Report No. 62955 - PA June 30, 2011 CURRENCY EQUIVALENTS Currency Unit = Panamanian Balboa US$1 = 1 Balboa (As of June 29, 2010) FISCAL YEAR January 1 – December 31 ACRONYMS AND ABBREVIATIONS ASMUN Ngobe Women’s Association BADEINSO Database of Social Statistics and Indicators (Base de Estadísticas e Indicadores Sociales) CSS Social Security Administration (Caja de Seguro Social) CCT Condicional Cash Transfer CEFACEI Community and Family Centres for initial Education CGK General Kuna Congress CIF Cost, Insurance and Freight COIF (Centros Integrales de Desarrollo Infantil) ECLAC Economic Commission for Latin America (Comisión Económica para América Latina) EIH Initial Education at Home EPH Permanent Household Survey (Encuesta Permanente de Hogares) ENV National Household Survey (Encuesta Nacional de Vida) FGT Foster Greer Thorbecke GDP Gross Domestic Product GIC Growth Incidence Curve GoP Government of Panama GNI Gross National Income IDAAN National Sewers and Aqueducts Institute (Instituto de Acueductos y Alcantarillados Nacionales) IDB Inter-American Development Bank IMF International Monetary Fund INADHE National Institute for Human Resource Development (Instituto Nacional de Formación Profesional y Capacitación para el Desarrollo Humano) INAFORP National Institute of Vocational Training (Instituto Nacional de Formación Profesional) Name changed to INADHE INEC National Statistics and Census Institute (Instituto Nacional de Estadísticas y Censos) INFAD/FIDA International Fund for Agriculture Development IFARHU Instituto para la Formación y Aprovechamiento de Recursos Humanos IPEA Institute of Applied Economic Research (Instituto de Pesquisa Econômica Aplicada) LA Latin America LAC Latin America and the Caribbean LPG Liquified Petroleum Gas LSMS Living Standards Measurement Study M&E Monitoring and Evaluation MEDUCA Ministry of Education (Ministerio de Educación) MEF Ministry of Economy and Finance MICI Ministry of Commerce and Industries MIDA Ministry of Agricultural Development (Ministro de Desarrollo Agropecuario) MIDES Ministry of Social Development MIC Middle Income Countries MINSA Ministry of Health (Ministerio de Salud) MIVI Ministry of Housing (Ministerio de Vivienda) NAS Panama National Accounts NGO Non-Governmental Organization OECD Organization for Economic Cooperation and Development PARVIS Programa de Ayuda Rápida de Viviendas de Interés Social PMT Proxy Means Testing PER Public Expenditure Review PRAF Family Allowance Program (Programa de Asignaciones Familiares). PROMEBA Integral Improvement Neighborhood Program PROINLO Program of Local Investments PROVISOL Housing Solidarity Program SA Social Assistance SC Social Cabinet SENAPAN National Secretariat for Food and Nutrition SENADIS Secretaría Nacional para la Integración Social de las Personas con Discapacidad SI Social Insurance SIF Social Investment Fund SP Social Protection SPS Social Protection System (Sistema de Protección Social) TSF Tariff Stabilization Fund UNDP United Nations Development Programme UNFPA United Nations Population Fund UNICEF United Nations Children’s Fund WDI World Development Indicators Vice President: Pamela Cox Country Director: Carlos Felipe Jaramillo Director PREM: Rodrigo A. Chaves Lead Economist: J. Humberto Lopez Sector Manager PREM: Louise Cord Country Manager: Ludmilla Butenko Task Manager: Pedro Olinto Co-Task Manager: Carolina Diaz-Bonilla TABLE OF CONTENTS Introduction .................................................................................................................................................. 1 Economic Growth, Poverty and Inequality ................................................................................................ 2 The Evolution of Poverty in Panama .......................................................................................................... 6 Migration and the shifts in the geographic incidence of poverty ............................................................15 Human Opportunity Index .........................................................................................................................18 Introduction .................................................................................................................................................27 Methodology.................................................................................................................................................27 Model Specifications ....................................................................................................................................28 Results ..........................................................................................................................................................29 Introduction .................................................................................................................................................37 Education .....................................................................................................................................................37 Health ...........................................................................................................................................................42 Malnutrition .................................................................................................................................................44 Conclusion and Policy Implications ...........................................................................................................46 Introduction .................................................................................................................................................47 The Current Social Protection System in Panama ...................................................................................48 Conditional Cash Transfer: The Red de Oportunidades Program .........................................................50 Simulating the Targeting and Incidence of Panama’s new Transfer Programs: Bono 100 a los 70 and Beca Universal ..............................................................................................................................................58 Conclusions and Policy Implications .........................................................................................................63 Policy Considerations ..................................................................................................................................66 References ....................................................................................................................................................68 Annex 1.1 Poverty statistics by Areas ........................................................................................................69 Annex 1.2. Poverty statistics by Provinces ................................................................................................79 Annex 1.3 Selected Inequality Statistics ....................................................................................................81 Annex 1.4. Growth Incidence Curves ........................................................................................................84 Annex 2.1 Basic Descriptive Statistics .......................................................................................................88 Annex 2.2 Consumption models .................................................................................................................90 Annex 3.1 Education Statistics ...................................................................................................................92 Tables Table 1.1: Mean Real Per Capita Expenditure by Urban, Rural and Indigenous Areas and by Quintile of Expenditure Distribution ................................................................................................................................ 4 Table 1.2: Key Socio-Economic Indicators in Panama 1997-2008 ................................................................ 6 Table 1.3: Characteristics of Rural-to-Urban and Urban-to-Rural Recent Migrants in 2003 and 2008 ........17 Table 1.4: Occupation of Rural to Urban Recent Migrants ...........................................................................18 Table 1.5: Inequality of Opportunity Profile – Specific D-Indices................................................................22 Table 1.6: Decomposing changes in the HOI ................................................................................................24 Table 1.7: Characteristics of the vulnerable – 2008 ......................................................................................25 Table 2.1: Mobility out of extreme poverty, 2003-2008 ...............................................................................34 Table 2.2: Mobility out of poverty, 2003-2008 .............................................................................................35 Table 3.1: Malnutrition in Panama (children younger than 5 years) .............................................................45 Table 3.2: Incidence of Diarrhea in children under 5 ....................................................................................46 Table 4.1: International Comparison of Social Spending (percent of GDP) ................................................49 Table 4.2: Distribution of Social Assistance Spending, 2008 ......................................................................50 Table 4.3: Amount of Conditional Money Transfers, by Year .....................................................................51 Table 4.4: Coverage of CCT Program (based on 2008 data) ........................................................................55 Table 4.5: Coverage of CCT Program by Region (2008 data): % of extreme poor households that are beneficiaries...................................................................................................................................................56 Table 4.6: Coverage of CCT Program by Region (2008 data) ......................................................................57 Table 4.7: Propensity Score Matching Estimation of the Impact of Red de Oportunidades on Log- consumption per capita ..................................................................................................................................58 Figures Figure 1.1: Per capita GDP evolution in LAC, 1990-2009............................................................................. 2 Figure 1.2:: Changes in growth composition in Panama ................................................................................ 3 Figure 1.3: Gini Coefficient for Consumption ............................................................................................... 5 Figure 1.4: Growth and Poverty Reduction in Panama and LAC between 1997 and 2008 ............................ 7 Figure 1.5: Poverty Measures by Area –Headcount Ratio ............................................................................. 8 Figure 1.6: Poverty Incidence and Contribution by the Indigenous and Non-indigenous .............................. 9 Figure 1.7: Poverty Gap Incidence and Contribution by the Indigenous and Non-indigenous .....................11 Figure 1.8: Incidence and Contribution to National Poverty by Ethnicity and Region .................................12 Figure 1.9: Poverty by Gender ......................................................................................................................13 Figure 1.10: Poverty by Gender of Head of Household ................................................................................13 Figure 1.11: Share of Households Headed by Females .................................................................................14 Figure 1.12: Poverty by Gender of Head of Household ................................................................................15 Figure 1.13: Teenage Pregnancy in Panama ..................................................................................................15 Figure 1.14: The Human Opportunity Index in Panama, Central America, and LAC (2008) .......................20 Figure 1.15: The Human Opportunity Index in Panama, Central America, and LAC (1997) .......................20 Figure 1.16: Changes in the HOI: Panama 1997 - 2008 compared to LAC and Central America ................21 Figure 1.17: Predicted probability of access in various parts of the country .................................................23 Figure 2.1: Region of residence and movements into and out of extreme poverty........................................31 Figure 2.2: Education and movements into and out of poverty .....................................................................33 Figure 3.1: Schooling Attainment by Region of Residence ..........................................................................38 Figure 3.2: Tertiary completion rates by birth cohort and gender .................................................................38 Figure 3.3: School Enrollment by Age ..........................................................................................................39 Figure 3.4: Enrollment by Region and Poverty Status ..................................................................................40 Figure 3.5: Schooling and Poverty ................................................................................................................41 Figure 3.6: Percent returns to one additional year of education ....................................................................41 Figure 3.7: Quality of Education Services: Public versus Private Schools....................................................42 Figure 3.8: Key Health Indicators 1990-2009 ...............................................................................................43 Figure 3.9: Path to Meeting Malnutrition MDG ............................................................................................43 Figure 3.10: Path to Meeting Infant Mortality MDG ....................................................................................44 Figure 3.11: Percentage of Pregnant Women with at Least One Prenatal Care Consultation .......................44 Figure 4.1: Transfers from 100 para los 70 by consumption decile .............................................................59 Figure 4.2 4.4: Share of transfers in total consumption by decile .................................................................59 Boxes Box 1.1: Measuring Welfare in Panama ........................................................................................................26 Box 4.1: Conditional Cash Transfers.............................................................................................................54 This Poverty Assessment is the product of a collaborative effort between the World Bank and Panama’s Ministry of Finance. From the World Bank, Adriana Cardozo, Carolina Diaz-Bonilla (Co-TTL), Will Durbin, Amer Hasan, and Pedro Olinto (TTL) participated under the overall guidance of Humberto Lopez (LCC2C), David Gould (LCSPE) and Louise Cord (LCSPP). From Panama’s Ministry of Finance Rogelio Alvarado and his team participated providing guidance and suggestion for topics. The Peer Reviewers were Jose Antonio Cuesta Leiva (PRMPR), Gabriela Inchauste (PRMPR), and Kinnon Scott (DECRG). While Panama is one of the fastest growing economies in the Latin America and Caribbean region (LAC), translating Panama’s growth into more rapid poverty reduction remains a major challenge. The country’s GDP per capita doubled in real terms between 1990 and 2009, and while the 2008 crisis has impacted growth negatively, Panama was one of the few countries in the region to grow in 2009 (Figure 1). However, despite its superb growth performance, Panama has not been able to exhibit a corresponding progress in poverty and extreme poverty reduction in the past. Compared to the rest of LAC, poverty reduction in Panama was less responsive to growth (Figure2). The region grew at a slower pace in terms of real per capita GDP (22.4 versus 56.1 percent), but was able to reduce poverty and extreme poverty by 47 and 52 percent, respectively. In contrast, Panama has only experienced drops of 12 and 23 percent in poverty and extreme poverty during the same period. Nevertheless, since 2006, the country has started to implement targeted Conditional Cash Transfers (CCTs) programs that have been shown to be effective in reducing poverty in other countries of the region. Figure 1: Per capita GDP evolution in LAC, 1990-2009 Note: Constant 2005 international PPP $ Source: World Development Indicators (2010), World Bank The observed low growth elasticity of poverty reduction in Panama is likely associated with the large and growing contribution of its ethnically indigenous population to both poverty and extreme poverty. The analysis in this report indicates that there are two very distinct groups of poor people in Panama, and that they respond very differently to economic growth: the indigenous poor, who have very low levels of human capital and skills, and the non-indigenous poor, who are relatively more educated and better able to take advantage of growth in the non- agricultural sectors. Both groups are concentrated in rural areas, but, induced by the fast growth in the services and industry sectors, rural-to-urban migration has increased significantly between 2003 and 2008, especially for the non-indigenous. As indicated by the analysis of mobility in Chapter 2, compared to the non-indigenous, the indigenous were significantly less likely to move out of poverty during this period of high growth. Consequently, while contributing to only nine percent of the country’s population, the contribution of the indigenous to national poverty and extreme poverty has increased steadily since 1997. They now represent 26 percent of the poor and 47 percent of the extreme poor, up from 21 and 36 percent in 1997. More importantly, their i contribution to the national extreme poverty gap increased even faster, going from 48 percent in 1997 to 65 percent in 2008.1 This is an indication that they are lagging further and further behind the non-indigenous, and respond less to economic growth in removing themselves out of poverty. The more educated non-indigenous rural poor, on the other hand, seem to have been more able to take advantage of fast growth to lift themselves out of poverty. As a result, rural poverty for the non-indigenous has been falling faster than for anyone else in the country. Moreover, as the analysis in Chapter 2 indicates, they were five times more likely than the indigenous to exit poverty between 2003 and 2008, and nine times less likely to move back into poverty. The migration analysis in Chapter 1 suggests that the upward mobility rates for the rural non-indigenous may have been helped by increased rural-to-urban migration. Because they are more educated and skilled than the indigenous, non-indigenous rural-to-urban migrants were more likely to obtain jobs in the fast growing services and construction sectors. Accelerating the rate of human capital accumulation by the indigenous is likely needed if they are to benefit more from economic growth to move out of poverty. As indicated in Chapter 2, upward mobility is closely linked to human capital. The results in Chapter 3 suggest that although the rate of human capital accumulation by the indigenous has increased between 1997 and 2008, through increased access to education and health services, they are still lagging far behind the non-indigenous in terms of schooling and health outcomes. A main concern is the growing incidence of chronic malnutrition in indigenous areas. Because most indigenous households earn considerably less than the national extreme poverty line, the average indigenous household is not able to afford half of the daily caloric needs for healthy living. As a result, the prevalence of chronic malnutrition in indigenous areas is extremely high, affecting 62 percent of children under five years of age. This rate has been increasing since 1997 (from 48 percent) and is now comparable to levels of stunting in countries with less than one tenth of Panama’s per capita GDP. Such high levels of chronic malnutrition are known to have long term negative impacts on cognitive development, productivity and general well being. To help them take greater advantage of their growing access to education services, chronic malnutrition among the indigenous needs to be addressed. As the country expands the coverage of its social assistance programs to poor households, extreme poverty is likely to decrease. There was considerable mobility out of poverty in Panama between 2003 and 2008. For instance, of those who were extremely poor in 2003, at least 17 percent were able to climb out of destitution by 2008. However, at least 0.7 percent of the non- destitute fell into extreme poverty during the same period. Therefore, as the country rapidly expands programs like Red de Oportunidades (RdO), Beca Universal and Bono 100 a los 70, the ratio of upward to downward mobility should increase, and growth should become more closely linked to poverty reduction. Moreover, since the rate of extreme poverty is much larger for the indigenous, the recent expansion of the coverage of RdO to all indigenous areas is likely to narrow the gap between the indigenous and non-indigenous poor. 1 If extreme poverty in Panama was to be eradicated via lump-sum transfers, because of their contribution to the national extreme poverty gap, the indigenous would need to receive almost two thirds of the total amount transferred. This is of course a hypothetical and impractical example, but it provides us with a clear view of why public resources should be targeted mostly to the indigenous if Panama is to accelerate both poverty and extreme poverty reduction. ii Figure 2: Growth and Poverty Reduction in Panama and LAC between 1997 and 2008 Moderate Poverty Extreme Poverty 50 50 Honduras Chile % change in extreme poverty rates Honduras % change in poverty rates Colombia Colombia 0 0 Dom.Rep Panama Bolivia Panama Peru Bolivia Peru Dom.Rep El Salvador -50 LAC -50 Argentina Venezuela, RB Ecuador LAC Paraguay Brazil Costa Rica El Salvador Venezuela, RB Brazil Ecuador Mexico Chile Argentina Mexico Paraguay Costa Rica Uruguay -100 Uruguay -100 0 20 40 60 0 20 40 60 % change in per capita GDP % change in per capita GDP Note: LAC average is the population weighted average of the sixteen countries included in the analysis. These countries were selected because they had poverty and extreme poverty measurements circa 1997 and 2008. Source: World Development Indicators (2010), World Bank The government’s decision to target its new generation of poverty alleviation programs to the extreme poor, and especially the indigenous, seems to be a move in the right direction. Since 2006 the country has embraced targeted CCT programs as its main strategy for fighting extreme poverty and to help increase the human capital accumulation of the poor. CCTs, when well targeted and monitored, have shown to be effective in reducing poverty in the short and long runs in other countries in LAC. Short run poverty falls due to the increased purchasing power of beneficiary families. Long run poverty is reduced through the acceleration of human capital accumulation by their children. The international evidence on the impact of CCTs on chronic malnutrition is especially encouraging for