages will take effect in stages through 2019. Another important piece of wage legislation is the Domestic Workers Act or Kasambahay Law, which regulates wages given to household employees and enforces the provision of social and other benefits. Wages of those working in private households have grown, on average, by 4.2 percent annually since the law was passed in January 2013. 60 Figure 3.12. Labor market indicators, urban/rural Figure 3.13. Daily wage, urban/rural 80% 400 60% 315 40% 300 Unit: Real PHP 2006 20% National Average 0% 210 200 100 Urban Rural Philippines Urban Rural Source: Staff estimates using LFS, 2006–2015 Source: Staff estimates using LFS, 2006–2015 Variation by Area Variation by Gender The labor market environment differs in rural and Women have much lower labor participation and urban areas. Urban areas are characterized by a lower employment ratios than men (Figure 3.14). higher active population rate, a lower participation Three out of four men of working-age participate in rate, a lower employment rate, a higher the labor market, but only 50 percent of women do. unemployment rate, a lower underemployment rate, The ratio of employment to working-age population and a higher daily wage (Figure 3.12). The difference is nearly 60 percent—but it is about 70 percent in unemployment and underemployment between for men and 45 percent for women. Among those urban and rural areas is large. Unemployment employed with positive wages, women have lower in urban areas is 50 percent higher than that in wages for any given education level (Figure 3.15). For rural areas (9 percent versus 6 percent), while workers with less than a tertiary education, female underemployment in rural areas is 50 percent wages are only 65 percent to 80 percent of those for higher than that in urban areas (24 percent versus males with similar education; for workers with a 16 percent). While the national wage average is ₱263 tertiary education, the wage gap between female and (2006 pesos), it is ₱315 in urban areas and ₱210 in male narrows to 92 percent. rural areas (Figure 3.13). M A K I N G G ROW T H WO R K F O R T H E P O O R Figure 3.14. Labor market participation and Figure 3.15. Daily earnings, 2006–2015 employment ratios 600 80% 531 488 76% 70% 400 2006 PHP 60% 300 232 248 180 185 200 151 163 49% 126 115 116 124 40% 46% 95 0 20% 0% Labor force participation rate Employment to working-age Population Ratio Male Female Male Female Source: Staff estimates using LFS, 2006–2015 Source: Staff estimates using LFS, 2006–2015 61 Education and Labor than the average time spent by workers with lower education levels. About 80 percent of unemployed Market Status workers have completed secondary education or higher. The skill mismatch can also be an indication of the poor quality of education (World Bank The educational level of an individual is closely 2017b). associated with his or her labor market status. While unemployment is lower for individuals The education level of the labor force varies widely with the least education—the poor cannot afford by region and gender. The labor force in NCR to be idle and not working—those with higher stands out as the most educated (Figure 3.16). For educational attainment have significantly higher example, 27 percent of the labor force in NCR had wages than those with little or no education (Table completed a tertiary education, and only 10 percent 3.3). The share of college-educated individuals who have an elementary school education or less. This are underemployed is only half that of those with can be compared to the two least-educated regions, lower educational attainment, and their daily ARMM and V. In ARMM, only 8 percent of the wage is nearly 250 percent of those with a high labor force had completed a tertiary education, school education, over three times of those with an and 56 percent had an elementary school education elementary school education, and over four times of or less; in V, these numbers are 12 percent and 36 those with no schooling. percent. The female labor force is more educated—32 percent had completed a tertiary education, which is A recent World Bank report suggests that there is double the rate for males. a mismatch between supply and demand of skills, including traditional technical and cognitive skills Employment growth came at the expense of the lack as well as socioemotional skills, in the labor market. of real wage growth in the Philippines. The marginal About one-third of employers reported having wage increase of only 4 percent over 2006–2015 unfilled vacancies because of a shortage of applicants masks significant differences between some with the necessary skills. The mismatch is more subgroups. Overall, workers with no schooling saw acute for workers in skill-intensive occupations. a 16 percent wage increase, while those with a high Workers with completed tertiary education spend school education experienced a 2 percent decline, an average of 5.5 weeks searching for a job, far longer and those with tertiary education a 2 percent Table 3.3. Employment, unemployment, and daily earnings, by educational attainment M A K I N G G ROW T H WO R K F O R T H E P O O R Employment-to- Region working-age population Underemployment rate Unemployment rate Daily wage (2006 pesos) ratio No schooling 52% 22% 3% 115 Some elementary 68% 27% 3% 141 Elementary graduate 66% 24% 4% 158 Some high school 46% 24% 7% 166 High school graduate 60% 20% 9% 206 Some college 47% 17% 10% 280 College graduate 67% 11% 8% 506 Source: Staff estimates using various rounds of LFS 62 Figure 3.16. Share of labor force with each grade completed by region 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% NCR CAR I II III IVA IVB V VI VII VIII IX X XI XII ARMM XIII No grade completed Elementary ungraduate Elementary graduate Highschool undergraduate Highschool graduate College undergraduate College graduate Source: Staff estimates using LFS-FIES, 2015 increase (Figure 3.17).30 The U-shape relation of wage in recent years (Figure 3.19 and Figure 3.20). In the growth rate is similar across wage levels.31 period 2006–2012, wage growth was negative for The poorest two quintiles registered a 5–7 percent all but the richest quintile, but over 2012–2015, the wage increase, the third and fourth quintiles a 3 pattern of wage growth was the opposite. The poorest percent decline, and the richest quintile a 5 percent two quintiles registered 12–13 percent wage growth, increase (Figure 3.18). while the richest quintile had a 1 percent decline. The patterns are similar across education groups— Interestingly, the pattern of wage growth changed the less educated had higher wage increases in more over time. After 2012, it became more inclusive. recent years (Figure 3.21 and Figure 3.22). The shift of Workers with lower wages have had greater increases employment out of agriculture and the increase in the Figure 3.17. Wage growth by education level Figure 3.18. Wage growth by wage quintile 800 20.0% 800 20.0% 16% 16.0% 16.0% 600 600 12.0% 12.0% Unit: 2006 PHP Unit: 2006 PHP 7% M A K I N G G ROW T H WO R K F O R T H E P O O R 400 6% 8.0% 400 8.0% 6% 5% 5% 4% 2% 4.0% 4.0% 200 200 -1% 0.0% 0.0% -2% -3% -3% 0 -4.0% 0 -4.0% No schooling Elementary High school College graduate First quintile Second Third quintile Fourth Fifth quintile Graduate graduate quintile quintile 2006 2015 growth rate 2006-15 2006 2015 growth rate 2006-15 Source: Staff estimates using LFS, 2006–2015 Source: Staff estimates using LFS, 2006–2015 30 Wage increase also varied by industry and occupation. According to the World Bank Employment and Poverty Report (World Bank2016c), drawing from the “Structure of Earning Survey,” highly skilled workers such as engineers and accountants saw considerable real salary increases, while some low-skilled workers, such as freight handlers, saw their wages fall. 31 As noted earlier, the data in the LFS do not reflect earning of the self-employed. The results are to be interpreted with caution. 63 Figure 3.19. Real daily wage in 2006–2012 Figure 3.20. Real daily wage in 2012–2015 by wage income groups by wage income groups 800 20% 800 20% 12% 13% 600 600 10% 10% Unit: 2006 PHP Unit: 2006 PHP 6% 6% 6% % change % change 400 400 0% -1% 0% 200 200 -6% -7% -8% -8% 0 -10% 0 -10% First quintile Second Third Fourth Fifth quintile First quintile Second Third Fourth Fifth quintile quintile quintile quintile quintile quintile quintile Wage (2006) Wage (2012) growth rate Wage (2012) Wage (2015) growth rate Source: Staff estimates using LFS, 2006, 2012 Source: Staff estimates using LFS, 2006, 2012 minimum wages have contributed to the increase in from 14 percent in 2006 to 17 percent in 2015. the wage at the lower end in recent years. In 2015, 24 percent of the 25–34 age group had The pattern of wage growth in recent years might completed a tertiary education, double the share of have contributed to a more rapid decline in the tertiary education for the 55–64 age group. poverty rate. However, in the long run, the limited wage increase might result in emigration in pursuit of better job opportunities—many Filipino Younger cohorts and more educated workers are workers are already overseas—and dampen the more likely to find employment with higher wages. competitiveness of the economy. Compared with the older generation, younger workers are more likely to employed in jobs with Over time, the labor force has become more higher wages, such as employment in private educated. The younger cohorts are more educated establishments or government, while some have or than the older. The share of the labor force with work for pay on their own family-operated farm or complete tertiary education has gradually increased business. These jobs are typically better paid than Figure 3.21. Real daily wage in 2006–2012 Figure 3.22. Real daily wage in 2012–2015 by education groups by education groups M A K I N G G ROW T H WO R K F O R T H E P O O R 800 20% 800 20% 20% 600 600 10% 8% 10% 7% Unit: 2006 PHP Unit: 2006 PHP 5% 5% % change % change 4% 400 400 0% 0% -1% -2% -2% 0% 0% -3% 200 200 -5% -7% 0 -10% 0 -10% No Some Elementary Some HS High Some College No Some Elementary Some HS High Some College schooling elementary Graduate school college graduate schooling elementary Graduate school college graduate graduate graduate Wage (2006) Wage (2012) growth rate Wage (2012) Wage (2015) growth rate Source: Staff estimates using LFS, 2006, 2012 Source: Staff estimates using LFS 2012, 2015 64 work in private households, work without pay, or However, individuals in poorer households remained self-employment. For example, among the younger much less educated. Only 2 percent of the labor workers (ages 15–25), 62 percent were employed in force in poor households had completed a tertiary the above-mentioned higher-wage jobs (employed education, compared with 20 percent in non-poor in private establishments or government, have their households. In particular, youth from poor families own family-operated farm or business, or work for remained less educated compared with the rich. pay in their own family-operated farm or business), In 2015, in the bottom income quintile, some 60 while only 46 percent of older workers (ages 50–65) percent of the youth (20–29 years old) did not have were in the higher-wage categories. Similarly, across full secondary education, compared with only 5 educational groups, three-quarters of those with no percent of the youth in the richest income quintile schooling worked in private households, worked (Figure 3.23). Similarly, only 2 percent of the youth without pay, or were self-employed; while for those from the poorest quintile and 7 percent from the with tertiary education, 83 percent were employed second-poorest quintile have completed tertiary in jobs with higher wages (in private establishments education, compared with nearly 60 percent from the or government, have their own family-operated richest quintile. As observed, workers with less than farm or business, or work for pay in their own secondary education have significantly lower earning family-operated farm or business). The average wage and higher chance of falling into poverty. The large for the workers with no school was ₱115, but that for gaps in educational attainment of the youth from workers with a tertiary education was ₱506 (or over the poor and non-poor households might perpetuate four times the wage of the workers with no school). their earning ability and income status. Figure 3.23. Youth (20–29 years old) education level across income groups M A K I N G G ROW T H WO R K F O R T H E P O O R 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1st quintile 2nd quintitle 3rd quintile 4th quintile 5th quintile No schooling Some elementary Elementary graduate Some high school High school graduate Some college College graduate Source: Staff estimates using LFS 2015 65 It is worth noting that nearly one-quarter of the young workers (ages 15–24) are not employed, in school, or in training, (NEET) (Figure 3.24). The ratio Box 3.1. Minimum wage is high in the Philippines, but its coverage is low is slightly higher for youth from poorer households. This may raise significant challenges in poverty The Regional Tripartite Wage and Productivity reduction going forward. The increase in the ratio of Boards set regional minimum wages for NEET in 2009–2012 for the youth from the poorest employees in private establishments and quintile is largely related to the increase in the domestic workers in private households. Those number of youth not in the labor force. The decline wages vary by administrative region as well as by sector and type of establishment. Minimum of the ratio of NEET in 2012–2015, for the youth from wage generally is high relative to the median the poorest households as well as for the population, wage based on several measures, both relative is likely related to the increase in the number of to Filipino workers productivity and to minimum youth in school. wage rates in other countries with similar levels of economic development (World Bank 2013; Betcherman 2014). Nine out of 17 regions have Figure 3.24. Share of youth not in employment, a minimum wage that is higher than the median education, or training wage (World Bank 2016). For most of the years in 30% the past decade, minimum wage rates increased 28% consistently at 3.4 percent on average per year. 26% The minimum wage for private firms is set at an 24% amount that would cover the needs of workers and their families. To account for these needs, 22% the government introduced the two-tier wage 20% system in 2012, whereby the first tier is the mandatory regional wage floor, while the second 18% tier is an amount that is a guide for employers to 16% adjust wages above the floor. The latest reform 2006 2009 2012 2015 aimed to set the wage floors close to the poverty Youth Youth from Bottom Quintile thresholds so the minimum wage would serve Source: Staff estimates using LFS, 2006–2015. as a social safety net among wage workers. Consequently, the number of minimum wages below the poverty threshold was greatly reduced. But in fact, informality severely limits the actual Minimum Wage in the Philippines coverage of minimum wage policy. Less than half (45 percent) of wage workers in private firms are employed in formal firms (World Bank 2016). The Government of the Philippines is actively using Of these wage workers, about 75 percent are minimum wage as a policy tool to address in-work paid equal or above the minimum wage. In the M A K I N G G ROW T H WO R K F O R T H E P O O R poverty. Minimum wage is high relative to the informal sector, the minimum wage accounts for median wage of all workers in most regions in the about 115 percent of the sector’s average wage, Philippines, which is set at an amount that would which is so high that it is likely to discourage cover the needs of workers and their families and informal firms from formalizing their activity. is meant to serve as a social safety net. However, In setting minimum wage, it is advisable to consider the widespread informality means that the poor wage distribution not only in the formal sector, but neither benefit from the minimum wage policy nor also in the informal sector, and set it at a level that from employment protection legislation (Box 3.1). does not cut deep into the overall wage distribution. Only about one-third of workers in private firms Admittedly, this is a difficult to do in a two-tier labor are covered by the minimum wage policy. Hence, market, like the one in the Philippines. The wage distributions are very different in the upper, formal the potential for minimum wage to reduce in-work tier, and (continued next page) poverty in the Philippines is limited. Aligning 66 to become employed (particularly in wage employment in private establishments, government Box 3.1. Minimum wage is high in the Philippines, but its coverage is low (continued) or government corporations) but also affects his or her wage earnings. This section empirically discusses in the lower, informal tier. The minimum wage estimated returns to education using the Mincer set based on the wage distribution in the formal (1974) method for education level, gender, rural/ sector, as it is currently the case, is too high urban areas, and island groups, and the role of to be used in the informal sector, where labor educational attainment of those who worked for productivity is low. On the other hand, if the minimum wage were set based on the wage private establishments, worked for government or distribution in the informal sector, it would be government corporations, employer in own family- too low to be meaningful for formal workers. operated farm or business, or worked for pay in own Some compromise is necessary to strengthen family-operated farm or business in the Philippines the incentives for employers to hire low-skilled (see Annex G for details). As the majority of workers formally. An empirically informed workers who worked for private households, discussion among social partners is needed to find a middle ground (World Bank 2013, 2016). worked without pay in own family-operated farm or business, or were self-employed without paid Sources: Betcherman, Gordon (2014), “Labor Market Regulations: What Do We Know about Their Impacts in Developing Countries?”, employees do not report wage earnings in the World Bank Policy Research Working Paper 6819. World Bank (2013), “Philippine Development Report: Creating More and Better Jobs,” LFS, the subsection includes only wage earners World Bank, Manila. World Bank (2016), “Labor Market Review: in the private establishments and government or Employment and Poverty,” World Bank, Manila. government corporations. Having another year or level of education is strongly minimum wage with worker productivity could associated with better wage employment in private improve the chances of low-skilled workers being establishments and government or government hired formally and benefit from minimum wage (see corporations. Those returns vary by level of Annex F). education, however. The returns in terms of wages to an additional year of high school are modest: Returns to Education 6 percent per year. But completing high school also opens up the possibility of attending college Labor markets in the Philippines offer significant or completing technical/vocational education returns to education. Educational attainment (TVET), which has much higher returns (Figure plays a key role not only in an individual’s ability 3.25). On average, each year of college boosts wages M A K I N G G ROW T H WO R K F O R T H E P O O R 67 Figure 3.25. Rate of return for education Figure 3.26. Rate of return for another year of education by education level 18% 16% College 19% 16% 14% 12% 12% 11% Technical and Vocational 11% 10% 10% 9% 8% High School 6% 6% 4% Elementary 2% 2% 0% 0% 5% 10% 15% 20% All Male Female Urban Rural Source: Staff estimation using LFS 2015 Source: Staff estimation using LFS 2015 by 19 percent, and returns to TVET are 11 percent Asia average (Figure 3.28). They are slightly lower per year. For example, the rate of return for one than those in China and Malaysia, but higher than additional year of education in completing a college those in Indonesia and Mongolia. education is about 19 percent, while it is 6 percent for completing a high school education. These Educational attainment is also positively associated premiums particularly favor females, who are, as with wage employment in private establishments noted earlier, disadvantaged in the labor market. and in government or government corporations.32 The gap in the rate of returns between rural and The probability of wage employment increases by urban areas is not as high as that between genders 2.4 percentage points with another year of schooling (Figure 3.26). (Figure 3.29). Again, this favors females and those in rural areas. More significant, completion of Analysis of subsamples of this data provides additional insights. Consistently higher returns accrue for educational attainment among females, Figure 3.27. Rate of return for education but if they do not even complete basic education, by island group the returns are the smallest of all. For females, completion of basic education is a fundamental College condition for success in the labor market. Returns are higher for urban areas up to high M A K I N G G ROW T H WO R K F O R T H E P O O R Technical and Vocational school completion, but attainment beyond the postsecondary level is more important for rural High school wage workers. Finally, only slight differences are observed in island groups, but the differences are larger for tertiary education (in favor of Mindanao Elementary and Visayas) (Figure 3.27). 0% 5% 10% 15% 20% 25% NCR Luzon Visayas Mindanao The returns to another year of schooling and tertiary Source: Staff estimation using LFS 2015 education in the Philippines are higher than the East 32 The estimation of returns to education omits those who are not wage earners in private establishments, government, or government corporations. 68 Figure 3.28. Rate of return for education by additional years of schooling China (2002) Malaysia (2010) Philippines (2015) Colombia (2012) Indonesia (2010) East Asia & Pacific * Mongolia (2011) Peru (2012) 0% 5% 10% 15% 20% 25% Returns to education total college Return to another year of schooling Source: Staff estimation using LFS 2015 and Montenegro, Patrinos, and Anthony (2014) tertiary education is particularly important for jobs in rural areas in comparison with urban areas. wage employment (Figure 3.30). It may increase the The effect of tertiary education is not significantly probability of employment by 4 percentage points. different in wage employment for males and The highest marginal increase in wage employment females. However, high school completion exhibits probability is observed among those in rural areas. a big difference between males and females. It is This may be related to scarcity of workers with a particularly important for women to at least complete college education and greater competition for wage high school if they mean to look for wage jobs. Figure 3.29. Marginal effects of probability of wage Figure 3.30. Marginal effects of probability of wage employment with an additional year of schooling employment with an additional year of schooling by education level 4.0% 3.5% College 3.0% 2.5% Technical and Vocational M A K I N G G ROW T H WO R K F O R T H E P O O R 2.0% High School 1.5% 1.0% Elementary 0.5% 0.0% 0% 1% 2% 3% 4% 5% 6% All Male Female Urban Rural All Male Female Urban Rural Source: Staff estimation using LFS 2015 Source: Staff estimation using LFS 2015 69 Map 3.1. Annual percentage reduction or increase in stunting rate, 1995–2015 M A K I N G G ROW T H WO R K F O R T H E P O O R Source: Galasso and Wagstaff (2016) 70 71 M A K I N G G ROW T H WO R K F O R T H E P O O R C HA PTE R FOUR Interplay Between Income and Human Capital Accumulation • The wide inequality in access and quality of education and health care services has led to inequality in human capital outcomes across regions and socioeconomic groups. This, in turn, has led to inequality in income. The vicious cycle of inequality of opportunity and inequality of outcomes are mutually reinforcing. • Education public spending has increased in recent years. Access to basic education (K–12) has improved and is now broadly the same as countries at a similar income level. Despite recent progress in basic education, including the rollout of universal kindergarten and senior high school, two principal challenges remain: learning remains limited and secondary school attendance and completion rates are low among the poor. • Increasing access to basic education, particularly the dropout rate beyond elementary among the poor, remains a challenge. Differences in school attainment and learning between children from poor and wealthier families result in differences in their earning power as adults, perpetuating inequality across generations. • The total fertility rate (TFR) in the Philippines, at 3.0 births per woman (recent DHS 2017 showed that TFR has declined to 2.7 births per woman), is among the highest in East Asia and higher than the total wanted fertility rate of 2.2. The TFR is three times as high for women in the poorest quintile as for those in the M A K I N G G ROW T H WO R K F O R T H E P O O R wealthiest quintile and it has not fallen among the poor in recent years. Unmet needs for family planning are highest in poorer families. Over 80 percent of married women want to either space their births or limit childbearing. Teen pregnancy has increased since 1998, the Philippines ranks third highest in Southeast Asia in terms of adolescent fertility rate with 57 births per 1,000 girls aged 15–19. • Pro-poor policies and health insurance changes introduced in recent years had some positive effects on the poor, especially in increased health service usage. Health outcomes for the poor improved little, however, and the quality of health care services remained uneven. Infant mortality rates improved slightly in the most recent data but are higher for the poorer quintiles compared to the richer quintiles. Household spending on health, for both regular and catastrophic needs has 72 remained high. The share of the population pushed into poverty by health spending has doubled over the past decade. • Rates of child malnutrition have shown little improvement, and wide gaps remained across regions and income groups. Malnutrition of young children hampers their economic success as adults. The unequal start of lives contributes to an intergenerational cycle of poverty. The returns from investments to reduce malnutrition are extremely high in the Philippines. This chapter assesses the accumulation of human Education capital in the Philippines. Specifically, it examines the current state of education and health care access Access to education has improved for all over and quality. Using data from the Family Income the past decade and most markedly in access to and Expenditure Survey (FIES), the Demographic elementary school for children from the poorest and Health Survey (DHS), and other administrative families. However, wide differences in access to sources, it documents continuing disparities good-quality education remain across socioeconomic between the poor and non-poor in educational groups, genders, and regions. The poor are attainment and health care status, and the unequal struggling more than wealthier families to complete access and quality of education and health services a full cycle of basic education. This, in part, accounts between the poor and non-poor. for the large disparities in educational attainment levels among Filipinos in the labor force discussed in the previous chapter. Disparities in Access and Figure 4.1. Government expenditure on education to GDP ratio, compared with other countries Quality of Education and 9% Health Care 8% 7% M A K I N G G ROW T H WO R K F O R T H E P O O R 6% Disparities in access and quality of education and 5% health care in the Philippines are large, and the 4% share of public expenditure devoted to education 3% and health remains low despite the increase in 2% public spending in education and health in recent 1% years (Figure 4.1 and Figure 4.2). Household out-of- 0% pocket expenditure accounts for a high share of total expenditure on education and health.33 This could suggest significant disparities in service access and quality across the different segments of income level. Source: WDI 33 Due to lack of data availability, the analysis of this report does not capture the impact of the recent reports, such as the new K–12 and Senior High School program and the recent PhilHealth programs with coverage expansions. 73 Figure 4.2. Government expenditure on health to GDP ratio, compared with other countries 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% Source: WDI E D UC AT I O N ACCESS across the country.35 Restructuring the country’s educational system has been a tough but strategic Over the past decade, the Philippines has carried move on the government’s part to ensure that the out a series of ambitious basic education sector system produces competent graduates who can serve reforms to reduce poverty and increase national as the backbone of a highly skilled and employable M A K I N G G ROW T H WO R K F O R T H E P O O R competitiveness.34 The reforms set out concrete and workforce. ambitious targets for basic education to eliminate teacher and infrastructure deficits. The reform Primary education continued expanding, putting program has been backed up by large increases the country only a few percentage points from in public basic education spending. As a result, achieving universal primary enrollment by 2015. the learning environment has started to improve In contrast, participation in secondary education 34 In 2006, the government introduced a package of policy reforms to improve the access and quality of basic education service, called the Basic Education Sector Reform Agenda, aiming to achieve Education for All by 2015. Following the introduction of the Mother-Tongue Based Multilingual Education from kindergarten to grade 3 in 2009 and the Universal Kindergarten Act of 2011, which stipulates that every Filipino child at least five years of age shall attend one year of preparatory education as a prerequisite for grade 1, the Basic Education Act of 2013 introduced reforms that expanded the basic education cycle from 10 years to 13 years by introducing kindergarten as a prerequisite for entry into elementary school, as well as adding senior high schools. 35 The K–12 basic education has been implemented in sequence since 2011, with the first cohort under the new system is completing the final grade (grade 12) in the school year 2017–2018. In tertiary education, the higher education road map introduced strategies to improve efficiency, upgrade quality, and expand access. 