Panama. Moreover, CCTs may also help reduce rural- to-urban migration flows, and therefore, even when targeted to rural areas, they may help prevent increases in urban poverty. The analysis in Chapter 4 indicates that, while still in the early stages of its rollout in 2008, the Panamanian CCT program RdO seems to be very well targeted to the extreme poor. Moreover, RdO seems to have had a positive impact on per capita expenditures of indigenous households. However, more thorough evaluations of the long run impacts of the program are still needed. As the country also continues to expand the coverage of education services, enhanced schooling by the poor is likely to further boost upward mobility. As suggested by the results in Chapter 2, education is closely linked to mobility in Panama. Almost 15 percent of poor households in 2003 for which the heads had completed secondary education were able to climb out of poverty by 2008. On the other hand, for those with no primary education, upward mobility was substantially lower at 1.4 percent. Likewise, downward mobility is also strongly associated to education. While downward mobility was estimated at only 1.4 percent for those with complete secondary schooling, for those with incomplete primary it was much higher at 5.4 percent. As average schooling of adults continues to increase in the country, greater upward mobility relative to downward mobility should continue to contribute to accelerating poverty reduction. In addition to expanding health and education services and transfer programs, policy interventions specifically targeted to female headed households and female teenagers are iii also likely to further boost poverty reduction in Panama. The share of the country’s population living in female headed households has increased by 30 percent (from 20 to 26 percent) between 1997 and 2008. This trend appears to be pervasive in all regions of the country. Since poverty incidence is on average higher for female headed households, programs that provide gender specific interventions, as job training and access to affordable childcare, may be needed to further accelerate poverty reduction. Likewise, policies designed to reduce the prevalence of teenage pregnancy are also likely to have a positive impact in reducing poverty in the short and the long run, especially for the indigenous population for which 35 percent of teenagers have already had at least one pregnancy. Lack of fiscal resources is unlikely to be a binding constraint to developing and expanding effective poverty reduction programs in Panama. The country still has a large program of untargeted indirect subsidies. It spends approximately 1.2 percent of GDP on subsidies for water, electricity, fuel, and mortgage interest rates. Together these amount to almost US$280 million a year and have been shown to be vastly regressive. If this sum were to be transferred directly to those below the extreme poverty line, this would amount to a per capita transfer of approximately US$1.60 a day, which would completely eradicate extreme poverty, even if almost 40 percent of the transfers leaked to the non-destitute. Therefore, to further support its efforts to increase the coverage of its social assistance programs, the country may consider reallocating what it currently spends in untargeted price subsidies to its new generation of direct conditional transfers programs. The heavy investments that Panama undertook to mitigate the disparities in access to education in the past appear to be bearing fruit. Children are entering the school system earlier and staying longer. Most importantly, all of this expansion seems to have happened in rural and in indigenous areas where in the past lack of services had severely restricted pre- primary and secondary education opportunities for the poor (Figure 3). Furthermore, women seem to be leaving men behind in terms of attainment: females born in the 1980s surpass males of the same age in terms of average years of education and secondary completion rates. Most notably, almost 60 percent more women than men have completed some post secondary degree. This stands in stark contrast to the cohorts of those born in the 1930s, in which almost twice as many males had completed some post secondary education. If this trend continues, there will likely be twice as many women with college degrees than men in the near future. Nevertheless, if the country does not eradicate malnutrition, improve the quality of its schools, and to attract more investments in high productivity sectors, the positive impacts of further expanding access to education are likely to diminish. Private returns to additional years of education seem to have already dropped by almost 16 percent between 1997 and 2008 (Figure 4). There are plausibly three reasons behind this decline: First, the share of workers employed in low productivity sectors, especially in the construction and service sectors, has increased. Second, as the country strives to rapidly increase access to education to the poor in rural and indigenous areas, the average quality of education may suffer since almost all of this increase is obtained by the expansion of the public system. While there is indication that relative to private services the quality of public education has improved, important differences still remain. Finally, the level of cognitive development of new students entering the school system may be falling. As school access expands to areas where the prevalence of stunting is high, student ability may drop substantially since chronic malnutrition is closely associated with low cognitive development and school performance. Therefore, policies aimed at attracting investment in high productivity sectors, combined with interventions to reduce chronic malnutrition and to improve the quality of education, are likely to be needed to revert the trend of falling returns to schooling. iv Figure 3: School Enrollment by Region and Poverty Status % Enrolled in Pre-school Programs 25 25 25.8 23.5 20 20 19.9 20.8 19.2 18.9 15 15 17.1 15.7 16.4 15.2 13.4 12.5 10 10 11.8 10.6 8.7 7.5 5 5 6.3 5.7 4.0 0 0 1997 2008 1997 2008 1997 2008 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private % Enrolled in Secondary School 80 85.1 87.8 80 83.6 84.7 75.0 71.7 60 60 65.0 67.1 62.8 64.9 61.7 60.1 53.4 50.6 40 40 40.8 39.5 34.1 33.7 20 20 19.8 19.4 0 0 1997 2008 1997 2008 1997 2008 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data Figure 4: Percent return in income to one additional year of education 10.5% 10.3% 10.2% 10.1% 9.9% 9.7% 9.5% 9.5% 9.3% 9.1% 8.9% 8.7% 8.6% 8.5% 1997 2003 2008 Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data Expanded access to sanitation and adequate prenatal and infant care services may also help in combating chronic malnutrition in indigenous areas. While access to prenatal care has increased substantially in urban and rural areas, covering almost all pregnancies, in indigenous areas, which contribute the most to stunting in the country, 40 percent of women go through their whole pregnancies without a single visit to a health center (Figure 5). Furthermore, because of its direct impact on infectious diseases, lack of access to adequate sanitation is one of the main causes of child malnutrition. However, as indicated by our analysis in Chapter 1, Panama lags well behind the rest of Latin America and Central America in terms of improvements in equality of opportunity as measured by access to sanitation for its poorest population (Figure 6). As of v 2008, while LAC was two thirds of the way in making access to sanitation equitably distributed across its population, Panama was only one third of the way to equality. Therefore, while improving the incomes of the extreme poor is vital for enhancing their diets, equalizing access to preventive care and sanitation will be crucial for tackling chronic malnutrition and boosting human capital accumulation in Panama. Figure 5: Percentage of Pregnant Women with at Least One Prenatal Care 100.0% Consultation 91.4% 90.0% 91.2% 80.0% 80.0% 66.0% 76.5% 70.0% 60.0% 60.6% 59.6% 50.0% 47.1% 40.0% 36.0% 30.0% 1997 2003 2008 urbana rural indígena Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data Figure 6: The Human Opportunity Index in Panama, Central America, and LAC (2008) Access to water 100 80 60 Sixth grade on time 40 Access to sanitation 20 0 School enrollment Access to electricity Panama 2008 Latin America and the Caribbean Central America vi INTRODUCTION 1.1 Panama is one of the richest and fastest growing economies in Latin America; however it is considered a country of stark contrasts and, for some of its citizens, abysmal poverty. Large disparities in extreme poverty, poverty, and in other measures of human development exist among its citizens. This chapter examines the trends in economic growth, inequality and poverty in Panama between 1997 and 2008 at both the national level and by region, ethnicity and gender. It presents characterization of the patterns of consumption growth across the consumption distribution (including whether the growth is “pro-poor�). Next, it examines the role of internal migration flows in explaining shifts in poverty and extreme poverty between rural and urban areas. Lastly, the chapter begins to analyze the inequality of access to basic opportunities among children using the Human Opportunity Index (HOI). The Human Opportunity Index (HOI) is an operational measure of opportunities that takes into account both coverage and the distribution of access to basic goods and services by children, who cannot be held accountable for pre-determined circumstances at birth such as their race, gender, family income, parents’ education level, or place of residence. 1.2 The study is based on nationally-representative Living Standards Measurement Surveys (LSMS), which were conducted in 1997, 2003 and 2008. The LSMS household questionnaire includes quantitative data on various aspects of living conditions, including household structure, housing, infrastructure, health, nutrition, education and training, economic activity (labor), migration, spending and consumption, income, savings, credit, independent business activities, and agriculture.2 Since the latest available information is for 2008, the poverty numbers presented here reflect the Panamanian situation after the end of the period of high growth but before the impact of the 2008-09 global financial crisis. With this latest dataset, there are now three comparable household surveys that allow for the study of the evolution of poverty in Panama between 1997, 2003, and 2008. 1.3 The new numbers will show a considerably better performance in poverty between 2003-2008 as compared to the previous 1997-2003 period (see Table 1.1). In terms of extreme poverty, the progress is more modest, but more consistent, decreasing steadily from 1997 to 2008. The highlight is the strong decrease in extreme poverty for the indigenous areas, although still remaining at very high levels. This may imply that the government’s strategy to use Conditional Cash Transfer programs to reach the most marginalized had a positive result (Chap 4), but much effort remains to be done to achieve a significant drop in extreme poverty since the contributions of the indigenous to the extreme poverty rate and the poverty gap have dramatically increased. In addition, compared to the Latin America region, the improvements in poverty in Panama have been relatively modest given the very high average GDP growth between 1997 and 2008, and especially between 2003 and 2008. The results will also show that inequality declined modestly between 1997 and 2008, although increasing during the high growth period between 2003 and 2008. Changes in the patterns of rural-to-urban and urban-to-rural (or “reverse�) migration may partially explain the changes in the geographic incidence of poverty and inequality. 1.4 The high levels of consumption inequality may originate at least partially in the existing inequality of access to basic opportunities among children. Therefore, the last set of results will focus on the HOI. While broad access to education and clean water services puts Panama among the countries with the highest equality of opportunity in Latin America, the country still lags behind in terms of access to 2 The final sample includes 7045 households; and 27162 individuals. 1 electricity and sanitation. In addition, in areas in which the country lags behind the region in terms of equality of opportunity, improvements have been mixed. ECONOMIC GROWTH, POVERTY AND INEQUALITY Growth Trends 1.5 Between 1990 and 2009, Panama was one of the fastest growing economies in Latin America. Its GDP per capita almost doubled in real terms during this period, reaching a 2009 level of $11,870 in purchasing power parity (constant 2005 international $; see Figure 1.1). Before the global economic crisis of 2008-09, many expected that the country would soon catch up with Chile and Mexico in terms of average output per capita. While the 2008 crisis has impacted per capita growth negatively in 2009, taking it down to 0.8 percent, Panama was one of the few countries in the region for which this figure was positive. Moreover, the country is expected to return to its high growth trajectory in 2011 by growing approximately 4 percent per capita (IMF, 2010). Figure 1.1: Per capita GDP evolution in LAC, 1990-2009 Note: Constant 2005 international PPP $ Source: World Development Indicators (2010) 1.6 Panama’s growth between 1997 and 2008 was mainly led by the services and the industry sectors. The services sector, which represents 75 percent of the country’s GDP, grew at an average 6.4 percent per year between 1997 and 2008. It was followed by industry, the second most important sector accounting for 17 percent of GDP, which grew at an average 4.7 percent per year. Agriculture, which employs a high percentage of the population, but represents only 7 percent of GDP, grew at a slower rate of 3.9 percent per year. Manufacturing, which accounts for 9 percent of GDP, was the worst performing sector in the economy. It grew at the slow rate of 0.5 percent per year between 1997 and 2008. 1.7 Growth patterns significantly shifted between the 1997-2003 and 2003-2008 periods, however. As seen in Figure 1.2, between 1997 and 2003 growth was highest in the agriculture sector (5.8 percent per year), followed by services (4.4 percent), industry (1.8 percent), and manufacturing which contracted at -1.8 percent annually. For the 2003-2008 period, growth patterns changed drastically, and agriculture took manufacturing’s place as the worst performing sector in the economy. It grew at the slow rate of 1.7 percent per year despite the significant increases in food prices in 2007 and 2008. On the other 2 hand, the two most important sectors of the economy boomed. Services and industry grew at the impressive rates of 8.7 and 8.0 percent per year, respectively. Manufacturing also experienced some recovery, growing at 3.9 percent annually. As discussed later in this chapter, this shifting in the economy’s growth composition is likely to be associated to the changing patterns in expenditure growth, poverty reduction and migration flows. Figure 1.2:: Changes in growth composition in Panama 8.7 8.0 8.0 6.0 5.8 4.4 4.0 3.2 2.0 1.7 1.8 0.0 Agriculture Services Manufacturing Industry -2.0 -1.8 1997-2003 2003-2008 Source: World Development Indicators (2010) 1.8 Panama’s high growth in the non-agricultural sectors between 2003 and 2008 was due primarily to the expansion of the Panama Canal and to high growth in the construction and financial intermediation sectors. The construction sector presented GDP increases of as much as 30 percent (in constant 1996 balboas) between 2003 and 2008, but accounted for less than 6 percent of GDP in 2008. The financial intermediation sector and the transportation and communication sector (in large part the Canal) increased by around 18 percent, and accounted for around 8.5 and 21 percent, respectively, of GDP in 2008. Panama’s largest economic sectors are transportation and communication, commerce (Colon Free Trade Zone), and financial intermediation. However, neither the Canal nor the banking sector is labor intensive, and therefore, the associate impacts on employment and poverty reduction were likely small. General trends in per capita expenditure 1.9 The observed changes in the patterns of household expenditure growth are consistent with the changes in the patterns of economic growth. As seen in Table 1.1, real per capita expenditure grew almost four times faster during the high growth period of 2003-2008, than in the low growth period of 1997-2008. Moreover, during the 1997-2003 period, when agriculture was the best performing sector in the economy, per capita expenditure grew more for rural households than for urban ones. During the 2003-2008, on the other hand, urban households outpace rural ones in terms of expenditure growth. This is consistent with the shift in growth from the agriculture sector to the more urbanized services, industry and manufacturing sectors. 3 1.10 For those living in indigenous areas, it is interesting to note that their expenditures grew much faster during the later 2003-2008 period when agricultural growth was considerably lower. This is likely due to the fact that most of the indigenous working in agriculture are subsistence farmers and do not benefit as much from agricultural sector growth. In addition, as discussed in Chapter 3, the expansion of the conditional cash transfers program Red de Oportunidades (RdO) since 2006 throughout most indigenous areas may have been partially responsible for the observed increase in the per capita consumption of indigenous households. 1.11 Even though mean per capita expenditure growth in indigenous areas was positive, the vast majority of their residents still consume much less than the urban and rural non-indigenous population. By 2008 mean per capita expenditure in indigenous areas was about 8 times lower than in urban areas and four times lower than in rural areas. The expenditure gap between the indigenous and non-indigenous increased drastically between 2003 and 2008. 1.12 Overall, growth seems to have been pro-poor between 1997 and 2008. As indicated in Table 1.1, the average per capita expenditure for those in the lowest two quintiles of the distribution was much larger than for those in the upper three quintiles. This overall outcome is largely due to the patterns of growth observed during the 1997-2003 period. After 2003, growth drastically changed to becoming more pro-rich. Even though those belonging to the first two deciles still exhibited positive growth in expenditures between 2003-2008, households in the upper three deciles performed relatively much better. This is again consistent with the shifting of economic growth from the agricultural sector, where most of the poor are employed, to the services and industry sectors. Table 1.1: Mean Real Per Capita Expenditure by Urban, Rural and Indigenous Areas and by Quintile of Expenditure Distribution Mean Annual Growth (%) 1997 2003 2008 1997-2008 1997-2003 2003-2008 Urban 3,207.90 2,889.00 3,080.30 -0.4% -1.7% 1.3% Rural 1,305.80 1,421.50 1,505.10 1.3% 1.4% 1.1% Indigenous 416.6 373.1 389.5 -0.6% -1.8% 0.9% Quintiles Lowest quintile 403.5 446.8 479.4 1.6% 1.7% 1.4% 2 946.1 980.7 1,056.90 1.0% 0.6% 1.5% 3 1,557.60 1,525.80 1,656.40 0.6% -0.3% 1.7% 4 2,502.80 2,407.30 2,536.20 0.1% -0.6% 1.0% Highest quintile 6,067.40 5,779.90 6,457.00 0.6% -0.8% 2.2% Total 2,296.40 2,229.00 2,437.90 0.5% -0.5% 1.8% Inequality Trends 1.13 Changes in national expenditure inequality also reflect the shifting of growth from being pro-poor to being pro-rich. As seen in Figure 1.3, overall, inequality as measured by the Gini 4 coefficient declined modestly between 1997 and 2008. Inequality declined between 1997 and 2003 and then rose slightly between 2003 and 2008. This is not surprising since growth was more pro-poor during the first period, driven by stronger growth in the agricultural sector which mostly affects rural areas where most of the poor live. Between 2003 and 2008, because of the much faster growth in the services and industry sectors, growth was less pro-poor and inequality rose again. 1.14 Changes in migration flows may partially explain the differences in inequality trajectories between the 1997/2003 and 2003/2008 periods in urban and rural areas. As seen in Figure 1.3, inequality first decreased in rural areas between 1907 and 2003, and then increased again between 2003 and 2008. Rural to urban migration induced by fast growth in non-agricultural sector may increase inequality simultaneously in both urban and rural areas. Those seeking better job opportunities in the cities usually need some minimal level of skills, and therefore are not the poorest of the poor relative to the overall rural population. On the other hand, they tend to be relatively poorer for the receiving urban areas. Thus, increased rural to urban migration is likely to lead to more polarization in the income distributions of both rural and urban areas. It is therefore not surprising that inequality increased between 2003 and 2008 when the services and industry sectors were booming in urban areas and agriculture was underperforming. On the other hand, because agriculture was the fastest growing sector between 1997 and 2003, migration flows were less intense, and inequality dropped. The analysis on migration flows later in this chapter supports this hypothesis. 1.15 The drastic increase in inequality in indigenous areas is likely due in part to the partial rollout of the Rdo conditional cash transfer program. While the program aimed to target all indigenous areas before moving to rural non-indigenous and urban areas, it was still in its infancy by 2008. It had only covered part of the population living in indigenous areas, and therefore, the observed drastic increase in expenditure inequality within these regions was to be expected since the amount being transferred at the time (U$35 per family per month) represents a substantial share of the pre-program income of indigenous households (see Chapter 4 for more details on the partial coverage of the RdO program in 2008). Figure 1.3: Gini Coefficient for Consumption 0.50 0.48 0.49 0.47 National, 0.48 0.46 0.44 Urban, 0.44 Gini Coefficient 0.41 0.42 0.42 Rural, 0.41 0.41 0.39 0.40 Indigenous, 0.40 0.41 0.38 0.36 0.35 0.34 0.32 0.30 1997 2003 2008 Year Note: Figures are calculated for individuals, based on per capita household consumption levels. Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data 5 THE EVOLUTION OF POVERTY IN PANAMA 1.16 Panama showed improvements in almost all of its socio-economic indicators between 1997 and 2008 (Table 1.1). Moderate poverty, extreme poverty, and inequality decreased between during this period. Moreover, education and health indicators also improved over the 11 year period. Nevertheless, some indicators highlight some weaknesses, such as increasing repetition rates in secondary school, or immunization rates that are lower than the 1997 figures. In addition, inequality shows a deterioration in the shorter period of high growth between 2003 and 2008, even though not reaching the levels of 1997. That is, during the period of faster growth since the 1990s (2003 to 2008), growth was not as balanced as the in the preceding slower growth period.3 Table 1.2: Key Socio-Economic Indicators in Panama 1997-2008 Year Change Indicator 1997 2003 2008 2003/2008 Moderate poverty rate (Headcount rate) 37.3 36.8 32.7 -4.1pp Extreme poverty rate (Headcount rate) 18.8 16.6 14.4 -2.2pp School enrollment, primary (% net) 96.5 98.3 98.3 0.0% School enrollment, secondary (% net) 60.1 61.1 65.6 7.4% Repeaters, primary, total (% of total enrollment) 6.4 5.4 5.3 -1.9% Repeaters, secondary, total (% of total enrollment) 4.7 5.1 5.9 15.7% Mortality rate, infant (per 1,000 live births) .. 21.5 19.3 -10.2% Mortality rate, under-5 (per 1,000) .. 26.3 23.3 -11.4% Immunization, measles (% of children ages 12-23 months) 92 83 85 2.4% Immunization, DPT (% of children ages 12-23 months) 95 98 82 -16.3% Internet users (per 100 people) 0.5 10 27.5 175.3% Source: World Development Indicators (2009), World Bank Has growth been effective in reducing poverty in Panama? 1.17 Despite its superb growth performance between 1997 and 2008, Panama has not been able to exhibit a corresponding progress in poverty and extreme poverty reduction. Compared to the rest of Latin America, the country has actually underperformed. As seen in Figure 1.2, between 1997 and 2008, GDP per capita in Latin America grew at less than half the rate of Panama (22.4 versus 56.1 percent). Nevertheless, the region was able to reduce poverty almost four times as fast (46.5 versus 12.3 percent) and extreme poverty more than twice as fast (52.3 versus 23.4 percent). To come closer to regions average, Panama will need to implement more effective poverty reduction policies. 1.18 In 2006, Panama launched the Red de Oportunidades (RdO) conditional cash transfer. RdO aims at attacking both long and short term extreme poverty by delivering cash transfers to beneficiary families and requiring in exchange that they invest in the human capital accumulation of their children. The 3 Headcount poverty rates are the percentage of the population with consumption below the poverty line—for 1997, 2003, and 2008, using both the moderate poverty line and the extreme poverty line (see Box 1.1 at the end of the chapter for an explanation of how welfare is measured in Panama). 6 program is explicitly targeted to the extremely poor. More recently, the country launched two new transfer programs that are also aimed at alleviating short and long-run poverty: Becas Universales and Bono 100 a los 70. While the 2008 data analyzed here does not allow one to capture the full extent of the medium and long term impacts of these programs, the analysis of Chapter 3 indicates that if continued to be expanded, RdO is likely to help Panama accelerate its poverty reduction. For this to happen under the same fiscal envelope, however, the country might need to redirect its large program of untargeted price subsidies, and target revenues to direct transfers via the RdO. Figure 1.4: Growth and Poverty Reduction in Panama and LAC between 1997 and 2008 Moderate Poverty Extreme Poverty 50 50 Honduras Chile % change in extreme poverty rates Honduras % change in poverty rates Colombia Colombia 0 0 Dom.Rep Panama Bolivia Peru Panama Bolivia Peru El Salvador Dom.Rep LAC -50 -50 Venezuela, RB Ecuador Argentina Paraguay Brazil Costa Rica LAC Mexico Chile El Salvador Argentina Venezuela, RB Brazil Ecuador Paraguay Mexico Uruguay Costa Rica -100 -100 Uruguay 0 20 40 60 % change in per capita GDP 0 20 40 60 % change in per capita GDP Note: LAC average is the population weighted average of the sixteen countries included in the analysis. These countries were selected because they had poverty and extreme poverty measurements circa 1997 and 2008. Source: World Development Indicators (2010), World Bank Who are the poor, and where do they live? 1.19 To increase the effectiveness of targeted policies aimed at accelerating poverty and extreme poverty reduction, one needs to know who the poor are and where they live. Therefore, a more disaggregated analysis of the evolution of poverty than the one presented in Table 1.1 is needed. As indicate in Figure 1.5, the overall positive trends in poverty and extreme poverty show some starkly different results across urban, rural and indigenous areas. Figure 1.3 shows the headcount poverty rates— the percentage of the population with consumption below the poverty line—for 1997, 2003, and 2008, at the national level and for the populations living in urban, rural and indigenous areas, using both the moderate poverty line and the extreme poverty line. 1.20 Regionally, the country shows markedly different patterns of changes in poverty. Urban areas, which traditionally have had the lowest poverty rates, saw a marked increase in both poverty and extreme poverty between 1997 and 2003, and some improvement between 2003 and 2008. Poverty rates jumped from 15.3 to 20.0 percent between 1997 and 2003, a period in which the non-agricultural sectors underperformed in terms of growth. Between 2003 and 2008, urban poverty decreased to 17.7 as expected due to the boom in the services and industry sectors. Nevertheless, despite the stellar growth in the urban sectors of the economy, poverty remained at a worse level than in 1997. Before concluding that policies targeted to the urban poor need to be implemented, however, one needs to look into the effects of rural to urban migration on increasing urban poverty. If this increase is mainly caused by rural and indigenous 7 destitute families seeking better opportunities in a growing urban economy, targeting poverty alleviation interventions to urban areas may induce more migration and further increase urban poverty. Later in this chapter we examine the likely effects that rural to urban migration may have had on increasing urban poverty. 1.21 Rural Panama experienced substantial drops in poverty between 1997 and 2003, and between 2003 and 2008. The largest decrease (approximately 9 percent drop) occurred during the period of high growth in the agricultural sector between 1997 and 2003. Between 2003 and 2008, rural poverty continued to fall, albeit at a slower pace (5.6 percent), despite the low growth in the agriculture sector. Rural extreme poverty, which had also experienced a substantial decline between 1997 and 2003, plunging from 28.7 to 22.0 percent, but remained relatively constant during the period of slow growth in agriculture and fast growth in the services and industry sectors between 2003 and 2008. The larger drop in moderate poverty relative to extreme poverty also suggests that rural-to-urban migration may be at play. The extreme poor often lack the necessary skills to take new job opportunities being created in urban areas, and therefore stay behind. Figure 1.5: Poverty Measures by Area –Headcount Ratio (i) Poverty (ii) Extreme poverty 100 95.4 98.4 96.3 100 90.0 86.3 84.8 80 Extreme poverty rate (%) 80 Poverty rate (%) 58.7 60 54.0 60 50.7 37.336.8 40 40 32.7 28.7 22.0 22.2 20.0 18.8 20 15.3 17.7 20 16.6 14.4 3.1 4.4 3.2 0 0 Urban Rural Indigenous National Urban Rural Indigenous National 1997 2003 2008 1997 2003 2008 Note: Extreme poor refers to the population with per capita consumption below the extreme poverty line value. Moderate poor refers to the population with per capita consumption below the poverty line value. Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 1.22 The already high poverty rate for Panamanians living in indigenous areas increased between 1997 and 2003 and decreased slightly during the high growth period of 2003-2008. Almost all (96.3 percent) of those living in indigenous areas lived in poverty in 2008, and close to 85 percent lived in extreme poverty. More remarkably, because most indigenous households earn considerably less than the national extreme poverty line, the average indigenous household is not able to afford half of the daily caloric needs for healthy living. As a result, as indicated in Chapter 3, the prevalence of chronic malnutrition in indigenous areas is extremely high, affecting 62 percent of children under five years of age. 1.23 Moreover, while the incidence of poverty for the non-indigenous Panamanians has decreased steadily between 1997 and 2008, it has increased for the ethnically indigenous regardless of where they live. As it can be seen in Figure 1.6, while poverty remained relatively constant for for the ethnically indigenous (rising from 90.3 to 91 percent), non-indigenous poverty dropped from 32.4 to 26.7 percent. In terms of extreme poverty, the incidence fell by 36 percent for the non-indigenous, going from 13.1 to 8.4 percent, while dropping by less than 9 percent for the indigenous. 8 Figure 1.6: Poverty Incidence and Contribution by the Indigenous and Non-indigenous (i) Poverty Incidence (ii) Extreme Poverty Incidence 79.76 79.73 100 80 94.78 90.97 72.88 90.26 80 60 % 60 % 40 40 32.35 30.59 26.72 20 20 13.10 9.81 8.43 0 0 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Indigenous Non Indigenous Indigenous Non Indigenous (iii) Contribution to Poverty (iv) Contribution to Extreme Poverty 20.75 25.01 25.91 36.35 46.66 47.04 100 100 79.25 80 80 74.99 74.09 63.65 60 60 percent percent 53.34 52.96 40 40 20 20 0 0 1997 2003 2008 1997 2003 2008 Non Indigenous Indigenous Non Indigenous Indigenous Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 1.24 Because the incidence of poverty has dropped much faster for the non-indigenous population, the contribution of the indigenous to national poverty and extreme poverty rates has increased steadily since 1997. As seen in the second row of Figure 1.6, in 1997 approximately 21 percent of the poor were indigenous. This share had increased to 26 percent by 2008. In terms of extreme poverty, the increase in the contribution by the indigenous population was even more drastic. It went from 36 percent in 1997 to 47 percent in 2008. That is, despite representing only 9 percent of the overall population, the indigenous contribute to almost half of the extreme poverty in Panama. Therefore, prioritizing the rollout of RdO in indigenous areas since its launching in 2006 seems to have been a move in the right direction by the government of Panama. 9 1.25 More striking than the increasing contribution of the ethnically indigenous to poverty and extreme poverty, is the rise in their contribution to the national poverty and extreme poverty gap.4 The poverty and the extreme poverty gaps give rough estimates of the national per capita cost, as a percentage of the poverty lines, of eliminating poverty or extreme poverty, respectively. As it can be seen in Figure 1.7, the poverty and extreme poverty gaps are much larger for the ethnically indigenous. Moreover, they have been falling much faster for the non-indigenous. Between 1997 and 2008, the poverty gap fell by 28 percent for the non-indigenous (going from 12.2 to 8.8 percent of the poverty line), compared to only 5 percent for the indigenous (which went from 61.5 to 58.5 percent). For the extreme poverty gaps the differences in trajectory are even more remarkable. For the non-indigenous, the extreme poverty gap fell by 50 percent, going from 4.33 to 2.15 percent. In contrast, it fell by only 9 percent for the indigenous, going from 43.4 to 39.3 percent of the extreme poverty line. As a consequence, the contributions of the indigenous to the national poverty and extreme poverty gaps have also been steadily increasing. In 1997, they contributed less than a third to the poverty gap. By 2008, their contribution had increased to 41 percent. For the extreme poverty gap, their contribution increased from 48.4 percent to 65.2 percent. 4 The poverty gap can be interpreted as the national per capita measure of the total shortfall o expenditure levels below the poverty (or extreme poverty) line. It is the sum of all the shortfalls of the poor divided by the national population and expressed as a percentage of the poverty line. 10 Figure 1.7: Poverty Gap Incidence and Contribution by the Indigenous and Non-indigenous (i) Poverty Gap Incidence (ii) Extreme Poverty Gap Incidence 61.53 62.27 43.35 41.04 60 58.53 39.31 40 30 40 20 % %20 10 12.19 10.16 8.76 4.33 2.63 2.15 0 0 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Indigenous Non Indigenous Indigenous Non Indigenous (iii) Contribution to the Poverty Gap (iv) Contribution to the Extreme Poverty Gap 32.13 39.75 40.69 48.43 62.71 65.22 100 100 80 80 67.87 60.25 59.31 60 60 percent percent 51.57 40 40 37.29 34.78 20 20 0 0 1997 2003 2008 1997 2003 2008 Non Indigenous Indigenous Non Indigenous Indigenous Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 1.26 In addition to becoming more indigenous, poverty has also become more urban and less rural in Panama. As indicated in Figure 1.8, while poverty is still mainly a rural phenomenon, its incidence has increased for both the indigenous and the non-indigenous living in urban areas. Consequently, between 1997 and 2008, the contribution of urban dwellers to national poverty climbed by almost 50 percent for the non-indigenous, going from 22 to 32 percent of the poor, and tripled for the indigenous, going from 0.8 to 2.4 percent. For extreme poverty however, there is a less clear trend. While the contribution to national extreme poverty of both the indigenous and non-indigenous urban residents increased between 1997 and 2003, it decreased between 2003 and 2008. 1.27 11 Figure 1.8: Incidence and Contribution to National Poverty by Ethnicity and Region (i) Poverty Incidence (ii) Extreme Poverty Incidence 95.4% 86.3% Indigenous Ind. Area 98.4% Indigenous Ind. Area 90.0% 96.3% 84.8% 74.0% 66.1% Indigenous Rural 84.8% Indigenous Rural 57.8% 88.5% 65.2% 42.6% 14.8% Indigenous Urban 78.7% Indigenous Urban 30.0% 62.8% 11.7% 58.5% 28.3% Non Indigenous Rural 53.3% Non Indigenous Rural 21.2% 49.4% 20.7% 14.9% 3.0% Non Indigenous Urban 18.7% Non Indigenous Urban 3.8% 16.8% 3.0% 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% 1997 2003 2008 1997 2003 2008 (iii) Contribution to Poverty (iv) Contribution to Extreme Poverty 19.2% 34.5% Indigenous Ind. Area 20.6% Indigenous Ind. Area 41.9% 20.9% 41.8% 0.7% 1.3% Indigenous Rural 1.6% Indigenous Rural 2.5% 2.5% 4.3% 0.8% 0.5% Indigenous Urban 2.7% Indigenous Urban 2.3% 2.4% 1.0% 57.3% 55.0% Non Indigenous Rural 44.9% Non Indigenous Rural 39.6% 41.8% 39.8% 22.0% 8.7% Non Indigenous Urban 30.1% Non Indigenous Urban 13.7% 32.3% 13.2% 0% 10% 20% 30% 40% 50% 60% 0% 10% 20% 30% 40% 50% 60% 1997 2003 2008 1997 2003 2008 Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. Gender and Poverty in Panama 1.28 Males are more likely to be poor or extremely poor than females in Panama. As indicated in Figure 1.9, in 1997, 2003 and 2008, the prevalence of poverty and extreme poverty was higher among men than among women. This is likely a reflection of the fact that women have outpaced men in terms of schooling (see Chapter 3). Nevertheless, the male/female gaps in the incidence of both poverty and extreme poverty have been decreasing. One possible explanation is the rapid expansion of access to education in rural and indigenous areas, where men tend to start working at earlier ages than women. As the benefit/cost ratio of attending school relative to working on agriculture increases, boys become more likely to stay in school. The results in Figure 1.10, and the analysis of access to education in Chapter 3, support this hypothesis. As it can be seen, the gap on the incidence of poverty between men and women has decreased faster in rural and indigenous areas than in urban areas. 12 Figure 1.9: Poverty by Gender Female Female 1997 1997 Male Male Female Female 2003 2003 Male Male Female Female 2008 2008 Male Male 0 .1 .2 .3 .4 0 .05 .1 .15 .2 Share of Poor Share of Extreme Poor Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. Figure 1.10: Poverty by Gender and Region 1997 1997 Urbana 2003 Urbana 2003 2008 2008 1997 1997 Rural 2003 Rural 2003 2008 2008 1997 1997 Indigena 2003 Indigena 2003 2008 2008 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Share of Poor Share of Extreme Poor Female Male Female Male Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 1.29 The share of the population living in female headed households has increased substantially in Panama. In 1997, about one in every five Panamanians lived in a household headed by a female member. By 2008, this rate had increased by more than 30 percent, reaching one in every four (Figure 1.11). This trend appears to be pervasive in all regions of the country. It can be observed in urban, rural and, to a lesser extent, indigenous areas. As the incidence of both moderate and extreme poverty has been historically higher for female headed households (Figure 1.12), this has important implications for the design and targeting of poverty reduction interventions. As women increasingly become the main bread winners in the country, programs that provide gender specific interventions, as job training and access to affordable childcare for example, may be needed to effectively reduce poverty among female headed households. 13 Figure 1.11: Share of Households Headed by Females National By Region 1997 1997 Urbana 2003 2008 1997 2003 Rural 2003 2008 1997 2008 Indigena 2003 2008 0 .05 .1 .15 .2 .25 0 .1 .2 .3 Share Female Headed Households Share Female Headed Households Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 1.30 The incidence of teenage pregnancy for the indigenous in Panama should be a main concern to the country’s policy makers. As indicated in Figure 1.13, depending on the survey year, 35 to 42 percent of indigenous females ageing 15 to 19 had had at least one pregnancy. This rate is two and half times higher than the rate for the non-indigenous. The differences between the indigenous and the non- indigenous are even more striking for the share of teens that have had at least two pregnancies by the age of 19 (13 versus 4 percent). The data also indicates that teen mothers are twice as likely to be poor or extremely poor as teens that have never had a pregnancy. The ongoing conditional cash transfer program RdO, which is targeted to the extremely poor, may have positive effects in reducing teenage pregnancy since it provides incentives for female teens to stay in school. However, a careful evaluation of the impacts of RdO on teenage fertility rates is needed, since the program may also induce destitute young females to try to become eligible to the transfers by establishing new families. 14 Figure 1.12: Poverty by Gender of Head of Household 1997 1997 Urbana 2003 Urbana 2003 2008 2008 1997 1997 Rural 2003 Rural 2003 2008 2008 1997 1997 Indigena 2003 Indigena 2003 2008 2008 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Share of Poor Share of Extreme Poor Female Head Male Head Female Head Male Head Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. Figure 1.13: Teenage Pregnancy in Panama % of Females 15 to 19 with at least one reported % of Females 15 to 19 with at least two reported pregnancy pregnancies 42.35 15 13.91 40 13.66 13.07 35.42 34.64 30 10 20 % % 14.81 15.38 14.61 5 4.04 3.83 4.03 10 0 0 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Indigenous Non Indigenous Indigenous Non Indigenous Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data MIGRATION AND THE SHIFTS IN THE GEOGRAPHIC INCIDENCE OF POVERTY 1.31 As indicated above, while still mainly a rural phenomenon, poverty is becoming increasingly an urban problem in Panama. Whether or not this trend implies that policy interventions should increasingly target the urban poor will depend on the extent that rural to urban migration is the main factor behind this steady shift in the location of poverty. In this section we analyze the migration information provided by the ENVs of 2003 and 2008 to help answer this question.5 In both surveys, 5 While the 1997 ENV also provides information on recent migration, the data does not allow one to determine if the recent migrant’s place of origin was rural or urban. 15 respondents were asked where they lived five years in the past. If respondents currently live in an urban area and declared that they lived in either a rural or indigenous area five years in the past, they are classified as rural-to-urban migrants. If at the time of the survey they lived in a rural or indigenous area, and stated that five years before they were living in an urban area, they are labeled urban-to-rural migrants. Unfortunately, for the 2003 survey it is not possible to determine if recent migrants to urban areas came from either a rural or an indigenous area. Therefore, to be able to compare changes in the rates of migration between the 2003 and 2008, we aggregate both indigenous and rural origins under a single rural origin category for both surveys. 1.32 Both rural-to-urban and urban-to-rural migration increased significantly between 2003 and 2008. As indicated in Table 1.3, individuals declaring that they were either rural-to-urban or urban-to- rural migrants increased significantly between the two survey years. These increases are observed for both the indigenous and the non-indigenous. Rural-to-urban migration increased considerably more for the non-indigenous than for the indigenous (47 versus 18 percent). On the other hand, urban-to- rural or “reverse� migration climbed significantly more for the indigenous (427 versus 10 percent). Since the indigenous are much more likely to be extremely poor and the non-indigenous are more likely to be moderately poor, this may partially explain: (1) the drop in the contribution of the indigenous to urban extreme poverty between 2003 and 2008, and (2) the concurrent increase in the contribution of the non-indigenous to moderate poverty in urban areas. That is, rural-to-urban migration by the non- indigenous may be causing moderate poverty to become more of an urban problem, while urban-to-rural or reverse migration by the indigenous may be helping reduce extreme poverty in urban areas and increase it in rural+indigenous areas. 1.33 Moreover, the incidence of extreme poverty among the indigenous migrating out of urban areas and into rural/indigenous areas increased by six fold between 2003 and 2008. As indicated in Table 1.3, in 2003, 11 percent of those that had recently migrated to rural areas were in extreme deprivation. In 2008, 66 percent of urban-to-rural migrants were extremely poor. It is interesting to note that the level of human capital, as measured by average years of schooling, also decreased substantially for the population of urban-to-rural indigenous migrants (from 9 to 6.1 years). This further supports the hypothesis that the poorest of the extreme poor among the indigenous left urban areas to return to their places of origin, perhaps because their level of skills was too low to benefit from high growth in the non-agricultural urban sectors. Moreover, by becoming beneficiaries of the RdO, they might have been able to attain a better standard of living in rural and indigenous areas. In fact, as indicated in Table 1.3, while only 0.6 percent of the indigenous that migrated to urban areas between 2003 and 2008 were beneficiaries of RdO, 37 percent of those migrating back to rural or indigenous areas were receiving the transfers in 2008. 