74 remains low, up from 57 percent in 2006 to only 64 the country’s key prolonged challenge. The gap in percent in 2015. The reality is that one out of three education access is widest between the rich and the children of junior high school age (12–15 years old) poor, particularly in tertiary education, but this gap are not attending high school on time. Tertiary starts to appear substantially at the secondary level enrollment reached 29 percent in 2015, up from 22 (Figure 4.4). The poorest groups have benefited most percent in 2006 (Figure 4.3) Regional comparators, from the improvement in access to elementary school such as China, Indonesia, Malaysia, and Vietnam, in recent years. Lower attendance in pre-primary have been more rapidly expanding basic education, among the poorest means that those children start and some have far surpassed the Philippines over the their primary education at a disadvantage. Even if past decade. they complete elementary school, they continue facing significant barriers to continuing secondary The overall improvements in school attendance and tertiary education. Their key constraints in conceal important differences between socioeconomic achieving higher educational levels include the high groups, genders, and regions, which constitute cost of attending schools and the opportunity costs Figure 4.3. Net enrollment rate by level, 2006 and 2015 (all) 100% 80% 60% 40% 20% 0% 2006 2009 2012 2015 2006 2009 2012 2015 2006 2009 2012 2015 2006 2009 2012 2015 Pre- school Elementary High school College Source: Merged FIES-LFS, various years Figure 4.4. Net enrollment rate by level, 2006 and 2015 (poorest and richest quintiles) 100% M A K I N G G ROW T H WO R K F O R T H E P O O R 80% 60% 40% 20% 0% 2006 2009 2012 2015 2006 2009 2012 2015 2006 2009 2012 2015 2006 2009 2012 2015 Pre- school Elementary High school College Bottom quintile Richest quintile Note: There is no data on the current educational level in the LFS data. The approximation assumed that for those who are currently in school, the current educational level is a level higher than the declared highest grade completed. Source: Merged FIES-LFS, various years 75 to poor families, in addition to the low quality of private high school was extended to new grade 11 learning, which limits the value of staying in school. senior high school students.37 School participation also varies by gender and Kindergarten enrollment (attended by children age region. While there is no significant difference five) doubled in absolute terms between 2006 and between the boys and girls until elementary school, 2015—over 50 percent of the 2.2 million five year disparities (in favor of girls) emerge in secondary olds attended kindergarten or some other early and tertiary education. By region, enrollment childhood education and development program rates outside of the NCR seem to have been in 2015. However, policy coordination for early catching up, and the gap between NCR and the childhood education before age five remained weak. rest of the country might have improved, though it Participation to any form of pre-primary education remains large (Map 4.1 and Map 4.2). However, the among children between three and five remained disparities between the NCR and the regions with around 30 percent, which is among the lowest in the the lowest rates are notable. East Asia Region and suggests potential needs for further expansion. Various interventions to address the supply-side challenge have sought to address the slow enrollment E ducational completion growths in secondary education. Over the last decade, the Department of Education (DepEd) has increased Despite the overall progress in basic education M A K I N G G ROW T H WO R K F O R T H E P O O R the number of public high schools by 30 percent and access, the dropout rate beyond elementary, improved the availability of textbooks and teachers, particularly among the poor, remains a challenge. and increased support for school operational expenses. In 2015, 82 percent of young adults from the richest Subsidies to private education through the Education quintile had attained at least elementary education, Service Contracting (ESC) scheme have been also compared with 67 percent from the poorest quintile. expanded.36 In 2016, this support for students at The gaps for secondary education were much 36 The support for private junior high schools was effective, mainly to alleviate the congestion of urban public high schools and to provide choices for students who wish to study in private junior high schools. In by 2016, 70 percent of private junior high school students (accounting for 17 percent of all junior high school students) were ESC grantees. 37 In June 2016, new senior high schools (SHS) opened their gates to 1.5 million new grade 11 students nationally for the first time. Of 1.5 million students, about a half benefited from the SHS voucher program, which covers tuition fees of students attending private high schools, since there were not sufficient spaces in the public system to accommodate the influx of grade 11 students. 76 Figure 4.5. Educational attainment rate among 22–24 Figure 4.6. Educational attainment rate among 22–24 year olds by income quintiles, 2006 and 2015 year olds by gender, 2006 and 2015 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 2006 2015 2006 2015 2006 2015 2006 2015 Primary Secondary Primary Secondary Bottom quintile Richest quintile Male Female Source: Merged FIES-LFS various years Source: Merged FIES-LFS various years wider. In 2015, 81 percent of young adults from the of personal interest in education” (Figure 4.7 and richest quintile had attained secondary education, Figure 4.8). This finding could reflect a perception compared with 41 percent from the poorest quintile by students and parents that the levels of learning (Figure 4.5). By gender, female students have been are limited. The reason may also be related to the able to complete elementary and high school more high opportunity costs and uncertainty of economic often than their male counterparts (Figure 4.6). returns to education (or the poor information about such returns), particularly among boys (Orbeta Children from households in the bottom income 2010). Also, in many cases, poor children need to quintile face the highest dropout incidence. The help meet the immediate welfare needs of their primary reason for dropout, cited in surveys families, which prevents them from staying in by about half of boys and a third of girls in the school. poorest quintile who have dropped out, is “lack Figure 4.7. Reasons for not attending elementary school Figure 4.8. Reasons for not attending high school among 6–11 year olds in the poorest quintile, 2014 among 12–15 year olds in the poorest quintile, 2014 M A K I N G G ROW T H WO R K F O R T H E P O O R Lack of personal interest High cost of education/Financial concerns Illness/disability Lack of personal interest High cost of education/Financial Marriage/Family matters concerns Illness/disability Too young to go to school Accessibility of school Accessibility of school Employment/looking for work 0% 10% 20% 30% 40% 50% 0% 10% 20% 30% 40% 50% 60% Boys Girls Boys Girls Source: APIS 2014 Source: APIS 2014 77 Students’ poor health conditions also appear Health Care concerning among children of elementary school age and is one of the major reasons for dropouts. Health access has been improving in recent About 20 percent of boys and girls who are not years, particularly among the poor, thanks to attending elementary schools cited health or the expansion of health insurance.40 Yet, health disability conditions. One out of every three outcomes improved little, and out-of-pocket children under five show stunted growth; one expenditure increased. The Philippines was not able out five is underweight.38 The DepEd has carried to meet the MDG targets for child and maternal out a school-based feeding program since 2010 to health in 2015. Immunization coverage declined. help malnourished children to keep their focus Wide differences in access to good-quality health in classroom, which reaches about 1.9 million care remain across socioeconomic groups, genders, student beneficiaries in the school year 2016–17. and regions. However, a recent independent impact evaluation revealed mixed results about the effectiveness and health care access sustainability of the program.39 The Philippines faced constraints in providing access Financial concerns and the high cost of education to affordable and quality health care, especially are also significant issues for those in the bottom for its poorer populations. Immunization coverage M A K I N G G ROW T H WO R K F O R T H E P O O R income quintile, particularly for attending high worsened slightly over the years (Figure 4.9). Use of school (Figure 4.8). About a half of the girls who health services for sick children across all income are not attending high schools said that financial groups has gone up but remained low. Only 50 concerns were the most significant reason, percent of children are taken to a provider when compared to about a third of male dropouts (Albert sick with fever. Few may receive appropriate care on 2016). time. Financial constraint remained the top reason for households not following through on advice to 38 National Nutrition Council. 39 Tabunda, et.al. 2016 40 The presentation of access to health care and health outcomes includes the insured and uninsured separately in a systematic manner to show the difference between the two. 78 Figure 4.9. Immunization coverage of children, 2008 and 2013 Immunization covergae of children 12 to 23 months of age, 2008 Immunization covergae of children 12 to 23 months of age, 2013 100 100 Percentage of children age 12 to23 months who received all Percentage of children age 12 to23 months who received all basic 94.5 88.8 87.6 88.4 88.5 80 84.5 80 85.9 86.5 80.4 83.2 78.2 79.2 79.4 80.6 basic vaccinations by quintile (%) 76.4 76.9 77.6 73.5 73.5 72.1 69.9 70.4 vaccinations by quintile (%) 60 60 61.3 57.9 40 40 20 20 0 0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Insured Noninsured Insured Noninsured Note: Includes BCG, measles, and three doses each of DPT and polio. Note: Includes BCG, measles, and three doses each of DPT and polio. Source: DHS (2008) Source: DHS (2013) seek inpatient care. Household income remains the Over 80 percent of married women in the major driver of inequities in the use of maternal and Philippines want to either space their births or child health services. limit childbearing (DHS 2013), indicating sizable unmet need. These unmet needs in reproductive The inclusion of outpatient benefits within health are higher among the poorest quintiles, the Philippine Health Insurance Corporation particularly among the uninsured. Although uptake (PhilHealth) made it more attractive for the insured remains low, the insured poor are more likely to poor to access outpatient care with no copayments. access family planning services than the uninsured The insured are more likely to use health services poor. Overall, unmet need was 18 percent for all, when their children are sick, and the insured and 21 percent among the poorest quintile in 2013 poor were more likely to get medicines (such as (Figure 4.10). Maternal mortality is significant, antibiotics) than the uninsured poor. The gaps in so birth spacing can result in improved maternal access to services between insured and uninsured health. Overall, family planning use is low. Having have widened. However, non-medical care remains a choice of family planning can be beneficial to the a financial constraint, as does travel distance to population. However, currently, usage of family facilities. planning is supported only in a limited way through the PhilHealth benefits package. These inequities remain, despite the recent national M A K I N G G ROW T H WO R K F O R T H E P O O R government subsidy for the health insurance of the While access to antenatal care (ANC) services41 has poor (Paredes 2016). According to a recent survey improved, poorer pregnant women are still less likely covering 21 communities with some of the poorest to get those services (Figure 4.11). When they do, households in the Philippines, access to and quality they tend to receive services of poorer quality than of health care services were uneven. In urban and the better-off (Figure 4.12). For this study, quality peri-urban areas, while many poor households can has been defined as having five services delivered at access routine care without charge, they reported a ANC clinics: weight measurement, blood pressure reluctance to seek treatment and difficulties obtaining measurement, height measurement, urine sampling, required medicines. In remote areas, poor households and blood sampling. Using a crude quality index reported that their access is constrained by the poor that captures these services and giving each equal quality of local facilities, transportation costs, and the weight, the poorer households seemed to have costs of treatment and medicine (World Bank 2013b). poorer quality of services compared to those in 41 ANC was a priority program with PhilHealth, and its coverage was almost universal. 79 Figure 4.10. Unmet needs among reproductive-age women, 2008 and 2013 Unmet need for family planning, 2008 Unmet need for family planning, 2013 35.0 25.0 Percentage of currently married women age 15-49 Percentage of currently married women age 15-49 30.0 23.2 30.2 20.0 by quintile and insurance status (%) by quintile and insurance status (%) 20.1 25.0 18.9 19.2 18.5 18.6 23.7 24.1 17.7 15.0 16.1 16.2 16.6 20.0 15.5 21.0 20.7 20.7 21.2 20.6 20.7 20.0 20.2 18.7 13.3 15.0 10.0 10.0 5.0 5.0 0.0 0.0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Insured Uninsured Insured Noninsured Note: Unmet need for family planning refers to fertile women who wish to Note: Unmet need for family planning refers to fertile women who wish to postpone the next birth (spacing) or stop childbearing altogether (limiting) postpone the next birth (spacing) or stop childbearing altogether (limiting) but who are not using contraception. but who are not using contraception. Source: DHS (2008) Source: DHS (2013) the richer quintiles. A survey regarding the service income groups (Figure 4.13). Fewer than 50 percent readiness of rural health units (RHUs) found of deliveries in the poorest quintile were assisted by relative readiness of RHUs for the delivery of ANC skilled attendants in both survey years (2008 and services,42 but the poorer local government units 2013), while the shares among the top three quintiles (LGUs)43 performed less well, which may explain are more than 70 percent (more than 80 percent in why poorer women get ANC of lesser quality. The 2013). The gaps in skilled birth attendance between the poor quality of services may hinder the achievement insured and uninsured narrowed in 2013, especially for of the desired outcomes for the poor. mothers in the second quintile, where there is a higher percentage of uninsured mothers. This narrowing of The proportions of skilled birth attendance have the gap is also evident in the national average. improved over time, but wide gaps remained across Figure 4.11. Use of health facilities for antenatal care, by quintile and insurance status, 2008 and 2013 Proportion of pregnancies who sought ANC Proportion of pregnancies who sought ANC at least 4 times in the last 5 years, 2008 at least 4 times in the last 5 years, 2013 100 100 Percentage of most recent pregnancies by quintile Percentage of most recent pregnancies by quintile M A K I N G G ROW T H WO R K F O R T H E P O O R 95.9 96.6 92.5 92.6 87.6 87.8 87.5 86.4 87.9 89.3 80 86.7 80 85.5 87.0 84.2 79.4 81.1 80.1 78.8 75.7 71.6 70.8 68.0 60 60 64.4 58.5 40 40 (%) (%) 20 20 0 0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Insured Noninsured Insured Noninsured Source: DHS (2008) Source: DHS (2013) 42 Rural health units (RHUs) are primary care facilities run by local government units (LGUs). Among the services they render are antenatal care, whether or not the RHU is accredited by PhilHealth as a maternal and child health package (MCP) provider. 43 LGUs are in six classes according to income status of the LGU. 80 Figure 4.12. Quality of antenatal care, 2008 and 2013 Quality of antenatal care in the last 5 years, 2008 Quality of antenatal care in the last 5 years, 2013 100 100 93.3 89.9 87.8 87.4 80 84.2 85.4 80 85.4 Average index by quintile (%) Average index by quintile (%) 83.8 82.1 81.3 77.3 78.9 77.2 76.7 76.6 75.5 71.4 68.5 67.6 60 63.9 60 63.3 62.3 60.0 48.5 40 40 20 20 0 0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Insured Noninsured Insured Noninsured Note: Quality of antenated care includes five services: weight measured, Note: Quality of antenated care includes five services: weight measured, height measured, blood pressure measured, urine sampled, and blood height measured, blood pressure measured, urine sampled, and blood sampled. Index is equally weighted among the five, with the perfect score of sampled. Index is equally weighted among the five, with the perfect score of 100 for all services done, and 0 if none of the services were done. Includes 100 for all services done, and 0 if none of the services were done. Includes only most recent pregancies with at least one ANC visit. only most recent pregancies with at least one ANC visit. Source: DHS (2008) Source: DHS (2013) health outcomes There is significant inequity by income in health from the Philippine DHS show self-reported illness outcomes. The outcomes of the poor are worse, is higher for poor children and adults than for the both over time and by quintiles. This is particularly non-poor. notable for the maternal and child health issues covered by MDGs 4 and 5, which the Philippines did The total fertility rate remains high in the not meet by 2015. Philippines, driven by multiple factors from the supply and demand sides (Box 4.1). TFR is much The salient indicators for child health have improved higher among poorer households than among the slightly over the past decade (Figure 4.14). Both non-poor (Figure 4.15). TFR has not budged among the infant mortality rate and the under-5 child the lowest quintile in the past five years (2008 to mortality rate remained high, particularly for the 2013); it is also at least three times higher than in the poorer segment of the population. In addition, data upper quintile. Figure 4.13. Skilled birth attendance, 2008 and 2013 M A K I N G G ROW T H WO R K F O R T H E P O O R Birth deliveries assisted by skilled attendants in the last 5 Birth deliveries assisted by skilled attendants in the last 5 years, 2008 years, 2013 100 100 Percentage of birth deliveries by quintile (%) Percentage of birth deliveries by quintile (%) 97.7 98.1 92.8 93.3 90.9 89.4 80 87.3 80 84.1 83.5 77.5 76.8 78.2 74.6 73.0 74.6 69.4 70.0 60 63.4 60 52.3 51.8 40 40 43.9 39.7 34.9 20 23.5 20 0 0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Insured Uninsured Insured Noninsured Note: Skilled birth attendants include doctors, nurses, and midwives. Note: Skilled birth attendants include doctors, nurses, and midwives. Source: DHS (2008) Source: DHS (2013) 81 Figure 4.14. Infant mortality and under-five mortality rates Infant mortality rate by quintile, 2008 and 2013 Under-5 mortality rate by quintile, 2008 and 2013 70 70 60 60 59 50 50 52 Per 1,000 live births Per 1,000 live births 40 40 40 36 38 30 30 33 32 29 27 20 24 23 20 23 22 21 22 17 17 17 10 15 10 13 0 0 Poorest Q2 Q3 Q4 Richest Poorest Q2 Q3 Q4 Richest 2008 2013 2008 2013 Note: Mortality rates are calculated for the 10-year period preceding each Note: Mortality rates are calculated for the 10-year period preceding each survey year. survey year. Source: DHS 2008 and 2013 Source: DHS 2008 and 2013 There are significant unmet needs for birth control girls aged 15–19 years are either pregnant or already and birth spacing. While the TFR, in 2013, is about has a baby. While overall fertility rate has declined three children per woman, the total wanted fertility over time, adolescent fertility rate increased from rate is estimated to be 2.2 births per woman, or 27 46 in 1998 to 57 in 2013 (PSA 2014). An increase in percent lower than the actual TFR (PSA and ICF adolescent pregnancy means more early high school International 2014). The high TFR translates into a dropouts, as well as higher infant and maternal high child dependency ratio, and therefore to low mortality. Parents with many children, especially household income per capita for a given amount of parents who themselves are still adolescent, have more income from the breadwinner, and it remained a key trouble in providing adequately for their children, constraint to poverty reduction. who are more likely to be malnourished, have worse health outcomes, and not enroll in school or dropout The Philippines ranks high in Asia in terms of early, which perpetuates the cycle of poverty. adolescent fertility rate (Figure 4.16). One in ten Vicious Cycle of Inequality Figure 4.15. Total fertility rates Total fertility rates of women age 15-49 of Income and Inequality of M A K I N G G ROW T H WO R K F O R T H E P O O R Education by quintile, 2008 and 2013 6.0 Avergae number of births per woman 5.0 5.2 5.2 4.0 4.2 3.7 Increased public education spending has led to 3.0 3.3 3.1 3.3 3.0 improvements in the access to education, but the 2.7 2.0 2.4 amount the country spends is still inadequate, and 1.9 1.0 1.7 spending needs to be more efficient and effective. 0.0 Thus far, increased spending has led to only modest Poorest Q2 Q3 Q4 Richest ALL improvements in learning outcomes, with lingering 2008 2013 disparities in learning. On the demand side, Note: Fertility rates are calculated for the three years preceding each survey year. Source: DHS 2008 and 2013 82 Figure 4.16. Adolescent fertility rates in EAP countries Lao PDR 66 Philippines 57 Timor Leste 52 Thailand 52 Indonesia 51 Cambodia 49 Viet Nam 36 Mongolia 27 China 7 0 10 20 30 40 50 60 70 Source: The numbers are estimates for 2010–2015. United Nations Population Division, World Population Prospects: 2017 Revision. Box 4.1. Determinants of fertility rate Fertility rate is affected by many factors. For example, improved child survival rates can result in fewer births. Improved access to reproductive health services and commodities can improve maternal health and help households voluntarily reduce births (Dumas and Lefranc 2016). Improved access to education and employ- ment can lead to delays in age at marriages and childbirths. Advocates of speeding the demographic transition emphasize the need to speed up the voluntary reduction in fertility rates. This can be done through public policies that assist households, particularly poorer households, to achieve such a reduction. Lessons from other East Asian countries point to three critical factors in transi- tioning from high to low fertility levels: health services, family planning, and education. Hence raising school enrollments and secondary school completion rates by the poor and improving access to quality health services and family planning are essential. While the Philippines has a reproductive health law, it has not been fully implemented, and an update of the strategy and implementation is much needed. The most recent annual review of the law recommended follow-through on measures to address the unmet need for modern and responsible family planning and the high level of adolescent pregnancy to help informed parents to make their own choices and achieve their desired family size. Abrigo et al. (2017) estimated the economic gains from a full implementation of the RPRH law and suggested helping couples achieve the desired number of children can potentially have substantial M A K I N G G ROW T H WO R K F O R T H E P O O R economic benefits in terms of more rapid economic growth. The top priority is to expand access to a wide range of modern and responsible family planning, especially for the poor. Among the other recommendations in the annual review is the need to ensure that budget alloca- tions for family planning at the central level should be linked to actual demand for family planning services at the local level. The report also calls for improving commodity logistics to avoid stock outages in rural health centers, creating a national-level family planning communications strategy, and reaching women with unmet needs through clinics that can provide services beyond the operating hours of government facilities, and ensuring that health centers have dedicated and trained family planning focal points. Reducing the incidence of adolescent pregnancy requires measures beyond expanding access to family planning. The reproductive health law mandated the creation of a curriculum for Comprehensive Sexuality Education (CSE). While the CSE has been integrated into the K–12 curriculum, implementation has lagged. (continued next page) 83 Box 4.1. Determinants of fertility rate (continued) Figure 4.17. Government basic education spending In addition, the annual report recommends steps 400 to provide adolescents with reproductive health 350 services and information on the risk associated 300 with early pregnancies. Foremost among these is Php Billion (2014 Prices) 250 ensuring that teachers are trained in CSE. Another obstacle to reducing rates of adolescent pregnancy 200 is the legal restrictions on access to family planning 150 services without parental consent. The annual 100 report recommends that mechanisms be insti- 50 tuted to ensure that while parental involvement is encouraged, minors are still entitled to specific 0 services even without parental consent. Local Government National Government Sources: Christelle Dumas and Arnaud Lefranc, 2016. “ ‘Sex in Marriage Is A Divine Gift’? Evidence on the Quantity-Quality Trade-Off from the Manila Contraceptive Ban.” The World Bank Economic Review, 17 Source: World Bank 2016g December 2016. lhw055, https://doi.org/10.1093/wber/lhw055. Department of Health and the Commission on Population, Philippines, 2017. The 3rd Report on the Implementation of the Responsible Parenthood and Reproductive Health Act of 2012, Manila. package of quality improvements as envisaged by DepEd (World Bank 2016g). On the government education remains very expensive for the poor. Non- education expenditure by education level, the shares tuition education expenses, which vary little across of postsecondary technical and vocational education socioeconomic groups, are a significant burden for and training and tertiary education spending have the poorest. been quite limited (Table 4.1). Public Education Spending Trends Efficiency and Effectiveness in Spending Compared with spending about 15 years ago, recent One way to consider the potential distributional public education spending has been favorable. The impact of government spending is via benefit Enhanced Basic Education Act of 2013 set concrete incidence analysis. Benefit incidence curves targets to eliminate teacher and infrastructure provide a graphical representation of the extent to deficits for basic education. Public basic education which the beneficiaries of a particularly form of spending increased by 125 percent in real terms government spending are from poorer or wealthier M A K I N G G ROW T H WO R K F O R T H E P O O R between 2005 and 2015 and has risen by 27 percent segments of society. This analysis for the Philippines in 2016 and by 25 percent in 2017. The DepEd shows that that spending on kindergarten and receives the largest share of the national budget primary education is pro-poor, that is, it flows for education—96 percent of the overall education disproportionately to poorer households. (This budget spent (Figure 4.17). reflects the fact that poorer households have more children on average.) Secondary education The Philippine government has been spending about spending is also pro-poor but to a lesser extent. The 2–3 percent of GDP on education. This is below distribution of technical and vocational spending is spending levels in Indonesia, Malaysia, Thailand, average. Spending on tertiary education is not pro- and Vietnam. Earlier studies of the country’s public poor, due to the low enrollment rates of those from education expenditure showed that it would need poorer households at this level (Figure 4.18). more than 6 percent of GDP to implement a broad 84 such as the student-teacher ratio and the student- The country has embarked on an ambitious classroom ratio. education sector reform program, led by the 2013 Basic Education Act, which extended the basic Further improving education in the Philippines education cycle from 10 to 13 years and backed up will require more effective use of the increased those reforms with increased budgets. Between resources available. Previous work has documented 2010 and 2015, public spending on basic education that the basic education system has opportunities increased by 60 percent in real terms. The reform to improve effectiveness in teacher deployment, program halted a long-term decline in public infrastructure provision, school-based management, basic education services. Large increases in school use of maintenance and other operating expenses infrastructure and teacher hiring have improved budgets provided to schools. To these ends, school conditions, as measured by basic indicators recommendations of an earlier study include M A K I N G G ROW T H WO R K F O R T H E P O O R Table 4.1. Public education spending and its share of total government expenditures, 2016 Percentage of total government ₱ billion expenditure Pre-primary and primary education 170 5.7% Secondary education 145 4.8% Postsecondary non-tertiary education 62 ... Tertiary education 37 1.2% Other 138 4.6% Total 491 16.3% Source: Source: Department of Budget and Management 85 Figure 4.18. Public spending and benefit incidence Pre-primary and primary education Secondary education 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0% 20% 40% 60% 80% 100% Vocational education Tertiary education 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% Source: Merged FIES-LFS various years, Department of Budget and Management. Note: This plots the cumulative share of benefits from public education (y axis) by education level against the cumulative share of the population ranked by consumption (x axis). The 45-degree line shows neutrality in the distribution. Any deviation from that line suggests that the concentration of benefits from education services varies by income. If the curve is above that line, the poorer groups receive greater share of benefits from government education spending than the richer groups. improving allocation mechanisms through better do not learn. The country’s learning outcomes planning, giving schools greater authority and are the weakest among major countries in East simplifying reporting requirements, improving the Asia, based on the most recent internationally transparency of fund allocation and resource use, comparable test data (from 2003). The country’s and strengthening the role of school governing education performance has likely improved since M A K I N G G ROW T H WO R K F O R T H E P O O R councils and parent teacher associations (World then, but the results from National Achievement Bank 2016). This agenda is capped by the ongoing Tests suggest that improvement has been modest, challenge of implementing the expansion of senior particularly not at the secondary level. Average high school. The first graduates of senior high scores at the primary level have increased modestly school received their diplomas in 2018, and much over time but remained roughly flat for secondary work remains as senior high school continues to school from 2004 up through 2015. expand. Fully realizing the benefits of the expanded education cycle can be expected to take some time. The Philippines has succeeded in raising school attendance and attainment, but some students learn Two major concerns highlighted in the earlier study little in school. This experience is similar to that of for basic education emerge from the analysis in this many countries around the world. The 2018 World Poverty Assessment. The first is that many students Development Report dubs this situation a “learning 86 crisis” (World Bank 2017j). Schooling without learning Figure 4.19. National achievement test of primary and secondary education—mean percentage scores is a wasted opportunity and a great injustice: society 80% is failing many of the children who most need an education to succeed in life. 70% 60% A principal factor for learning is teachers. Earlier 50% work found that teachers in the Philippines do 40% not have the knowledge, support, and materials 30% they need to teach effectively (World Bank 2016). 