16 Table 1.3: Characteristics of Rural-to-Urban and Urban-to-Rural Recent Migrants in 2003 and 2008 2008 2003 % Change % of total population that migrated from rural+indigenous to urban areas 2.5% 1.7% 46.8% % of indigenous population that migrated from rural+indigenous to urban areas 3.3% 2.8% 17.8% % of non-indigenous population that migrated from rural+indigenous to urban areas 2.4% 1.6% 50.2% % of total population that migrated from urban to rural+indigenous areas 1.4% 1.2% 15.1% % of indigenous population that migrated from urban to rural+indigenous areas 1.1% 0.2% 426.6% % of non-indigenous population that migrated from urban to rural+indigenous areas 1.4% 1.3% 10.0% Ratio of rural-to-urban/urban-to-rural migration rates for total population 1.81 1.42 27.5% Ratio of rural-to-urban/urban-to-rural migration rates for the indigenous 2.97 13.26 -77.6% Ratio of rural-to-urban/urban-to-rural migration rates for the non-indigenous 1.74 1.27 36.6% Share of the total poor that migrated from rural to urban areas in last 5 years 2.0% 1.3% 55.0% Share of the indegenous poor that migrated from rural to urban areas in last 5 years 2.7% 2.4% 9.5% Share of the non-indegenous poor that migrated from rural to urban areas in last 5 years 1.8% 1.1% 73.2% Share of the total extreme poor that migrated from rural to urban areas in last 5 years 0.2% 0.5% -57.3% Share of the indegenous extreme poor that migrated from rural to urban areas in last 5 years 0.1% 1.0% -94.1% Share of the non-indegenous extreme poor that migrated from rural to urban areas in last 5 years 0.4% 0.3% 11.3% Share of the total poor that migrated from urban to rural areas in last 5 years 1.1% 1.0% 9.3% Share of the indegenous poor that migrated from urban to rural areas in last 5 years 1.0% 0.1% 779.9% Share of the non-indegenous poor that migrated from urban to rural areas in last 5 years 1.2% 1.2% -6.5% Share of the total extreme poor that migrated from urban to rural areas in last 5 years 0.9% 0.7% 37.3% Share of the indegenous extreme poor that migrated from urban to rural areas in last 5 years 1.1% 0.0% 3598.9% Share of the non-indegenous extreme poor that migrated from urban to rural areas in last 5 years 0.8% 1.3% -38.7% Mean age of recent rural-to-urban migrants 31.7 27.8 14.0% Mean age of indigenous recent rural-to-urban migrants 34.3 25.1 36.6% Mean age of non-indigenous recent rural-to-urban migrants 31.4 28.2 11.3% Mean age of recent urban-to-rural migrants 32.9 32.3 1.9% Mean age of indigenous recent urban-to-rural migrants 28.6 37.7 -24.1% Mean age of non-indigenous recent urban-to-rural migrants 33.2 32.3 3.0% Mean years of education of recent rural-to-urban migrants 8.6 8.5 1.7% Mean years of education of indigenous rural-to-urban migrants 5.0 6.0 -16.3% Mean years of education of non-indigenous rural-to-urban migrants 9.0 8.8 2.6% Mean years of education of recent urban-to-rural migrants 8.5 8.4 1.3% Mean years of education of indigenous urban-to-rural migrants 6.1 9.0 -32.0% Mean years of education of non-indigenous urban-to-rural migrants 8.6 8.4 3.3% % of recent rural-to-urban migrants who are beneficiaries of RdO 1.81 -- -- % of indigenous rural-to-urban migrants who are beneficiaries of RdO 0.64 -- -- % of non-indigenous rural-to-urban migrants who are beneficiaries of RdO 1.95 -- -- % of recent urban-to-rural migrants who are beneficiaries of RdO 6.02 -- -- % of indigenous urban-to-rural migrants who are beneficiaries of RdO 37.41 -- -- % of non-indigenous urban-to-rural migrants who are beneficiaries of RdO 3.9 -- -- Source: World Bank staff calculations based on ENV 2003 and 2008 data. 17 1.34 Changes in the distribution of occupational choices indicate that jobs demanding relatively higher skills became more prevalent among recent rural-to-urban migrants. As seen in Table 1.4, there seems to have been a shift from lower to higher skill jobs taken by the population of recent rural-to- urban migrants as measure by the 2008 survey. While less of the migrants were employed on low skill activities, such as domestic help, micro and small enterprises (including street vendors) and the primary sector, more were working on construction, furniture making, manufacturing, hospitality services and urban transportation (e.g., taxi, micro-bus, and truck drivers). For the non-indigenous, the changes were more drastic than for the indigenous. In 2008, almost twice as many non-indigenous migrants were working on construction and urban transportation. This is not surprising since, as shown in Table 1.3, the level of human capital as measured by average years of schooling, is much higher for the non-indigenous migrants than for the indigenous migrants (9 versus 5 years of schooling). Table 1.4: Occupation of Rural to Urban Recent Migrants 2008 2003 Non-Indigenous Indigenous Total Non-Indigenous Indigenous Total Construction 12.42 4.34 11.67 5.77 2.29 5.4 Crafts and Small Manu 2.76 5.58 3.02 2.46 39.26 6.32 Domestic help and ser 16.74 13.51 16.44 21.22 20.37 21.13 Furniture making and 1.28 0 1.16 1.08 0 0.96 Hospitality Services 10.44 16.29 10.98 9.35 6.75 9.08 Large and Medium Reta 4.24 16.29 5.35 0.47 0 0.42 Micro and small busin 24.26 35.56 25.3 30.27 14.1 28.57 Other 4.92 0 4.46 1.71 0 1.53 Primary Sector (agric 3.39 5.58 3.59 5.81 4.43 5.66 Public administration 13.48 0 12.24 18.6 12.8 17.99 Urban transportation 6.07 2.85 5.77 3.28 0 2.94 Total 100 100 100 100 100 100 Source: World Bank staff calculations based on ENV 2003 and 2008 data. HUMAN OPPORTUNITY INDEX 1.35 The high levels of consumption inequality in Panama presented earlier may originate at least partially in the existing inequality of opportunities among children to access basic goods and services. This inequity can be explored for a given year in Panama or can be compared over time or to other countries. To do so, the chapter applies the Human Opportunity Index approach described in Barros et al. (2009) and updates the results presented in Molinas et al. (2010) using data from the 2008 round of the Panama LSMS. 1.36 The Human Opportunity Index (HOI) allows us to incorporate equity into discussions of coverage (access to services) directly. The HOI measures the availability of basic services necessary to lead a dignified life in modern society, discounted or “penalized� by how unfairly these services are distributed among the population. Two countries with identical coverage may have very different HOIs if the citizens that lack the service are all female, or black, or poor – or more generally, share a personal circumstance beyond their control. Simply put the HOI is an equity-sensitive coverage rate. 18 1.37 The HOI estimates how personal circumstances for which an individual cannot be held accountable, like birthplace, gender, or socioeconomic background, affect a child’s probability of accessing basic services that are necessary to succeed in life. Specifically this section focuses on five such basic services – access to school (for children 10-14 years old), completing sixth grade on time (for children 12-16 years old) as well as access to water, sanitation and electricity (for children 10 years of age or younger). For comparability, the analysis is limited to the same set of circumstances employed in Molinas et al. (2010). The analysis is organized as follows. First, 2008 levels of the HOI in Panama are compared to those in LAC and Central America. Next, the analysis explores how the evolution of the HOI in Panama has differed from what has been observed in LAC and Central America. The analysis then focuses on the sources of the inequality of opportunity observed in the data, followed by a presentation of the characteristics of the excluded. Lastly, the observed changes in the HOI are decomposed– an exercise that is very informative from a policy standpoint. The HOI can in theory change (increase) if: (i) people’s circumstances change (the “composition effect�); (ii) more service is provided to all (the “scale effect�); or (iii) service is distributed more fairly (the “equalization effect�). In turn these effects can each be thought of as reflective of various types of policies – both current and past. The composition effect may reflect sectoral policies which have induced changes in the distribution of circumstances; the scale effect may reflect broad-based (non-targeted) initiatives for increasing coverage, and the equalization effect may reflect targeted policies. Panama in a regional context 1.38 Panama’s HOI is above the LAC average for access to water, for completing sixth grade on time, and for school attendance, yet it lags far behind the region’s average HOI in providing equitable access to electricity and sanitation. For each of the five indicators mentioned above, Figure 1.5 compares Panama’s 2008 HOI to that of Central America and of the Latin America and the Caribbean (LAC) region overall. The results are mixed. Panama’s HOI for completing sixth grade on time and for access to water are both roughly 80, a value ahead of both the regional averages. Yet, even for these indicators, there exists substantial room for improvement to reach the goal of 100. Access to sanitation is a particular area for concern since the HOI of 33 is almost half that of the LAC region. The gap between the region and Panama in access to electricity is less stark, but the HOI value implies that only two-thirds of the opportunities for equitable access are available and equitably distributed. The HOI for school attendance is almost universal, suggesting an area where a level playing field does in fact exist among children. (However, note that universal access to attending school does not imply universal access to schools of high quality.) Finally, the results for the HOI using 1997 data show that Panama used to be even further ahead of the region in access to water and sixth grade on time (as compared to the 2008 HOI), while not so far behind in access to electricity and sanitation (Figure 1.6). In other words, over the decade between 1998 and 2008, Panama did not improve nearly as much as its neighbors in terms of providing universal access to basic goods and services for children. 19 Figure 1.14: The Human Opportunity Index in Panama, Central America, and LAC (2008) Access to water 100 80 60 Sixth grade on time 40 Access to sanitation 20 0 School enrollment Access to electricity Panama 2008 Latin America and the Caribbean Central America Figure 1.15: The Human Opportunity Index in Panama, Central America, and LAC (1997) Access to water 100 80 60 40 Sixth grade on time Access to sanitation 20 0 School enrollment Access to electricity Panama 1997 LAC Central America 1.39 A comparison of the annual rate of change of the HOI between 1997 and 2008 shows that Panama advanced more slowly than the LAC and Central America average for all five indicators, with especially worrying results for the sanitation HOI (Figure 1.7). This result, however, has different implications depending on the indicator. As shown above, Panama’s HOIs for school enrollment, completing sixth grade on time, and access to water were all closer to universality in 1997 than that of the region, and therefore could be expected to have a slower annual rate of growth. The water HOI, however, shows a particularly low rate of change, and worse, the growth in the 2003-2008 sub- period (not shown here) was practically zero – implying that most of the improvements in this indicator occurred in the late 1990s. In contrast, while Panama lags behind the region in providing universal and equitable access to electricity, it is catching up. Although growing somewhat slower than the rest of the region when the longer 1997-2008 period is used, Panama’s rate of increase of the HOI for access to electricity was roughly 1.3 percentage points per year for the 2003-2008 sub-period – larger than either Central America or LAC overall. The HOI for access to sanitation shows the worst results: it is not only far from universality and far behind the region, but its rate of growth is also very low in this context – a sign that the indicator is unlikely to improve in the near future. 20 Figure 1.16: Changes in the HOI: Panama 1997 - 2008 compared to LAC and Central America Access to water 1.5 1.0 Sixth grade on 0.5 Access to time sanitation 0.0 Access to School enrollment electricity Panama LAC Central America Sources of exclusion 1.40 As noted at the outset, the HOI is composed of two parts: a coverage rate and a “penalty� linked to the degree of inequality, which is sometimes called the D-index (or dissimilarity index). Also mentioned earlier, the HOI controls for a set of circumstances that should not have any influence on a child’s access to services. Therefore, the overall D-index represents the degree of inequality as a result of all circumstances considered simultaneously. Table 1.12 reports the results of an exercise that aims to assess the relative importance of each circumstance in determining the degree of inequality (as measured by the D-index). In this exercise, one circumstance at a time is allowed to vary while the remaining circumstances are held at a specific value (in this case, they are fixed at their average value), and the D- index is computed. Intuitively, considering only one circumstance at a time, the numbers in Table 1.12 represent the proportion of available slots of services that would have to be reallocated to achieve equality in the corresponding indicator. As such, the D index is a measure of the relative level of inequality of opportunity, when all circumstance groups are considered simultaneously.6 1.41 High D index scores indicate a systematic tendency for some social groups to have significantly lower chances of accessing basic services needed to advance in life, due to sources of exclusion that society considers to be morally wrong. High D-indices are evident in the case of indigenous residents when it comes to access to water, electricity and sanitation. In the case of access to water and electricity living in an indigenous area is among the top two most important circumstances. In the case of access to sanitation, the D-index for the indigenous circumstance is also large (12.5). The high level of inequality in these opportunities means that policymakers can make significant gains in equity by improving the targeting of basic services. By utilizing the information from the HOI toolkit, policymakers can design mechanisms to effectively expand basic service access to those children who are most disadvantaged. 6 A D index above 10 is considered a troublingly high level of inequality of opportunity. 21 Table 1.5: Inequality of Opportunity Profile – Specific D-Indices Water Sanitation Electricity Circumstance Panama LAC Panama LAC Panama LAC Parental education 2.0 4.3 18.1 7.5 4.2 1.6 Gender of child 0.2 0.4 1.6 0.2 0.0 0.1 Gender of head of household 0.4 1.2 1.8 1.5 0.3 0.5 Per capita income 1.6 5.8 23.9 8.8 6.0 2.1 Urban or rural 3.2 10.8 24.4 13.6 6.1 4.2 Presence of both parents 0.5 1.5 4.5 1.7 0.1 0.2 Number of siblings 0.1 1.2 6.4 1.6 0.8 0.4 Indigenous 2.5 - 12.5 - 6.5 - Overall D-Index 8.2 14.1 29.9 18.3 15.9 7.3 Sixth Grade Completion School Attendance 10-14 Circumstance Panama LAC Panama LAC Parental education 3.7 5.2 0.8 1.2 Gender of child 2.2 2.4 0.1 0.3 Gender of head of household 1.4 1.0 0.6 0.3 Per capita income 1.7 2.2 0.2 0.4 Urban or rural 0.1 2.0 0.2 0.4 Presence of both parents 1.9 1.0 0.8 0.4 Number of siblings 1.4 2.5 0.5 0.2 Indigenous 2.3 - 0.2 - Overall D-Index 7.6 10.2 2.0 1.9 1.42 The profile of inequality of opportunity is defined by the relative size of the specific D indexes related with only one circumstance at a time. The higher a specific D index, the more important is the circumstance associated with it as a source of exclusion from access to the service considered. These profiles can be complemented by typical probabilities of access for specific groups. In the case of Figure 1.8, we plot the probability that a typical child from urban, rural and indigenous parts of the country has access to each of the five services. 22 Figure 1.17: Predicted probability of access in various parts of the country Predicted Probability of Access 0.98 0.97 0.98 1 0.96 0.92 0.87 0.88 .9 0.81 0.81 .8 0.68 .7 0.64 0.62 .6 .5 .4 0.35 0.35 .3 .2 0.10 .1 0 Access to sewage Access to electricity Sixth grade on time Access to water School Attendance Typical child in indigenous areas Typical child in rural areas Typical child in urban areas Note: Each bar corresponds to the predicted probability of access for a typical child by area 1.43 The figure above clearly shows that the probability of access for a typical indigenous child is far lower than that of a typical urban or rural child. Even in the case of school attendance where the probability of access for an indigenous child is highest, there is still a 10 percentage point difference in probabilities. At the other extreme, an urban child is seven times more likely to have access to sewage than an indigenous child. Together, these exercises are designed to help policymakers identify where scarce resources for expanding basic opportunities would be most beneficial in reducing inequality as the country progresses towards universal coverage. 1.44 The overall inequality of opportunity in Panama is higher than that of the region when it comes to the indicators for access to electricity and sanitation while it is lower in the case of access to water and completion of sixth grade on time. The overall D-index for access to electricity is twice as large as the regional average, while the overall D-index for access to sanitation – despite not being twice as large as the regional average – has the highest D-Index of the five indicators. In the case of sanitation, electricity, and water, the circumstances of location of residence, per capita income and parental education are important factors in explaining inequality. For example, in the case of access to sanitation, 23 percent of available connections would have to be reallocated if only the location was considered as a circumstance while 18 percent of available connections would have to be reallocated if per capita income were considered alone. For both education indicators, parental education is the circumstance that matters most in the resulting inequality of opportunity. Characteristics of the vulnerable 1.45 The predicted coverage rates calculated with the HOI methodology also allow an analysis of the average characteristics of the population. Table 1.14 reports the average characteristics of those in the lowest and highest deciles of predicted coverage in 2008. Here, too, stark regional contrasts are evident. For all five indicators of opportunities, the results show that virtually all individuals in the highest decile of coverage are from urban areas while the bottom decile of coverage is comprised entirely of those in indigenous and rural areas. Differences across the top and bottom deciles are not as stark when one 23 considers the gender of the child, whether or not the head of the household is male and whether both parents are in the household. However the average child in the bottom decile of coverage (in any of the five indicators) is likely to have roughly three times as many siblings as the child in the top decile of coverage. The parents of children in the bottom decile have on average less than one year of education compared to the parents in the top decile who have on average what amounts to some college education. Income disparities are also strikingly large. Sources of change of the HOI 1.46 Understanding what is behind the changes in the HOI is important for policy makers interested in leveling the playing field for children. The sources of expansion of the HOI can be classified into two main groups: (i) changes in the distribution of circumstances (the “composition effect�); and (ii) changes in the group-specific coverage rates (the “coverage effect�), which can be further decomposed into the “equalization� and “scale� effects. The scale effect captures the impact of proportional changes in coverage rates for all circumstance groups, whereas the equalization effect captures improved coverage rates specifically for circumstance groups with below-average coverage rates. As Table 1.13 suggests, the scale effect dominates the equalization effect for every indicator – not only for Panama but also for the region as whole. Nevertheless, the equalization effect is at the heart of equality of opportunities. A society that wants to level the playing field will focus on expanding opportunities mainly for the vulnerable circumstance groups, and the equalization effect is a clear indicator of progress toward this goal. Table 1.6: Decomposing changes in the HOI Panama LAC Scale Composit Total Composit Equalization Effec Total ion Equalization Scale Change ion Effect Effect t Change Effect Effect Effect Water -0.3 0.1 -0.3 -0.1 1.1 0.7 0 0.4 Electricity 1.3 0.2 0.4 0.6 1.3 1 0 0.3 Sanitation 0.5 0.3 -0.1 0.3 1 0.3 0.2 0.5 School 0.5 0.0 0.2 0.3 0.5 0.2 0.1 0.2 attendance Sixth grade completion 1.0 0.2 0.1 0.7 1.3 0.7 0.2 0.4 on time 24 Table 1.7: Characteristics of the vulnerable – 2008 Indicator / Deciles of coverage Both parents Number Years of Per capita Male Head is male Urban in the of education of Indigenous income Water household siblings parents Decile 1 0.49 0.80 0.00 0.80 6.39 0.15 161.34 0.82 Decile 10 0.51 0.76 0.99 0.82 1.91 14.25 4313.75 0.00 Electricity Decile 1 0.51 0.83 0.00 0.92 6.85 0.41 147.14 0.82 Decile 10 0.51 0.75 0.97 0.81 1.87 13.67 4488.48 0.00 Sanitation Decile 1 0.53 0.86 0.00 0.97 7.45 0.76 180.63 0.78 Decile 10 0.49 0.71 0.99 0.76 1.83 13.94 4352.72 0.00 Sixth grade completion Decile 1 0.66 0.92 0.04 0.77 6.30 0.62 298.71 0.66 Decile 10 0.51 0.59 0.95 0.72 1.70 15.44 5212.49 0.00 School attendance Decile 1 0.54 0.93 0.05 0.81 6.82 0.88 274.5 0.64 Decile 10 0.51 0.53 0.95 0.73 1.81 14.83 4287.91 0.00 25 Box 1.1: Measuring Welfare in Panama The welfare measure used in Panama and throughout this study is per capita consumption. Consumption is preferred over income as a measure of household welfare for several reasons. First, consumption tends to be less variable than income over the course of time (due to consumption smoothing) and thus provides a better measure of long-term welfare. Second, household surveys in developing countries typically measure consumption more accurately than income. Third, consumption of the household’s own production, which is often a large portion of consumption for agricultural households, is usually not captured well (if at all) in income data. Ignoring home-produced food would greatly understate the consumption levels of rural households. In this report, consumption includes; (i) the value of all food consumption, whether the food is purchased, home produced or received as a gift or donation; (ii) the use value of durable goods, (iii) the use value of housing, (iv) expenditures on utilities, (v) expenditures for education, (vi) health expenditures, and (vii) expenditures on other consumption items and services. Total household consumption is divided by the number of household members to provide the per capita consumption measure of welfare. This measure is then adjusted for spatial cost of living differences by region to ensure comparability of the measure across the country. Poverty is defined as having per capita consumption below the poverty line, while extreme poverty or food poverty is defined as having per capital consumption below the level of the extreme poverty line. The extreme poverty line is set at the cost of obtaining the minimum requirement of calories in a form that is acceptable to local tastes and preferences. To calculate this poverty line, the first step was to determine the food consumption patterns of the population, specifically those in the 11-39th percentile who are expected to seek out a relatively inexpensive diet (compared to those in higher percentiles) but who are also not so constrained that their diet does not reflect preferences (as the diet of those in the bottom decile might). This ‘food basket’ is then analyzed for caloric content and adjusted to ensure that the minimum daily requirements of calories are obtained. Finally, the resulting basket is costed using price data from the household survey. The general poverty line is simply the extreme line plus an allowance for non-food consumption. This allowance is calculated by, first, determining the share of total consumption devoted to non-food consumption among those whose total consumption is at or near the extreme poverty line. This percentage is added to the value of the food poverty line. Several efforts were made to ensure the comparability of the poverty estimates between 1997, 2003 and 2008. First, the questions on consumption were kept the same in the three surveys. The consumption aggregate was also constructed in the same way, with only minor changes that reflected new items having come on the market in Panama since 1997. The same poverty lines from 1997 were used in 2008 updated for changes in prices. For the general poverty line, the non-food component was inflated using the regional consumer prices indices of the country given the difficulty of calculating this from the household survey data itself. In short, the comparison of poverty rates between the three surveys can be correctly done given the way in which both the welfare measure was constructed and the poverty lines were updated 26 INTRODUCTION 2.1. Despite the relative stability in Panama’s poverty rates as compared to LAC, there is considerable mobility out of poverty in the country. This chapter identifies whether there are segments of society that are more adept at moving out of poverty. It also highlights the confluence of factors that hamper movements out of poverty and, in extreme cases, force households into poverty. Historically, such analyses have relied on the existence of panel data. However, with the development of techniques that transform repeated cross-sections of household survey data into pseudo-panels, we can use the LSMS data to estimate upper and lower bounds of mobility in Panama from 2003 to 20087 – thereby building on the poverty trends depicted in the preceding pages. The next section presents the methodology, followed by a section on the specifications of the models. The final section presents the results of the mobility profiles. 