20% One critical need is to improve the professional 10% development opportunities for teachers, which will raise the quality of teaching and boost student 0% learning. Teacher training in the Philippines could be improved by moving from a mass-training model to one based on a personalized, coaching approach. Primary Secondary The second major concern emerging from the Pamilya could boost enrollment among the poor. analysis in this report is the persistently high For a significant number of poor girls, marriage or dropout rates for the poor. Only half of children family matters are often cited as the reason they in the poorest quintile enroll in secondary school. do not attend school. It is likely that this largely These children will face limited job and income captures the impact of adolescent motherhood on opportunities throughout their lives. Financial schooling. Expanding access to family planning and concerns are a principal reason cited by households implementing Comprehensive Sexuality Education for why their children do not attend secondary (CSE) could reduce teenage pregnancy and keep school. This suggests that increasing the value of the more girls in school. secondary school grants paid through the Pantawid M A K I N G G ROW T H WO R K F O R T H E P O O R 87 Figure 4.20. Household monthly education expenditure Figure 4.21. Per school-age child, tuition and non- by quintile of per capita total household expenditure tuition educational expenses by quintile of per capita in 2015 total household expenditure in 2015 Richest quintile Richest quintile 4th 4th 3rd 3rd 2nd 2nd Bottom quintile Bottom quintile 0 2 4 6 8 10 12 14 0 100 200 300 400 500 Thousands Thousands Household Non-Tuition education exp. per pupil (2015 Php) Household Education Expenditure (2015 Php) Household Tuition education exp. per pupil (2015 Php) Household Other Expenditure (2015 Php) Source: Merged FIES-LFS various years Source: Merged FIES-LFS various years Private Spending on Education Vicious Cycle of Inequality and Disparities of Income and Inequality of The amount families spend to educate their children Health Care differs considerably between the poorest and the richest groups.44 The richest quintile spends more than 10 times what the poorest quintiles spend Low public spending on health is one of the main for their children’s education on a per-child basis factors in the inequitable access and quality of (Figure 4.20). About two-thirds of the education health care services. Based on National Health expenditures for the richest quintile are for private Accounts (NHA) data, the Philippines spent about school tuition. In addition, the share of educational US$131 (₱5,852) per capita and 5 percent of GDP expenses in the household expenditure of the richest on health in 2014, slightly less than other countries quintile is about 5 percent, more than double that with the same income level. As a share of general for the poorest households (Figure 4.21). Non-tuition government expenditures, public spending on M A K I N G G ROW T H WO R K F O R T H E P O O R expenses are particularly significant burdens for the health was around 9 percent, below the regional poorest. The better-off households spend more for average for East Asia and Pacific Region (11.6 each child on education inputs such as additional percent) and for LMIC (10 percent) for 2014 (NHA). learning materials, private tutors, and afterschool classes that have a significant impact on learning The pattern of health spending is even less and on their future productivity as workers. inclusive: 56 percent of the total health spending 44 The FIES reports the amount (in pesos) of household-based expenditures and, though it does not itemize expenses for education, it does report overall educational expenditure and non-tuition educational expenditures broken down by income group. There is no tuition for public schools in the Philippines, so where there are tuition expenses for households, it means that children are enrolled in private schools. Non-tuition education expenses include costs for such things as transportation, uniforms, shoes, bags, stationary, exams, and school events. 88 Figure 4.22. Health spending in the Philippines against international comparators, 2014 A. Total health spending as share of GDP B. Out-of-pocket spending on health as share of Total Health Expenditures Note: Both x and y axes logged Note: Both x and y axes logged Source: WDI; WHO Global Health Expenditure Database Source: WDI; WHO Global Health Expenditure Database C. Public health spending as share of D. Public health spending as share of General Total Health Expenditure Government Expenditure Note: Both x and y axes logged Note: Both x and y axes logged Source: WDI; WHO Global Health Expenditure Database Source: WDI; WHO Global Health Expenditure Database was household out-of-pocket spending, while only health spending as a share of total health spending 32 percent was from the public sector. The share of increased to a peak of 58 percent in 2011, and out-of-pocket spending on health is higher than gradually declined after that (Figure 4.23). Even so, M A K I N G G ROW T H WO R K F O R T H E P O O R most countries in East Asia, including countries the household out-of-pocket spending share in 2014 with much lower income levels, such as Lao Peoples (56 percent) was still higher than it was in 2005 (52 Democratic Republic, Myanmar, and Vietnam and percent). only lower than Cambodia. The Philippines could make efforts to increase public health spending to Household income remains the major driver of improve financial protection and population health inequities in the use of maternal and child health (Figure 4.22). services, despite the recent national government-led subsidy for health insurance for the poor. Across The high out-of-pocket expenditure on health households at different income levels, per capita offered the population, particularly the poor, low spending on health increased significantly between financial protection against the costs of illness. This 2009 and 2012. The change in spending for the poor situation has remained largely the same over time. was far greater than that for the non-poor and Between 2005 and 2014, household out-of-pocket flattened afterwards. Per capita household spending 89 Figure 4.23. Household out-of-pocket spending on Figure 4.24. Household spending share for health by health as share of total health spending, 1995–2014 quintiles, 2009, 2012, and 2015 60 6 Percentage of total household expenditures (%) Share of Total Health Expenditure (%) 5 55 4 50 2009 3 45 2012 2015 2 40 1 35 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 1995 1996 1997 1998 1999 2010 2011 2012 2013 2014 0 Year Poorest Q2 Q3 Q4 Richest ALL Source: World Health Organization Global Health Observatory Note: Covers all households, including those with zero spending. Excludes spending on dental services. Source: FIES 2009, 2012, and 2015 for health was ₱9,510 in 2015, almost double the care in RHUs by addressing their infrastructure and sum in 2009. The largest share increases in health capital investment needs. spending—65 percent (between 2009 and 2015)—was among the poor households. The share of health The additional effort has resulted in significant spending to total household expenditure also pro-poor expansion in coverage of PhilHealth in increased sharply in 2009–2012 and flattened after recent years. Among the poor, nearly two-thirds of 2012 (Figure 4.24). While the share spent by poorer the population had enrolled in health insurance in households was lower (2.1 percent in 2015) than the 2013 (compared with only one-fifth in 2008). Due to national average (3.5 percent in 2015), there was a the increased coverage, the poor are more likely to significant increase over the years and among all use health services when sick. However, there was quintile groups. little improvement in health outcomes. The sharp increase in out-of-pocket health expenditure might be partly related to the increase in supply-induced Expansion of Health Care Coverage demand or medicine spending, because many people are still purchasing drugs out-of-pocket, while The Philippine Health Insurance Corporation also paying copayments (formal or informal) at M A K I N G G ROW T H WO R K F O R T H E P O O R (PhilHealth) was created in 1995 to implement hospitals. Medicines were the largest component universal health coverage in the Philippines and of health spending, comprising as much as 62 to improve financial protection of the population. percent. Inpatient services were a distant second at Over the past decade, the Department of Health 27 percent in 2015. For poor households, medicine has focused its attention on expanding the coverage accounts for nearly three-quarters of household for the poor.45 This has included (a) expansion health spending. of enrollment for indigents, (b) expansion of the benefits package that would address the health PhilHealth became more pro-poor over the past five needs of the indigents, and (c) efforts to strengthen years through expansion of the benefits package to primary and maternal and child health quality of include outpatient services. Enrollment increased 45 See Annex H for more details of the pro-poor health policies in the past decade. 90 from about 42 percent of the population in 2008 As indicated earlier, there were significant to 62 percent in 2013, according to the National differences in behavior among those who had Demographic and Health Surveys. Enrollment of the insurance compared to those who did not. The bottom quintile increased from around 21 percent in increase in health insurance coverage contributed 2008 to 62 percent in 2013 (Figure 4.25). The inclusion to improvement in the use of health care services, of the near poor under the subsidized premium was particularly for the poor. For maternal health, while introduced only in 2014, and its effect will probably the use of ANC services among poorer pregnant not be seen quickly. PhilHealth institutional data women was well below the national average, poor reported overall coverage of 78 percent in 2013, and pregnant women with insurance were more likely to of about 92 percent of the population in 2015, with use ANC services than the uninsured poor. For child the poor comprising 31 percent of the membership in health, across all income levels, the insured are more 2013 and 40 percent in 2015. likely to use health services when their children are Figure 4.25. Health insurance coverage by quintile, 2008 and 2013 Health insurance coverage, 2008 Health insurance coverage, 2013 80.0 Paying 80.0 Percentage of population per quinile (%) Paying Percentage of population per quinile (%) M A K I N G G ROW T H WO R K F O R T H E P O O R Indigent Indigent 60.0 60.0 Other Health Other Health Insurance Insurance 40.0 40.0 20.0 20.0 0.0 0.0 Poorest Q2 Q3 Q4 Richest ALL Poorest Q2 Q3 Q4 Richest ALL Source: DHS (2008) Source: DHS (2013) Note: PhilHealth (paying) refers PhilHealth coverage of the population that pays for its own premiums. PhilHealth (indigent) refers to the coverage of the population whose premiums are subsidized either by the national government or other sponsors, such as LGUs. Other insurance refers to the health insurance coverage other than PhilHealth. 91 sick. The inclusion of outpatient benefits within on health doubled during 2009–2015, from 3.8 PhilHealth made it more attractive for the insured percent in 2009 to 7.9 percent in 2015. Incidence of poor to use outpatient care with no copayments. For catastrophic spending47 on health tripled for the adult inpatient care, among those advised to seek it, poorest households, and it has continued to increase more insured poor households were likely to follow in recent years. The share of the population facing through compared with the uninsured poor. catastrophic health spending also doubled, from 3.8 percent in 2009 to 7.9 percent in 2015.48 For But despite these improvements in health care households in the poorest quintile, 3.7 percent of access brought about by expansion health insurance the population belonged to households that spent coverage, health care services remained limited and more than 10 percent of their total household gaps were wide between the poor and non-poor. consumption on health in 2015, compared with 1.3 Access to health units and quality of health care percent in 2009. remain poor in rural areas. As was pointed out earlier, ANC services available to poorer pregnant women Driven by the sharp increases in out-of-pocket and are of lesser quality than those available to better-off catastrophic health spending, a larger share of the pregnant women. The uninsured seemed to remain population was pushed into poverty in 2012 and vulnerable, and the gap in access to services between 2015 than in 2009 (Figure 4.26).49 In 2009, less than insured and uninsured has widened. Figure 4.26. Impoverishing impact of health spending by quintile, 2009 and 2015 Public and Private Spending on Health Household spending on health, for both regular and catastrophic needs remains high. The share of the population pushed into poverty by health spending has doubled over the past decade. Catastrophic spending on health occurs when a household’s total out-of-pocket health payments equal or exceed 10 percent of total household spending. Catastrophic spending on health care worsened between 2009 and 2015, especially for poorer households.46 Data from the FIES show that the share of households Note: Impoverishing impacts are analyzed using national poverty lines. spending more than 10 percent of total consumption Source: FIES 2009, 2012, and 2015 M A K I N G G ROW T H WO R K F O R T H E P O O R 46 See more details in Bredenkamp and Buisman (2015). 47 Out-of-pocket spending on health is considered catastrophic if it exceeds a certain fraction of total household expenditure. For the analysis here, we focus only on the 10 percent and 25 percent thresholds. 48 The concentration index, however, indicated that there was a greater tendency for richer households than for poorer households to spend out-of-pocket on health above each corresponding threshold level across all quintiles. This is an expected outcome since the non-poor may tend to opt for private sector and/or more expensive and/or more frequent visits to health providers than the poor. This is, however, below global benchmarks (10.7 percent in 2010) and regional East Asia benchmarks (13 percent in 2010) according to World Bank preliminary estimates. 49 Impoverishing impact analysis aims to measure the impact of health care payments on living standards and income inequality by focusing on households that may have been pushed into poverty—or further into it if the household is already poor—due to spending on health care. The basic idea is that out-of-pocket spending lowers the living standards of a household by reducing the amount of income available for other items the household would want to purchase. Out-of- pocket spending may be large enough to push a household below the poverty line. To measure impoverishing effect of health spending, we compare household expenditures with out-of-pocket and without out-of-pocket spending. If we generate a counterfactual measure by subtracting out-of-pocket from total health expenditures, this will provide us a sense of what the standard of living would have been if the household had not incurred health spending. See Wagstaff and Doorslaer (2003) for more discussion. See Annex I for more details. 92 Box 4.2. Public health spending and 1 percent (0.7 percent) of households were pushed health care services into poverty by health spending. This doubled in 2012 and 2015 to around 1.5 percent. To break the cycle of poor health and poor income, public investment in health care needs to be im- proved to ensure easy access to basic good-quality care. This is particularly important for the poor and vulnerable. Ensuring that all children receive a fair Vicious Cycle of Inequality start through quality health care will help them to from the Start of Life succeed later in life and break the intergenerational trap of poverty. Three areas need to be considered: access to care, Malnutrition is a critical contributing factor to the access to quality care, and financial protection cycle of intergenerational poverty in the Philippines. against illness costs. Via the Philippines Health Its drivers are many and varied (Box 4.3). Agenda (medium term strategy, 2016–22), the Malnutrition in the womb and during the first two government is embarking on various programs to years of life inhibits brain development, resulting help move toward universal health coverage. Some further strategizing may be required, including in lower levels of schooling, reduced cognitive in the Philippines 2012 policy on health service function, and lower earnings later in life. One entitlements for the poor (PhilHealth primary care in three children in the Philippines under five is benefits, PCB1 and MDG related package). stunted—the principal marker of malnutrition—and stunting rates have been stagnant for over a decade. Consideration also needs to be given to expand- ing upon the entitlements and to subsidizing an explicit essential benefits package, especially for indigents (poor and near poor) to include diagnos- Inequality in Child Malnutrition tics, medicines, and commodities that respond to Outcomes the needs of the poor. Consideration needs to be given to developing a national strategy for quality Child malnutrition, measured by stunting and of health care improvement. Philippines Health wasting of children under age five, has improved Insurance Corporation provider payment reforms may also be considered to incentivize service per- little (Figure 4.27). The prevalence of wasting has formance and cost controls. Finally, to alleviate the been flat over time. It was 7.1 percent in 2015. The burdens of out-of-pocket payment, medical costs rate of stunting fell through the early 2000s but and cost burden on the population need to be has been flat since then. Remarkably, the measured controlled and limited-balance billing assured. To rate of stunting in 2015—33.4 percent—was slightly fully use insurance to manage financial protection M A K I N G G ROW T H WO R K F O R T H E P O O R higher than in 2005.50 against illness costs, the targeting mechanism to reach the poor and near poor needs to be refined and enrollment of all the poor in the Philippines Malnutrition is strongly associated with income Health Insurance Program needs to be ensured. levels. At the regional level, malnutrition shows These interventions could have implications on wide variation by geography. There is a strong poverty reduction as it will respond to significant correlation between the poverty rate and the rate health care access and cost faced by the poor. of stunting at the provincial level (Figure 4.28). Stunting rates are notably higher in rural areas (35 50 Stunting indicates that a child is, loosely speaking, short for his or her age. In statistical terms, a child who is stunted has a height-for-age z-score more than 2 standard deviations below the median of a healthy reference population. Wasting indicates that a child has low weight for his or her age, specifically a weight-for- age z-score more than 2 standard deviations below the median, and is a sign of acute, short-term malnutrition. The discussion here focused on stunting, which is widely used worldwide as the principal proxy for malnutrition. 93 Box 4.3. Drivers of malnutrition Evidence from worldwide studies shows that malnutrition is driven by a complex mix of factors. Surprisingly, a comprehensive multivariate analysis of drivers of malnutrition specific to the Philippines has not been conducted. Several studies, however, have described factors that could be drivers of the high levels of stunting in the Philippines.a The following are brief observations around key factors that are known to influence child malnutrition: • Maternal malnutrition. One in four pregnant women in the Philippines were categorized as “nutritionally-at- risk” in 2015, and substantial numbers are anemic or have iodine deficiencies. • Lack of quality prenatal care. The share of women who complete the recommended four prenatal health visits is high (84 percent), but there are doubts about the quality of this care. • Child nutrient deficiencies. Forty percent of children aged six months to one year are anemic, with higher percentages among the poorer households. Vitamin A and zinc deficiencies also remains high among poor quintiles. • Lack of breastfeeding. The rate of exclusive breastfeeding among children under six months has risen over time, reaching 52 percent in 2013. However, this figure hides the rapid drop-off as the child ages. The rate falls from 66 percent in the first month to only 22 percent in the sixth month. • Low quality of complementary food. By the Minimum Dietary Diversity Score, a simple measure of dietary diversity, in 2013 just 17 percent of children ages 12–17 months had consumed a minimally diverse diet the previous day. • Food insecurity. In 2013, 34 percent of households were categorized as food insecure, meaning that they did not have access to sufficient food by some measure. • Low vaccination rates. Just 69 percent of children are fully immunized. • Lack of access to clean water and sanitation. Overall, 21 percent of households lack access to a protected source of clean water, and 11 percent lack adequate sanitation. • Adolescent pregnancy. Teenage pregnancy, a risk factor for birth outcomes, has increased in recent years. About 10 percent of women ages 15–19 have either had a child or are currently pregnant. Recent studies that discuss possible factors include Philippine Food and Nutrition Research Institute (2016) and Herrin (2016). Danaei and others (2016) M A K I N G G ROW T H WO R K F O R T H E P O O R estimates the importance of various drivers by applying relationships from international evidence to national data. 94 Figure 4.27. Malnutrition trends in the Philippines for Figure 4.28. Rates of malnutrition are highest in poor children under five areas: Under-five stunting versus poverty rates in the Philippines by province 50% 70% 45% 40% 60% 35% 33.4% 50% 30% 25% Stunting Rate of Stunting 40% of children Wasting under 5 20% 30% 15% 20% 10% 7.1% 10% 5% 0% 0% 1985 1990 1995 2000 2005 2010 2015 0% 10% 20% 30% 40% 50% 60% 70% 80% Poverty headcount rate Note: Poverty rates shown are using the national poverty line, calculated with the 2015 FIES data. Source: WDI Source: Philippine Food and Nutrition Research Institute (2016) and Bank staff analysis of FIES (2015) Table 4.2. Stunting rate in the Philippines for children percent) compared with urban areas (26 percent) under five by region, 2015 (Figure 4.29). Across regions, stunting rates are highest in the ARMM, where nearly one in two Region Stunting Rate children (45 percent) are stunted, and lowest Region NCR 25% III (Central Luzon), where one in four (23 percent) Region IV-A 28% of the children are stunted (Table 4.2) There is little difference in rates of stunting by gender. Rates of Region III 23% overall stunting and severe stunting are slightly Region II 29% higher for male children than for female children. Region I 31% Cordillera Administrative Region 37% Region XI 32% Figure 4.29. Rates of stunting for children under-five by Region IV-B 41% urban/rural and gender, 2013 Region VII 38% M A K I N G G ROW T H WO R K F O R T H E P O O R Rural Region VI 40% Region V 40% Urban Region IX 38% Severely Stunted Moderately Stunted Region X 37% Female Caraga 36% Region XII 40% Male ARMM 45% 0% 10% 20% 30% 40% Source: Philippine Food Nutrition and Research Institute (2015) Source: Philippine Food Nutrition and Research Institute (2016) 95 Figure 4.30. Rates of stunting for children under five, by Figure 4.31. Rates of stunting for children under five by income quintile, 2013 age group, 2013 50% 50% 45% 45% 40% 40% 35% 35% 30% 30% 25% 25% 20% 20% 15% 15% 10% 10% 5% 5% 0% 0% Poorest Second Middle Fourth Wealthiest 0-5 6-11 12-23 24-35 36-47 48-60 Age in Months Severely Stunted Moderately Stunted Severely Stunted Moderately Stunted Source: Philippine Food Nutrition and Research Institute (2015) Source: Philippine Food Nutrition and Research Institute (2015). Note: The overall stunting figure is broken down into severe stunting (more than 3 standard deviations below the median) and moderate stunting (between 2 and 3 standard deviations below the median). Across the segments of income distribution, Costs of Child Malnutrition stunting rates vary markedly by household wealth. Children in the poorest quintile of households Global evidence shows that the long-term costs are far more likely to be stunted than those from of child malnutrition51 are high: they include an wealthier households: 45 percent of the children additional 0.05–1.6 percent of GDP for health from the poorest quintile are stunted, and 17 percent costs alone. Furthermore, childhood stunting is are severely stunted. In comparison, 13 percent of associated with adverse outcomes throughout life. children in the wealthiest 20 percent are stunted, The malnourishment and disease that are the cause and less than 4 percent are severely stunted (Figure of widespread stunting impede the development of 4.30). The stunting rates for children under five young brains. The result is impaired cognitive and varied by more detailed age groups (Figure 4.31). The socioemotional skills and lower levels of schooling. broad research on malnutrition has demonstrated Children who are stunted are more likely to face that the most critical period for long-term growth health problems later in life, resulting in higher starts at conception and extends roughly through health care costs. In addition, children who are M A K I N G G ROW T H WO R K F O R T H E P O O R the first two years of life—often described in short- stunted are much more likely to have short stature hand as the first 1,000 days. The pattern by age as adults, and independent of other factors, adult in the Philippines is compatible with this general height is correlated with higher wages. Perhaps the finding. Stunting is 13 percent among newborns (less most convincing evidence comes from long-term than six months of age), and 16 percent among those cohort studies, which have followed children from between six and twelve months old. For children in birth to adulthood in Brazil and Guatemala. Both their second year of life, the rate jumps to 32 percent find substantial effects of stunting at age two on and remains roughly flat in the following years. adult income, for both men and women. 51 Galasso and Wagstaff (2016) provides one recent survey of this evidence. 