2.2. Several key messages arise from this analysis. Nationally, a significant number of Panamanians were able to move out of poverty between 2003 and 2008. The most conservative estimates indicate that of those who were poor in 2003, at least 9 percent were able to climb out of poverty by 2008. Similar estimates were obtained for upward mobility out of extreme poverty – at least 9 percent of the extreme poor in 2003 were able to leave destitution by 2008. 2.3. Drilling down, an examination of mobility profiles across key demographic characteristics reveals how these patterns varied across regions of the country and sub-groups of the population. Upward mobility was much stronger in urban areas, while downward mobility was most severe in indigenous areas. There was also a strong positive association between higher levels of education and the likelihood of exiting poverty – both moderate and extreme. Similarly there was a strong negative association between higher levels of education and the likelihood of entering poverty. Households that reported a female head were slightly better shielded from falling into extreme poverty – as a fraction of the extreme poor in 2003, more women were able to exit extreme poverty than men and fewer non-poor women fell into extreme poverty. The same was not true in the case of moderate poverty. METHODOLOGY 2.4. The procedure followed in this chapter allows an analysis of welfare dynamics by permitting the analyst to treat two or more rounds of cross-sectional data as a pseudo-panel. The procedure is described fully in Lanjouw et al. (2011) and builds on an out-of-sample imputation methodology described in Elbers et al. (2002, 2003) for small-area estimation of poverty (poverty maps). 2.5. This approach has a number of advantages for describing and understanding mobility. Traditional pseudo-panel techniques construct panels of cohort averages and track these cohorts through multiple rounds of cross-section survey data. While this allows the researcher to consistently identify a measure of mobility based on demographic factors or cohort-level shocks it does not account for mobility within those cohorts (see Antman and Mckenzie, 2007). By contrast, the technique of Lanjouw et al. (2011) predicts the welfare of the same household in a different period to construct the panel. An additional advantage of this approach is that, by estimating bounds of mobility and focusing on discrete movements 7 The lower bound method assumes that the prediction error for household i in round 1is the same as it is for round 2 – there is perfect positive autocorrelation. The upper bound method assumes that the prediction errors across rounds are independent of each other. 27 into and out of poverty or the middle class, we abstract from movements due to transitory shocks or measurement error.8 2.6. In the case of Panama, a model of consumption is estimated in the first round of the cross-section data (2003), using a specification which includes only time-invariant covariates. Parameter estimates from this model are then applied to the same time-invariant regressors in the subsequent survey round (2008). This allows one to estimate the (unobserved) first period’s consumption for the households surveyed in this second round. Analysis of mobility can then be done based on comparisons of the actual consumption observed in the second round along with this estimate from the first round. Depending on the assumptions made about the distribution of the residuals, one can derive an upper bound (residuals across rounds are independent) and a lower bound (residuals across rounds are perfectly positively correlated) for individual mobility. We refer the reader to Lanjouw et al. (2011) for a complete description of the methodology. MODEL SPECIFICATIONS 2.7. Two data constraints guide our selection of variables as we specify models to predict consumption. First the variables must be available in comparable form in both rounds of the survey and second they must be time-invariant. An additional restriction imposed is that the sample is restricted to households where the age of the head is between 25 and 55 years of age. We build up a sequence of progressively more richly specified prediction models. Appendix 1 presents basic descriptive statistics that sketch a profile of the poor in Panama. On average the poor report having half as many years of education as the non-poor. They are ten times more likely to report that their mother tongue is indigenous than the non-poor. Overwhelmingly the poor are more likely to report being employed in the agriculture sector than the non-poor. 2.8. With these characteristics of the poor in mind, the starting point is a parsimonious specification that includes variables that are most obviously time-invariant individual level characteristics: the gender of the head of the household, the years of education of the head (and its square), as well as a variable for the birth year of the head of the household (and its square). 2.9. The second model adds to the first a series of indicator variables for the education level of the mother of the household head. We also have information on the education level of the father of the household head. We choose to report the results that use mother’s education though the results using father’s education are similar. We also include a series of indicator variables that characterize the head’s mother tongue – indictors for indigenous and other language are included.9 2.10. Lastly, we take advantage of a retrospective module on migratory patterns to compute whether the respondent currently resides in the same location as he/she did five years ago. This information is collected at the level of the district. 2.11. Ideally a more complete model would include information on assets owned not only currently but also retrospectively. However given the wide variation in reported asset ownership across urban, rural and indigenous areas, this was not feasible with these data. The results of two models for consumption in 2008 are reported in Appendix 2 – with and without the indicator variables describing whether or not a respondent resided in the same district in 2003 and 2008. 8 We refer the interested reader to Lanjouw et al. (2010) for a detailed discussion of the validation results as well as of the assumptions inherent in the lower and upper bound techniques they describe. 9 The excluded reference category is mother tongue is Spanish. 28 RESULTS 2.12. Based on the full model, which controls for time-invariant individual, household and community variables we analyze mobility profiles. This allows us to identify sub-groups of the population that have seen significant upward or downward mobility – an exercise that is of particular interest to policymakers. In Tables 2.1 and 2.2 below, we report the results from both the upper and lower bound procedures described in Lanjouw et al. (2011).10 We limit the set of mobility profiles we report on to those where sample sizes allow for meaningful interpretation. 2.13. The pseudo-panel methodology allows us to compute poverty status in both periods. In columns 1-4 of tables 2.1 and 2.2 we report the percent that was always poor, exited poverty, entered poverty and that was never poor. These four columns add up to 100. In columns 5 and 6, respectively, we compute the fraction of the poor in 2003 that exited poverty by 2008 and the fraction of the non-poor in 2003 that entered poverty by 2008. These results are reported for both the lower bound (left panel) and upper bound methodologies.. 2.14. Several key messages arise from this analysis: there was considerable movement out of extreme poverty between 2003 and 2008. The data suggest that between 17 to 47 percent of the extreme poor exited extreme poverty (Table 2.1) between 2003 and 2008. However, this was mostly offset by entry into extreme poverty. Nationally between 0.7 – 6.9 percent of the non-extreme poor population joined the ranks of the extreme poor. Looking across regions of the country it appears that movements into and out of extreme poverty were largely offsetting in urban areas: between 41 to 82 percent exited extreme poverty while between .3 and 2.6 percent entered. In rural areas our estimates suggest that entry rates into extreme poverty were 2 to 15 percent while exit rates were 8 to 52 percent. For indigenous areas we find estimates of exit from extreme poverty ranging from 9 to 15 percent and of rates of entry into extreme poverty between 52 to 98 percent. 2.15. In the case of moderate poverty at the national level (Table 2.2) our estimates of the exit rates range from 11 to 40 percent – while entry rates were a bit lower at 0.9 – 18 percent. In urban areas the estimates of exit from poverty range from 3 – 10 percent while those of entry into poverty range from 0.5 to 7 percent. In rural areas exit rates from poverty were slightly higher (3 – 16 percent) than exit rates (1 – 16 percent). For indigenous areas we also find higher estimates of exit from moderate poverty (3 – 5 percent) than we do of entry (0.6 – 3 percent). 2.16. As a percentage of poverty in 2003, the lower bound estimates (the most conservative estimates) indicate that of those who were poor in 2003, at least 9 percent were able to climb out of poverty by 2008 (Table 2.2, Column 5). Similar estimates were obtained for upward mobility out of extreme poor – at least 9 percent of the extreme poor in 2003 were able to leave destitution by 2008. 2.17. By definition the upper bound estimates of mobility are larger. These suggest that of those who were poor in 2003, almost 44 percent were able to move out of poverty by 2008. The data suggest that upward mobility for the extreme poor was even stronger. More than half of the extreme poor (52 percent) were able to climb above the extreme poverty line. There was relatively less movement into poverty. Since the non-poor population is larger this was to be expected. However, the downward movements were sufficiently high to prevent poverty rates from falling further. Nationally, approximately one out of every ten non-poor households in 2003 became poor in 2008. Of the non-extreme poor households in 2003 approximately one out of every twenty fell into extreme poverty in 2008. Therefore, if the country had been able to prevent this fall into destitution, the extreme poverty rate would have dropped by even more than was observed between 2003 and 2008. 10 The upper bound estimates are obtained from 5 repetitions. 29 2.18. Upper bound estimates reveal that patterns of mobility varied across regions of the country. Upward mobility was much stronger in urban areas, while downward mobility was most severe in indigenous areas. We estimate that almost 62 percent of the urban poor in 2003 were able to climb out of poverty by 2008. In indigenous areas, only five percent of the poor were able to do so. On the other hand, 60 percent of the non-poor indigenous fell into poverty by 2008, compared to only 9 percent of the non- poor in urban areas. In rural areas, the rates of upward and downward mobility were much closer to each other – 41 percent of the poor in 2003 exited poverty according to our estimates while 26 percent of the non-poor entered. On balance in urban and rural areas the fractions of the population that exited poverty was greater than the fraction that entered. In indigenous areas, the reverse was true – a much smaller fraction was able to exit poverty than entered into poverty. 2.19. Lower bound estimates display the same overall pattern. Upward mobility was more prevalent in urban areas, while downward mobility was stronger in rural areas: 17 percent of the urban poor in 2003 were able to climb out of poverty by 2008. In rural and indigenous areas the estimates suggest that six and three percent, respectively, were able to do so. 2.20. In terms of extreme poverty, upper bound estimates suggest a small fraction (0.4 percent) of those that were poor and lived in urban areas in 2003 remained poor by 2008. The vast majority of the urban poor in 2003 (4 percent) exited extreme poverty. Two percent of the non-extreme poor fell into extreme poverty. Given the low levels of extreme poverty in urban areas in 2003 implied by the pseudo-panel methodology, it appears that 90 percent of the extreme poor in 2003 who were living in urban areas were able to extricate themselves from destitution, while in rural and indigenous areas only 60 and 14 percent respectively were able to accomplish the same. 2.21. Lower bound estimates are far more conservative. The fraction of the extreme poor that climbed above the extreme poverty line by 2008 was 35 percent in urban, 12 percent in rural, and 6 percent in the indigenous areas. Again, the effect of upward mobility on poverty reduction was lessened by downward mobility. In rural areas, 2 percent of the non-poor became poor and 1 percent of the non-extreme poor fell into destitution. In indigenous areas, 13 percent of the few non-poor became poor, and 27 percent of the non-extreme poor dropped below the extreme poverty line. 2.22. Urban and indigenous areas present stark contrasts in mobility patterns (See Figure 2.1). Upward mobility was much stronger in urban areas, while downward mobility was most severe in indigenous areas. Movements into extreme poverty were considerably more prevalent in indigenous areas. Over half of the few non destitute indigenous fell into extreme poverty between 2003 and 2008. For urban and rural dwellers, the rates of downward mobility into extreme poverty were of two and 11 percent, respectively. Thus in this instance too, greater fractions of urban and rural area residents exited extreme poverty than entered it. In indigenous areas however, the reverse was true. These trends are in line with the changes in poverty across regions reported in Chapter 1. 30 Figure 2.1: Region of residence and movements into and out of extreme poverty Panama 2003-2008: Movements into and out of poverty by region (Lower bound) 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Exiting Entering Exiting Entering Extreme Poverty Moderate Poverty Urban Rural Indigenous Panama 2003-2008: Movements into and out of poverty by region (Upper bound) 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Exiting Entering Exiting Entering Extreme Poverty Moderate Poverty Urban Rural Indigenous Note: The top panel is based on estimates from the lower bound methodology. The bottom panel is based on estimates from the upper bound methodology. Estimates as reported in columns 5 and 6 of Tables 2.1 and 2.2 31 2.23. An examination of mobility profiles across key demographic characteristics helps shed some light on the channels that were associated with mobility in this period. Households that reported the head was a woman were slightly better shielded from falling into extreme poverty – as a fraction of the poor in 2003 more women were able to exit poverty than men and fewer non-poor women fell into poverty. This was not true in the case of moderate poverty. 2.24. Just as there is a clear association between region of the country and mobility patterns, there is a strong positive association between higher levels of education and the likelihood of exiting poverty – both moderate and extreme. Similarly there is a strong negative association between higher levels of education and the likelihood of entering poverty (See Figure 2.2). Two key features stand out. Some poor households whose heads had not completed primary school were able to exit poverty in this period. However the fraction of such households entering poverty was twice as high as the fraction exiting it. In contrast, for those with higher education levels the rates of exit were higher and the rates of poverty entry were substantially smaller. Therefore, Panama’s efforts to expand access to secondary education in rural and indigenous areas are likely to result in significant poverty reduction in the medium run. 2.25. Almost all the poor households in which the heads had completed secondary education in 2003 were able to move above the poverty (72 percent) and extreme poverty (93 percent) lines. Of poor households where the heads had completed only primary school, 66 percent were able to move out of extreme poverty, and 50 percent were able to move out of poverty. Households where the head had not completed primary school experienced much lower upward mobility (22 and 30 percent for poverty and extreme poverty respectively). In the case of poverty downward mobility was almost twice as high for such households (38 percent) while in the case of extreme poverty downward mobility was smaller (18 percent). 32 Figure 2.2: Education and movements into and out of poverty Panama 2003-2008: Movements into and out of poverty by education (Lower bound) 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Exiting Entering Exiting Entering Extreme Poverty Moderate Poverty Head has not completed primary education Head has completed primary education Head has completed secondary education Panama 2003-2008: Movements into and out of poverty by education (Upper Bound) 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Exiting Entering Exiting Entering Extreme Poverty Moderate Poverty Head has not completed primary education Head has completed primary education Head has completed secondary education Note: The top panel is based on estimates from the lower bound methodology. The bottom panel is based on estimates from the upper bound methodology. Estimates as reported in columns 5 and 6 of Tables 2.1 and 2.2 33 Table 2.1: Mobility out of extreme poverty, 2003-2008 Lower Bound Observed Upper Bound Fraction Fraction Fraction of Fraction of Always Exited Entered Never of poor Headcount Headcount Always Exited Entered Never of poor non-poor non-poor Poor Poverty Poverty Poor who 2003 2008 Poor Poverty Poverty Poor who who entered who entered exited exited Overall 14.3 3.1 0.6 82.1 17.6 0.7 17.3 14.9 9.2 8.2 5.7 76.9 47.2 6.9 Regions: Urban 3.2 2.2 0.3 94.4 40.7 0.3 5.4 3.4 1.0 4.4 2.5 92.2 82.2 2.6 Rural 20.9 1.9 1.6 75.6 8.3 2.1 22.8 22.5 10.9 12.0 11.7 65.5 52.4 15.1 Indigenous 81.3 8.3 5.4 5.0 9.2 51.9 89.5 86.7 76.4 13.1 10.3 0.2 14.6 98.1 Demographic Characteristics: Male 14.7 3.0 0.9 81.4 16.8 1.1 17.7 15.6 9.6 8.1 6.0 76.3 45.7 7.3 Female 11.8 4.3 0.8 83.0 26.7 0.9 16.2 12.6 7.5 8.6 5.1 78.7 53.4 6.1 Has not completed primary 47.6 3.7 2.9 45.8 7.1 6.0 51.3 50.5 37.9 13.4 12.6 36.1 26.1 26.0 Has completed primary 12.6 3.5 1.1 82.8 21.8 1.3 16.1 13.7 6.1 10.0 7.6 76.3 62.4 9.1 Has completed secondary 1.2 0.7 0.0 98.0 37.2 0.0 2.0 1.2 0.3 1.7 1.0 97.0 87.4 1.0 Mother tongue indigenous 8.3 1.9 0.6 89.2 18.6 0.7 10.2 8.9 3.5 6.7 5.4 84.4 65.8 6.0 Mother tongue other 65.5 13.5 2.5 18.5 17.1 12.1 79.0 68.0 58.5 20.5 9.5 11.5 25.9 45.4 Employment Characteristics: Formal 5.8 2.2 0.3 91.6 27.4 0.4 8.0 6.2 3.1 4.9 3.1 88.9 61.4 3.3 Informal 25.2 1.7 1.3 71.8 6.5 1.8 26.9 26.5 17.6 9.3 8.9 64.2 34.5 12.2 Unemployed 16.3 3.1 0.9 79.7 16.0 1.1 19.4 17.2 10.1 9.3 7.0 73.6 47.7 8.7 Inactive 15.1 5.0 2.4 77.4 25.1 3.0 20.1 17.5 9.9 10.2 7.6 72.3 50.6 9.5 Employed in Agriculture 46.5 2.5 9.3 41.7 5.0 18.2 49.0 55.8 38.9 10.1 16.9 34.1 20.6 33.2 Employed in Manufacturing 6.6 5.7 1.2 86.4 46.3 1.4 12.3 7.9 2.6 9.7 5.3 82.4 78.8 6.0 Place of birth: Large city 2.0 0.5 0.0 97.5 20.3 0.0 2.5 2.0 0.4 2.1 1.6 95.9 83.4 1.6 Medium city 1.1 3.5 0.7 94.7 75.8 0.7 4.7 1.8 0.5 4.1 1.3 94.0 89.0 1.4 Other urban center 5.0 2.6 0.9 91.4 34.4 1.0 7.7 5.9 2.4 5.3 3.6 88.7 69.2 3.9 Rural area 24.4 4.9 1.6 69.1 16.7 2.3 29.3 26.0 16.8 12.5 9.2 61.5 42.8 13.1 Source: Authors’ calculations 34 Table 2.2: Mobility out of poverty, 2003-2008 Lower Bound Observed Upper Bound Fraction Fraction Fraction of Fraction of Always Exited Entered Never of poor Headcount Headcount Always Exited Entered Never of poor non-poor non-poor Poor Poverty Poverty Poor who 2003 2008 Poor Poverty Poverty Poor who who entered who entered exited exited Overall 33.4 4.3 0.6 61.7 11.4 0.9 37.7 34.0 22.5 15.2 11.5 50.8 40.4 18.5 Regions: Urban 19.0 2.7 0.8 77.4 12.6 1.1 21.8 19.9 10.3 11.5 9.6 68.6 52.6 12.2 Rural 48.7 5.3 1.3 44.7 9.8 2.8 54.0 50.0 33.1 20.9 16.9 29.2 38.7 36.7 Indigenous 96.4 1.7 0.5 1.4 1.8 25.7 98.1 96.9 92.8 5.3 4.1 -2.2 5.4 216.0 Demographic Characteristics: Male 33.7 4.4 0.5 61.5 11.5 0.8 38.0 34.2 22.6 15.4 11.6 50.4 40.6 18.7 Female 31.9 4.6 1.4 62.1 12.6 2.2 36.5 33.3 22.0 14.5 11.3 52.2 39.7 17.8 Has not completed primary 75.5 1.1 1.3 22.1 1.4 5.4 76.6 76.8 63.6 13.0 13.2 10.2 17.0 56.2 Has completed primary 37.9 4.7 0.5 56.9 11.0 0.9 42.5 38.4 23.4 19.2 15.0 42.4 45.1 26.2 Has completed secondary 10.3 1.8 1.2 86.7 15.1 1.4 12.1 11.5 4.4 7.6 7.1 80.9 63.3 8.0 Mother tongue indigenous 27.2 4.0 0.7 68.0 12.9 1.1 31.2 27.9 16.3 15.0 11.7 57.1 47.9 17.0 Mother tongue other 90.2 4.5 0.4 4.9 4.7 8.0 94.7 90.6 83.8 10.8 6.8 -1.4 11.4 126.9 Employment Characteristics: Formal 21.4 1.4 1.2 76.0 6.0 1.6 22.8 22.6 13.1 9.7 9.5 67.7 42.4 12.3 Informal 47.4 5.1 0.7 46.8 9.7 1.5 52.5 48.1 34.8 17.7 13.4 34.1 33.8 28.1 Unemployed 40.6 2.1 1.0 56.3 4.9 1.7 42.7 41.6 29.8 12.9 11.8 45.5 30.3 20.6 Inactive 34.8 7.1 9.1 49.0 16.9 15.7 41.9 44.0 31.2 10.7 12.7 45.4 25.4 21.9 Employed in Agriculture 73.5 1.2 3.4 21.9 1.6 13.5 74.6 76.9 63.6 11.1 13.3 12.0 14.8 52.6 Employed in Manufacturing 29.5 2.8 3.8 64.0 8.5 5.6 32.2 33.2 17.2 15.0 16.0 51.8 46.5 23.6 Place of birth: Large city 12.9 0.9 2.5 83.8 6.3 2.9 13.7 15.4 6.9 6.9 8.5 77.7 50.1 9.9 Medium city 14.1 12.0 0.4 73.5 46.1 0.6 26.1 14.5 6.8 19.3 7.7 66.2 73.9 10.4 Other urban center 23.5 3.0 1.2 72.4 11.2 1.6 26.5 24.7 13.7 12.8 11.0 62.5 48.3 15.0 Rural area 48.7 5.0 0.6 45.7 9.3 1.4 53.7 49.3 35.3 18.4 14.0 32.3 34.3 30.3 Source: Authors’ calculations 35 2.26. The evidence on mobility from examining household heads whose mother tongue is an indigenous language presents an interesting contrast to the evidence found in the case of regions of residence. Of those who were poor in 2003 and whose mother tongue was an indigenous language – 9 percent had successfully exited moderate poverty. In contrast 0.7 percent had entered poverty. Movements into and out of extreme poverty were similar (14 percent of the poor exited while only 0.3 percent entered). This is the same pattern as demonstrated by those whose mother tongue is a language other than Spanish – suggesting that indigenous location is a greater barrier to mobility than is indigenous language. 