96 Box 4.4. High costs of childhood malnutrition in the Philippines The rich literature on the high costs of childhood Several studies quantify the high cost of early malnutrition in the Philippines (Box 4.4) found that childhood malnutrition in the Philippines, drawing the overall cost of child malnutrition in education from the Cebu Longitudinal Health and Nutrition and productivity channels are equivalent to 2.8 Survey (CLHNS) in the Metropolitan Cebu area. percent of the Philippines GDP. Glewwe, Jacoby, and King (2001) found that children who were better nourished in the early The returns from investments to reduce years performed significantly better in school. malnutrition are extraordinarily high in the Those children entered school earlier, and thus Philippines. Fully implementing existing nutrition had more schooling, but also had higher learning initiatives would reduce poverty. productivity per year of schooling. Daniels and Adair (2004, 2005) found that early childhood Given the combination of high levels of stunting growth deficiencies are associated with later age at entry into school, more grade repetition, and and the low costs of nutrition-specific interventions, lower completed school attainment. Mendez and the rate of return to investments for a package Adair (1999) found that early childhood stunting of such initiatives is extraordinarily high. (Figure was associated with lower IQ at age eight. Carba, 4.32) shows benefit-cost ratios for such investments Tan, and Adair (2009) found that low height-for- across a set of countries. The estimated benefit in age at young ages was associated with reduced the Philippines is 44 pesos for every 1 peso invested. likelihood of working in formal wage jobs as adults in a follow-up round to the study in 2005, when This is the second-highest rate of return across all participants were approximately age 21. countries analyzed. Note: The CLHNS was carried out in the Metropolitan Cebu area on the island of Cebu, Philippines. It has tracked over time a sample of 3,289 children born between May 1, 1983, and April 30, 1984. Figure 4.32. Rates of return to investments to reduce stunting, by country Indonesia 48 Philippines 44 India 39 Vietnam 35 Pakistan 29 Yemen 29 Nigeria 24 Sudan 23 Bangladesh 18 M A K I N G G ROW T H WO R K F O R T H E P O O R Burma 17 Kenya 15 Tanzania 15 Uganda 13 Nepal 13 Ethiopia 11 Madagascar 10 Democratic Republic of Congo 4 0 10 20 30 40 50 60 Source: Hoddinott and others (2013) 97 Map 4.1. Secondary education enrollment by province M A K I N G G ROW T H WO R K F O R T H E P O O R 98 Map 4.2. Tertiary education enrollment by province M A K I N G G ROW T H WO R K F O R T H E P O O R 99 M A K I N G G ROW T H WO R K F O R T H E P O O R 100 C HAPTER FI V E Role of Private and Public Transfers on Poverty and Inequality • Transfers from public and private sources, which represent 15 percent of total household income, have been key drivers of poverty reduction. Over 2012–2015 the effect of government transfers on poverty reduction has been almost three times that of transfers from family members. • Two-thirds of Filipinos, 15 million households, receive domestic and foreign remittances. Both transfer types have similar impacts on reducing the poverty rate (reducing it by about 4 percentage points), but domestic remittances reduce inequality, while foreign remittances increase it. • Remittance-recipient households generally spend more on health and education than non-recipients. Children of recipients tend to be enrolled in school and work less than children of non-recipients. Labor force participation tends to be lower among recipients than non-recipients, especially for those receiving foreign remittances. • The government’s Pantawid Pamilya Pilipino Program (the conditional cash transfer program) has become the primary government social assistance program for the poor. The conditional cash transfer program (CCT) extends cash grants directly to 77 percent of the poorest households and covers 20 percent of the income gap of M A K I N G G ROW T H WO R K F O R T H E P O O R poor beneficiaries. • Pantawid Pamilya contributes to reducing poverty and inequality and helps influence behavior change and build beneficiary families’ human capital. It improved school enrollment of older children and encouraged early childhood education. It also increased the health-seeking behaviors of beneficiaries. • Listahanan, the national database for the identification of poor households for the Pantawid Pamilya Program, can also be used to channel cash assistance to the poor for other purposes (such as dealing with a post-disaster emergency). 101 This chapter analyzes the distribution and impact Figure 5.1. Foreign and domestic remittances of domestic and foreign transfers. It considers 14% 70 both foreign and domestic remittances and 13% public transfers. It also examines the private 12% 65 and public transfers on building human capital Share of Households 11% Share of Income 60 through education and health care. It discusses 10% the government’s social protection programs with 9% 55 8% a focus on the national CCT, Pantawid Pamilya, 7% including its targeting mechanism and effectiveness 50 6% in promoting behavioral changes. 5% 45 2006 2009 2012 2015 Share of Income from Remittances Share of Households with Remittances Patterns and Distribution Source: Calculations based on FIES 2006 and 2015 of Domestic and Foreign types of remittance: while those from domestic Remittances sources only constitute 11 percent of total household income, those from foreign sources constitute about a quarter of total household income, and those from The Philippines is among the top destinations for both foreign and domestic sources constitute around migrant remittances in the world. Foreign migrant 32 percent. In general, domestic remittances are more remittances were US$30 billion in 2016,52 below common among households in rural areas, while only China and India in the developing world, and foreign remittances are more common in urban equivalent to 10 percent of GDP (Box 5.1).53 Foreign areas. While coverage rates of foreign remittances remittances have constantly trended upward since among urban and rural residents are 49 percent and 1998, which was important in mitigating the risks in 55 percent, respectively, they are 30 percent and 26 the 2008–2009 global financial crisis. The six million percent for foreign remittances. Filipino migrants constitute the seventh-largest diaspora in the world (surpassed by Bangladesh, Remittances from foreign and domestic sources China, India, Mexico, Pakistan, and the Russian have been increasingly important in the Philippines, Federation). In 2015, 15 million households, or two- particularly for the poor. The share of households thirds of the total household population, received receiving remittances increased from 59 percent remittances—12 million households (53 percent of in 2006 to 66 percent in 2015. As a share of total the total households) from domestic sources, 6.3 household income, they increased from 11 percent in M A K I N G G ROW T H WO R K F O R T H E P O O R million from foreign sources (28 percent of the total 2006 to 13 percent in 2015. households), and 3.2 million from both sources (14 percent of the total households) (Figure 5.1). Domestic Remittances Remittances are an important source of household income for recipients. For migrant households, 22 In terms of incidence, a large percentage of the poor percent of total household income comes from this receive remittances from domestic sources. In 2015, source. But there are substantive differences in the 23 percent of recipients of domestic remittances 52 Top origin countries for international remittances to the Philippines are the United States, United Arab Emirates, and Saudi Arabia. 53 There is no information on the aggregate size of domestic remittances because tracking these transactions is often considered unnecessary from the national accounts’ point of view (Sander 2003). Also, domestic remittances are likely to come through informal channels that are difficult to capture in official data (Castaldo and others 2012). 102 Figure 5.2. Incidence of domestic remittances by Box 5.1. Data sources for remittances income quintile 100% Estimates for remittances are based on household 90% survey data, primarily the FIES for 2006, 2009, 80% 2012, and 2015. In FIES, the amount of foreign 70% remittances received is the sum of the values 60% reported of cash from family members who are 50% overseas contract workers (OCW); cash from 40% family members who are working abroad other 30% than OCW; cash gifts, support, relief, etc. from 20% abroad. The amount of domestic remittances 10% received is cash, any gift, assistance, or relief from 0% other families domestically. However, estimates 2006 2015 from household surveys underreport actual Poorest 2 3 4 Richest foreign remittances reported in national accounts Source: Calculations based on FIES 2006 and 2015 (Ducanes 2010). In 2015, foreign remittance estimates from household data are only US$12 billion, while data from national accounts on foreign remittances report about US$24 billion. were in the bottom 20 percent of the income Similar underreporting is also present for 2012: the distribution, and just 13 percent were in the top 20 household data estimate is US$11 billion, while percent. This pattern is surprisingly stable, with that from national accounts is US$18 billion. virtually no differences with that registered back in 2006 (Figure 5.2). In terms of coverage, domestic remittances have increased across all income among those in the poorest income quintile, and quintiles, with the largest one registered among the lowest for those in the richest income quintile. The bottom 20 percent (from 51 percent in 2006 to 61 share of domestic remittance as part of household percent in 2015), while among those in the richest income also increased for those in the poorest income quintile, the increase is much more modest, quintile—from 6.2 percent in 2006 to 7.8 percent in from 30 percent in 2006 to 35 percent in 2015 (Figure 2015. This increase is not observed among those in 5.3). the richest decile, which remained at 3 percent from 2006 to 2012. The domestic remittances also account for a larger share of total household income among the poor By geographic location, Luzon had the highest compared with the rich (Figure 5.4). This indicates share of households with domestic remittances that the poor rely more on domestic remittances. from 2006 to 2015. However, the NCR had fastest From 2006 to 2015, the share of domestic remittance increase in the share of households with domestic to total household income was consistently highest remittances—from 29 percent in 2006 to 46 percent M A K I N G G ROW T H WO R K F O R T H E P O O R Figure 5.3. Share of households with domestic Figure 5.4. Share of domestic remittance income to remittances by income quantile total household income by income quantile 70% 9% 8% 60% 7% 50% 6% 40% 5% 30% 4% 3% 20% 2% 10% 1% 0% 0% Poorest 2 3 4 Richest Poorest 2 3 4 Richest 2006 2009 2012 2015 2006 2009 2012 2015 Source: Staff estimates using various rounds of FIES Source: Staff estimates using various rounds of FIES 103 Figure 5.5. Share of households with domestic Figure 5.6. Share of domestic remittance income to remittances by island group total household income by island group 60% 8% 7% 50% 6% 40% 5% 30% 4% 3% 20% 2% 10% 1% 0% 0% 2006 2009 2012 2015 2006 2009 2012 2015 NCR Luzon Visayas Mindanao NCR Luzon Visayas Mindanao Source: Staff estimates using various rounds of FIES Source: Staff estimates using various rounds of FIES in 2015. This was followed by Mindanao, which had 3.6 percent in 2006 to 5.7 percent in 2015 (Figure 5.6) an increase in the share of households with domestic remittances rising from 36 percent in 2006 to 48 At the regional level, not all the poorest regions percent in 2015 (Figure 5.5). received a large share of domestic remittances. For example, coverage of domestic remittances was Based on share of total household income, highest in one of the poorest regions in 2015, Region those living in Visayas relied more on domestic V (Bicol Region), at 69 percent (Table 5.1). At the remittances compared to those living in Luzon and same time, coverage was lowest in the two poorest Mindanao. The share of domestic remittance as part regions, ARMM and Region 12. This is consistent of total household income was consistently highest with studies that found migration from the poorest for those in Visayas from 2006 to 2015. However, areas difficult due to their remoteness and the lack Mindanao had the highest increase in share—from of human and social capital.54 Table 5.1. Regions with highest and lowest coverage of domestic remittances, 2015 Coverage Poverty incidence Region (Philippines = 22%) With highest coverage Region V – Bicol 69% 36% Region IV-B – Mimaropa 61% 24% M A K I N G G ROW T H WO R K F O R T H E P O O R Region IX – Zamboanga Peninsula 59% 34% Region I – Ilocos 58% 13% Region III – Central Luzon 57% 11% With lowest coverage Autonomous Region in Muslim Mindanao 29% 54% Region XII – Soccsksargen 39% 37% Region II – Cagayan Valley 40% 16% Cordillera Administrative Region 43% 20% National Capital Region 46% 4% Source: Calculations based on FIES 2015. 54 See de Haan and Yaqub (2008); Deshingkar and Start (2003). 104 Figure 5.7. Share of households Foreign Remittances with foreign remittances by income quintile 60% Unlike domestic remittances, foreign remittances 50% are more important to the rich than the poor. From 2006 to 2015, those in the richest income quintile 40% consistently had the highest share of households 30% with foreign remittances, and those in the poorest income quintile consistently had the lowest share. In 20% 2015, however, the share of households with foreign 10% remittances in the poorest quintile significantly increased, and they decreased in the households in 0% Poorest 2 3 4 Richest the richest quintile (Figure 5.7). 2006 2009 2012 2015 Source: Staff estimates using various rounds of FIES The share of foreign remittances in total household income is also higher among the rich compared with Figure 5.8. Share of foreign remittance income to the poor. However, from 2006 to 2015, the share total household income by income quintile of foreign remittances in total household income 20% increased for all but the highest quintile. Those in 18% the richest quintile experienced a decrease in the 16% share of foreign remittances from 15 percent in 2006 14% 12% to 13.6 percent in 2015 (Figure 5.8). 10% 8% Over time, the NCR and Luzon had the highest 6% shares of households with foreign remittances; Visayas 4% and Mindanao consistently had the lowest shares. 2% Despite this, Mindanao had the highest increase in 0% Poorest 2 3 4 Richest 2006 2009 2012 2015 Source: Staff estimates using various rounds of FIES M A K I N G G ROW T H WO R K F O R T H E P O O R 105 Figure 5.9. Share of households with foreign Figure 5.10. Share of foreign remittance income to total remittances by island group by income quintile household income by island group 18.0% 40.0% 16.0% 35.0% 14.0% 30.0% 12.0% 25.0% 10.0% 20.0% 8.0% 6.0% 15.0% 4.0% 10.0% 2.0% 5.0% 0.0% 0.0% Poorest 2 3 4 Richest 2006 2009 2012 2015 2006 2009 2012 2015 NCR Luzon Visayas Mindanao Source: Staff estimates using various rounds of FIES Source: Staff estimates using various rounds of FIES the share of households with foreign remittances— 2015. The rest of the regions saw a net increase of from 15 percent in 2006 to 21 percent by 2015 (Figure only 1 percent in the share of foreign remittances 5.9). The share of households with foreign remittances from 2006 to 2015 (Figure 5.10). also increased in Visayas from 21 percent in 2006 to 25 percent in 2015. NCR and Luzon were relatively Foreign remittances have much higher value than stable across time, increasing from 2006 to 2009 then domestic ones, though they are substantially less decreasing in 2012. pro-poor in distribution. Less than 1 percent of foreign remittances go to the bottom 20 percent Likewise, those living in the NCR and Luzon relied of the population, while two-thirds flow to the on foreign remittances more than those living richest quintile of the distribution (Figure 5.11). This in Visayas and Mindanao. The share of foreign pattern is slightly less unequal than in 2006, when remittances as part of total household income was the proportion of remittances going to the richest 20 consistently highest for Luzon from 2009 to 2015. percent was three-quarters of the total. The amount of The increase in share of foreign remittance as part of foreign remittances received by the upper 20 percent total household income was slow for all geographic was around 17 times that received by the bottom 20 regions from 2006 to 2015. It even decreased for the percent in 2015, a total that is lower than in 2006, NCR, from 8.3 percent in 2006 to 6.6 percent in when the difference reached 26 times (Figure 5.12). Figure 5.11. Distribution of foreign remittances Figure 5.12. Annual foreign remittances M A K I N G G ROW T H WO R K F O R T H E P O O R by income quintile by income quintile 100% 45,000 90% 40,000 80% 35,000 70% 30,000 60% 25,000 50% 20,000 40% 30% 15,000 20% 10,000 10% 5,000 0% 0 2006 2015 2006 2015 (2006=100) Poorest 2 3 4 Richest Q1 Q2 Q3 Q4 Q5 Source: Staff estimates using various rounds of FIES Note: In pesos, nominal value peso. Source: Staff estimates using various rounds of FIES. 106 Impact of Remittances on compared to 3.3 percent in 2006, but its impact on the poverty gap is negligible. In contrast to domestic Poverty and Human Capital remittances, a peso of foreign remittance only closes the national poverty gap by 6 centavos in 2015 and 4 centavos in 2006. Moreover, foreign remittances The generally pro-poor nature of remittances in the increased income inequality (measured by the Gini Philippines was important in reducing poverty and coefficient) by 0.3 points in 2015. Nevertheless, these supporting human capital building. figures are better than was the case in 2006, when the impact on the poverty incidence and the poverty gap were even smaller. Poverty Reduction Domestic remittances reduce poverty incidence Other Impacts by up to 3.8 percentage points in 2015 compared to 2.8 percentage points in 2006, and every peso Poverty reduction is not the only effect of remittances. of domestic remittances can close the poverty gap The effects on human and social capital building can by up to 18 percent in both years.55 In other words, help ensure sustained impact from remittances. As in without domestic remittances, the poverty incidence many other countries, remittances play an important in the Philippines would have been 25.4 percent role as a safety net, providing additional income for instead of 21.6 percent in 2015, while the poverty consumption and investment. gap would have been ₱1 instead of ₱0.82. Back in 2006, the reduction in poverty incidence would The impact of remittances on labor participation have been lower, of 2.8 percentage points, but would is usually a subject of debate. A direct comparison M A K I N G G ROW T H WO R K F O R T H E P O O R produce a similar impact on the poverty gap. Using (controlling for other individual characteristics such similar assumptions, domestic remittances reduced as age, gender, education, and area of residence) seems inequality by 1.2 percentage points in 2015. to suggest that adults (18 years old and up) from households receiving remittances (particularly those The effect of foreign remittances on the poverty receiving foreign remittances) have much lower labor incidence is similar to that of domestic remittances, force participation than non-recipients.56 In 2015, the lifting 3.8 percent of people out of poverty in 2015 conditional average labor force participation rate 55 The estimated poverty impact refers to the difference between the poverty incidence based on reported income and the poverty incidence excluding domestic remittances from reported income. A caveat in interpreting poverty impact is that the analysis does not account for the income the migrant worker would earn if he or she remained in their home. This could overstate the transfers’ impact on poverty reduction. In addition, information on remittances is only available for receiving households and not for sending households, hence the impact of remittances on consumption or welfare for sending households cannot be directly estimated. 56 See more details of the impact of remittances on labor participation, children school attendance, and consumption spending patterns in Annex J. 107 (the mean of the household labor force participation Figure 5.13. Differences in consumption patterns by rate, controlling for other factors) for non-recipients remittance-recipient status of foreign remittances was 71 percent, while that Food Clothing Housing Health Education Others 1.5% for recipients was 53 percent. The difference is less 1.0% noticeable by domestic remittance status, since the reduction in labor force participation is just 3 0.5% percentage points for recipients (65 percent) versus 0.0% non-recipients (68 percent). This difference by -0.5% remittance origin is expected, given the much larger -1.0% value amount for foreign remittances. Employment -1.5% rates and hours worked were also lower for recipients -2.0% than non-recipients. Still, some studies that have done -2.5% a more careful counterfactual analysis (that is, taking by foreign remittance status by domestic remittance status into account that migration is not a random event, but Source: Calculations based on FIES 2015 an intrahousehold labor reallocation decision) suggest that the difference in labor participation between remittance recipients and non-recipients is insignificant Overview of Social Protection if the characteristics of the migrant households are considered (see, for example, Ducanes 2012). Programs in the Philippines Children between the ages of 5 and 18 from households receiving remittances are also more likely The Philippine social protection system consists to attend school than non-recipients, while the margin of programs to provide for social welfare, social is small. The conditional average school attendance safety nets, labor market interventions, and social rate among non-recipients of foreign remittances insurance. The concept of social protection was (controlling again for other individual factors) is formally recognized in the Philippines in 2007, 89 percent, while it is 91 percent among recipients. when the Department of Social Welfare and The difference is again smaller (1 percentage point) Development (DSWD) initiated reforms in social according to domestic remittance status. Child labor57 protection to align resources to priority programs is also less common among recipient households as and projects that offered high impact in coverage, their children spend more time in school. cost-effectiveness, sustainability, and efficiency. The Pantawid Pamilya, a CCT, then became the In terms of spending patterns, there are also centerpiece of the government’s social protection noticeable differences in household behavior by framework. The national household targeting remittance-recipient status. Controlling for several system for poverty reduction, Listahanan, was M A K I N G G ROW T H WO R K F O R T H E P O O R household characteristics (income; gender, age, and piloted through the CCT program (Box 5.2). The education of household head, area of residence), 2012 enhanced social protection framework took recipients spend less on food as a share of total a broader focus on the multiplicity of risks faced spending with respect to non-recipient, while by Filipinos and called for a more effective and spending more on health, education, and housing convergent social protection operational strategy (Figure 5.13). Differences are again more noticeable (Philippines, DSWD and NEDA 2012). The CCT for foreign remittance recipients, especially in the program was counted as a social safety net program higher allocation to housing and education expenses, alongside social insurance, labor market, and social while those receiving domestic remittances tend to welfare programs and interventions. increase the most expenses related to health. 57 Child laborers are household members below age 15 who worked at least one hour in the past week. 108 Box 5.2. The Pantawid Pamilyang Pilipino Program and Listahanan The Pantawid Pamilyang Pilipino Program (Pantawid Pamilya) is the Philippines’ national conditional cash transfer program. With the DSWD as lead agency, the program was piloted in 2007 with 6,000 household ben- eficiaries. It was formally launched in 2008 with coverage of 320,000 households. Since then, the program has expanded—to 4.4 million households in 2015, or 100 percent of all poor households with children as identified by the national household targeting system for poverty reduction—Listahanan. Covering about 21 percent of the population, Pantawid Pamilya is currently the third-largest CCT program in the world, following Brazil, which covers 29 percent of its population, and Mexico, which covers 27 percent. Eligibility for Pantawid Pamilya requires that a household be poor, have either a pregnant mother or at least one child, and agree to comply with the program conditions. P antawid P amil y a P rogram C onditions Education conditions Health conditions 1. Children three–five years old enroll in preschool or 1. Children below five years of age go for monthly day care facilities and maintain school attendance of visits to health stations to receive age-appropriate at least 85 percent of school days per month. health checks and services as prescribed by the Department of Health. 2. Children 6–18* years old enroll in elementary or high school and maintain school attendance of at 2. Children 6–14 years old take deworming pills twice least 85 percent of school days per month. a year in school. 3. Pregnant women go for trimestral consultations during pregnancy. 4. Pregnant women have delivery attended by a skilled health worker. 5. Grantee and/or spouse attend/s monthly Family Development Sessions (FDS). Source: Pantawid Pamilya Operations Manual (as of September 2014). *Extension of program conditions and benefits to 15–18 year-old school children began in January 2014. M A K I N G G ROW T H WO R K F O R T H E P O O R Listahanan is a proxy means test–based targeting system. Listahanan estimates the income level of the house- hold and compares it against the government’s official income poverty threshold to determine the house- hold’s poverty status, using a standard set of household information that is easy to collect, measure, and verify. It was created in 2007 to initially identify poor households that could benefit from Pantawid Pamilya. It has since grown to become a nationwide household-based targeting system that is used to identify beneficiaries for almost all government programs targeted at the poor and vulnerable population. Listahanan currently has 11 million households in its database (out of 20 million households nationwide), of which 5.2 million house- holds were classified as poor, among which 4.4 million had children 0–14 and/or pregnant women, and were gradually enrolled in Pantawid Pamilya as its beneficiaries. Source: Excerpt from Acosta and Velarde (2015). 109 The objective of Pantawid Pamilya is twofold: system and data constraints, this chapter focuses to reduce poverty and to build human capital. on analyzing the impact of the CCT rather than It provides income support to poor households discussing the impact of other social protection through a CCT, helping them afford to meet their programs. basic needs in the short-term, and incentivizing investment in the well-being of children so they can be more productive citizens in the future and break the cycle of poverty. Impact of Pantawid Pamilya on Poverty Reduction Government spending on social protection increased rapidly as Pantawid Pamilya expanded, but it remained only a small share of GDP. A study This section presents the results of an analysis estimated that government spending on social for the Philippine CCT program using the latest protection was only 0.4 percent of GDP in 2007 National Household Survey data, including three (Manasan 2006), and this mostly went to untargeted rounds of the FIES (2009, 2012, and 2015) and the in-kind (rice) subsidies that benefited the poor and 2013 APIS of the PSA, complemented by available non-poor almost equally. In 2017, social protection administrative data on the program’s budget and accounted for 4.5 percent of the national budget, implementation. It uses benefit incidence analysis nearly tripling from a nominal ₱59 million in 2005 (BIA)58 to evaluate the targeting performance and to ₱143 million in 2017. Pantawid Pamilya used 38 progressivity of the CCT program by looking at percent of the social protection allocation in 2017 how beneficiaries and benefits (cash grants) are (Figure 5.14). However, the share of spending on distributed between poor and non-poor households social protection remained low compared with or across income or consumption groups. BIA the average in lower-middle-income countries (1.6 assesses the poverty effect of the program by percent) and most East Asian countries (only higher comparing standard poverty and inequality than the Lao Peoples Democratic Republic, Papua indicators (such as changes in poverty status, New Guinea, and Vanuatu). Given the importance poverty gap, and income distribution) with and of Pantawid Pamilya in the social protection without a program, assuming all other components of a household’s income or spending patterns remain Figure 5.14. Budget for social protection unchanged. It is important to keep this assumption 300,000 in mind because it implies that results of a BIA do not account for possible changes in behavior 250,000 due to the program intervention (for example, 200,000 potential reduction in labor income for families in a program). M A K I N G G ROW T H WO R K F O R T H E P O O R 150,000 100,000 The coverage of Pantawid Pamilya expanded rapidly, from 11 percent of the eligible poor population in 50,000 2009 to 59 percent in 2015 (Figure 5.15). Of every - four beneficiary households, three are from the 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 “bottom 40,” or those who belong to the poorest Pantawid Pamilya Total Social protection 40 percent of the income range (Q1 + Q2 in Figure Note: “Total social protection” is as defined in the Budget of Expenditures and 5.16). This means that the program was able to Sources of Financing and includes social security, welfare, and employment. maintain its poverty focus over the years. The Source: Department of Budget and Management, Budget of Expenditures and Sources of Financing (various years) CCT was also able to reach a greater share of the 58 See Annex K for details about BIA. 110 Figure 5.15. Coverage of the poor Figure 5.16. Distribution of program beneficiaries 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% 2009 2012 2013 2015 2009* 2012 2013 2015 Coverage of the Poor Undercoverage of the Poor Poorest 2 3 4 Richest Source: Staff estimates using various rounds of FIES and APIS Source: Staff estimates using various rounds of FIES and APIS bottom 20 (poorest 20 percent of the population) of to cover 15–18-year-old children of beneficiary in comparison with other CCTs around the world households was meant to augment the program’s (Figure 5.17). cash assistance. The amount of a monthly grant from Pantawid Actual grants received may be this low for various Pamilya is low (Figure 5.18). Beneficiary households reasons. One is that national surveys show that received an average monthly grant of ₱117 (US$2.43) Pantawid Pamilya households, on average, only have per person in 2015. This means that for an average two eligible children 3–18 years old. This means that beneficiary household, with six members, a Pantawid many of them can receive a maximum of ₱12,000 Pamilya household received ₱701 (US$14.60) each per year—₱6,000 for health (₱500 for 12 months) month, or about ₱8,408 (US$175.17) for the full and ₱6,000 for education (₱300 for 10 months for year of 2015. This corresponds to only 6 percent an average of 2 eligible children). Another reason of beneficiary households’ pre-transfer income may be delays in payment, which could arise from in 2015 (9 percent for households in the bottom unreported changes in household information, 20), and barely half (49 percent) of the maximum such as transfer of residence or school of eligible program entitlement of ₱17,000 per year if the children. Throughout 2016, some 7 million household has two children in elementary school transactions related to beneficiary information and one child in high school. In 2014, the expansion updates were received by program implementers. Figure 5.17. Percentage of beneficiaries in the bottom Figure 5.18. Generosity per quintile M A K I N G G ROW T H WO R K F O R T H E P O O R 20 percent 60% 25% 50% 20% 40% 15% 30% 10% 20% 5% 10% 0% 0% Philippines Poorest 1 2 3 Richest Philippines Philippines Brazil Mexico Jamaica Colombia Ecuador 2009 2012 2013 2015 (2015) (2013) (2009) (2010) (2010) (2010) (2012) Note: Percent to pre-transfer income. Source: FIES 2015, APIS 2013, ASPIRE (accessed 23 June 2017) Source: Staff estimates using various rounds of FIES and APIS 111 Figure 5.19. International comparison: generosity Pantawid Pamilya is progressive, with the greatest 25% 22.5% share of benefits going to the poorest households. The distribution of program benefits mirrors how well 20% 18.0% the program has been able to reach poor households 15.8% 15% 14.0% in the bottom 20 or 40 percent. The biggest share of 11.6% program benefits went to the poorest households (45 10% 8.7% 8.5% 8.4% percent to Q1), and the share drops to nil in higher income groups (Figure 5.20). The Lorenz curve of the 5% Pantawid Pamilya shows that the program is highly 0% progressive, that is, the poorest are receiving a higher share of program benefits than their actual share in the national income distribution (Figure 5.21). Note: Percent to pre-transfer income, among bottom 20 only. While the program remains progressive, its targeting Source: Authors calculations based on FIES 2015 and World Bank ASPIRE performance has declined over the years, because databases for other countries it is still using an old targeting database. The share A great majority (80 percent) of these transactions of program beneficiaries from the bottom 20 or were requests to update schools where children’s 40 percent and the share of benefits they receive attendance should be monitored. has continuously declined over the years. Some of the beneficiary families were living just above The Pantawid cash grants continued to decline in the poverty line (including some who crossed the real value. The grant schedule has remained fixed poverty line thanks to the CCT). Being part of the since the pilot phase in 2007, so its value has been Pantawid program as well as receiving other social eroded by inflation over the years. While it started assistance, including livelihood assistance can help at about the same level of generosity as CCTs in lower the risk that they will fall back to poverty. In other countries in its early phase of implementation part, the decline in non-poor beneficiaries reflects (estimated at 23 percent of beneficiary households’ the rollout plan (phases of expansion) of the program, income), today it lags far behind the others, even which began by covering the poorest areas where the after the additional ₱500 for high school children poorest households were located (such as in 2009), that was introduced in 2014 (Figure 5.19).59 and gradually expanding to other areas in the country where potentially fewer poor households could be Figure 5.20. Distribution of benefits, by income group Figure 5.21. Progressivity of the Pantawid Pamilya 100% 6.2% M A K I N G G ROW T H WO R K F O R T H E P O O R 98.9 100 90% 10.5% 13.1% 100 93.3 17.6% 17.1% 80% Cumulative % of benefits received 70% 28.5% 76.2 29.0% 80 60% 31.4% 50% 60 44.9 40% 73.8% 40 30% 57.3% 53.9% 20% 44.9% 20 10% 0% 0 2009 2012 2013 2015 0 20 40 60 80 100 Poorest 2 3 4 Richest Cumulative % of population Source: Staff estimates using various rounds of FIES and APIS Source: Staff estimates using various rounds of FIES and APIS 59 The 2015 estimates do capture the additional cash assistance (rice subsidy) that started in January 2017. 