2.27. Whether or not the head of the household was employed in the formal or informal sector also impacted the mobility transitions experienced by households in Panama. The majority of those who were poor in 2003 and who belonged to the formal sector exited poverty by our calculations (55 percent) while in the case of the informal sector this number was somewhat smaller (36 percent). The data suggest that 20 percent of poor households employed in the informal sector entered poverty in this period compared to 9 percent of those employed in the formal sector. 2.28. The head of household’s sector of employment was also associated with substantially different mobility experiences over this time frame. The difference in the fractions entering poverty relative to exiting it was 16 percentage points for agriculture – with a net implied increase in those in poverty. In contrast, there was a 36 percentage point difference between entry and exit into poverty rates for those who were poor in 2003 and employed in manufacturing. In the case of those employed in manufacturing the net implication was of a decrease in poverty. 2.29. Those who were born in large or medium sized cities had very similar rates of entry and exit into poverty. The data suggest that there was substantial exit from poverty for these groups as well as for those born in smaller urban centers (though this rate of exit was smaller). Those born in rural areas also saw a higher fraction exit poverty than enter it though the differences in these two were substantially smaller than those seen in other groups based on place of birth of the household head. 36 INTRODUCTION 3.1. To what extent has Panamanians’ access to public services changed? Previous analyses of Panama (World Bank 2000 and 2007) depicted a country with a large degree of inequality in individuals’ access to public services, depending on their geographic location or welfare status. The purpose of this chapter is to examine the evolution of education, health and malnutrition indicators between 1997 and 2008 in Panama. To what extent has this changed? Clearly, understanding the changes that have occurred is the first step to identifying means to further improve existing policy and the nation’s pace of human capital accumulation. 3.2. The chapter is organized as follows: in the next section we examine education. We look at changes in educational outcome indicators, and changes in disparities in access between the poor and the non-poor. We conclude that while educational outcome indicators have substantially improved in Panama, striking inequalities still persist between the indigenous and the non-indigenous. 3.3. In the following section, we look at changes in health outcomes, and disparities between the poor and the non-poor. Health indicators have not changed significantly, despite substantial increases in spending and in the supply of health care services. Inequities in access to services between wealth and ethnic lines also remain largely unchanged. 3.4. In the final section, we look at changes in malnutrition which has remained extremely high in Panama, especially in indigenous areas. EDUCATION 3.5. The formal education system in Panama consists of basic education, secondary and higher education. Basic education is free and compulsory and comprises two years of pre-primary, six years of primary (grades 1-6) and three years of lower secondary education (grades 7-9). Upper secondary education is also free and consists of three years of studies in diversified careers for those that want to proceed to higher education or to enter the labor market. Primary education consists of six grades and currently serves 450,000 students. Ninety percent of these students are in public schools. Of the total number of students in public schools, two-thirds are in single-grade schools and the other third in multi- grade schools. The latter modality is offered mostly in rural and indigenous areas. Schooling Attainment Overtime 3.6. Panama has one of the most highly educated population in the region, and the country continues to advance in schooling attainment. About 92 percent of countries adult population is able to read, and approximately 60 percent of them have had some secondary education. In Mesoamerica, only Costa Rica has better literacy rates, and no other country has higher net enrollment rates in secondary school. Relatively few people in the country have no schooling at all, except for the indigenous that have not been able to accompany the rest of the country in its strong trajectory of human capital accumulation. 3.7. Average schooling of adults in the country has increased dramatically across generations. As seen in Figure 3.1, while adults born in the 1930s exhibit on average five years of schooling, those born in the 1980s and who have entered the labor force during the last decade have accumulated almost 37 twice as much education. The country’s young adults have on average close to 10 years of schooling. More than 85 percent of them have completed primary school (compared to 45 percent of those born in the 1930s), and about 40 percent have completed some secondary education (compared to 10 percent of the 1930 birth cohort). Figure 3.1: Schooling Attainment by Region of Residence Years of Schooling Primary Completion Secondary Completion .6 1 10 Secondary Completion Rate .8 Primary Completion Rate Years of Schoolint .4 .6 5 .4 .2 .2 0 0 0 1930 1940 1950 1960 1970 1980 1930 1940 1950 1960 1970 1980 1930 1940 1950 1960 1970 1980 Year of Birth Year of Birth Year of Birth National Urban National Urban National Urban Rural Indigenous Rural Indigenous Rural Indigenous Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 3.8. Women have been the main drivers of the increase in average schooling in the country. Females born in the 1980s surpass males in terms of average years of education (9.5 versus 9 years), and secondary completion rates (42 versus 35 percent). Most notably, almost 60 percent more women than men have completed some post secondary degree (19 percent for women versus 12 percent for men). In contrast, for the cohorts of those born in the 1930s, almost twice as many males had completed some post secondary education (Figure 3.2). At this current trend, there will likely be twice as many women with college degrees than man in the near future. This implies that labor market institutions will have to adapt to women’s needs in the high productivity sectors if the country is to take advantage of its high skills female labor force. Figure 3.2: Tertiary completion rates by birth cohort and gender .2 Terciary Completion Rate .15 .1 .05 0 1930 1940 1950 1960 1970 1980 Year of Birth Female Male Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. Access to Education Services over Time and Across Regions 3.9. The heavy investments that Panama undertook to mitigate the disparities in access to education in the last decade appear to be bearing fruit. As indicated in Figure 3.3, children are entering the school system earlier and staying longer. The proportion of children aged between 13 and 18 enrolled has risen significantly as access to secondary education has increased dramatically in the last decade. Most importantly, all of this expansion seems to have happened in rural and in indigenous areas where lack of services had severely restricted secondary education opportunities in the past. As a 38 result, poverty for the next generation of adults is likely to decrease substantially as education is closely linked to upward mobility and helps prevent downward mobility. Figure 3.3: School Enrollment by Age National 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1997 2008 Urban Rural Indigenous 1 1 1 0.9 0.9 0.9 0.8 0.8 0.8 0.7 0.7 0.7 0.6 0.6 0.6 0.5 0.5 0.5 0.4 0.4 0.4 0.3 0.3 0.3 0.2 0.2 0.2 0.1 0.1 0.1 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 1997 2008 1997 2008 1997 2008 Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 3.10. Efforts to expand access to pre-school programs in the country are also noteworthy. As seen in Figures 3.3 and 3.4, the proportion of children under six years of age enrolled in pre-school programs has increased substantially between 1997 and 2008. As in the case of secondary school, most of this increase in enrollment has occurred in rural and indigenous areas. Therefore, in relative terms, the increase has been the greatest among the extreme poor, for whom enrollments rates in pre-school have increased almost four-fold. For all poor, enrollment rates have more than doubled in pre-school during the same period. As there is mounting evidence that early childhood education and pre-school programs have significant impacts on cognitive skills and future student performance, this outcome is likely to affect human capital accumulation and poverty in the long run. Returns to Education and Its Relationship with Poverty 3.11. Education is a key determinant of welfare. As shown in Figure 3.5 below, there is a strong association between years of schooling of the head of the household and the probabilities of the household being either poor or extremely poor. Households with heads that have no schooling at all have a probability of being extremely poor four times higher than households with heads that have approximately 12 years of schooling. 3.12. Nevertheless, the associaten between schooling and poverty has weakened in the last decade. As shown in Figure 3.5, the slope of the association has substantially decreased, especially for extreme poverty. Decreasing returns to education are likely behind this change. As seen in Figure 3.6, returns to education have indeed dropped by almost 16 percent between 1997 and 2008. In 1997, one additional year of schooling yielded in average 10.2 percent more income. These returns have steadily decreased to 9.5 percent in 2003 and 8.6 percent in 2008. 39 Figure 3.4: Enrollment by Region and Poverty Status % Enrolled in Pre-school Programs 25 25 25.8 20 23.5 20 19.9 20.8 19.2 18.9 15 15 17.1 15.7 16.4 15.2 13.4 12.5 10 10 11.8 10.6 8.7 7.5 5 5 6.3 5.7 4.0 0 0 1997 2008 1997 2008 1997 2008 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private % Enrolled in Primary School 100 80 85.7 87.5 86.1 87.1 86.2 85.2 80 87.4 89.1 88.4 85.7 86.9 84.9 84.8 84.5 81.3 79.7 67.7 69.4 60 60 68.3 69.7 40 40 20 20 0 0 1997 2008 1997 2008 1997 2008 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private % Enrolled in Secondary School 80 85.1 87.8 80 83.6 84.7 75.0 71.7 60 60 65.0 67.1 62.8 64.9 61.7 60.1 53.4 50.6 40 40 40.8 39.5 34.1 33.7 20 20 19.8 19.4 0 0 1997 2008 1997 2008 1997 2008 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private % Enrolled in University or other Post Secondary Programs 40 30 32.5 30.9 33.7 34.0 30 23.6 20 24.8 19.5 20 18.0 21.2 10 12.4 10 9.4 8.0 7.5 6.0 0 0 2.8 2.7 1.5 1997 2008 1997 2008 1997 0.8 2008 0.8 1997 2008 1997 2008 Urban Rural Indigenous Non Poor Poor Public Private Public Private Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 40 3.13. The expansion of secondary education in rural and indigenous areas may be one of the causes behind the diminishing returns to schooling. As Panama has strived to rapidly increase access to education to the poor in rural and indigenous areas, the average quality of education in the country may have suffered. As indicated in Figure 3.4, between 1997 and 2008, secondary enrollment increased by 40 percent in rural areas, and 90 percent in indigenous areas. As a result, enrollment rates among the poor increased by more than 50 percent. Moreover, almost all of this increase has been obtained by the expansion of the public school system. Therefore, the proportion of secondary enrollment in public versus private schools has increased substantially at the national level due to the expansion of the public system in rural and indigenous areas. Figure 3.5: Schooling and Poverty Poverty Extreme Poverty .5 .3 Probability of being Extremely Poor .4 Probability of being Poor .2 .3 .2 .1 .1 0 0 0 5 10 15 20 0 5 10 15 20 Head of Household's Years of Education Head of Household's Years of Education 1997 2008 1997 2008 Source: World Bank staff calculations based on ENV 1997 and 2008 data. Figure 3.6: Percent returns to one additional year of education 10.5% 10.3% 10.2% 10.1% 9.9% 9.7% 9.5% 9.5% 9.3% 9.1% 8.9% 8.7% 8.6% 8.5% 1997 2003 2008 Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 3.14. While there is indication that the quality of public education services has improved relative to private services over time, important differences still remain. As indicated in Figure 3.7, absenteeism rates in public secondary schools, while dropping over the years, are still almost twice as large as in private schools. Moreover, in public schools teachers’ absences are the main reason given by students to explain their own absence from school. In private schools, students’ absences are most likely to occur for personal reasons. Finally, the proportion of students without text books in public secondary schools has almost doubled between 1997 and 2008. While it has also increased in private schools, the incidence is 37 percent higher in public schools. 41 Figure 3.7: Quality of Education Services: Public versus Private Schools Percentage of Students Absent 2 or More Weeks/Year Percentage of Students without Text Books 13.0% 34.0% 12.6% 31.0% 12.0% 26.0% 11.0% 9.6% 10.0% 21.0% 9.0% 14.6% 9.2% 16.0% 13.7% 8.0% 11.0% 6.7% 8.0% 7.0% 7.2% 9.8% 6.9% 6.1% 6.0% 6.0% 1997 2003 2008 1997 2003 2008 Public Private Public Private Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. 3.15. There may also be a decline in the cognitive ability of students entering the school system, a trend that may also be contributing to the drop in returns to education. As public school access expands to the rural and indigenous areas, where the prevalence of chronic malnutrition is high, the level of cognitive development of students may fall. Chronic malnutrition, or stunting, is known to be highly associated with low cognitive skills and school performance. In indigenous areas, stunting rates have been increasing since 1997 and have reached 62 percent for children under five. HEALTH 3.16. Panama’s public health spending is significantly greater than most countries in the Latin America and Caribbean region with similar per capita income levels. During 1990-2008, the upper middle-income countries in LAC devoted an average of 3.1 percent of GDP to health spending, while Panama spent almost twice as much. Only Costa Rica comes close to Panama in terms of health spending, while Chile spends less than one-half as much. 3.17. Because of its high expenditures on health Panama exhibits better indicators than the average LAC country on infant, child and maternal mortality (Figure 3.8). These have declined since 1990, but this decline has not been as dramatic as in other LAC or Central America countries. Relative to the LAC average this may be explained by better initial indicators. 3.18. Nevertheless, Panama is the only country in Central America that is not on track to meet some of its most important Millennium Development Goals (MDGs) in 2015, namely child malnutrition, infant-mortality and under-5 mortality. As seen in Figures 3.9 and 3.10, most countries in the sub-region have already reached their goals in malnutrition and infant mortality, or are well on track to meet them (see Figure 3.8 for under-five mortality). 42 Figure 3.8: Key Health Indicators 1990-2009 Panel 1: Infant Mortality Panel 2: Under 5 Mortality Panel 3: Maternal Mortality 40.00 140 50.00 35.00 120 30.00 40.00 100 25.00 30.00 80 20.00 60 15.00 20.00 10.00 40 10.00 5.00 20 0.00 0.00 0 1990 1995 2000 2005 2009 1990 1995 2000 2005 2009 1990 1995 2000 2005 2009 LAC Central America Panama LAC Central America Panama LAC Central America Panama Source: UNICEF in ECLAC BADEINSO Note: Infant mortality is rate per 1000 live births as is under-five mortality. Maternal mortality is per 100,000 live births. LAC is an average of 33 countries in Latin American and the Caribbean. Figure 3.9: Path to Meeting Malnutrition MDG Costa Rica El Salvador Guatemala 12 35 6 10 30 5 25 4 8 20 3 6 15 2 4 10 1 2 5 0 0 0 Honduras Nicaragua Panama 25 14 8 12 7 20 10 6 15 5 8 4 10 6 3 4 2 5 2 1 0 0 0 Source: World Bank WDI 2010 3.19. Lack of access to adequate prenatal and infant health care is a likely contributor to chronic malnutrition in rural and indigenous areas. Preventive prenatal and infant health care are known to protect infants from stunting by providing immunization against diphtheria, pertussis and tetanus (DPT) and measles, and by informing mothers about good hygiene and nutrition practices. While access to prenatal care has increased substantially in urban and rural areas, covering almost all pregnancies, in indigenous areas, which contribute the most to stunting in the country, 40 percent of women go through their whole pregnancies without a single visit to a health center (Figure 3.11). Therefore, while improving the incomes of the indigenous is crucial for enhancing their diets, it may not be enough to tackling chronic malnutrition if access to preventive prenatal and infant care is not also expanded 43 Figure 3.10: Path to Meeting Infant Mortality MDG Costa Rica El Salvador Guatemala 20 60 70 18 60 50 16 50 14 40 12 40 30 10 30 8 20 20 6 10 10 4 2 0 0 0 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 Honduras Nicaragua Panama 50 60 30 45 40 50 25 35 40 20 30 25 30 15 20 15 20 10 10 10 5 5 0 0 0 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 1990 2007 2008 2009 2010 2011 2012 2013 2014 2015 Source: World Bank WDI 2010 Figure 3.11: Percentage of Pregnant Women with at Least One Prenatal Care Consultation 100.0% 91.4% 90.0% 91.2% 80.0% 80.0% 66.0% 76.5% 70.0% 60.0% 60.6% 59.6% 50.0% 47.1% 40.0% 36.0% 30.0% 1997 2003 2008 urbana rural indígena Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data. MALNUTRITION 3.20. Indicators of malnutrition provide a somewhat mixed picture of what has happened between 1997 and 2008 in Panama. The overall levels of malnutrition have remained high during the period. While acute malnutrition as measured by weight for age has decreased from 6.7 to 3.9 percent, chronic malnutrition has remained high in the country as a whole, at approximately 20 percent. In indigenous areas it has increased by almost 27 percent between 1997 and 2008, going from 48.5 to 61.9 percent of children under five years of age. These rates are comparable to levels of stunting in countries with less than one tenth of Panama’s GDP per capita, such as Burundi and Ethiopia. 44 3.21. As indicated in Table 3.1, stunting prevalence in rural and indigenous areas are much higher than in urban areas. Most notably, in indigenous areas, stunting rates have actually been increasing since 1997 and have reached 62 percent for children under five. Stunting reflects not only inadequate current nutritional intake, but also the cumulative effects of malnutrition in past. 3.22. These extremely high levels of stunting in indigenous areas are not surprising since the expenditures of most indigenous households are well below the national extreme poverty line. The average per capita consumption in indigenous areas amounts to less than half of this line. That is, the average indigenous household is not able to afford half of their daily caloric needs. Table 3.1: Malnutrition in Panama (children younger than 5 years) Wasting (Weight for Height). Percentage of children under five years 1997 2003 2008 Urban 0.89 1.26 1.46 Rural 1.07 1.37 0.93 Indigenous 1.78 1.18 0.97 Total 1.07 1.28 1.25 Underweight (Weight for age) Percentage of children under five years 1997 2003 2008 Urban 2.79 4.07 2.48 Rural 7.12 5.6 3.12 Indigenous 21.04 21.45 12.27 Total 6.7 6.76 3.94 Stunting (Height for age) Percentage of children under five years 1997 2003 2008 Urban 5.66 13.81 10.48 Rural 14.5 18.36 17.22 Indigenous 48.5 56.73 61.88 Total 14.26 20.58 19.14 3.23. In addition to low income, other important determinants of malnutrition are inadequate water supply and sanitation, given their direct impact on infectious disease. Repeated or persistent diarrhea is known to be one of the main causes of malnutrition in poor areas. As indicated in Table 3.2, the incidence of diarrhea in children under five years of age has increased in Panama from 21 percent in 1997 to approximately 23 percent in 2008. Child diarrhea is particularly high and rising in indigenous areas where stunting is also prevalent; over this time period, child diarrhea has increased from approximately 37 percent to 41 percent. The rising rates of stunting in these areas are likely closely linked to the increasing incidence of diarrhea. As discussed in Chapter 1, Panama lags well behind Latin and Central America in terms of improvements in access to sanitation. It is no surprise then that 45 the incidence of child diarrhea and malnutrition has been increasing in the country as whole and especially in indigenous areas. Table 3.2: Incidence of Diarrhea in children under 5 1997 2003 2008 Urbana 18.4% 18.3% 20.5% Rural 20.6% 17.6% 18.8% Indigenous 36.9% 40.1% 40.5% National 21.5% 21.0% 22.7% CONCLUSION AND POLICY IMPLICATIONS 3.24. Our analysis of human capital accumulation and access to schooling in this chapter indicates that Panama should continue to have one of the most highly educated populations in the LAC region. The stock of human capital has grown consistently for a number of generations, and given the tremendous investments being made in the expansion of basic education it should continue to grow in the future. 3.25. However, the overall growth in human capital belies the striking disparities between the growth of the general population and that of the indigenous. While rural workers have been converging with their urban peers in terms of average years of schooling and primary and secondary completion rates, the indigenous are lagging further and further behind. A concerted effort to improve access to basic and secondary education for the indigenous people is likely needed if the country is to eradicate extreme poverty and reduce its high levels of inequality. 3.26. However, more access to schools will not produce the expected outcomes if indigenous students continue to suffer from chronic malnutrition. Stunting in indigenous communities reaches levels comparable to countries that have less than one-tenth the per capita GDP of Panama. A parallel concerted effort to eradicate chronic malnutrition will therefore be required to ensure that schooling investments do pay off in terms of poverty reduction and growth. Policy toward this goal should include not only programs to increase the incomes of the poor, but also investments in sanitation infrastructure which would allow the country to catch up with the rest of the region. 3.27. Finally, despite being by far the biggest per capita spender in health in Latin America, Panama has extremely weak health outcomes. It lags behind other countries with similar per capita incomes in several important health indicators, including infant mortality, maternal mortality rate, and malnutrition. The declining coverage of immunization among the poor and the extreme poor is of particular concern. 46 INTRODUCTION 4.1. Panama’s underperformance in poverty and inequality reduction cannot be attributed to the lack of social spending. The country spends more than 16% of its GDP in the social sector. This level of social spending is higher than the average of a sample of representative countries in Latin America (13%) and matches Costa Rica, a country known for its considerable investment in social programs and for having achieved substantial poverty reduction in the past. In fact, if the overall amount currently spent on the social sectors were to be distributed in cash to the whole population, no one in Panama would live with less than $3 dollars a day, that is, poverty would be completely eradicated. 4.2. Social spending in Panama is not only high but has been growing faster than GDP since 2005. Between 2005 and 2008, spending in education and health grew in real terms by 40 and 58 percent respectively. Even during a period of fast growth, as a percent of GDP, spending in education grew from 3.8 to 4 percent and in health from 5.3 to 5.4 percent. During the same period, spending in social protection grew from 6.7 to 6.9 percent of GDP (Silverio, 2009). 4.3. In addition to high spending on the social sectors, Panama also has a large program of untargeted indirect subsidies. The country spends approximately 1.2 percent of GDP on subsidies for water, electricity, fuel, and mortgage interest rates. Together these amount to almost US$280 million a year. If this sum were to be transferred directly to those below the extreme poverty line, this would amount to a per capita transfer of approximately US$1.60 a day, which would completely eradicate extreme poverty in the country, even if one takes into account targeting errors. Figure 4.1: Expenditure on Goods with Price Subsidies by Income Decile Electricity Cooking Gas Fuel and Lubricants Source: World Bank staff calculations based on ENV 2008 data. 4.4. More than three quarters of these indirect price subsidies ($214 million) go to electricity, cooking gas and fuels, and, as indicated in Figure 4.1, these are likely to be highly regressive since the poor consume these items considerably less than the non-poor. In fact, if these subsidies were eliminated and the resources saved were to be distributed to the lowest four deciles, the transfers would more than compensate the poor for the losses due to the price increases. 4.5. As a response to the global financial crisis, the Government of Panama augmented some of these untargeted subsidies, but also expanded direct monetary transfers targeted to the poor. It continued to subsidize public transportation and introduced additional subsidies for electricity, water, cooking fuel, and mortgage interest rates. Nonetheless, recognizing that price subsidies are not the most cost-effective approach to offset the impact of the crisis on the poor, the GoP took a number of steps to expand its safety net system with direct transfers. It increased the Red de Oportunidades transfers to 47 eligible families from B/35 to B/50 per month in early 2009, created a noncontributory pension for the elderly poor and significantly strengthened scholarships and vocational training programs. This response to the crisis by the Government is a sign of a paradigm shift in the country. A new view of social protection, one in which ineffective price subsidies are eliminated and the effect of crises and fiscal reforms on the poor is mitigated with direct transfers, is emerging. 4.6. In this chapter we examine social protection spending in Panama. The first section of the chapter presents a broad assessment of Panama’s Social Protection (SP) System. It focuses on the major public social insurance (SI) and social assistance (SA) programs.11 Other smaller assistance programs, particularly those implemented by NGOs, are not covered. 4.7. In the second part of the chapter we assess the Conditional Cash Transfer (CCT) program Red de Oportunidades (RdO). We evaluate its targeting performance and its impact on consumption per capita in indigenous areas. 4.8. In the final section we assess the likely incidence and progressivity of two new transfer programs launched in 2010, namely Beca Universal and Bono 100 a los 70. THE CURRENT SOCIAL PROTECTION SYSTEM IN PANAMA 4.9. While most social spending in Panama goes to health and education (about nine percent of GDP), the rest (seven percent of GDP) goes to social protection (SP). Social protection spending encompasses spending on both social insurance (SI) and social assistance (SA). As in most countries in Latin America, social protection spending in Panama is mainly limited to social insurance programs, which are typically aimed at mitigating unemployment, health and old age poverty risks (e.g., health insurance, unemployment insurance and old age pension). Eligibility to SI in Panama requires participation in the formal labor market and corresponding contributions to the SI programs via payroll taxes. 4.10. Because the majority of the poor work in the informal sector (Galiani, 2006), they have de facto been excluded from formal SI programs in Panama. Thus, as is typical in most Latin American countries, Panama has developed a variety of social assistance (SA) programs to help the poor, regardless of whether they are unemployed or not, healthy or ill, old or young. These programs range from untargeted price subsidies to targeted food-based programs. More recently the GoP has followed other countries like Brazil, Mexico, and Colombia, and launched a targeted CCT, Red de Oportunidades, in 2007. The RdO provides cash assistance to poor families in exchange for beneficiary compliance with key human development actions such as school attendance, vaccines, prenatal care and child growth monitoring. Since the middle of 2008, the RdO has also provided transfers to extremely poor elderly (over 62) who have no pension. 4.11. In recent years, social spending has grown rapidly. Between 2005 and 2008, spending on education and health increased by 40 and 58 percent, respectively. Although GDP grew at very high rates, in terms of GDP spending increased from 3.8 to 4.0 percent on education and from 5.3 to 5.4 percent in health. Social protection expenditure also grew very rapidly, over 40 percent in real terms. As a percentage of GDP, it increased from 6.7 percent in 2005 to 6.9 percent in 2008. In 2008, 11 Social Assistance programs are aimed at lifting households out of poverty, are usually targeted to the poor, and are not linked to previous contributions to an insurance pool. Social Insurance programs, on the other hand, linked to previous contributions, are designed to mitigate the impact of unexpected income shortfalls due to unemployment, health problems, disability and old age. 48 spending on social assistance reached 2.6 percent of GDP and social security (pension payments by the CSS) reached 4.3 percent. 4.12. Although not exempt from difficulties, international comparisons of spending on social sectors in general and on social protection in particular provide a first approximation to the relative importance that countries attach to these sectors.12 Panama’s total spending in social protection (i.e., SP=SI+SA) is relatively high when compared to other countries in Latin America, and even when compared to the United States. The country spends 6.7 percent of GDP on social protection, with 5.0 percent spent on SI and 1.7 percent on SA. The average in Latin America is 5.9 percent of GDP for total SP, 4.7 percent for SI, and 1.1 percent for SA (see Table 4.1). The United States spends 8.3 percent of GDP in total, but has a much larger elderly population (12 percent aged 65 or above) that absorb a greater share of resources per capita than the younger population in Panama where only seven percent of its inhabitants are elderly citizens. 4.13. More impressive perhaps is the 2.6 percent of GDP that Panama spends on social assistance. This is 136 percent higher than the Latin American average, and is substantially higher than countries like Mexico, Chile and Costa Rica, known for large and effective social programs. It is even higher than the level of social assistance spending in Continental Europe, according to one estimate. Table 4.1: International Comparison of Social Spending (percent of GDP) Year Social Protection Education Health Total Country Social Social Total Insurance Assistance Social Protection Panama a/ 2005 5.0 1.7 6.7 3.8 5.2 15.7 Panama a/ 2008 4.3 2.6 6.9 4.0 5.3 16.2 Argentina 2004 7.7 1.5 9.2 3.8 4.3 17.3 Chile 2003 6.9 0.7 7.6 4.1 2.9 14.6 Costa Rica 2004 4.3 1.5 5.8 4.9 5.1 15.8 México 2004 2.6 1.0 3.6 4.1 2.1 9.8 Venezuela 2000 2.1 1.0 3.1 4.9 1.5 9.5 LAC Average b/ N/A 4.7 1.1 5.9 4.4 3.2 13.4 USA 2001 7.9 0.4 8.3 4.8 6.2 19.3 Continental 2001 14.8 1.5 16.3 6.9 6.4 29.6 Europe Source: World Bank reports, OECD, and staff estimates for Panama. a/ Education and health spending is adjusted to eliminate double counting with SA. b/ The average of the five LAC countries listed above. 4.14. Given the relatively large amounts spent on social assistance in Panama, it is remarkable that poverty, and especially extreme poverty and malnutrition remain at high levels. This is a clear indication that social protection spending in Panama is ineffective. Either programs are not being well targeted to the most in need, or, when well targeted, they are not effective in the sense that they do not generate the expected impacts on beneficiary outcomes. (Although new, the Red de Oportunidades program already shows signs that it is very well targeted and efficient, as discussed below.) 12 For a discussion of some of these difficulties see Marques, José Silvério “Central America, Cross-Country Evaluation of Social Safety Net Assessments (SSNAs)- Issues Paper�, paper prepared for the World Bank, November 2002. 49 4.15. Social Assistance spending is spread throughout the population, with nearly three-fifths of SA spending untargeted (i.e., spent on the general population). Of the remaining 40 percent, most goes towards children and adolescents: Table 4.2: Distribution of Social Assistance Spending, 2008 Age Group Annual Cost (B/ 000) % of Social Assistance % of GDP 0-5 51,154 8.5 .2 6-17 157,882 26.2 .7 18-61 21,828 3.6 .1 62+ 5,000 0.8 0 General Population 365,664 60.8 1.6 Total 601,528 100 2.6 CONDITIONAL CASH TRANSFER: THE RED DE OPORTUNIDADES PROGRAM 4.16. As argued in the previous section, Panama stands to gain substantially in terms of poverty and inequality reduction from improving the effectiveness of its social expenditures, especially its social assistance spending. In this section we analyze the Red de Oportunidades program or RdO. The RdO is a conditional cash transfer program that is being targeted at the extreme poor following the molds of Oportunidades in Mexico and Bolsa Familia in Brazil. 4.17. Conditional Cash Transfer (CCT) programs have become pervasive in Latin American and the Caribbean. They currently reach approximately 60 million people representing approximately 60 percent of the extremely poor in LAC (Lindert, Skoufias and Shapiro, 2005). In Mexico and Brazil alone, Oportunidades and Bolsa Familia take approximately 0.35 percent of these nations’ GDP. Empirically solid impact evaluations have demonstrated that these programs are cost effective in terms of reducing poverty and malnutrition and in increasing human capital accumulation by the poor (see Box 4.1). CCT programs originated as substitutes for untargeted subsidies for food, cooking gas, water and electricity, which were phased out in most adopting countries as a result of economic reforms. They have shown to be considerably more progressive and effective in reducing poverty and inequality than non-targeted subsidies (World Bank, 2006). 4.18. Red de Oportunidades was established with the goal of providing benefits to all extremely poor households living in all 621 jurisdictions of the country. By April 2009, the RdO had reached 75,157 households in all 621 jurisdictions – a total of over 412,000 people including 208,000 children and 6,000 handicapped. Two-thirds of the jurisdictions with RdO are rural, 13% are urban and 9% are indigenous. The most recent data (from 2008) show that the program is well-targeted, but is still not reaching all the people it is intended to reach (see table 4.4, below). Since 2008, the program has expanded, so it is likely that at least some of the people in the 2008 survey who were not beneficiaries at that time have since become so. 4.19. Transfers are provided to mothers of children under 18, provided each beneficiary family complies with one or two co-responsibilities in education and health (discussed below). The unit receiving the transfer is the household in urban and non indigenous rural areas, and the mother-children unit in indigenous comarcas. Families are eligible to stay in the program during a five-year period so long as they comply with their co-responsibilities. Accordingly, the program is expected to benefit about 20 percent of the Panamanian population including more than 70 percent of the indigenous communities. 50 4.20. Mothers are the recipients of the transfer because targeting them is believed to be the most effective way to improve children’s wellbeing. International evidence on intra-household allocation and on the use of CCTs in other countries in the region has repeatedly shown that increases in women’s income translate to more expenses for food, children’s clothing, education supplies and other children’s goods (shoes, medicines, etc.). In addition, female beneficiaries have reported increases in self-esteem and improved relationships – with less incidence of domestic violence – with their partners or other decision-makers (in-laws) as they are more able to contribute to the household’s economy (Mexico, Brazil). In the case of education, mothers feel more motivated to interact with teachers and get involved with children’s homework. These findings hold in a variety of settings – urban/rural, across income levels and cultures – as shown by the evaluations of Oportunidades (Mexico), Familias en Acción (Colombia), Red de Protección Social (Nicaragua). 4.21. The total amount spent on the RdO was approximately B/43.5 million in 2008 and B/55 million in 2009. The amount of the RdO transfer per beneficiary is fairly generous by international standards. The B/50 per month represents approximately 17.6% of the average monthly consumption among the poorest 20% of the country. By comparison, the Bolsa Familia program in Brazil gives about 5% (2004), the Jamaican PATH program 8% (2004), Colombia’s Familias en Acción 17% and Mexico’s Oportunidades program 22%. Table 4.3: Amount of Conditional Money Transfers, by Year 2005 2006 2007 2008 2009 (est) Districts 15 81 475 591 621 Beneficiaries (Households) 3,900 20,519 48,589 70,599 76,590 Beneficiaries (Individuals) N/A N/A 283,758 386,717 N/A Expenditure (millions B/) 0.6 17.2 28.4 43.5 55.0 Note: includes Red de Oportunidades and Bonos para Alimentos Source: World Bank staff calculations with assistance from José Silvério; World Bank reports. 4.22. To decide how to roll out the program, the administrators of the program ranked jurisdictions geographically according to the vulnerability index, constructed by the Ministry of Economy and Finance using a poverty map and recently updated on the basis of the 2003 Living Standards Measurement Survey. In urban and rural non-indigenous jurisdictions the program applies a proxy means test to identify potential beneficiary households. This proxy means test was specifically designed for use by the Red de Oportunidades and has been carefully calibrated to reduce both inclusion and exclusion errors. In indigenous jurisdictions, the program is intended to cover households with children younger than 18 so long as the poverty rate is at least 95 percent – which (unfortunately) applies to most indigenous communities. 4.23. The Government has adapted international experiences throughout Latin America to design the basic features of the CCT that support the RdO program. The first such feature is that the RdO is designed to create incentives for beneficiary families to invest in the human capital of their children, and thus to improve their education and health status. The program disburses cash transfers to households conditional on fulfilling at least one co-responsibility: (i) education: children older than four and younger than 17 must attend school at least 85 percent of school days; and/or (ii) health: pregnant women and children under five must regularly visit service providers according to the national health protocol. In particular, the health co-responsibility requires the family to visit health providers at least a certain number of times per year, have updated vaccinations, receive regular growth monitoring, and undertake other such activities. A household may not receive two transfers: if a household has both 51 school-aged children and either a pregnant woman or children under five, it will receive only one transfer (i.e., the same transfer as a household which has only school-aged children). 4.24. Second, the motivation behind the program concerns not only building social capital, but also increasing the monthly incomes of very poor households. Families will receive the transfer during five years provided that they comply with their co-responsibilities. The transfer amount was calibrated to be high enough to motivate the behavior changes necessary to comply with co-responsibilities and low enough to avoid creating distortions regarding adult labor force participation decisions. The transfer itself, apart from creating incentives to invest in human capital, provides significant additional income to increase household consumption, mitigate effects of low income and to reduce the poverty gap. 4.25. Operationally speaking, mothers receive the benefits through a combined mechanism – the commercial banking system when available and the public telecommunication agency (COTEL) in remote areas. The program also relies on a comprehensive monitoring and evaluation system to follow program progress, verify completion of co-responsibilities, reconcile payments and quickly detect process problems. As part of this monitoring, co-responsibilities are verified on a by-monthly basis by MIDES in coordination with the education and health system. 4.26. It is clear from international experience that families need additional support to take full advantage of CCT programs. Complying with a conditional cash transfer program may imply major behavioral change among the poor, especially among those who have not traditionally had access to services such as indigenous people and the extreme poor in remote rural areas. Families may not be used to accessing services on a regular basis and receiving a certification for doing so, nor to entering a Bank/Post office to collect their transfer. Similarly, education and health providers and bank teller staff may be little prepared to deal with a new diverse range of clients. Recent experiences have shown that dealing with the complexities very poor families face when participating in CCT programs constitutes a key element for program success. In Panama, this may be a paradigmatic change, because the Red de Oportunidades program entails a completely different relationship between the public sector and most beneficiary households living in rural and indigenous jurisdictions, which will require that education and health staff in participating areas provide services in an appropriate way. 4.27. Several other countries’ CCT programs have implemented complementary activities to encourage beneficiary families’ participation in the program and to increase their demand for health and education services so that they obtain the maximum benefits they are entitled to under the program. Some programs train individual beneficiary mothers, such as the promotoras in Mexico’s Oportunidades, or the madres-líderes in Colombia’s Familias en Acción. Some programs, such as El Salvador’s Red Solidaria, also involve NGOs specializing in community development, which work in close coordination with local service providers. Programa Puente in Chile uses professional social workers who work with the families and follow their progress in implementing a family development plan and fulfilling a related individual menu of co-responsibilities over the two years of the program. Panama’s Red de Oportunidades program is aware that the program needs specific strategies to ensure that the large proportion of indigenous beneficiaries can fully benefit from the program. 4.28. For this reason, the Government of Panama designed a strategy to support beneficiary families. This support consists of a package of training and assistance services to eligible families and communities, as well as support to program operations at the local level. Specifically, the support aims at (i) helping beneficiary families who participate in the program (a) understand the importance of complying with their commitments and (b) obtain the maximum benefit from the conditional cash transfers; (ii) supporting targeted communities in creating or strengthening local organizations to participate in the program’s social audit and control; (iii) disseminating information about culturally adequate social services; (iv) assisting beneficiary households in accessing other available 52 complementary social services and programs which promote increased skills and income generation; and (v) supporting MIDES’ local and provincial operations. This support is outsourced to specialized non public organizations. 4.29. The RdO also aims at strengthening interventions aimed at promoting child growth and securing access to health services for program beneficiaries. The GoP has and will continue to strengthen and expand an existing package of preventive health and nutritional services to close coverage gaps in indigenous areas. These improved services would better meet the needs of the increased demand generated by RdO among poor and indigenous households and thus would help them build human capital. This component delivers interventions to prevent chronic malnutrition through mechanisms and organizational arrangements already established by the existing basic health outreach program, the Paquete de Atención Integral de Servicios de Salud (PAISS). 4.30. Moreover, because the current supply of social services in Panama is far from adequate, especially in indigenous and rural areas, the GoP is in the process of working in a number of areas, in particular to (i) strengthen capacity for improved policy and decision making, coordination, and monitoring in the social sectors; (ii) enhance knowledge generation for informed policy making; and (iii) strengthen the capacity of MIDES to formulate social assistance strategies and policies, and design, implement, monitor and evaluate social assistance programs. In the context of this broader reform, the GoP is in the process of reorganizing its social transfer programs to increase public funds for priority social services and assistance programs. 