112 Figure 5.22. Distribution of beneficiaries, by poverty status Figure 5.23. Distribution of benefits, by poverty status 100% 100% 90% 90% 24% 32% 35% 29% 32% 80% 80% 35% 49% 50% 70% 70% 60% 60% 50% 50% 40% 40% 76% 71% 68% 65% 68% 66% 30% 30% 51% 50% 20% 20% 10% 10% 0% 0% 2009* 2012 2013 2015 2009 2012 2013 2015 Poor Non-Poor Poor Non-Poor Source: Staff estimates using various rounds of FIES and APIS Source: Staff estimates using various rounds of FIES and APIS enrolled in the program (such as in 2015). In part, it was Pantawid Pamilya has made an important also due to the use of Listahanan as a targeting system.60 contribution to reducing poverty among its 4.4 Over time, a larger share of its beneficiaries were the million beneficiary households. While the real non-poor, and a larger share of the benefits went to value and progressivity of the program’s cash grants the non-poor. In 2009, 76 percent of the program may have declined over time, the latest National beneficiaries were poor compared with only 51 percent Household Survey shows that it remains an M A K I N G G ROW T H WO R K F O R T H E P O O R in 2015 (Figure 5.22); in 2009, 71 percent of the program important resource for poor households. The ₱701 benefit went to the poor, compared with only 50 average grant received by beneficiary households percent in 2015 (Figure 5.23). This also resulted in every month allowed them to deal with one-fifth increasing leakage rates (or the share of program grants of their current income shortfall to afford to meet that went to non-poor households). While 74 percent their basic needs. Without the cash assistance from of program grants went to the bottom 20 in 2009, this Pantawid, poverty among beneficiaries would declined to 45 percent by 2015 as increasing numbers of have been higher by 6 percentage points. Thus, the less-poor households were covered by the program. program kept nearly 1.5 million poor beneficiaries out of poverty in 2015. 60 The Listahanan was intended to be updated every four years so that the list of poor households can be revalidated and refreshed, and can be used to update beneficiaries of the Pantawid. The database was updated in 2015, but, it remains unused by the government. DSWD will launch a post-implementation review of the Listahanan 2015 experience in 2018, which includes a pilot of tablet-assisted enumeration to explore the potential of using a more dynamic system in the next rounds. 113 Figure 5.24. Impact on national poverty rate Figure 5.25. Impact on national income gap 35% 35% 3.8% 30% 30% 3.4% 3.7% 1.2% 1.4% 25% 25% 1.5% 20% 20% 15% 15% 29.5% 30.9% 26.4% 26.4% 27.6% 23.1% 10% 10% 5% 5% 0% 0% 2012 2013 2015 2012 2013 2015 pre-transfer post-transfer Post-transfer Pre-transfer Source: Staff estimates using various rounds of FIES and APIS Source: Staff estimates using various rounds of FIES and APIS At the national level, the cash grants were able to Resolution No. 13 in 2013 to allow flexibility in fill 3.7 percent of the income gap of poor households extending needed financial support to poor families. and resulted in a reduction in the poverty rate by When schools, health facilities, and services may not 1.5 percentage points in 2015 (Figure 5.24 and Figure be available due to disaster, cash grants are given to 5.25). The program’s poverty focus also helped reduce CCT beneficiaries without requiring compliance national income inequality by 0.6 percentage point with program conditions for a period of one to (Figure 5.26). Among the beneficiaries, the poverty six months, depending on the extent of calamity rate was 5.6 percentage point lower with the cash or crisis. This is consistent with the practice of transfer (Figure 5.27). other country CCTs being used as a crisis-response measure.61 This facility has been used by DSWD in In addition, Pantawid Pamilya provided households many instances and was used most extensively after with a cushion against sudden income shocks in times Typhoon Yolanda in November 2013. DSWD was able of disaster and crisis. The program has a built-in to quickly release ₱550.5 million (US$12.5 million) mechanism to waive the application of education of “unconditional” Pantawid Pamilya cash grants and health conditions in times of disaster or crisis in to Yolanda-affected CCT beneficiaries between a project area. This was introduced into the program November 2013 and February 2014—just three months design through National Advisory Committee after the disaster struck. Figure 5.26. Impact on national income inequality Figure 5.27. Impact on poverty rate among beneficiaries M A K I N G G ROW T H WO R K F O R T H E P O O R 70% 70% 5.8% 60% 60% 6.5% 0.5% 0.6% 0.6% 50% 50% 5.6% 40% 40% 30% 30% 61.9% 58.0% 47.6% 49.2% 45.9% 20% 45.5% 20% 10% 10% 0% 0% 2012 2013 2015 2012 2013 2015 pre-transfer post-transfer Post-transfer Pre-transfer Source: Staff estimates using various rounds of FIES and APIS Source: Staff estimates using various rounds of FIES and APIS 61 Fiszbein, et al. (2011); Fiszbein and Schady (2009). 114 Impact of Pantawid Pamilya on Figure 5.28. Impacts on school enrollment Human Capital Building Breaking the cycle of inequality of opportunity and inequality of outcomes and the consequent intergenerational poverty trap is crucial. This section draws on the results of two impact evaluations62 of the impact of the CCT program on human capital building. Overall, it can be seen that Source: Orbeta, et al. (2014) Pantawid Pamilya has contributed to building long- term human capital in recipient households. Figure 5.29. Impacts on school attendance Pantawid Pamilya improved school enrollment of older beneficiary children (above 12 years old). The World Bank impact evaluation (World Bank 2013a) found school enrollment is higher by 5 percentage points among Pantawid children aged 12–14 relative to their comparator group. Orbeta and others (2014) found that school enrollment among Pantawid children aged 12–15 is higher by 6 percentage points than among non-Pantawid children. These years are an especially important time for children to continue schooling and transition from grade school to high Source: Orbeta, et al. (2014) school, rather than dropping out to find work. Because near universal enrollment in elementary school has been achieved in most locations, the children, Pantawid children work fewer days enrollment impact for children under age 11 in 2014 compared with non-Pantawid children. Orbeta and was limited (Figure 5.28). others showed that Pantawid children worked 7 fewer days a month than non-Pantawid children. Pantawid also has a positive impact on early childhood education. Orbeta and others showed There is no evidence that Pantawid discourages labor that, while enrollment is the same for Pantawid participation of adult members in the household. M A K I N G G ROW T H WO R K F O R T H E P O O R and non-Pantawid children in preschool, Pantawid Orbeta and others showed that the proportion of preschool children attend classes more regularly working-age household members who are employed (Figure 5.29). Meanwhile, regular attendance rates and want to have extra work is higher among for children aged six years and older are already too Pantawid households (17 percent) than among near universal to show discernible program impacts. their non-Pantawid counterparts (11 percent). For beneficiaries, this reflects high compliance with Labor force participation, employment rates, and program conditions. average number of working hours among working household members do not differ between Pantawid While the incidence of child labor is similar and non-Pantawid households. between Pantawid children and non-Pantawid 62 World Bank (2013) used a randomized control design. Orbeta, and others (2014) used a regression discontinuity design. See Annex L for details. 115 Regarding the effect of the CCT on remittances, Figure 5.30. Reported utilization of Pantawid cash grants analysis indicates no significant impact. While there might be a slight negative correlation between Medical Housing being a recipient of the CCT and the receipt of 7% 1% Debt Invest remittances, the small amount of the CCT cash Education Clothing 2% Savings 0% Alcohol Recreation 0% 25% 11% 0% grant means the effect of the grants on the amount 2% of the remittances received is limited.63 There is Other no evidence that the expansion of Pantawid will 3% Other use 3% significantly reduce remittances. Food 49% Pantawid Pamilya increased the health-seeking behavior of beneficiaries, and mothers seek better maternal care services. Orbeta and others found that Source: FIES 2012 in the past five years, around 70 percent of Pantawid mothers deliver in Department of Health–accredited health facilities, compared with only 56 percent stunting among young Pantawid children aged among non-Pantawid mothers. Both evaluations found 6–36 months. However, Orbeta and others found that the program increased the rate of postnatal no difference in overall nutritional status between check-ups (10 percentage points in the World Bank Pantawid and non-Pantawid children. study and 21 points in Orbeta and others). Both evaluations also found that more Pantawid children At the same time, Pantawid Pamilya does not induce below age six received iron supplements and vitamin beneficiaries to gamble or consume more alcohol A than non-Pantawid children, more Pantawid and tobacco. Orbeta and others found no statistical children underwent regular weight monitoring, and difference between the expenditures of Pantawid more Pantawid children aged 6–11 also received at least and non-Pantawid households for these goods, one deworming pill per year. while the World Bank study found that Pantawid households spend less on alcohol by 39 percent than Beneficiary households prioritized spending for non-Pantawid households. This is consistent with education and health. Both evaluations found beneficiaries’ reported use of the cash grants as that Pantawid Pamilya beneficiaries spend more captured in the National Household Survey in 2012 for education and health than the comparator (Figure 5.30). group. The most recent evaluation found that the annual educational expenditure per school-aged Women in the program were encouraged to try child among Pantawid households is higher by 82 modern reproductive health methods at least once percent than among non-Pantawid households.64 through family development sessions (FDS).65 M A K I N G G ROW T H WO R K F O R T H E P O O R Similarly, the World Bank study found that Pantawid Orbeta and others found that the program had a households spend more on education and health per significant impact on encouraging women to do so. capita by 34 percent and 38 percent, respectively, The program also led to an increased use of modern than non-Pantawid households. The same study and responsible family planning methods by 6 also found that Pantawid households spend more percentage points, from the baseline of 68 percent by 38 percent on protein-rich food, such as dairy among 15–49-year-old women who gave birth in the products and eggs, a behavior observed widely in last 5 years. However, the study found that the FDS CCT evaluations around the world. Consequently, leads only to increased trial use, not in sustained use the World Bank study found a reduction in severe of these methods. 63 See Annex J for details. 64 World Bank (2013) used a randomized control design. Orbeta, and others (2014) used a regression discontinuity design. See Annex K for details. 65 Discussions on family planning during an FDS seeks to inform parents about the benefits of modern and responsible family planning that could lead to 116 changes in behavior concerning reproductive health. 117 M A K I N G G ROW T H WO R K F O R T H E P O O R C HA PTE R S I X Constraints on Poverty Reduction and Potential Policy Remedies The Philippines has solid macroeconomic Making the pattern of growth more inclusive, fundamentals, and its growth prospects remain particularly providing more well-paying jobs, will positive. With a healthy current account, strong help people to achieve higher and more stable international reserves, significant fiscal space, and incomes. The government can help end the vicious low and stable inflation, the economy is in a strong cycle of unequal opportunity and outcomes that position to apply multiple policy tools to seize trap people in poverty, as well as establish mutually opportunities, mitigate regional and global shocks, reinforcing positive cycles that will create a and provide the basis for productive job creation growing middle class that is well-integrated with and poverty reduction. other groups. It can help improve service delivery for all and increase non-farm wage employment Despite economic growth over the past decade, the opportunities through increased demand for rate of poverty reduction in the Philippines has manufacturing goods and services. Finally, more lagged that of many of its East Asian neighbors. progressive and better-administered taxes can help The pace of extreme poverty reduction in the finance needed investments in both physical and Philippines averaged 0.9 percentage points per year human capital. between 2006 and 2015, less than half the 1.4 points per year decline in the developing world overall and Strong economic growth will be the basis for M A K I N G G ROW T H WO R K F O R T H E P O O R much slower than China, Indonesia, or Vietnam. In productive job creation and poverty reduction. In part the sluggish pace of poverty reduction could the long run, productivity will be fundamental. be ascribed to long-standing policy distortions, Addressing the key factors identified in this including a protracted implementation of the land report—creating more well-paying jobs; improving reform programs and unclear property rights, as productivity in all sectors, including agriculture; well as competition and labor market regulations reducing income and wealth inequality through that undermine the potential to make growth more more investment in people and skill development; inclusive. However, the main reasons poverty in rebuilding conflict-affected areas of Mindanao; the Philippines did not decline as fast as in other and better managing risk and protecting the countries in the East Asia and Pacific Region vulnerable—can help accelerate poverty reduction. include: lower pace and less pro-poor pattern of growth, high inequality of income and wealth, and disasters and conflict. 118 Constraints on Poverty Wealth is highly concentrated in a small share Reduction of the population, giving them a strong interest in maintaining the status quo. This could hinder implementation of the reforms needed to facilitate Three constraints have trapped people in a vicious more inclusive growth and poverty reduction. In cycle of “low skills and low wages” and “low the past, elite capture and corruption have been investment and low-quality jobs.” A significant corrosive, not only to public service delivery but also share of the poor are working poor in unproductive to overall political and economic performance. The jobs or involuntary underemployment. In the past risk of elite capture has limited the attractiveness decade, employment grew at roughly the same rate of long-term (foreign and domestic) investment, as the working-age population, but only a fraction particularly infrastructure investment, which has of the jobs created were well-paying jobs. During aggravated regional disparities. Long-overdue 2006–2015, the average real wage for the overall land reforms and unclear property rights have workforce66 increased only 4 percent over a period similarly discouraged investment in agriculture. of 10 years, and real wages for the highly skilled With a low rate of investment (20 percent of GDP), (with complete college educations) increased only the economy is largely driven by consumption, 2 percent, which indicates low effective demand which limits the potential for rapid structural in the labor market, particularly for the skilled. transformation and increased productivity. Addressing these challenges will require solutions on the demand side, to move up the value chain and Constraint #2: Low and inequitable distribution M A K I N G G ROW T H WO R K F O R T H E P O O R create more gainful employment opportunities, and of human capital. The learning outcomes in on the supply side, to equip the labor force with the Philippines are the weakest among major better education and skills. countries in East Asia. The country failed to meet MDG targets for child and maternal health in Constraint #1: Inequality of incomes and wealth. 2015. The poor segment of the population suffers The highly unequal distribution of incomes and disproportionately from the low endowment of wealth may have negatively affected the business human capital. Twenty percent of children under environment, limiting long-term investment, age five are malnourished and stunted.67 Children inclusive growth, and productive job creation. from poor households have limited education. In the 66 Due to the data limitation, the analysis of real wage covers the workers who reported positive wage only. The earning of those self-employed and work without paid are not included in the statistics. 67 Source: FNRI (2013 estimate based on 2013 National Nutrition Survey). 119 labor force, for the poor households, only 31 percent consumption and increased both in real terms of the poor have completed secondary education and as a share of total household income in 2009, and 2 percent tertiary education, compared with 59 despite the crises; World Bank Group 2012b)—as in percent and 15 percent of the non-poor, respectively. many other countries, the poor and vulnerable in The low level of education and skills renders the the Philippines suffered more from external shocks poor uncompetitive for productive jobs in formal than the rich, and the poverty rate spiked following sectors, such as high-end services or business process these shocks. The global financial crisis, the food and outsourcing (BPO) jobs, which require tertiary fuel crises, and several highly destructive typhoons education. This constrains the total supply of skilled in 2008–2009 increased poverty by an estimated labor, which dampens the business environment for 4 percentage points, or an additional 3 million investors, perpetuating the cycle of inequality of people, and Typhoon Yolanda alone pushed millions opportunity and inequality of outcomes. into poverty. As weather patterns shift the path of M A K I N G G ROW T H WO R K F O R T H E P O O R seasonal natural disasters, and with the possible Constraint #3: Natural disasters and conflict. intensification of the El Niño, the poorest regions of Frequent natural disasters, including deadly the country, where agriculture is the predominant typhoons that disproportionately hit poor regions, economic activity and the capacity to manage risk persistent conflicts in parts of Mindanao, and global is particularly weak, face increased vulnerability economic crises continually push vulnerable groups to shocks. The high level of natural disaster risks into poverty and jeopardize long-term human capital reduced the risk taking of households and firms, development. While the economy has been resilient through lower investment or the selection of “safer” and recovered swiftly from global crises—due to and less promising technologies, which in turn would sound macroeconomic fundamentals and strong lead to a reduction in growth and job creation.68 flows of remittances (which cushioned household 68 See Annex C for details and a review of literature. 120 Potential Policy Remedies • Improve the business environment to attract more investment. Underinvestment in human and physical capital has been a major Addressing the three key constraints and tackling constraint to improved labor productivity their adverse consequences can help clarify how and has resulted in the low quality and high best to achieve faster poverty reduction. Six ways informality of jobs. Compared with most are proposed to achieve more rapid and inclusive high-performing countries in East Asia, growth, tackle inequality of opportunity and the Philippines investment-to-GDP ratio is outcomes, reduce conflict and vulnerability, and low. Investment in productive capacity, in protect the poor. particular, has lagged in the manufacturing sector. To attract more private investment, the business environment needs to be improved, Facilitate the creation particularly through addressing institutional of more well-paying jobs. constraints, strengthening competition in key sectors, securing property rights, providing A significant share of the poor is working in jobs risk management solutions, and simplifying with very low wages or are mired in involuntary business regulations. To attract foreign and underemployment. In the past decade, employment domestic investment, the government can play grew at roughly the same rate as the working-age a key role by improving infrastructure and population, but a large portion of those jobs are basic services delivery, as well as by providing poorly paid. Nearly 95 percent of the population targeted support to the self-employed or those in the labor force is employed. However, some 20 working in small and medium-size enterprises, percent is underemployed, and to the extreme, where large numbers of the poor are employed some household might earn as little as 50–100 pesos (US$1–2) a day. Many urban poor are trapped in • Upgrade value chains to support strong low-wage and low-productivity jobs in the informal and sustainable growth. Improve labor service sector. Support for the creation of more well- productivity and moving up the value chains paying jobs, particularly semi-skilled jobs, for the are a proven basis for creating more well- majority of today’s labor force who have less than a paying jobs. The Philippines has gone from high school education, can help reduce poverty and being an agricultural economy to a (low- address inequality through higher wage incomes. M A K I N G G ROW T H WO R K F O R T H E P O O R 121 end) service economy, without developing other service-based sectors, such as tourism. a manufacturing sector. Labor productivity This, in turn, could contribute to successful growth mainly stems from within-sector transformation by creating more productive productivity growth. This is contrary to the employment opportunities, including development patterns of many neighboring opportunities with skill requirements countries in East Asia, where booming compatible with those of individuals from manufacturing sectors created large numbers poor households. of labor-intensive jobs, absorbing the surplus labor from agriculture. It is an ongoing debate whether manufacturing can still deliver Improve productivity in all sectors, the same productivity gains and well-paid especially agriculture. employment opportunities for the unskilled workers as in the past. The Philippines needs The Philippines is a middle-income country whose to find its specific niches in the services sector economy is becoming less dependent on agriculture and in regional and global value chains to for output and employment. Nevertheless, capitalize on its growing services sector and agriculture remains important for poverty reduction enhance the productivity gains from structural and employment as well as sustainable and equitable transformation. growth. Compared with many countries in the region, the sector performs below its potential • Strengthen backward and forward linkages for contributing to growth, employment, and to build on the comparative advantages of poverty reduction. Improvements in productivity, skilled labor and create jobs for the unskilled. diversification, and value-addition are crucial, as Linkages between the services sector and well as progress in making agriculture more resilient manufacturing and agriculture are critical to natural disasters and climate change. to upgrading the domestic value. This would include proficiency in English and good • Increase agricultural productivity. Over information technology skills, as well as taking the past decade, productivity growth in advantages of the time zone. In doing so, the the Philippines has lagged that of the best Philippines could leverage strong performance performers in East Asia, including China, in business process outsourcing to expand Indonesia, and Vietnam. Agricultural M A K I N G G ROW T H WO R K F O R T H E P O O R 122 productivity has been low and stagnant for 30 • Support agribusiness and broader value years. Farmers and fisherfolk remain among chain development. Within the structural the poorest in the rural areas. Reasons for transformation agenda, the role of agriculture the persistent low productivity of agriculture is evolving, although slowly. The share of include high input costs; small land sizes; agribusiness in the GDP of several countries in insufficient ability to manage rainfall variability the region undergoing structural transformation and other natural hazards; limited and untimely is higher than that of agriculture (agribusiness access to finance, applied research, and accounts for 33 percent of GDP in Indonesia, extension services; and limited connectivity 43 percent in Thailand, and 15 percent in the and links to market outlets. As evidenced in Philippines, which is higher than the agriculture other middle-income countries in the region, share in GDP).69 As agriculture’s share of GDP M A K I N G G ROW T H WO R K F O R T H E P O O R structural transformation will attract workers continues to fall and incomes and urbanization out of agriculture as the manufacturing and rise, the composition of agricultural output service sectors expand. However, agriculture changes as part of agricultural diversification. continues to be a large employer and absorption To reduce poverty in rural areas, support of surplus labor by manufacturing and service will be needed for agricultural development sectors is not undertaken at a fast pace, at and diversification through support for the least in the short run. Improving income development of agribusiness, bringing in from agriculture will help address persistent various input providers and agro-processors, poverty issues and contribute to employment distributors, and retailers for value chain opportunities in rural areas. development. 69 See Oyelaran-Oyeyinka et al. (2017) for details. 123 completing high school. Improving education Ensure that Filipinos acquire the skills quality principally requires equipping teachers they need for the 21st century economy. with the tools they need via effective training and materials. Improvements of quality In recent years, the Philippines has made admirable will help address the second challenge, by strides in education. Critical advances have been the attracting more students to stay in school. creation of both universal kindergarten and senior Other critical priorities are continuing efforts high school education, with the first cohort of grade to improve budget execution and the effective 12 students graduating in 2018. Key challenges now use of public education funds. Strengthening include making sure students in school are learning, collection of learning outcome data including reducing high dropout rates for the poor, and participation to the international standardized developing socioemotional skills. students’ assessments and use of the data to determine the direction of the ongoing basic • Boost learning in basic education overall education reform will be important. and increase secondary enrollment and completion among the poor. To close gaps • Develop socioemotional skills in addition in education, two principle challenges to traditional technical skills and cognitive remain. The first is that despite a high level skills. A recent World Bank report shows of commitment by teachers and improved the growing importance of socioemotional learning environment, learning outcomes are skills for competitiveness in the global weak. The Philippines’ experience is similar economy. A higher level of socioemotional to that of many countries around the world skills is associated with greater probability that have boosted school completion rates but of being employed and with higher daily still face quality challenges, which globally earning. Therefore, worker competitiveness constitutes what the 2018 World Development increasingly requires not only traditional Report (World Bank 2017j) terms a “learning technical and cognitive skills but also crisis.” The second challenge is that secondary improved socioemotional skills. Moreover, enrollment is low and dropout rate remain such skills are associated with the greatest high among the poor beyond primary level. wage differential among workers with low The returns to education are high at college educational levels. As a substitute for, instead levels, but many among the poor are not of complement to, traditional technical and cognitive skills, socioemotional skills can offer M A K I N G G ROW T H WO R K F O R T H E P O O R 124 a route to higher earnings for workers with (PhilHealth). However, the scope and quality limited formal education. To take advantage of care available in public facilities remains of this insight it will be necessary to develop limited and uneven. To break the cycle of poor teacher preparedness and training to actively health and poor income, public investment in foster these skills in all education and training, health care needs to be improved to ensure easy including early childhood education, K–12 access to basic good-quality care and alleviate education, and tertiary education, as well as the burdens of out-of-pocket payment. The top regular and vocational training. policy priority is to expand the essential health benefits package available to the poor. The next priority is to develop a national strategy for Invest in health and nutrition. quality of health care improvement. A third is to ensure that all of those who qualify for M A K I N G G ROW T H WO R K F O R T H E P O O R Although the Philippines aims to achieve universal PhilHealth coverage are enrolled and are aware health coverage, it still has weakness in de facto that they are insured. Limited and uneven health access and quality, rates of child malnutrition access and quality of health care contribute remain high, and the country has faced challenges to the general health challenges of the poor in implementing its reproductive health policies. A as well as to weaknesses in reproductive series of efforts in these areas are needed to boost health and nutrition as well as general health human capital and make possible a demographic challenges of the poor. dividend. • Reduce child stunting. One in three children • Boost health care quality and equity. in the Philippines under age 5 is stunted— The Philippines has made great progress the principal marker of malnutrition—and in expanding access to health care via the stunting rates have been stagnant over a Philippines Health Insurance Program decade, even as other socioeconomic indicators 125 have seen progress. Malnutrition in the womb dividend” of the sort that has been important and during the first two years of life inhibits in economic development across East Asia. The brain development, resulting in lower levels total wanted fertility rate for the Philippines of schooling, reduced cognitive function, is 2.2 births per woman, 27 percent lower than and lower earnings later in life. The returns the actual fertility rate of 3.0 (recent DHS 2017 from investments to reduce malnutrition are shows that total fertility rate has declined to 2.7 extraordinarily high in the Philippines: each births per woman). One important measure is peso invested results in a return of 44 pesos. to help households meet their need for family The Philippine Plan of Action for Nutrition planning. A recent study based on a natural provides a solid framework for tackling the experiment in Manila shows that reducing challenge. The critical needs are to focus health access to family planning increases family size interventions on the “first 1000 days” of a and decreases education attainment. Following child’s life from conception through the first through on the commitments of the 2012 RPRH two years of life, combined with multisector Law will allow informed parents to make their efforts involving education, social protection, own choices and achieve their desired family agriculture, and water and sanitation. size. A recent study estimated the economic gains from a full implementation of the RPRH • Fully implement the Responsible Parenthood law and suggested helping couples achieve the and Reproductive Health (RPRH) Law. Filipino desired number of children can potentially have women in the poorest quintile have more than substantial economic benefits in terms of more five children on average and the fertility rate rapid economic growth. Critical aspects of the has been steady in the past decades. One in ten law that need to be fully implemented include girls age 15-19 is either pregnant or already a expanding access to a wide range of modern mother. An increase in adolescent pregnancy and responsible family planning, especially for means higher maternal and infant mortality, the poor, as well as Comprehensive Sexuality as well as more school dropouts. At a macro Education to reduce teen pregnancies. level, the slow decline of fertility has robbed the Philippines the opportunity for a “demographic M A K I N G G ROW T H WO R K F O R T H E P O O R 126 Focus poverty reduction efforts on Mindanao. needed to build human capital in Mindanao and strengthen local governance. As the region is home to two-fifths of the poor, little progress on poverty is possible without inclusive • Support efforts to resolve conflict and growth in Mindanao. Five decades of violence bring peace to Mindanao. Breaking the has depressed growth and poverty reduction. cycle of insecurity and reducing the risk of Conflict has affected over 60 percent of Mindanao’s its recurrence requires a virtuous spiral of population. Over 50 percent of the population in restoring confidence in collective action ARMM lives below the poverty line. Economic between groups who have been in conflict progress and poverty reduction in the Philippines and transforming institutions to provide a will depend on the success of development in sustained level of security, justice, and jobs. Mindanao. This will mean drawing on the region’s This can be accomplished through two key untapped potential, linking lagging areas to growth steps: 1) Creating productive employment centers, and strengthening peace-building efforts opportunities, particularly for youth, who in conflict-affected areas to break the cycles of might otherwise be tempted to join extremists’ M A K I N G G ROW T H WO R K F O R T H E P O O R insecurity. armed groups or organized crime; 2) Delivering government programs and basic services • Increase broad public investment in more effectively, which could help anchor Mindanao. Increasing public investment in stabilization; and 3) Increasing programs to Mindanao to boost development in areas build human capital by expanding coverage of where the bulk of the poor live would provide basic services, including health, education and the basis for generating opportunities. As skills development. Ultimately, enduring peace three-fifths of Mindanao’s production and and development will hinge on the success of employment is driven by agricultural value a political solution that addresses the causes chains, investment is particularly needed to of violence—injustice, weak governance, support the agriculture sector and improve land dispossession, discrimination, and connectivity. Complementary efforts are sociocultural marginalization. 127 Manage risks and protect the of post-disaster support systems, including vulnerable. social safety nets, remittances, insurance, and other financial instruments can mitigate Poor people are more vulnerable to negative shocks. the well-being losses of the poorest Filipinos They are more exposed to the risks through lack from natural disasters, even without directly of resources, more sensitive to the impacts due reducing asset losses. to an inability to cope with them, and lack the capacity needed to adapt to potential risks and • Strengthen social protection systems. therefore suffer repeated setbacks. Children from The Pantawid Pamilya conditional cash poor families are particularly vulnerable not getting transfer program has helped to provide poor the needed education and health care. Providing households with much-needed financial targeted support to the poor and vulnerable to augmentation to meet basic needs, and it has mitigate shocks, build up human capital, and provided an incentive to keep poor households’ provide an effective safety net for those times children in school and healthy. It is important when it is needed, is crucial. Managing risks and to continue the cash assistance to poor cover protecting the vulnerable not only protects public all poor households with children and to investments in individuals and private assets, it also increase the amount of transfers to sustain and supports broader growth and capital accumulation enhance the gains, and to keep the convergence through reducing repeated losses of physical of government efforts—in raising demand- and human capital, and through increasing the side pressures and supply-side responses—to acceptable thresholds of natural risks for investors. maintain the program’s effectiveness in achieving outcomes. To ensure that the • Improve natural disaster risk management program keeps up with the evolving needs of systems. Poor people are more exposed to poor beneficiaries, several improvements need negative shocks—they are more likely to live to be considered. First, targeting efficiency in flood-prone areas in fragile housing, with can be improved through regular updating a large share of their meager income spent on of the roster of potential conditional cash staples—and are more vulnerable given their transfer beneficiaries in the Listahanan and lack of capacity for prevention and limited by using the most updated database. Second, ability to cope with and recover from shocks. to strengthen the impact on building human Effective disaster prevention measures can capital, it is important to move beyond access yield high returns, especially when they are to measure and monitor quality (that is, correctly designed and implemented as part monitor learning as well as school of a larger program of poverty reduction. attendance, and measure M A K I N G G ROW T H WO R K F O R T H E P O O R Early warning systems, improved access to improved nutrition as personal banking, insurance policies, and social well as growth). assistance (such as cash transfers and public works programs) can improve the capacity of individuals to cope with and recover from shocks and avoid well-being losses three-to-five times greater than their costs. Development 128 References Abrigo, Michael, Aniceto Orbeta, and Alejandro Herrin. 2017. “Analyzing the impact of the RPRH Law Implementation on Poverty Reduction in the Philippines.” Philippines Institute for Development Studies. Acosta, P., and S. Velarde. 2015. “Sa Pantawid, Malapit nang Makatawid! (With Pantawid, We Are Closer to Getting Out of Poverty!) An Update of the Philippine Conditional Cash Transfer’s Implementation Performance.” Philippines Social Protection Note, Number 8. Washington, DC: World Bank Group. Aghion, Philippe, and Peter Howitt. 1992. “A Model of Growth Through Creative Destruction.” Econometrica 60 (2): 323–51. https://doi.org/10.2307/2951599. Asian Development Bank (2014), “Republic of the Philippines National URBAN Assessment,” Asian Development Bank. Al-Samarrai, Samer. 2016. Assessing Basic Education Service Delivery in the Philippines: Public Education Expenditure Tracking and Quantitative Service Delivery Study. World Bank Working Paper AUS6799. Washington, DC: World Bank Group. Albert, Jose Ramon G. 2016. “What Do Statistics Say about Basic Education in the Philippines?” Philippines Institute for Development Studies Policy Note. No. 2016-19. Quezon City, Philippines. Anttila-Hughes, Jesse Keith, and Solomon Hsiang. 2013. “Destruction, Disinvestment, and Death: Economic and Human Losses Following Environmental Disaster.” Goldman School of Public Policy Working Paper. Berkeley, CA. M A K I N G G ROW T H WO R K F O R T H E P O O R Ballestros, M. 2010. “Linking Poverty and Environment: Evidence from Slums in Philippine Cities.” PIDS Discussion Paper Series No. 2010-33. Makati City, Philippines, Philippine Institute for Development Studies. Betcherman, Gordon (2014), “Labor Market Regulations: What Do We Know about Their Impacts in Developing Countries?” World Bank Policy Research Working Paper 6819. Bowen, Thomas. 2016. “Typhoon Yolanda (Haiyan) and the Case for Building an “Emergency Cash Transfer” Program in the Philippines.” Philippines Social Protection Note, No 10. Washington, DC: World Bank Group. 129 Bredenkamp, C., and L. Buisman. 2015. “Universal Health Coverage in the Philippines: Progress on Financial Protection Goals.” World Bank Policy Research Working Paper no. 7258. Washington, DC: World Bank Group. ———. 2017. Twenty Years of Progress on Maternal and Child Health in the Philippines: An Equity Lens. Washington DC: World Bank Group. Bredenkamp, C., V. Gomez, and S. Bales. 2016. Health Insurance Expansion in the Philippines through an Equity Lens. Washington DC: World Bank Group. Callen, Michael, Mohammad Isaqzadeh, James D. Long, and Charles Sprenger. 2014. “Violence and Risk Preference: Experimental Evidence from Afghanistan.” American Economic Review 104 (1): 123–48. https://doi.org/10.1257/aer.104.1.123. Cameron, Lisa, and Manisha Shah. 2015. “Risk-Taking Behavior in the Wake of Natural Disasters.” Journal of Human Resources 50 (2): 484–515. https://doi.org/10.3368/jhr.50.2.484. Carba, Delia B., Vivencia L. Tan, and Linda S. Adair. 2009. “Early Childhood Length-for-Age Is Associated with the Work Status of Filipino Young Adults.” Economics & Human Biology 7 (1): 7–17. Cassar, Alessandra, Andrew Healy, and Carl von Kessler. 2015. “Trust, Risk and Time Preferences after a Natural Disaster: Experimental Evidence from Thailand.” Castaldo, A., Deshingkar, P. and McKay, A. 2012. “Internal Migration, Remittances and Poverty: Evidence from Ghana And India.” Working Paper 7, Migrating out of Poverty Research Programme Consortium. Sussex: University of Sussex. Credit Suisse Research Institute. Various years. Global Wealth Report. Zurich. Danaei, Goodarz, Kathryn G. Andrews, Christopher R. Sudfeld, Günther Fink, Dana Charles McCoy, Evan Peet, Ayesha Sania, Mary C. Smith Fawzi, Majid Ezzati, and Wafaie W. Fawzi. 2016. “Risk Factors for Childhood Stunting in 137 Developing Countries: A Comparative Risk Assessment Analysis at Global, Regional, and Country Levels.” PLoS Med 13 11): e1002164. Daniels, Melissa C., and Linda S. Adair. 2005. “Breast-Feeding Influences Cognitive Development in Filipino M A K I N G G ROW T H WO R K F O R T H E P O O R Children.” The Journal of Nutrition 135 (11): 2589–2595. Daniels, Melissa C., and Linda S. Adair. 2004. “Growth in Young Filipino Children Predicts Schooling Trajectories Through High School.” The Journal of Nutrition 134 (6): 1439–1446. Datt, Gaurav. 2017. Multidimensional Poverty in the Philippine, 2004–13: Do Choices for Weighting, Identification, and Aggregation Matter? Policy Research Working Paper No. WPS 8099. Washington, DC: World Bank Group. Demery, Lionel. 2000. Benefit Incidence: A Practitioner’s Guide. Washington, DC: World Bank Group. Deshingkar, Priya, and Start, D. 2003. “Seasonal Migration for Livelihoods in India: Coping, Accumulation 130 and Exclusion.” ODI Working Paper 220. Retrieved from: http://www.odi.org.uk/publications/ working_papers/wp220.pdf Ducanes, Geoffrey. 2010. The Case of the Missing Remittances in the FEIS: Could It Be Causing Us to Mismeasure Welfare Changes? UP School of Economics Discussion Paper No. 2010-04. Quezon City. http://www.econ.upd.edu.ph/dp/index.php/dp/article/viewFile/4/4 ———. 2012. “Are Remittances Inducing Laziness in Households? A Reexamination of the Evidence.” Philippine Review of Economics 49 (2): 1-24. (http://pre.econ.upd.edu.ph/index.php/pre/article/ view/883/782 de Haan, Arjan, and Shahin Yaqub. 2008. “Migration and Poverty: Linkages, Knowledge Gaps and Policy Implications.” Retrieved from: https://www.unicef.org/socialpolicy/files/Migration_and_Poverty_ Linkages_Knowledge_and_Policy_Implications.pdf. Eeckhoudt, Louis, Christian Gollier, and Harris Schlesinger. 1996. “Changes in Background Risk and Risk Taking Behavior.” Econometrica 64 (3): 683–89. https://doi.org/10.2307/2171866. Ellsberg, Daniel. 1961. “Risk, Ambiguity, and the Savage Axioms.” The Quarterly Journal of Economics 75 (4): 643–69. https://doi.org/10.2307/1884324. Fiszbein, A., and N.R. Schady. 2009. Conditional Cash Transfers: Reducing Present and Future Poverty. World Bank Policy Research Report. Washington, DC, United States: World Bank. Fiszbein, Ariel, Dena Ringold, and Santhosh Srinivasan. 2011. “Cash Transfers, Children and the Crisis: Protecting Current and Future Investments.” Development Policy Review 29 (5): 585 – 601. Galasso, Emanuela, and Adam Wagstaff. 2016. The Economic Costs of Stunting and How to Reduce Them. World Bank Group Policy Research Note. Washington, DC: World Bank Group. GFDRR (Global Facility for Disaster Reduction and Recovery). 2014. Philippines Country Program Update. Washington, DC: World Bank Group. Glewwe, Paul, Hanan G. Jacoby, and Elizabeth M. King. 2001. “Early Childhood Nutrition and Academic Achievement: A Longitudinal Analysis.” Journal of Public Economics 81 (3): 345–368. M A K I N G G ROW T H WO R K F O R T H E P O O R Gollier, Christian, and John W. Pratt. 1996. “Risk Vulnerability and the Tempering Effect of Background Risk.” Econometrica 64 (5): 1109–23. https://doi.org/10.2307/2171958. Gollier, Christian, and Edward Schlee. 2006. “Increased Risk-Bearing with Background Risk.” Working Paper. Department of Economics, W. P. Carey School of Business, Arizona State University. http:// econpapers.repec.org/paper/asuwpaper/2132848.htm. Grossman, Gene M., and Elhanan Helpman. 1991. “Quality Ladders in the Theory of Growth.” The Review of Economic Studies 58 (1): 43–61. https://doi.org/10.2307/2298044. 131 Guiso, Luigi, and Monica Paiella. 2008. “Risk Aversion, Wealth, and Background Risk.” Journal of the European Economic Association 6 (6): 1109–50. Guiso, Luigi, Paola Sapienza, and Luigi Zingales. 2013. “Time Varying Risk Aversion.” Working Paper 19284. National Bureau of Economic Research. http://www.nber.org/papers/w19284. Hallegatte, Stephane, Mook Bangalore, and Marie Agnes Jouanjean. 2016. “Higher Losses and Slower Development in the Absence of Disaster Risk Management Investments.” World Bank Policy Research Working Paper, no. 7632. Heaton, John, and Deborah Lucas. 2000. “Portfolio Choice in the Presence of Background Risk.” The Economic Journal 110 (460): 1–26. Herrin, Alejandro N. 2016. Putting Prevention of Childhood Stunting into the Forefront of the Nutrition Agenda: A Nutrition Sector Review. Philippine Institute for Development Studies Discussion Paper Series No. 2016–21. Quezon City, Philippines. Hoddinott, John, Harold Alderman, Jere R. Behrman, Lawrence Haddad, and Susan Horton. 2013. “The Economic Rationale for Investing in Stunting Reduction.” Maternal & Child Nutrition 9 (S2): 69–82. ICF International. 2014. National Informal Settlement Upgrading Strategy, Comprehensive Assessment Report. INQUIRER. 2014. Forbes’ 10 richest in PH for 2014. Retrieved from: https://business.inquirer.net/165351/ forbes-10-richest-in-ph-for-2014 International Energy Consultants. 2012. Regional/Global Comparison of Retail Electricity Tariffs. Manila: Meralco. Kahneman, D. 2003. “A Perspective on Judgment and Choice: Mapping Bounded Rationality.” American Psychologist 58 (9): 697. Kahneman, Daniel, and Amos Tversky. 2013. “Prospect Theory: An Analysis of Decision under Risk.” In Handbook of the Fundamentals of Financial Decision Making: Part I, 99–127. World Scientific. M A K I N G G ROW T H WO R K F O R T H E P O O R Kandpal, Eeshani, Harold Alderman, Jed Friedman, Deon Filmer, Junko Onishi, and Jorge Avalos. 2016. “A Conditional Cash Transfer Program in the Philippines Reduces Severe Stunting.” The Journal of Nutrition 146 (9): 1793–1800. Keenan, J. 2013. The Grim Reality behind the Philippines’ Economic Growth. Retrieved from: http://www. theatlantic.com/international/archive/2013/05/the-grim-reality-behind-the-philippines-economic- growth/275597/ Liu, Elaine M. 2012. “Time to Change What to Sow: Risk Preferences and Technology Adoption Decisions of Cotton Farmers in China.” Review of Economics and Statistics 95 (4): 1386–1403. https://doi. org/10.1162/REST_a_00295. 132 Lusk, Jayson, and Keith Coble. 2008. “Risk Aversion in the Presence of Background Risk: Evidence from an Economic Experiment.” In Risk Aversion in Experiments, 12:315–40. Research in Experimental Economics 12. Emerald Group Publishing Limited. http://www.emeraldinsight.com/doi/abs/10.1016/ S0193-2306%2808%2900006-9. Malmendier, Ulrike, and Stefan Nagel. 2011. “Depression Babies: Do Macroeconomic Experiences Affect Risk Taking?” The Quarterly Journal of Economics 126 (1): 373–416. https://doi.org/10.1093/qje/qjq004. Manasan, R. (2006) “Review of Government Programs and Spending Priorities for Social Welfare Protection and Social Development: Phase I.” Manasan, Rosario G., Janet S. Cuenca, and Eden C Villanueva-Ruiz. 2008. Benefit Incidence of Public Spending on Education in the Philippines. Philippine Institute for Development Studies Discussion Paper series no. 2008-08. Quezon City. Mapa, Dennis S. 2015. “Demographic Sweet Spot and Dividend in the Philippines: The Window of Opportunity is Closing Fast.” http://www.bsp.gov.ph/events/pcls/downloads/2015/BSP_2b_mapa_ presentation.pdf Mendez, Michelle A., and Linda S. Adair. 1999. “Severity and Timing of Stunting in the First Two Years of Life Affect Performance on Cognitive Tests in Late Childhood.” The Journal of nutrition 129 (8): 1555–1562. Mincer, J., 1974. Schooling, Experience and Earnings. New York: Columbia University Press. Montenegro, Claudio E., and Harry Anthony Patrinos. 2014. Comparable Estimates of Returns to Schooling Around the World. World Bank Policy Research Working Paper No. WPS 7020. Washington, DC: World Bank Group. Morimoto, T. et. al. 2017. Supply Side Readiness of Primary Health Care in the Philippines. Washington DC: World Bank Group. Orbeta, Aniceto C. 2010. “The Filipino Child. Global Study on Child Poverty and Disparities. A Glimpse at the School Dropout Problem.” Philippine Institute for Development Studies Filipino Child Policy Brief; no. 4. Quezon City. M A K I N G G ROW T H WO R K F O R T H E P O O R Orbeta, A., A. Abdon, M. del Mundo, M. Tutor, M. Valera, M. Yaris. 2014. Keeping Children Healthy and in School: Evaluating the Pantawid Pamilya Using Regression Discontinuity Design. Second Wave Impact Evaluation Results. Department of Social Welfare and Development, Republic of the Philippines. Oyelaran-Oyeyinka et al. 2017. “Structural Transformation and Economic Development: Cross Regional Analysis of Industrialization and Urbanization.” Routledge Studies in Development Economics. Paredes, Karlo Paolo P. 2016. “Inequality in the Use of Maternal and Child Health Services in the Philippines: Do Pro-Poor Health Policies Result in More Equitable Use of Services?” International Journal for Equity in Health 15:181. 133 Philippine Health Insurance Corporation. 2010. “PhilHealth Adopts the Means Test Protocol of the NHTS- PR as the Source of Poor Families to be Enrolled under the Sponsored Program.” PhilHealth Board Resolution No. 1417/2010. Philippine Health Insurance Corporation. 2011. “New PhilHealth Case Rates for Selected Medical Case and Surgical Procedures and the No Balance Billing Policy.” PhilHealth Circular No. 0010/2012. Philippine Health Insurance Corporation. 2012. “Implementing Guidelines for PCB 1 Transition Period.” PhilHealth Circular No. 0020/2011. Philippines, NEDA (National Economic and Development Authority). 2013. Reconstruction Assistance on Yolanda: Build Back Better. Pasig City, Philippines: NEDA. Philippines, Department of Budget and Management. Various years. Budget of Expenditures and Sources of Financing. Republic of the Philippines. Philippines, Department of Education. Various years. Basic Education Information System. Republic of the Philippines. Philippines, Department of Health. 2011. The MNCHN Manual of Operations. Republic of the Philippines. Philippines, Department of Social Welfare and Development. 2014. Pantawid Pamilya Operations Manual: Extension of Program Conditions and Benefits to 15–18-Year-Old School Children Began in January 2014. Republic of the Philippines. Philippines, DSWD (Department of Social Welfare and Development and NEDA). 2012. Enhancing the Social Protection Operational Framework and Strategy of the Philippines. Republic of the Philippines. Philippines Food and Nutrition Research Institute. 2015. An Assessment of the Nutrition Situation in the Philippines. Taguig: Republic of the Philippines. ——. 2016. Presentation of 2015 Updating Survey Results. Taguig: Republic of the Philippines. PSA (Philippines, National Statistics Authority). 2014. Annual Poverty Indicators Survey. Republic of the Philippines. M A K I N G G ROW T H WO R K F O R T H E P O O R ———. Various years. Family Income and Expenditure Survey. Republic of the Philippines. ———. Various years. Labor Force Survey. Republic of the Philippines. PSA and ICF Macro. 2009. National Demographic and Health Survey 2008. Calverton, Maryland: PSA and ICF Macro. PSA and ICF International. 2014. Philippines National Demographic and Health Survey 2013. Manila, Philippines, and Rockville, Maryland, United States: PSA and ICF International. PISA (Program for International Student Assessment). A Triennial International Survey that Aims to Evaluate Education Systems Worldwide by Testing the Skills and Knowledge of 15-Year-Old Students. Retrieved from: http://www.oecd.org/pisa/ 134 Romer, Paul M. 1990. “Endogenous Technological Change.” Journal of Political Economy 98 (5): S71–102. Rosenzweig, Mark R., and Oded Stark. 1989. “Consumption Smoothing, Migration, and Marriage: Evidence from Rural India.” Journal of Political Economy 97 (4): 905. https://doi.org/10.1086/261633. Sander, C. 2003. Migrant Remittances to Developing Countries. Retrieved from: https://www. microfinancegateway.org/sites/default/files/mfg-en-paper-migrant-remittances-to-developing- countries-2003.pdf Save the Children Philippines. 2016. Cost of Hunger: Philippines. The Economic Impact of Child Undernutrition on Education and Productivity in the Philippines. Makati City, Philippines: Save the Children Philippines. Shaw, Kathryn L. 1996. “An Empirical Analysis of Risk Aversion and Income Growth.” Journal of Labor Economics 14 (4): 626–53. Tabunda, Ana Maria L. Albert, Jose Ramon G. Angeles-Agdeppa, Imelda. 2016. Results of an Impact Evaluation Study on DepED’s School-Based Feeding Program. Philippines Institute for Development Studies. Taruc, P. 2015. “A Tale of Two Economies: Exclusive Growth in the Philippines.” CNN Philippines. Retrieved from: http://cnnphilippines.com/business/2015/02/26/a-tale-of-two-economies-exclusive-growth.html Tella, Rafael Di, Sebastian Galiant, and Ernesto Schargrodsky. 2007. “The Formation of Beliefs: Evidence from the Allocation of Land Titles to Squatters.” The Quarterly Journal of Economics 122 (1): 209–41. https://doi.org/10.1162/qjec.122.1.209. TIMSS (Trends in International Mathematics and Science Study). A Survey Assessing in Grade 4, 8, and the Final Year of Secondary Schools. Retrieved from: https://timssandpirls.bc.edu/about.html Trostel, Philip, Ian Walker, and Paul Woolley. 2002. “Estimates of the Economic Return to Schooling for 28 Countries.” Labour Economics 9: 1–16. Tversky, A., and D. Kahneman. 1974. “Judgment under Uncertainty: Heuristics and Biases.” Science 185 (4157): 1124–1131. M A K I N G G ROW T H WO R K F O R T H E P O O R United Nations University and Alliance Development Works. 2014. World Risk Index, 2014. Berlin and Bonn. UNSDR (United Nations Office for Disaster Risk Reduction) 2015. Global Assessment Report on Disaster Risk Reduction 2015—Making Development Sustainable: The Future of Disaster Risk Management. Geneva: UNSDR. Voors, Maarten J., Eleonora E. M. Nillesen, Philip Verwimp, Erwin H. Bulte, Robert Lensink, and Daan P. Van Soest. 2012. “Violent Conflict and Behavior: A Field Experiment in Burundi.” American Economic Review 102 (2): 941–64. https://doi.org/10.1257/aer.102.2.941. Wagstaff, A., and E. van Doorslaer. 2003. “Catastrophe and Impoverishment in Paying for Health Care: with Applications to Vietnam, 1993-1998.” Health Economics 12: 921-34. 135 Wealth-X. 2014. Wealth-X and UBS Ultra Wealth Report 2014. Retrieved from: http://www.wealthx.com/ articles/2016/the-wealth-x-world-ultra-wealth-report-2014 World Bank Group. 2001a. Philippines Poverty Assessment, Volume 1, Main report. Washington, DC: World Bank Group. Retrieved from: http://documents.worldbank.org/curated/en/558641468780279884/Main-report ———. 2001b. Philippines Poverty Assessment, Volume 2, Methodology. Washington, DC: World Bank Group. Retrieved from: http://documents.worldbank.org/curated/en/964981468776393595/ Methodology ———. 2012a. Philippines: Basic Education Public Expenditure Review. Washington, DC: World Bank Group. ———. 2012b. Philippines Development Report 2012. Washington DC: World Bank Group. ———. 2013a. Philippine Conditional Cash Transfer Program Impact Evaluation 2012. Washington DC: World Bank Group. ———. 2013b. Philippine Development Report: Creating More and Better Jobs. Washington, DC: World Bank Group. ———. 2015. Enterprise Surveys: Philippines 2015 Country Profile. Washington, DC: World Bank Group. ———. 2016a. Breaking the Link Between Extreme Weather and Extreme Poverty. Washington, DC: World Bank Group. ———. 2016b. East Asia and Pacific Update: Reducing Vulnerabilities. Washington, DC: World Bank Group. ———. 2016c. Employment and Poverty Report. Washington, DC: World Bank Group. ———. 2016d. “Metro Manila Slum Survey.” Washington, DC: World Bank Group. ———. 2016e. Philippines Economic Update: Outperforming the Region and Managing the Transition. Washington, DC: World Bank Group. ———. 2016f. Republic of the Philippines Labor Market Review: Employment and Productivity. Washington, M A K I N G G ROW T H WO R K F O R T H E P O O R DC: World Bank Group. ———. 2016g. Assessing basic education service delivery in the Philippines: public education expenditure tracking and quantitative service delivery study. Washington, DC: World Bank Group. ———. 2017a. Atlas of Sustainable Development Goals 2017. Washington, DC: World Bank Group. ———.2017b. Developing Socioemotional Skills for the Philippines’ Labor Market. Washington, DC: World Bank Group. ———. 2017c. Doing Business 2017: Equal Opportunity for All. Washington, DC: World Bank Group. 136 ———. 2017d. Philippines Economic Update. Washington, DC: World Bank Group. ———. 2017e. Navigating Informality: Perils and Prospects in Metro Manila’s Slums. Washington, DC: World Bank Group. ———. 2017f. Unbreakable: Building the Resilience of the Poor in the Face of Natural Disasters. Washington, DC: World Bank Group. ———. 2017g. World Bank East Asia and Pacific Economic Update: Sustaining Resilience. Washington, DC: World Bank Group. ———. 2017h. World Development Indicators. Retrieved from: http://databank.worldbank.org/data/reports. aspx?source=world-development-indicators (Accessed: May 2017) ———. 2017i. Philippine Mindanao Jobs Report: a strategy for Mindanao Regional Development. Washington, DC: World Bank Group. ———. 2017j. World Development Report 2018: Learning to Realize Education’s Promise. Washington, DC: World Bank Group. ———. 2018. The State of Social Safety Nets 2018, Washington, DC: World Bank Group. ———. Forthcoming a. Case Studies of Extreme Poverty in the Philippines. Washington, DC: World Bank Group. ———. Forthcoming b. Richer and Fairer: An East Asian Miracle for the 21st Century, Washington, DC: World Bank Group. ———. Various years. Philippine Economic Update. Washington DC: World Bank Group. World Health Organization, Government Health Expenditures Database. Retrieved from: http://apps.who. int/nha/database (Accessed: May 2017). Yamauchi, Futoshi, Takiko Igarashi, Nicholas M. Tenazas, and Marites Manlangit Tiongco. 2016. Alternative and Inclusive Learning in the Philippines. Washington, DC: World Bank Group. M A K I N G G ROW T H WO R K F O R T H E P O O R 137 M A K I N G G ROW T H WO R K F O R T H E P O O R 138 A nnex A Intersectoral Labor Allocation Compared with other East Asian countries, including China, Indonesia, Malaysia, and Thailand, labor productivity growth in the Philippines disproportionately relies on within-sector productivity growth (Figure A.1). In other words, the reallocation of labor toward sectors with higher productivity (or “static reallocation”) or faster productivity growth (or “dynamic reallocation”), including from agriculture toward non-agricultural activities such as manufacturing, construction, and services, was more limited in the Philippines over the past decade than in many other East Asian countries. Within-sector productivity growth was consistently the major driving force for labor productivity growth (Figure A.2). In 2015, roughly 100 percent of the labor productivity growth stemmed from within-sector productivity growth. The role of resource reallocation or structural transformation was negligible. Over the seven-year period 2006–2012 (when comparable data are available), the contribution of dynamic reallocation to labor productivity growth was negative in four years. In other words, resources were reallocated to the sectors with slower growth. Figure A.1. Intersectoral labor allocation in selected East Asian countries 10 Static reallocation Dynamic reallocation Within-sector growth 8 6 4 M A K I N G G ROW T H WO R K F O R T H E P O O R 2 0 -2 -4 -6 1999-2007 1999-2007 1999-2007 1999-2007 1999-2007 1990-96 1996-99 2007-10 1990-96 1996-99 2007-12 1990-96 1996-99 2007-11 1990-96 1996-99 2007-12 1990-96 1996-99 2007-11 China Indonesia Malaysia Philippines Thailand Source: cited from EAP update, April 2017, page 66. Staff estimates based on data from the Groningen Growth and Development Centre 10-Sector Database; www. rug.nl/ggdc/productivity/10-sector 139 Figure A.2. Intersectoral labor reallocation in the Philippines 2006 2007 2008 2009 2010 2011 2012 100% 80% 60% 40% 20% 0% -20% -40% -60% -80% -100% Within-sector growth Dynamic reallocation Static reallocation Source: reproduced drawing from the data used in EAP Update, April 2017 M A K I N G G ROW T H WO R K F O R T H E P O O R 140 A nnex B Income Structure of Agriculture Households and Agriculture Sector Income and Employment Shares Income structure of agriculture households Agriculture accounts for nearly three-quarters of total household income. Of this, three-fifths is from enterprise activities. Other income sources that matter are remittances (mostly from domestic sources), salaries from non-agriculture sources and rental value of dwellings. Figure B.1. Components of Agriculture Household Incomes Others 2% Non-agri enterprise 3% Agri salaries Non-agri salaries 24% 7% Pensions & retirement benefits M A K I N G G ROW T H WO R K F O R T H E P O O R 0% Rental value of owner- Agriculture occupied dwelling 73% Agri enterprise 6% 43% Transfers and remittances Subsistence and net share of crops 9% 6% 141 Table B.1. Components of household income Poor agri Agri in rural Income source Agri HHs Rural HHs Poor HHs HHs HHs Salaries and wages Non-agriculture 7% 36% 26% 5% 6% Agriculture 24% 6% 15% 30% 22% Entrepreneurial incomes Non-agriculture 3% 13% 9% 2% 3% Agriculture 43% 12% 18% 36% 44% Transfers and remittances Foreign 2% 10% 2% 1% 2% Domestic Government 3% 2% 6% 6% 4% Private institutions 0% 0% 0% 0% 0% Other households 4% 5% 7% 4% 4% Rental value of owner-occupied dwelling 6% 7% 7% 6% 6% Pensions and retirement benefits 0% 3% 1% 0% 0% Other agriculture-related sources 6% 3% 5% 7% 6% Others 2% 4% 3% 2% 2% Agriculture sub-sector income shares Incomes reported in FIES can be classified into sub-categories. The following table shows the respective share of every sub-sector for income grouping. For wages, over 40 percent comes from seasonal jobs in agriculture, followed by regular agricultural labor sources with 36 percent. For agriculture enterprises, majority comprises crop farming (70 percent) and fisheries (21 percent). M A K I N G G ROW T H WO R K F O R T H E P O O R Trade and manufacturing account for almost three-quarters of non-agriculture enterprises. They are very relevant particularly for poor agricultural households with 56 percent engaged in trading and 17 percent in manufacturing (processing and/or marketing of their agircultural products). Other significant source are transportation and communication which are closely linked to the primary source of household income. Subsistence farming among agriculture households accounts for a small share of total income (about 4 percent) of which half is from logging and a third from cultivated crops. Most farm households are selling, processing, and marketing some of their produce. 142 Table B.2. Breakdown of income shares by household Agric. HHs Rural HHs Poor HHs Agri-Poor HHs Agri-rural HHs Wages agriculture regular 36% 6% 14% 34% 33% agriculture seasonal 42% 8% 22% 52% 45% non-agriculture regular 13% 70% 36% 6% 13% non-agriculture seasonal 9% 17% 28% 9% 9% Entrepreneurial Income - Agriculture crop farming 70% 71% 69% 70% 70% livestock 7% 10% 6% 5% 7% fisheries 21% 17% 20% 22% 21% Entrepreneurial Income - Non-agriculture trade 62% 59% 44% 56% 62% manufacturing 11% 8% 11% 17% 12% community, social, recreation and 7% 9% 7% 6% 7% personal services transport and communication 16% 19% 33% 19% 15% mining and quarrying 1% 1% 2% 1% 1% construction 1% 1% 1% 1% 1% Subsistence crop farming 25% 30% 28% 25% 25% livestock 13% 11% 11% 12% 13% fisheries 8% 6% 6% 7% 8% logging 54% 52% 55% 55% 54% hunting 0% 0% 0% 1% 0% Sub-sector employment shares M A K I N G G ROW T H WO R K F O R T H E P O O R Characteristics of employed members of agriculture-dependent households can be extracted from the corresponding Labor Force Survey of the FIES. About one-fifth of the employed are rice farmers, followed by corn farmers, coconut farmers, vegetable growers, and fisherfolk. Agriculture services account for about 15 percent of employment. Non-agriculture related jobs comprise about 16 percent of employment shares. Sectors related to agriculture account for about 5 percent while the rest are low-skilled jobs such as construction and domestic services. 143 Figure B.2. Employment shares of agricultural household members by sub-sector activity, 2015 livestock fishing 3% 11% agri services 15% other crops 4% agroprocessing 1% banana 3% trade of agri goods 4% sugar 4% domestic service 2% Other 16% vegetables construction 7% 2% others coconut 7% 7% corn rice 12% 18% M A K I N G G ROW T H WO R K F O R T H E P O O R 144 A nnex C Natural Disaster Risk Effects on Investment Managing risks and protecting the vulnerable does not merely protect public investments in individuals and private assets, but also contributes to broad growth and capital accumulation in several important ways. First, effective disaster risk management programs reduce the repeated losses of capital that occur every year in the Philippines and hinder the accumulation of assets and the development of resilience by individuals and firms. Second, they increase productivity by protecting human capital from the secondary impacts of frequent shocks, including health, the costs of which may exceed direct asset losses. Third, effective disaster risk management strategies promote investments by providing investors with acceptable levels of natural risks and visibility on the support they would receive should they be affected by a disaster. The contribution of risk taking (for example, through investments, innovation, or entrepreneurship) to economic growth is well-established in the economic literature and was grounded on the theory of endogenous technical change (Aghion and Howitt 1992, Grossman and Helpman 1991, Romer 1990). If the presence of natural risk leads to a reduction in risk taking by households and firms, through lower investment or the selection of “safer” and less promising technologies, then it would lead to a reduction in growth and job creation. Risk aversion has been linked to lower investment in physical and human capital (Rosenzweig and Stark 1989), wage growth (Shaw 1996), and technology adoption (Liu 2012) thereby reducing growth and economic development potential. If high natural risks lead individuals to become less inclined to take risks through M A K I N G G ROW T H WO R K F O R T H E P O O R innovation, education, or entrepreneurship, growth and development will suffer. Gollier’s seminal work (Eeckhoudt, Gollier, and Schlesinger 1996; Gollier and Pratt 1996; Gollier and Schlee 2006) finds, under general conditions, that a higher level of “background risk” (here, the risk of flood or drought) makes individuals less willing to take risks in other domains, such as innovation or entrepreneurship. In other words, being exposed to one risk increases an individual’s risk aversion regarding other categories of risk. These results suggest that households consider their vulnerability to natural risks like floods and droughts when making other risk-related decisions in other domains, such as creating a business or migrating to a city. Empirical work finds that higher levels of background risk are associated with increased risk aversion in financial decisions (Guiso and Paiella 2008, Heaton and Lucas 2000, Lusk and Coble 2008). More 145 recent literature also finds evidence of risk vulnerability with regards to land reform (Tella, Galiant, and Schargrodsky 2007), early life financial experiences (Malmendier and Nagel 2011), stock market crises (Guiso, Sapienza, and Zingales 2013), and violent trauma (Callen et al. 2014 and Voors et al. 2012). There are two mechanisms through which an increase in the background risk can lead to high risk aversion and lower investment in growth and development. The first is perfectly rational: there is a possibility that the two independent risks (one related to disasters, the other to risk taking in general) materialize together (Gollier and Pratt 1996). This combined risk—and the non-linearity in the utility function—increases risk aversion because a large income shock changes not just an individual’s location on the utility function, but also the shape of that function (Cassar, Healy, and von Kessler 2015). The second mechanism is behavioral. A shock such as a flood can lead to an overestimation in an individual’s perceived likelihood of future natural shocks occurring. Cameron and Shah (2015) find, after a flood in Indonesia, that an individual’s expectation of future flood occurrence is an order of magnitude higher than the true probability. Emotional responses can lead individuals to have greater fear of any negative event, reducing risk taking (Cassar, Healy, and von Kessler 2015). The result of this effect is that people, firms, and investors will tend to reduce their risk taking and investments in location exposed to large natural risks, reducing economic growth and job creation. It means that actions to reduce natural risks—or to provide better tools and instruments to manage them—will likely increase investments and growth (Hallegatte, Bangalore, and Jouanjean 2016). Compounding these effects is the role of “aversion to ambiguity.” Ambiguity refers to situations when there is no appropriate data available to support decision making (Ellsberg 1961). It is the case for instance in flood- prone areas, when the probability of occurrence of a flood is unknown. Or when firms and households are uncertain about how much support they would get from the government and community if a flood occurs. People usually show a large aversion to ambiguity and try to avoid ambiguous situations (Ellsberg 1961, D. Kahneman 2003, Daniel Kahneman and Tversky 2013, Tversky and Kahneman 1974). Practically it means that firms or investors having to choose an investment location will tend to select a location with less ambiguity, that is a better knowledge of the level of risk and of the contingent plans in case of disasters. In a world where locations are in tough competition to attract investments, reducing ambiguity and risk with good data on natural risks and appropriate instruments to manage disasters can be M A K I N G G ROW T H WO R K F O R T H E P O O R an important comparative advantage. 146 A nnex D The Poor Suffered Greater Loss of Well-Being for any Given Asset Loss A socioeconomic resilience assessment conducted by the government found that the Philippines suffers asset loss of around ₱182 billion, and well-being losses (or impact on quality of life) of around ₱208 billion per year due to natural disasters. However, while the asset losses of the poorest Filipinos account for only 7 percent of total asset losses (₱12.2 billion per year), they suffer 27 percent of the total well-being losses (₱56 billion per year). The well-being of the poorest Filipinos is disproportionately affected by natural disasters because their livelihoods depend on fewer assets and their consumption is closer to subsistence levels. They cannot rely on savings to smooth the impacts of losses, placing their health and education at greater risk and potentially requiring more time to recover and reconstruct. For these reasons, the same peso amount of asset losses has a greater impact on the well-being of the poor than of the non-poor. For example, a once-every-25-year typhoon in Manila causes ₱2,700 in asset losses per capita for the poorest quintile, while the wealthiest quintile loses assets worth ₱16,600 per capita. However, these losses affect the poorest and wealthiest residents of the capital very differently: equivalent well-being are nearly four times higher than asset losses (₱10,200 per capita) for the poorest quintile, while the wealthiest quintile experiences well-being losses of roughly a third of asset losses (₱4,600 per capita). Disaster Figure D.1.Disaster losses losses from afrom in Manila in Manila a once-every-25 once-every-25 year typhoon year typhoon M A K I N G G ROW T H WO R K F O R T H E P O O R 18,000 16,000 Poorest 14,000 Disaster losses (Php per capita) quintile 12,000 Q2 10,000 Q3 8,000 Q4 6,000 4,000 Richest quintile 2,000 0 Asset loss Well-being loss 147 Socioeconomic capacity, defined as the ratio of asset to well-being losses, measures the capacity of individuals to minimize the effects of natural disasters on their well-being. For example, a population with socioeconomic capacity twice as large as another will experience half the well-being losses for the same asset losses. The metric is defined for each province in the Philippines and varies widely across regions. Due to factors that condition the resilience of a region, such as quality of housing and infrastructure, financial inclusion, social protection, diversification, early warning systems, and remittances, regions in eastern Visayas and Mindanao are characterized by lower socioeconomic capacity. Despite their relative ability to cope with disasters, well-being losses in Luzon and the Eastern Visayas are high due to those regions’ elevated exposure to typhoons and earthquakes. M A K I N G G ROW T H WO R K F O R T H E P O O R 148 A nnex E Seasonality and Employment Dynamics Seasonality of employment is an important part of the employment dynamic in the Philippines: and individual can have four possible employment statuses during a year. Workers likely shift from one status to another, including shifts in the sector of employment. A panel70 is constructed from the Labor Force Survey to trace employment dynamics in four points during the year. This is beyond reporting primary employment in cross-sectional analysis which only captures one point. This analysis provides a better understanding of the annual incomes reported in FIES. In analyzing the panel, the following operational definition is applied to evaluate employment status (Table E.1). Table E.1. Employment Categories Previous Quarter Previous Quarter Definition July January of July of January Persistent Employed (Employed in all periods) Employed Employed Employed Employed Transient Employed (At least one period of being not in the Employed Not in LF Employed Employed Labor Force, but employed in other periods) Transient Unemployed (At least one period of being un- Unemployed Employed Not in LF Unemployed employed, but employed or not in the labor force in other periods) M A K I N G G ROW T H WO R K F O R T H E P O O R Persistent Unemployed (Unemployed in all periods) Unemployed Unemployed Unemployed Unemployed Source: Piza, Edillon and del Mundo (2016). Persistent and transient employment (defined above as transient employed or transient unemployed) is the focus of the analysis. Based on the panel created for 2015, about three-quarters are persistently employed. The proportion is lower for the bottom quintile with about three tenths. This reflects the bigger number of seasonal workers. 70 This occurs in years the FIES is collected. Since the FIES is a rider to the LFS, respondents are visited twice where each visit covers a semester. The other two data points in the panel are based on recall of employment status in the previous quarter which are asked in LFS. The observations pertain to the first months of the quarter. The corresponding July and January LFS rounds were used to create the panel. Unique household identifiers, age, gender and relationship to household head were used to extract the unique member specific observations. 149 A dissection of the panel shows the following characteristics of these type of employment (see Table E.2). Location. Transient employment is more prevalent in rural areas. Likewise, there are more persistently employed in rural areas compared to urban. The National Capital Region and peripheral regions of CALBARZON and Central Luzon have the highest shares of persistent employment. Transient employment is evenly spread in regions with vibrant non-agriculture sector but still with a significant dependence on agriculture (CALABARZON, Western Visayas, Central Luzon, Ilocos Region). Transient unemployed is also highest in regions with persistent employment. These are mostly short-term contracts offered in retail and services. Table E.2. Characteristics of the employed  Characteristic Persistent Employed Transient Employed Transient Unemployed Urbanity Urban 40.4 29 43.3 Rural 59.6 71 56.7 Region Region I - Ilocos Region 5.3 8.5 5.7 Region II - Cagayan Valley 4 5.4 2.9 Region III - Central Luzon 10.7 8.5 11.8 Region V- Bicol 5.8 6.9 6.1 Region VI - Western Visayas 8.6 9.2 7.8 Region VII - Central Visayas 6.5 4.5 6.4 Region VIII - Eastern Visayas 4.3 6.3 4.8 Region IX - Zamboanga Peninsula 3.5 2.6 3.3 Region X - Northern Mindanao 4.9 6.5 4.6 Region XI - Davao 4.2 4 4.4 Region XII - SOCCSKSARGEN 4.5 5.9 3.7 National Capital Region 11.8 6.1 12.2 Cordillera Administrative Region 1.7 3.5 1.3 Autonomous Region in Muslim 4.1 2.5 5.4 Mindanao M A K I N G G ROW T H WO R K F O R T H E P O O R Region XIII - Caraga 2.9 4.1 2.8 Region IVA - CALABARZON 13.7 10.3 14 Region IVB - MIMAROPA 3.6 5.3 2.9 Gender Male 63.84 40.1 34.6 Female 36.2 59.9 65.4 Age Group 15 to 24 12.7 23.5 45.8 25 to 34 24.8 23.3 19 35 to 44 27.5 19.5 13.5 45 to 54 22.4 18.4 10.9 55 over 10.8 13.5 9.9 150 Table E.2. Characteristics of the employed (continued)  Characteristic Persistent Employed Transient Employed Transient Unemployed Education No Grade Completed 1.5 1.6 1.6 Elementary Undergraduate 14.6 15.6 8.1 Elementary Graduate 14.1 15.5 9.7 High School Undergraduate 12.4 17.9 21.4 High School Graduate 32.3 32.6 32.2 College Undergraduate 8.9 9.8 18.5 College graduate 16.1 7 8.5 Income Quintile Poorest 17.4 24.6 19.1 2nd 18.8 23 20.1 3rd 19.8 21.8 21.3 4th 21 19 21 Richest 23 11.7 18.5 Demographic characteristics. Persistently employed males outnumber females three to two. The opposite is the case for transient employment. With regards to age, those in their prime (35-44) have the higher share of persistent employment. Transient employed are predominantly with the younger age groups (15-24 and 25-34) and those with high school diploma. Economic status. Higher shares of transient employment are concentrated in the bottom quintiles. Conversely, there are more persistent employment among those in the higher quintiles. Among those employed in all quarters (see Table E.3), about 90 percent are employed in the same sector and majority of which are in services (about 37 percent in 2015). In contrast, agriculture is still the leading sector among those in the bottom quintile. Note however, that the share has dropped considerably from 49 percent in 2006 to 40 percent in 2015. The share of those in services have increased but not as much as the decline in agriculture. M A K I N G G ROW T H WO R K F O R T H E P O O R In the bottom quintile, 2-percentage point increase in the mixed sector with agriculture happened between 2012 and 2015. This suggests that those engaged in agriculture seized opportunities to engage in non- agriculture employment to augment their incomes during off season. Moreover, the increasing share of employment in services sector among transient workers is an indication of better employment prospects for the poor. Recall that we have sustained economic growth during the latter period. The increasing human capital may have paved for more employment options. Collectively, these changes in the employment structure have contributed to the change in the composition of household incomes. 151 Table E.3. Employment classification by sector 2006 2009 2012 2015 All households Persistent Agriculture 26% 24% 22% 19% Industry 9% 9% 9% 9% Services 33% 36% 36% 37% Mix w/ agriculture 6% 6% 6% 6% Mix w/o agriculture 2% 2% 4% 4% Transient Agriculture 8% 7% 7% 7% Industry 4% 3% 3% 4% Services 11% 11% 12% 13% Mix w/ agriculture 1% 1% 1% 1% Mix w/o agriculture 1% 0% 1% 1% Bottom quintile Persistent Agriculture 49% 48% 45% 40% Industry 4% 5% 4% 5% Services 11% 13% 13% 14% Mix w/ agriculture 9% 9% 9% 11% Mix w/o agriculture 1% 1% 2% 2% Transient Agriculture 16% 14% 15% 14% Industry 2% 2% 2% 3% Services 6% 7% 8% 9% Mix w/ agriculture 2% 2% 2% 2% Mix w/o agriculture 0% 0% 0% 1% Source: Staff estimates from constructed panel based on several LFS rounds M A K I N G G ROW T H WO R K F O R T H E P O O R 152 A nnex F Minimum Wage Labor regulations in the Philippines are comprehensive and strict, but they cover only a relatively small fraction of the workforce. Minimum wage is high relative to the median wage in most regions in the Philippines (the minimum wage varies by administrative region as well as by sector and type of establishment). It is found to be high by several measures, both relative to Filipino workers’ productivity and to minimum wage rates in other countries with similar levels of economic development (World Bank 2013 and Betcherman 2014). Nine out of 17 regions have a minimum wage that is higher than the median wage (World Bank 2016). Minimum wage is set at a high level because it is meant to serve as a social safety net. The minimum wage for private firms is set at an amount that would cover the needs of workers and their families. To account for these needs, the government introduced the two-tier wage system in 2012, whereby the first tier is the mandatory regional wage floor while the second tier is an amount that is a guide for employers to adjust wages above the floor. The latest reform aimed to set the wage floors close to the poverty thresholds in order for the minimum wage to serve as a social safety net among wage workers. Consequently, the number of minimum wages below the poverty threshold was greatly reduced. But in fact, informality severely limits the actual coverage of minimum wage policy. Less than half (45 percent) of wage workers in private firms are employed in formal firms (World Bank 2016).71 Of these wage workers, about 75 percent are paid equal or above the minimum wage. Therefore, only about one-third of workers in private firms are actually covered by the minimum wage policy. In the informal sector, the minimum wage accounts for about 115 percent of the sector’s average wage, which is so high that it is likely to discourage informal firms from formalizing their activity. Aligning minimum wage with worker productivity could improve the chances of low-skilled workers being hired formally. In setting minimum wage, it is advisable to consider wage distribution not only in the formal sector but also in the informal sector and set it at a level that does not cut deeply into the overall wage distribution. Admittedly, this is difficult in a two-tier labor market, like the one in the Philippines. The wage distributions are M A K I N G G ROW T H WO R K F O R T H E P O O R very different in the upper, formal tier, and in the lower, informal tier. The minimum wage set based on the wage distribution in the formal sector, as it is currently the case, is too high to be used in the informal sector, where labor productivity is low. On the other hand, if the minimum wage were set based on the wage distribution in the informal sector, it would be too low to be meaningful for formal workers. Some compromise is necessary to strengthen the incentives for employers to hire low-skilled workers formally. An empirically informed discussion among social partners is needed to find a middle ground (World Bank 2013, 2016). 71 The World Bank 2016 report defines formal employment as follows: In the case of wage employment, the criteria used to distinguish between formal and informal employment are: a) having a written employment contract, b) payment of social security contributions by the employer, and c) protection from job dismissal. This definition considers a job formal when at least two of the three criteria are met. In the case of the self-employed, the enterprise is considered formal when it maintains proper bookkeeping and accounting practices. By assumption, unpaid family workers are considered informally employed. 153 A nnex G Returns to Education Estimation approach People make decisions on their own schooling or children’s schooling in anticipation of benefits that can be realized in the future. Benefits include both economic ones such as labor wage and non-economic ones, such as satisfaction in life, additional skills, and other individual values. In the Philippines, not all people are completing basic education, and limited numbers go to vocational training and college education. As discussed before, education attainment is particularly low among the poorest. In this section, we focus on the returns to education in the Philippines using the latest Labor Force Survey data to review returns to education using the Labor Force Survey. We estimate the private returns based on using the conventional Mincer (1974) model of earnings (the human capital earnings function), which has log wage rates determined by years of schooling or level of education (elementary, secondary, postsecondary and non-tertiary vocational, and tertiary education0, age or experience and other explanatory variables. Then, we estimate marginal probability for adults to work as wage earners. Both estimations are conducted first with all adults and then with sub-groups such as females and males, people residing in rural areas and urban areas, or different island groups (Luzon except NCR, NCR, Visayas, and Mindanao). The private rate of return compares the costs and benefits of schooling as incurred and realized by the individual student who undertakes the investment. Mincer (1974) has provided a great service and convenience in estimating returns to schooling by means of the semi-log earnings function (see also Becker and Chiswick (1966). The standard method to estimate private returns per year of schooling is to estimate log earnings equations of the form: M A K I N G G ROW T H WO R K F O R T H E P O O R ln(wi) = a + β1Si + β2 Xi + β3 Xi2 + μi (1) where ln(wi) is the natural log (of hourly or annual, depending on data) earnings for the ith individual; Si is years of schooling (as a continuous variable); Xi is labor market potential experience (estimated as agei - Si- 6); Xi^2 is potential experience-squared; and μi is a random disturbance term reflecting unobserved abilities. Therefore, β1 can be viewed as the average rate of return to years of schooling to wage employment. The list of control variables is kept deliberately small to avoid overcorrecting for factors that are correlated with years of schooling. This is also known as the “Mincerian” method (Mincer, 1974). The earnings function method can be used to estimate returns at different schooling levels by converting the continuous years of schooling variable (S) into a series of dummy variables, say Dp, Ds and Dt (where p is primary schooling, s is secondary schooling and t is tertiary) to denote the fact that a person has achieved that level of schooling. The omitted level is people with no schooling and that dummy is not in the equation 154 to avoid matrix singularity. The estimation equation in this case is of the form: ln(wi ) = a + βp Dpi + βs Dsi + βt Dti + β1 Xi + β2 Xi2 + μi (2) After fitting this “extended earnings function” (using the above dummies instead of years of schooling in the earnings function), the private rate of return to different levels of schooling can be derived from the following formulas: rp = (βp)/(Sp) (3) rs = (βs - βp)/(Ss - Sp) (4) rt = (βt - βs)/(St - Ss) (5) where Sp, Ss and St stand for the total number of years of schooling for each successive level. The sample for wage estimation only includes wage earners formally employed and omits unemployed and informal sector workers. Regression results from this estimation are summarized below: Regression output tables Table G.1. Returns to education to another year of schooling (OLS) (1) (2) (3) (4) (5) Wage (ln_daypay) Total Male Female Urban Rural Years of schooling 0.110*** 0.0887*** 0.159*** 0.121*** 0.102*** (116.1) (86.72) (82.31) (83.15) (82.29) Work experience 0.0232*** 0.0257*** 0.0184*** 0.0209*** 0.0256*** (32.73) (31.28) (14.47) (19.10) (27.29) Work experience (squared) -0.000320*** -0.000353*** -0.000221*** -0.000290*** -0.000355*** (-22.36) (-21.75) (-8.245) (-12.52) (-19.42) Urban - dummy -0.218*** -0.238*** -0.186*** M A K I N G G ROW T H WO R K F O R T H E P O O R (-37.74) (-36.65) (-17.11) Female - dummy -0.109*** -0.114*** -0.109*** (-17.53) (-13.03) (-12.43) Constant 4.767*** 4.959*** 4.089*** 4.453*** 4.376*** (287.2) (264.8) (125.3) (224.8) (269.0) Observations 35,675 23,714 11,961 16,112 19,563 R-squared 0.350 0.332 0.419 0.313 0.277 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 155 Table G.2. Returns to education by education level (1) (2) (3) (4) (5) Wage (ln_daypay) Total Male Female Urban Rural Attained primary education 0.140*** 0.174*** 0.0405 0.183*** 0.127*** (13.94) (17.12) (1.481) (9.745) (10.85) Attained secondary education 0.255*** 0.221*** 0.406*** 0.301*** 0.211*** (34.11) (28.13) (22.75) (24.96) (22.04) Attained vocational training 0.221*** 0.175*** 0.286*** 0.195*** 0.245*** (16.87) (10.97) (12.77) (11.12) (12.36) Attained tertiary education 0.745*** 0.638*** 0.833*** 0.665*** 0.853*** (104.0) (69.56) (70.65) (68.56) (79.47) Work experience 0.0252*** 0.0273*** 0.0224*** 0.0237*** 0.0262*** (37.04) (34.13) (18.14) (22.08) (29.76) Work experience (squared) -0.000401*** -0.000411*** -0.000392*** -0.000386*** -0.000411*** (-29.26) (-25.99) (-15.00) (-16.98) (-23.96) Urban - dummy -0.235*** -0.245*** -0.217*** (-42.57) (-38.87) (-20.76) Female - dummy -0.158*** -0.140*** -0.183*** (-26.30) (-16.31) (-21.79) Constant 5.408*** 5.398*** 5.197*** 5.134*** 4.945*** (363.7) (325.0) (158.9) (263.2) (355.0) Observations 35,675 23,714 11,961 16,112 19,563 R-squared 0.404 0.371 0.460 0.344 0.362 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table G.3. Returns to education to another year of schooling (OLS) - island groups M A K I N G G ROW T H WO R K F O R T H E P O O R (1) (2) (3) (4) (5) Wage (ln_daypay) Total NCR Luzon Visayas Mindanao Years of schooling 0.110*** 0.116*** 0.107*** 0.104*** 0.104*** (116.1) (43.77) (74.82) (48.28) (56.61) Work experience 0.0232*** 0.0111*** 0.0226*** 0.0265*** 0.0286*** (32.73) (6.043) (22.09) (15.71) (19.10) Work experience (squared) -0.000320*** -0.000167*** -0.000312*** -0.000384*** -0.000366*** (-22.36) (-4.203) (-15.06) (-11.73) (-12.32) Female - dummy -0.109*** -0.132*** -0.115*** -0.0877*** -0.107*** (-17.53) (-9.672) (-12.95) (-5.574) (-7.838) 156 Table G.3. Returns to education to another year of schooling (OLS) - island groups (continued) Urban - dummy -0.218*** -0.180*** -0.103*** -0.136*** (-37.74) (-21.07) (-6.947) (-11.07) Constant 4.767*** 4.802*** 4.770*** 4.509*** 4.505*** (287.2) (131.0) (192.1) (112.4) (136.6) Observations 35,675 4,958 15,654 6,381 8,682 R-squared 0.350 0.289 0.319 0.309 0.325 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table G.4. Returns to education by educational level (OLS) – island groups (1) (2) (3) (4) (5) Wage (ln_daypay) Total NCR Luzon Visayas Mindanao Attained primary education 0.140*** 0.0566 0.0940*** 0.116*** 0.116*** (13.94) (1.338) (6.034) (5.465) (6.448) Attained secondary education 0.255*** 0.253*** 0.226*** 0.238*** 0.223*** (34.11) (12.31) (21.72) (12.84) (14.51) Attained vocational training 0.221*** 0.231*** 0.230*** 0.215*** 0.188*** (16.87) (8.131) (12.65) (6.491) (5.995) Attained tertiary education 0.745*** 0.577*** 0.755*** 0.810*** 0.849*** (104.0) (39.04) (71.98) (44.10) (54.66) work experience 0.0252*** 0.0134*** 0.0247*** 0.0268*** 0.0295*** (37.04) (7.383) (25.36) (16.81) (21.12) work experience (squared) -0.000401*** -0.000247*** -0.000399*** -0.000445*** -0.000432*** (-29.26) (-6.282) (-20.14) (-14.35) (-15.62) female -0.158*** -0.143*** -0.164*** -0.150*** -0.193*** (-26.30) (-10.64) (-19.25) (-9.965) (-14.90) M A K I N G G ROW T H WO R K F O R T H E P O O R Urban/Rural -0.235*** -0.194*** -0.0996*** -0.156*** (-42.57) (-23.77) (-7.109) (-13.71) Constant 5.408*** 5.630*** 5.448*** 5.082*** 5.121*** (363.7) (131.9) (239.9) (144.1) (178.8) Observations 35,675 4,958 15,654 6,381 8,682 R-squared 0.404 0.313 0.383 0.384 0.415 t-statistics in parentheses *** p<0.01, ** p<0.05, * p<0.1 157 Table G.5. Marginal probability for wage employment by years of schooling (Probit) (1) (2) (3) (4) (5) Variables Total Male Female Urban Rural Years of schooling 0.024*** 0.017*** 0.034*** 0.020*** 0.026*** -0.001 -0.001 -0.001 -0.001 -0.001 Work experience 0.017*** 0.015*** 0.020*** 0.015*** 0.019*** 0 0 -0.001 -0.001 0 Work experience (squared) -0.000*** -0.000*** -0.000*** -0.000*** -0.000*** 0 0 0 0 0 Female - dummy 0.051*** 0.066*** 0.038*** -0.003 -0.005 -0.004 Urban - dummy 0.057*** 0.057*** 0.055*** -0.004 -0.005 -0.007 Observations 78,355 47,336 31,019 28,673 49,682 -287.2 -131 -192.1 -112.4 -136.6 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Table G.6. Marginal probability for wage employment by education level (Probit) (1) (2) (3) (4) (5) Variables Total Male Female Urban Rural Attained primary education 0.038*** 0.013** 0.098*** 0.047*** 0.034*** -0.005 -0.006 -0.009 -0.01 -0.006 Attained secondary education 0.072*** 0.057*** 0.097*** 0.049*** 0.083*** -0.004 -0.005 -0.008 -0.007 -0.006 Attained vocational training 0.045*** 0.035*** 0.051*** 0.023* 0.065*** -0.009 -0.013 -0.013 -0.012 -0.013 Attained tertiary education 0.156*** 0.148*** 0.156*** 0.102*** 0.220*** M A K I N G G ROW T H WO R K F O R T H E P O O R -0.005 -0.007 -0.007 -0.007 -0.008 Work experience 0.018*** 0.016*** 0.020*** 0.015*** 0.020*** 0 0 -0.001 -0.001 0 Work experience (squared) -0.000*** -0.000*** -0.000*** -0.000*** -0.000*** 0 0 0 0 0 Female - dummy 0.047*** 0.065*** 0.031*** -0.003 -0.005 -0.004 Urban - dummy 0.056*** 0.057*** 0.051*** -0.004 -0.005 -0.007 Observations 78,355 47,336 31,019 28,673 49,682 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 158 A nnex H Pro-Poor Health Policies Introduced by the Government, 2005–2015 • In the 2008, DOH instituted major reforms for reduction of maternal and neonatal mortality, known as the Maternal, Neonatal, and Child Health and Nutrition (MNCHN) strategy. A major component of this was a behavior change strategy that will enable all pregnant women to see antenatal care and postnatal care and deliver in a facility equipped with emergency obstetric care. • In 2010, the formal adoption by the PhilHealth program list of those identified as the “poor” and the “near poor” from national household targeting system for poverty reduction (NHTS-PR) of the Department of Social Welfare and Development (DSWD). • In 2011, introduction of the Department of Health policy to provide PhilHealth registration for all indigents through subsidized premiums, and in 2014 to include within the category of “indigents” those considered near poor. • In 2011, implementation of a no-balance billing policy wherein indigent members of PhilHealth are not to pay anything beyond what is paid by PhilHealth when confined in government hospitals. • In 2012, significant increase in the budget (through revenue from earmarked tobacco taxes) for subsidizing PhilHealth registration of the indigents. • In 2012, expansion of PhilHealth’s outpatient primary care benefit package72 delivered by rural health units for the indigents (serving urban and rural poor population). • In 2012, DOH embarked on a Health Facilities Enhancement Program (HFEP) to accelerate the supply-side readiness to provide health services, including maternal and neonatal care. The HFEP involved resources to be poured into infrastructure and equipment for facilities based according to health facilities plans based on mapping and needs assessment checklists. M A K I N G G ROW T H WO R K F O R T H E P O O R • The MNCHN strategy was complemented by changes in PhilHealth’s packages for maternity care such as introducing expansion of coverage from first two deliveries up to the fourth delivery (2006 and 2008), introducing a Newborn Care Package (2006), expanding the coverage from the first two deliveries up to the fourth delivery (2008), paying a bigger reimbursement for deliveries in non-hospital settings than in for hospital deliveries (2011), unbundling antenatal care as a separate package from delivery (2014), and reintegrating pregnant women who are not yet members or with inactive membership by providing immediate coverage with no requirement for waiting time. • The conditional cash transfers program introduced by the DSWD included within it for conditions for increasing antenatal care and delivery under supervision of skilled birth attendants. 72 This was expanded since 2012 into the current primary care benefit (PCB1) covers a range of essential outpatient services such as screening for non- communicable diseases (NCDs) and diagnosis and treatment for common infectious disease conditions (e.g. asthma, acute gastroenteritis, upper respiratory tract infection), including dispensing of some basic drugs for these conditions. 159 A nnex I Benefit Incidence Analysis BIA is a standard approach for examining who benefits and how much from public expenditure on various programs, in particular in social protection, health, and education. Tt is a statistical method for computing the distribution of public expenditure across different population groups, such as poor and non-poor, men and women, children and elderly, and so on. BIA does this by combining the unit cost of providing the service with information on the use or receipt of the service. In some countries, BIAs for social assistance programs have been effective tools to justify reforms that eliminate ineffective programs and replace them with better-targeted ones. Early applications of BIA in government-wide reforms include education and water and sanitation systems in Colombia and health systems in Malaysia and Ghana in the 1990s. Based on empirical evidence from BIAs, Indonesia in 2005 initiated dialogue to reform ineffective general subsidies, such as those for petroleum, and reallocated funds to health, education, and a new cash transfer program (Indrawati 2005). Like any other research methodology, BIA has advantages and limitations. Among the advantages is that BIA does not require specialized and usually expensive data collection. BIA uses existing national household surveys for the data needed to undertake the analysis, which at the minimum requires the following information: (i) indicator of household participation to the program, and (ii) amount of assistance received from the program. If these two pieces of information are available from national household surveys, BIA can be a very straightforward exercise, which is its second most appealing trait. By comparing certain indicators with and without the program—or post-transfer versus pre-transfer scenarios—BIA can estimate program coverage of various segments of the national population as well as answer important questions for policy makers such as the program’s impact on overall poverty and income distribution. Findings from a BIA thus apply to the entire country unlike many highly-specialized surveys. Unlike rigorous impact evaluation, which assesses a program’s impact on specific development outcomes M A K I N G G ROW T H WO R K F O R T H E P O O R (such as education completion, child stunting, and wasting) due to intended behavioral changes induced by a particular program, BIA does not account for behavioral changes. It assumes that, except the provision of cash assistance to beneficiaries, everything else remained unchanged in the living conditions of both beneficiaries and non-beneficiaries.73 This is the third in a planned series of BIA for the Pantawid Pamilya that was conducted by the World Bank. The first BIA was done in 2012 and updated in 2015.74 DSWD conducted its own Pantawid BIA in 2013. The most recent available national household surveys that could be used for this study is the 2015 FIES produced by the Philippine Statistical Authority. The 2015 FIES collected information on the amount of cash grant 73 For more information on benefit incidence analysis, see Demery (2000). 160 received by beneficiary households for the reference year (2015). It captured a nationally representative sample of the CCT beneficiaries. The FIES 2015 captured an unweighted sample of 9,366 household beneficiaries of Pantawid Pamilya or about 23 percent of the total survey sample, which fully represented 4.4 million Pantawid household beneficiaries nationwide. For this BIA, results presented are mostly from the FIES 2015. It will be complemented by results from previous rounds of national surveys and available administrative data, as needed. FIES 2009 rider question: “Is any member of your household a beneficiary of the Pantawid Pamilyang Pilipino Program?” FIES 2012 and 2015 Income module: “During the period (Jan–Dec 2012), did you or any member of your family receive in cash any gift, support, assistance, or relief from the Pantawid Pamilyang Pilipino Program in cash.” APIS 2013 (social protection module): “During the period January–June 2013, has any member of your family received benefits/payments from the Pantawid Pamilyang Pilipino Program? How much was received in cash from the Pantawid Pamilyang Pilipino Program in the last 6 months” Weighted CCT Year Sample households CCT sample households Weighted households households 2015 41,544 9,366 21,980,141 4,479,955 2013 10,864 1,845 21,892,397 3,360,409 2012 40,171 6,722 20,056,813 2,902,477 2009 18,451,414 420,096 Source: PSA (FIES 2009, 2012, 2015; APIS 2013). M A K I N G G ROW T H WO R K F O R T H E P O O R 74 Fernandez and Velarde (2012); Acosta and Velarde (2015). 161 A nnex J Impact of Remittances Estimation approach Remittances increase the income of recipients and reduce liquidity constraints that may affect their decisions on labor supply, school attendance, and consumption. Household recipients are expected to increase their consumption of goods and services, and investments in human capital, such as sending their children to school. Meanwhile, remittances may encourage dependency among recipients and reduce labor supply. We estimate the impact of remittances on household decision outcomes using Family Income and Expenditure Survey – 2015 and Labor Force Survey – January 2016. Generally, the models include an indicator for remittance recipients, indicator for recipients of a conditional cash transfer (CCT), and characteristics of the household, household head and members, which vary depending on what is appropriate for the sample population. For binary outcome variables, we estimate using the probit model: Pr (Y(i,j)=1│X(i,j),Hj,Rj ) = ϕ (X(i,j)^’ β1 + Hj^’ β2 + Rj’ β3 + u(i,j)) where Y(i,j) takes the value 1 if member j in household i is part of the labor force, currently working, or attending school; X(i,j) is the set of characteristics of the household member (for example, age, sex, birth order, etc.); Hj is the set of characteristics of the household (for example, recipient of CCT, urban resident, etc.); Rj takes the value 1 if household receives foreign or domestic remittances; and u(i,j) is the error term. For direct interpretation of coefficients, marginal probabilities are calculated. Meanwhile, for continuous outcome variables, we estimate a model using ordinary least squares (OLS): M A K I N G G ROW T H WO R K F O R T H E P O O R Y(i,j) = β1 X(i,j) + β2 Hj + β3 Rj + u(i,j) where Y(i,j) is the number of hours worked or expenditure share of a consumption item. Labor supply We estimate the impact of foreign and domestic remittances on labor force and employment using a probit model (reported are in marginal fixed effects) and on number of hours worked using OLS. For labor supply outcomes, individuals analyzed are those aged 18 – 64; for employment outcomes, only those in labor force; while for number of hours worked, only those who are employed. Aside from dummy variables for foreign and domestic remittance recipients, the model includes controls for household characteristics: recipients 162 of CCT, geographic characteristics, income deciles, number of members aged below 15 and above 64; and characteristics of the individual of interest: age, marital status, educational attainment, and whether he/she is the household head). Child outcomes For child outcomes, we also used a probit model (also in marginal fixed effects). For school attendance, individuals analyzed are those aged 5 – 17; while for child labor, only those aged 5 – 14. There are children reported to be working and also attending school. Controls used are household characteristics: recipients of CCT, geographic characteristics, income deciles, number of members aged below 15 and aged 18 – 64; characteristics of the HH head: age, educational attainment, and marital status; characteristics of the individual of interest: age and whether he/she is the oldest child. Expenditure behavior We estimate the impact of foreign and domestic remittances on household expenditures, measured by the share of expenditure on a particular commodity to total expenditure. In particular, we looked at expenditures shares of food, clothing, housing, health, education, and others. Share of others is calculated by subtracting from 1 the total share of food, clothing, housing, education, and health. These include expenditures on durable (except clothing and footwear) and non-durable goods, recreation, utilities, transportation, house operations, among others. Controls used are household characteristics: recipients of CCT, geographic characteristics, family size, and income deciles; and household head characteristics: age, educational attainment, and sex. Table J.1: Regression estimates for labor supply indicators (Adults 18 – 64 years old) Probit (marginal fixed effects) OLS Model Labor force participation Employment Hours work HH receives domestic remittances -0.026 -0.010 -1.655 (0.003)** (0.002)** (11.08)** HH receives foreign remittances -0.191 -0.028 -1.774 M A K I N G G ROW T H WO R K F O R T H E P O O R (0.004)** (0.002)** (10.04)** HH receives CCT 0.040 0.009 -1.465 (0.004)** (0.002)** (7.44)** HH head 0.203 0.023 1.761 (0.004)** (0.002)** (8.94)** Male 0.233 -0.011 0.367 (0.003)** (0.002)** (2.05)* Age 0.042 0.007 0.116 (0.001)** (0.000)** (2.71)** Age squared -0.001 -0.000 -0.003 (0.000)** (0.000)** (5.95)** 163 Table J.1: Regression estimates for labor supply indicators (Adults 18 – 64 years old) (continued) Probit (marginal fixed effects) OLS Model Labor force participation Employment Hours work Married -0.048 0.033 0.564 (0.004)** (0.002)** (3.09)** College grad -0.228 -0.006 0.620 (0.006)** (0.003)* (2.20)* Number of members aged below 0.001 0.002 0.699 15 and above 64 (0.001) (0.001)** (13.79)** Urban area -0.018 -0.013 3.095 (0.004)** (0.002)** (16.84)** Upper 10% 0.091 0.035 10.242 (0.008)** (0.002)** (24.50)** Constant 32.031 (35.91)** Regional controls Yes Yes Yes N 54,509,985 37,022,147 34,801,349 R2 0.18 0.14 0.07 * p<0.05; ** p<0.01 Table J.2: Regression estimates for school attendance (Children 5 – 17 years old) and child labor (Children 5 – 14 years old) Model School attendance Child labor HH receives domestic remittances 0.006 0.001 (0.002)** (0.000)** HH receives foreign remittances 0.019 -0.002 (0.002)** (0.000)** M A K I N G G ROW T H WO R K F O R T H E P O O R HH receives CCT 0.010 0.002 (0.002)** (0.000)** Male -0.031 0.001 (0.002)** (0.000)** Age 0.090 0.010 (0.002)** (0.000)** Age squared -0.004 -0.000 (0.000)** (0.000)** Oldest child -0.010 0.000 (0.003)** (0.000)** Number of members aged below 15 -0.006 0.001 (0.001)** (0.000)** 164 Table J.2: Regression estimates for school attendance (Children 5 – 17 years old) and child labor (Children 5 – 14 years old) (continued) Model School attendance Child labor Number of members who are aged 18-64 0.001 -0.001 (0.001) (0.000)** Age of HH head 0.000 0.001 (0.001) (0.000)** Age of HH head squared -0.000 -0.000 (0.000) (0.000)** HH head is college grad 0.022 0.002 (0.004)** (0.000)** Upper 10% 0.057 0.001 (0.002)** (0.000)** HH head is married 0.024 0.000 (0.004)** (0.000) Urban area -0.006 -0.001 (0.003)* (0.000)** Regional controls Yes Yes N 30,140,767 23,386,243 R 2 0.18 0.14 * p<0.05; ** p<0.01 Table J.3: Regression estimates for Household Spending Patterns (as Share of Total Expenses) Model Food Clothing Housing Health Education Others HH receives for- -0.020 0.003 0.011 0.008 0.010 -0.011 eign remittances (18.66)** (13.29)** (12.05)** (11.45)** (17.66)** (10.97)** M A K I N G G ROW T H WO R K F O R T H E P O O R HH receives -0.008 -0.000 0.002 0.013 0.002 -0.009 domestic remit- tances (8.25)** (1.05) (2.55)* (21.01)** (4.31)** (9.72)** HH receives CCT 0.015 0.002 -0.012 0.001 0.002 -0.007 (10.90)** (6.69)** (10.63)** (0.92) (3.48)** (5.62)** Decile10 -0.288 0.010 -0.004 0.040 0.032 0.211 (105.67)** (19.36)** (1.59) (23.06)** (22.60)** (80.88)** HH head is college -0.076 0.003 0.036 0.004 0.015 0.019 graduate (21.67)** (3.98)** (11.92)** (1.62) (8.13)** (5.76)** Urban area 0.000 -0.003 0.026 -0.002 -0.004 -0.018 (0.17) (12.84)** (26.67)** (2.89)** (7.03)** (15.99)** Family size -0.003 0.000 -0.011 0.000 0.004 0.010 (11.62)** (6.16)** (51.99)** (1.29) (29.17)** (41.23)** 165 Table J.3: Regression estimates for Household Spending Patterns (as Share of Total Expenses) (continued) Model Food Clothing Housing Health Education Others Male (HH head) 0.002 0.000 -0.010 0.003 -0.003 0.006 (2.12)* (1.97)* (9.89)** (4.48)** (4.43)** (5.70)** Age (HH head) -0.000 -0.000 -0.001 -0.001 0.002 0.001 (2.18)* (6.08)** (4.38)** (8.73)** (17.94)** (3.58)** M A K I N G G ROW T H WO R K F O R T H E P O O R Age squared (HH -0.000 0.000 0.000 0.000 -0.000 -0.000 head) (1.26) (1.57) (11.28)** (13.71)** (19.96)** (7.53)** _cons 0.688 0.032 0.237 0.010 -0.042 0.076 (98.45)** (24.81)** (39.38)** (2.18)* (11.74)** (11.31)** Regional controls Yes N (households) 22,730,410 22,730,410 22,730,410 22,730,410 22,730,410 22,730,410 R 2 0.56 0.09 0.29 0.08 0.10 0.30 * p<0.05; ** p<0.01 166 A nnex K Impact Evaluation Designs for the Pantawid Pamilya Since the program was launched in 2008, DSWD has already launched two impact evaluations of the Pantawid Pamilya. The first impact evaluation (World Bank 2012) used a randomized control trial (RCT) design. Data collection was done in 2011 and the report released in 2013. The second impact evaluation (Orbeta et al. 2014) used a regression discontinuity design (RDD). Data collection was done in 2013 and the report was released in 2014.75 This World Bank IE used tested the use of an alternative approach for subsequent evaluations, the RDD. Orbeta et al. fully employed the RDD approach in assessing the impacts of the program. The RCT compared outcomes between treatment villages with households that received the CCT and control villages with eligible households but did not receive the CCT. On the other hand, the RDD compared outcomes in households that are just above and below the poverty line, with the premise that these households exhibit similar characteristics. The World Bank RCT. RCT is generally considered the “gold standard” of evaluation methods. For social programs like Pantawid Pamilya, the most rigorous approach to IE assigns treatment/control status on a randomized basis. An RCT estimates program impact by comparing outcomes among eligible households in the “treatment” localities—meaning those that received the program—with outcomes among households in the “control” localities who would have been eligible if the program had been in operation there. A prior statistical assessment (power calculation) ensured that the evaluation study included enough households to assess the impact of the program effectively. An RCT does not require baseline data for impact indicators as randomization will fully suffice to cancel out all other factors that could affect differences in measured outcomes except the program itself. The key step in RCT, therefore, is ensuring that assignment of the M A K I N G G ROW T H WO R K F O R T H E P O O R treatment is fully random in the study sample. This method was feasible in the World Bank IE because the program was just starting and it was possible, with the authorization of DSWD, to delay enrollment of the “control group” into the program until the study is completed. For succeeding IEs, RCT is no longer possible as Pantawid Pamilya intended to scale up and exhaust all potential households eligible be enroll in the program. The Orbeto et al. RDD. RDD is a quasi-experimental method of evaluating program impact that is applicable when observation units (households) can be sorted using some continuous metric (income). Program eligibility is defined using a predetermined threshold or cutoff point of the sorting metric, for which the 75 World Bank (2013) “Philippine Conditional Cash Transfer Program Impact Evaluation 2012”; Orbeta, et al. (2014) “Keeping Children Healthy and in School: Evaluating the Pantawid Pamilya Using Regression Discontinuity Design. Second Wave Impact Evaluation Results.” 167 population has no direct control. This sorting metric is often referred to as the assignment, running, or forcing variable. In RDD, observations just below the cutoff are similar to, and therefore, compare well to those just above the cutoff. In the absence of the program, one would expect that any shifts in outcome variables would happen smoothly alongside minor changes in the running variable. Thus, a large jump in the outcome variable, observed precisely at the threshold value of the running variable, after program intervention can be attributed to the program itself. In recent years, use of RDD in evaluating the impacts of development programs has been growing. One of the strengths and advantages of RDD includes the weaker assumptions required for its validity compared to other non-experimental impact evaluation methods. The main caveat is that because program impact is estimated locally, or using observations very close to the cutoff, the generalizability of RDD estimated effect is limited. While the evaluation results using RDD has strong internal validity properties considered by many as next only to RCT, it needs to be recognized that its external validity is limited to observation units near the eligibility threshold. For more details, refer to World Bank (2013) “Philippine Conditional Cash Transfer Program Impact Evaluation 2012;” Orbeta, et al. (2014) “Keeping Children Healthy and in School: Evaluating the Pantawid Pamilya Using Regression Discontinuity Design. Second Wave Impact Evaluation Results.” M A K I N G G ROW T H WO R K F O R T H E P O O R 168 A nnex L Impact of Conditional Cash Transfers on Remittances Given the small amount of the CCT cash grant, we test the null hypotheses that the CCT grant does not crowd out remittances. We draw from the following literature on the impact of public transfers on private transfers: • Schoeni (1996): private assistance in the form of cash and time-help were crowded out by Aid to Families with Dependent Children benefits in the United States. • Schoeni (2002): unemployment insurance crowds out interfamily transfers. • Cutler and Gruber (1996): extension of Medicaid to pregnant women and children in the United States crowds out private insurance coverage. • Cox, Exer, and Jimenez (1998): Social Security benefits crowd out prevalence of private transfers in Peru. However, estimator may be biased downwards because recipients of those benefits are less likely to receive private transfers. Formal workers with the benefit tend to have more access to credit and savings mechanisms and thus, more able to mitigate shocks and therefore, less likely to need private transfers. • Attansio and Rios-Rull (2000): found weak evidence supporting crowding out for Mexican CCT. • Teruel and Davis (2000): Being in PROGRESA program has no influence over the incidence or the level of either monetary or non-monetary private inter-household transfers. We use two models to test whether CCT crowds out private transfers such as remittances using the FIES 2015 data. First, we use a probit model to determine whether receiving CCT increases the likelihood of receiving remittances. Second, we use a linear regression to test whether the amount of CCT received by M A K I N G G ROW T H WO R K F O R T H E P O O R the beneficiary decreases the amount of remittances received. The results show that there is a slight negative correlation between being recipient of CCT and being recipient of remittances, though when we looked at amounts the effect disappears. 169 Table L.1: Effects of CCT on remittances Effect of CCT participation on incidence of Effect of size of CCT received on the size of remittances (Probit) remittance (Linear regression) HH receives domestic HH receives foreign Amount of domestic Amount of foreign Model remittances remitances remittances received remittances received HH receives CCT -0.007 -0.02 (0.000)** (0.000)** Amount of CCT received -0.017 -0.001 -0.49 -0.02 (mean) poorhh 0.05 -0.009 -539.385 -256.775 (0.001)** (0.001)** -0.7 -0.2 Upper 10% -0.421 0.507 2,133.09 78,489.59 (0.001)** (0.001)** -0.74 (16.62)** HH head is college grad 0.039 0.104 3,860.91 -1,944.23 (0.001)** (0.001)** (2.91)** -0.89 Urban area -0.004 -0.031 -258.618 102.364 (0.000)** (0.000)** -0.62 -0.15 Family size -0.022 0.03 274.626 1,502.41 (0.000)** (0.000)** (3.48)** (11.57)** HH head is male -0.099 -0.113 -7,929.89 -3,724.81 (0.000)** (0.000)** (17.06)** (4.87)** Age of HH head -0.007 -0.006 -118.843 -478.707 (0.000)** (0.000)** -1.44 (3.53)** Age of HH head squared 0 0 1.913 5.567 (0.000)** (0.000)** (2.41)* (4.26)** Constant 9,496.24 4,375.58 M A K I N G G ROW T H WO R K F O R T H E P O O R (3.71)** -1.04 N 22,730,410 22,730,410 18,140,760 18,140,760 R2 0.08 0.11 0.1 0.14 * p<0.05; ** p<0.01 170 M A K I N G G ROW T H WO R K F O R T H E P O O R 172