4.31. A future pillar of the RdO will focus on enhancing skills and capacity building of beneficiary families, as a way to ease their graduation once cash transfers conclude. This pillar is aimed at improving families’ economic self-sustainability and avoiding possible negative effects resulting from the expiration of the benefit (e.g., dropping-out of school or stopping visits to health facilities). For this purpose, the available public agencies in charge of adult education, skills development and individual capacity building (i.e., INADEH, Ministerio de Trabajo, and IFARHU and extension services of MIDA and other special departments of MEDUCA) will increase their efforts and supply of services in the RdO-targeted jurisdictions. As a result, families living in those areas will face increased opportunities to enhance their knowledge and acquire tools to raise their income generating capacity. 4.32. In this section we evaluate RdO’s targeting performance and its impact on consumption per capita in indigenous areas. Note that the data used in the analysis below are from a 2008 survey, before the RdO had been fully rolled out. As a result, this analysis includes households not benefiting from the program who may have become beneficiaries since the time of the survey. It is also worth noting that the RdO has its own Monitoring and Evaluation system, which is unusual for Panamanian social programs – but clearly a welcome development. 53 Box 4.1: Conditional Cash Transfers Over the past decade, numerous countries in LAC have introduced “conditional cash transfers� (CCTs), which have the dual objectives of (a) reducing current poverty and inequality through the provision of cash transfers to poor families (redistributive effect); and (b) reducing the inter-generational transmission of poverty by conditioning these transfers on beneficiary compliance with key human capital investments (structural effect). Initiated in Brazil at the municipal level in the mid-1990s, Mexico developed the first large-scale CCT program, originally called Progresa, now Oportunidades. Brazil then expanded its municipal programs to the national level, first as Bolsa Escola, which focused on school attendance, then with Bolsa Alimentaçao, which introduced health-related conditionalities. In 2003, these programs were merged with two others to form the Bolsa Família Program, which integrated these transfers, as well as the health and education conditionalities for greater synergies. CCTs have spread to other countries in LAC, including: Argentina, Chile, Colombia, Dominican Republic, Ecuador, Honduras, and Jamaica. 13 Interest extends beyond the region, with similar schemes emerging in countries such as Turkey, the West Bank and Gaza, Pakistan, Bangladesh, Cambodia, Burkina Faso, Ethiopia, and Lesotho. Eligibility rules vary, but most programs seek to channel CCT benefits to poor families, with significant efforts to develop strong targeting mechanisms, usually combining geographic targeting with some sort of household assessment mechanisms, such as proxy means testing (using multi-dimensional indicators that are correlated with poverty as a way to screen for eligibility). Conditionalities vary, but usually include minimum daily school attendance, vaccines, prenatal care, and growth monitoring of young children. Mexico’s Oportunidades has also added bonuses for school graduation and participation in health- awareness seminars. The programs range in size. Brazil’s Bolsa Familia is now the largest, covering 8 million families (32 million people, or close to a fifth of its population), followed by Mexico’s Oportunidades (5 million families). Others are smaller, such as Chile’s Solidario Program, which covers over 200,000 families, and Colombia’s Familias en Acción program, which covers about 400,000 families. All are fairly lean, in terms of resource use. CCTs in both Mexico and Brazil represent about 0.37% of GDP. With higher unit transfers, Argentina’s Jefes claims a slightly larger share of GDP (0.85%), though still less than one percent. Programs in Chile (0.08% of GDP) and Colombia (0.1%) claim an even smaller share. As discussed below, administrative costs of these programs are fairly low, averaging about 5% of total program outlays (for mature programs; start-up costs are higher), as compared, say, with an average of 36% for food-based programs. Despite their relative economies, CCTs are showing impressive impacts. This paper demonstrates that, as a class of programs, CCTs are by far the best targeted to the poor (vis-a-vis: all other social assistance programs, utilities subsidies, social insurance, and public spending on health and education). With the majority of CCT benefits actually reaching the poor (no small feat in LAC), their redistributive impacts are muted only by the relatively small size of the unit transfers in most countries, which dampens their potential impact on current poverty. Moreover, their structural impact on breaking the inter-generational transmission of poverty is impressive. Experimental and quasi-experimental evaluations suggest important impacts, well beyond the redistributive impacts discussed in this paper:14  On health and nutrition: (a) increased total and food expenditures (Brazil BA, Mexico, Honduras, Nicaragua); (b) increased calorie intake and improved dietary diversity (Brazil BA, Mexico, Nicaragua); (c) improved child growth (Mexico, Nicaragua, Brazil BA); (d) increase in use of prenatal care and reduced maternal mortality (Mexico); (e) reduced incidence of smoking and alcohol consumption (Mexico); and (f) improved treatment of diabetes (Mexico).  On education: (a) improved primary enrolment among the poor who were not previously enrolled (Nicaragua, Honduras, Brazil); (b) increased secondary enrolment (Mexico, Colombia); (c) reduced drop-out rates and repetition (Mexico, Nicaragua, Honduras); and (d) reduced child labor (Mexico-boys, Nicaragua, Honduras-boys, Colombia, Brazil). Source: World Bank (2006) 13 Argentina’s Jefes de Hogares program is a bit different in that the “conditionalities� involve work-related and labor training actions on behalf of beneficiaries rather than school attendance and health care. Argentina also operates a smaller CCT, called the Income for Human Development Program (IDH), which conditions cash transfers on schooling and health care. 14 See: Maluccio (2004), Olinto (2004), Rawlings and Rubio (2004) and Rawlings (2004) for summaries of the impacts of CCTs. 54 Assessing Red de Oportunidades’ Targeting 4.33. The first step in designing a CCT program is to define its target population. In the case of Panama, the government has decided to target all families living under the annual extreme poverty line of B.\640 per capita consumption. Therefore, 14 percent of the population should be targeted to receive RdO. CCT coverage remained low in 2008 (the most recent year of data) as the program was still in its infancy. According to these data, the program is still a far way from reaching all those who need it: 72% of those in extreme poverty are not part of the program, and 93% of moderate poor are not (Table 4.4). However, more recent numbers show a marked improvement (Table 4.5). Table 4.4: Coverage of CCT Program (based on 2008 data) Receive CCT Do NOT receive CCT # % # % Extreme poor 135,654 28.2% 345,598 71.8 Moderate (not extreme) Poor 43,045 7.1% 566,211 92.9 Poor (moderate + extreme) 178,699 16.4% 911,809 83.6 Not poor (moderate or extreme) 6,325 0.3% 2,237,655 99.7 Total, Panama 185,024 5.5% 3,149,464 94.5% Source: authors’ calculations based on 2008 LSMS data. 4.34. More recent data show a better targeting, as the following table demonstrates: 55 Table 4.5: Coverage of CCT Program by Region (2008 data): % of extreme poor households that are beneficiaries Province/ Comarca Beneficiaries % of Households in Beneficiaries % of extreme Expenditure % of (individuals) individual extreme (households) poor (B/ 000) Expendit beneficiaries poverty households ure that are beneficiaries Provinces Bocas de Toro 27,440 7.1 6,077 4,086 67.2 3371 7.7 Coclé 47,548 12.3 9,128 8,274 90.6 5935 13.6 Colon 14,404 3.7 3,665 2,564 70 947 2.2 Chiriquí 38,554 10 7,422 6,760 91.1 3,608 8.3 Darién 13,288 3.4 3,364 2,830 84.1 1,993 4.6 Herrera 9,338 2.4 930 1,810 194.6 916 2.1 Los Santos 3,516 0.9 1,361 892 65.5 508 1.2 Panamá 42,885 11.1 10,956 7343 67 2,430 5.6 Veraguas 53,812 13.9 8,130 11,104 136.6 5,549 12.7 Province Sub-Total 250,785 64.8 51,033 45,663 89.5 25,257 58 Comarcas Kuna Yala 24,479 6.3 1635 3979 243.4 3327 7.6 Emberá 7,443 1.9 6381 1506 23.6 1118 2.6 Nägbe Buglé 104,011 26.9 17541 19451 110.9 13,828 31.8 Comarca Sub-Total 135,933 35.2 25,557 24,936 97.6 18,273 42 National Total 386,717 100 76,590 70,599 92.2 43,530 100 Source: MIDES; authors’ calculations with help from José Silvério . 56 4.35. Coverage rates are not uniform by region. Because the CCT program targets the poorest, it is directed largely at indigenous and rural communities since they have much higher poverty rates. As a result, the proportion of the population receiving CCT transfers is higher in indigenous and rural areas. Furthermore, a higher fraction of CCT beneficiaries hail from indigenous and rural communities, despite urban communities having a much larger general population. 4.36. Though the coverage of the RdO appears small (in 2008), the 2008 data demonstrate that the transfers are extremely well targeted to the neediest. Nearly three-quarters of the program’s beneficiaries are the targeted audience, the extreme poor. Of the rest, almost all are moderately poor. Only about 3% of the funds go to the non-poor. While it would be ideal for all the funds to go to the extreme poor, the Panama program does very well by international comparisons in preventing leakage – and in leaking almost entirely to the moderate poor. In short, the high incidence shows that the funds used for the CCT are extremely well-targeted Table 4.6: Coverage of CCT Program by Region (2008 data) Poverty Rates Receive CCT Extreme Moderate + Extreme % # Urban 3.2% 17.7% 0.4% 8,082 Rural 22.2% 50.7% 9.0% 85,490 Indigenous 84.8% 96.3% 38.6% 91,452 Total, Panama 14.4% 32.7% 5.6% 185,024 Source: authors’ calculations based on 2008 LSMS data Figure 4.2: Benefit Incidence of CCT Programs Source: authors’ calculations based on World Bank reports 57 The Impact of Red de Oportunidades 4.37. Evidence from 2008 suggests that the program has made an impact, particularly among the indigenous households: those receiving the CCT have experienced an estimated 8-9% increase in per capita income (see table 4.7, which shows the “Average Treatment Effect� of RdO using a number of different econometric estimations). During the 2003-2008 period, extreme poverty rates among indigenous decreased from 90% to 85%. Approximately one-quarter of this drop is attributable to the CCT program. This period also witnessed a decrease in the poverty gap among the indigenous population (the poverty gap measures the average amount of additional income a person below the poverty line would need to reach the poverty line). In short, the CCT program has been extremely effective at raising the incomes of the indigenous population, who make up the largest share of the extremely poor in Panama. Table 4.7: Propensity Score Matching Estimation of the Impact of Red de Oportunidades on Log- consumption per capita Average Treatment Effect # # Regression method (SE) t-score Treatment Control kernel-based 0.091 3.281 1197 1691 matching (0.028) kernel-based 0.091 3.405 1197 1691 matching (logit) (0.027) nearest neighbor 0.077 2.792 1204 1834 matching (0.028) nearest neighbor 0.077 2.758 1204 1834 matching (logit) (0.028) 0.077 Radius matching 2.768 1204 1834 (0.028) Stratification 0.082 2.861 1174 1714 matching (0.029) All regressions were run using a Probit model unless otherwise specified. For the bootstrapping, all estimates were replicated 100 times, using 3,038 observations SIMULATING THE TARGETING AND INCIDENCE OF PANAMA’S NEW TRANSFER PROGRAMS: BONO 100 A LOS 70 AND BECA UNIVERSAL 4.38. The Government of Panama has continued to expand its safety net system by introducing two new targeted social transfers programs, namely, the program Bono 100 a los 70 and Beca Universal. The former was launched in 2009. It transfers US$100 to seniors aged 70 and above who do not receive a formal pension. There is no targeting on income or any other social indicator besides age and not being a beneficiary of the pension system. 4.39. In 2010, Panama launched the Beca Universal program. The program rewards successful students in public school throughout the country with $20 a month. It started in secondary public schools in 2010. In 2011 it will expand to 4th to 6th grades, and moving to lower grades in primary school. MIDES and IFARHU, the two institutions in charge of running this program, did receive increased budget allocations for ensuring payments, but received only limited resources for the operation of the programs. 58 Targeting of Bono 100 a los 70 4.40. As the program is currently structured, transfers from Bono 100 a los 70 accrue roughly equally to all deciles of the consumption distribution (Figure 4.3). As a share of their pre-transfer per capita consumption however, the transfer is larger for lower deciles than for higher ones (Figure 4.4). For the lowest two deciles, the transfer corresponds to roughly 10 percent of consumption. For deciles three and four the program accounts for about 5 percent more consumption. For higher deciles, the program’s contribution to consumption is minimal. Figure 4.1: Transfers from 100 para los 70 by Figure 4.2 4.4: Share of transfers in total consumption decile consumption by decile Size of transfer from 100 para los 70 100 50 90 80 Balboas per capita 40 70 60 Balboas per capita 50 30 40 30 20 20 10 10 0 1 2 3 4 5 6 7 8 9 10 Consumption per capita 0 Transfer from 100 para los 70 1 2 3 4 5 6 7 8 9 10 4.41. Targeting Bono 100 a los 70 could make the incidence of the program much more progressive. In figure 4.5, we simulate the effects of imposing eligibility restrictions on the benefit incidence of the program. To carry out this exercise we use the predicted probability of being extremely poor from the same Proxy Means Test (PMT) employed in the Red de Oportunidades program. The first exercise (dashed line) is done by simulating the exclusion of those with a predicted probability of being extremely poor smaller than 10 percent. The second exercise (solid black line) is done by excluding those with the probability of being extremely poor greater than 20 percent. The dotted line shows the baseline scenario, that is, the scenario without any PMT targeting restriction. 4.42. The benefit incidence in the baseline case reflects the underlying distribution of the population of 70 year olds in the country. 70 year olds are less likely to be represented in the bottom deciles of the distribution. In fact it appears that there are roughly one-third fewer such individuals in the bottom decile than in any other deciles. Thus the budget incidence is lowest in the bottom deciles and roughly even across the remaining deciles. As the figure demonstrates, without targeting, only 15 percent of the benefits accrue to the bottom two deciles of the income distribution. 4.43. When proxy-means targeting criteria are imposed, the benefit incidence of the program becomes considerably more progressive. Excluding those with a probability of being extremely poor of less than 10 percent results in 50 percent of the program’s benefits going to the bottom two deciles of the income distribution. Restricting eligibility even further (that is, requiring that beneficiaries exhibit a predicted probability of being extremely poor of 20 percent or more) would result in more than 70 percent of the benefits accruing to the bottom two deciles. Therefore, by simply adopting the same targeting 59 methodology already being applied in the RdO program, the Bono 100 a los 70 is likely to become considerably more cost effective in reducing poverty at old age. Figure 4.5: Budget incidence of 100 para los 70 Budget Incidence of 100 para los 70 using PMT targetting 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 9 10 0 0.1 0.2 4.44. The Beca Universal program appears to be progressive when all four phases are taken into account, even though some phases (especially the final phase, in which private schools are included) are regressive. Over the course of the four years from 2010 to 2013, consumption for the bottom decile will increase by 20 percent and for the second and third deciles by roughly 10 percent provided that all children meet the Beca’s performance requirement (Figure 4.6). The transfers will account for roughly 5 percent of the gains to the consumption of the fourth decile. For deciles 5 and above, the Beca Universal program will constitute a small share of total consumption. 60 Figure 4.6: Beca Universal as a share of consumption per capita Share of Beca Universal transfers in consumption per capita 100 90 80 Percentage 70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 Consumption before Beca Universal Beca Universal in 2010 Beca Universal in 2011 Beca Universal in 2012 Beca Universal in 2013 Source: Staff calculations using the Panama 2008 LSMS 4.45. However, not all of the phases are equally progressive. In particular, the bottom four deciles will benefit from the transfers from 2010 to 2012, but will receive little benefit from the 2013 (private school) phase. constitute virtually the entirety of the benefits they will receive from the program. For instance, for the bottom decile, the 2010 roll-out presents 20 percent of the total transfers from the program, the 2011 roll-out represents 30 percent, and 2012 the remaining 50 percent. and In contrast, the transfers from the 2013 phase are entirely concentrated in the top half of the consumption distribution. While the transfers constitute a small fraction of this group’s consumption (Figure 4.6) it is the largest component of the Beca Universal program that they receive (Figure 4.7). Figure 4.7: Each phase of transfers as a share of the overall Beca Universal program Share of transfers per capita from the phase in of Beca Universal 100 90 80 Percentage 70 60 50 40 30 20 10 0 1 2 3 4 5 6 7 8 9 10 Beca Universal in 2010 Beca Universal in 2011 Beca Universal in 2012 Beca Universal in 2013 Source: Staff calculations using the Panama 2008 LSMS 61 4.46. The benefit incidence graph (Figure 4.8) reveals a similar story . The 2013 phase of Beca Universal is concentrated largely in the top half of the distribution. The phases from 2010 to 2012 are more evenly distributed across the first seven deciles of the consumption distribution. It is unclear how the budget incidence would change if the performance standard required to receive the Beca was also incorporated into these estimates. The current figures are premised on the assumption that all children meet this performance standard, which is unlikely to be the case. Furthermore, it is quite possible that a higher rate of children in lower-income households will fail the performance standard, making the Beca Universal less progressive. Figure 4.8: Each phase of transfers as a share of the overall Beca Universal program Budget Incidence of Beca Universal Phases 1-3 compared to Phase 4 30% 25% 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 9 10 Public School Students (2010-2012) Private School students (2013) 62 CONCLUSIONS AND POLICY IMPLICATIONS Panama has historically spent a substantial amount of resources on the social sectors in general, and on social protection in particular, but in the past, the results have not been commensurate with this spending. However, early signs indicate that the fledgling Red de Oportunidades (RdO) and the recently launched Beca Universal and Bono 100 a los 70 programs may be different. Panama has a large program of indirect subsidies, which accounts for almost two-thirds of spending in Social Assistance, but these subsidies mostly benefit the non-poor. Poor infants and mothers and poor seniors receive a much smaller share of the SA resources. Consequently, there is a need to develop a clear Social Protection strategy with specific targets which should drive the process of resource allocation in the sector. A more comprehensive and in-depth review of existing programs should be undertaken, cost ineffective practices eliminated, and available resources targeted to the most effective programs. On the positive side, Red de Oportunidades is extremely well targeted and is likely behind the drop in extreme poverty among the indigenous population. Of the beneficiary families in the survey sample, three quarters are extremely poor, and nearly all the rest are moderately poor. Less than 4% of the funds are leaking to the non-poor. This compares very favorably to similar programs in the region. Moreover, almost half of the beneficiaries of Red de Oportunidades are indigenous, the group that contributes the most to extreme poverty. For this group, the program was reaching almost 39 percent of the targeted population according to a 2008 survey. While half of the CCT recipients are indigenous, only 5 percent of the transfers go to households in urban areas. The remaining 45 percent goes to rural dwellers, the second largest contributors to extreme poverty in Panama. However, by 2008, the program was covering only about 28 percent of its intended population of the estimated 100,000 extremely poor families in the country. Continued expansion of RdO is likely to be effective in reducing extreme poverty and mal nutrition. The Government of Panama renewed its commitment to address the key risks affecting vulnerable seniors by increasing the coverage of non-contributory pensions via the Bono 100 a los 70 program. However, as indicated above, this program is likely to be regressive. Employing the same proxy-means targeting methodology currently being used in RdO could considerably improve targeting and the progressivity of the program. 63 5.1. While Panama is one of the fastest growing economies in the Latin America and Caribbean (LAC) region, translating Panama’s growth into more rapid poverty reduction remains a challenge. The impact of Panama’s stellar growth on poverty is still timid when compared to the rest of LAC. The region grew at a much slower pace, but was able to reduce poverty and extreme poverty in the last decade at a faster pace. 5.2. Without addressing the problem of severe destitution in the indigenous areas, it will be difficult for Panama to translate its stellar growth into rapid poverty reduction. While contributing to only nine percent of the country’s population, the indigenous represent 47 percent of the extreme poor. In addition, they contribute to more than 65 percent of the national extreme poverty gap. The average per capita consumption in indigenous areas amounts to less than half the national poverty line. That is, the average indigenous household is not able to afford half of the daily caloric needs for healthy living. Not surprising, chronic malnutrition in indigenous areas has increased by almost 27 percent between 1997 and 2008, going from 48.5 to 61.9 percent of children under five years of age. 5.3. Despite the country’s low elasticity of growth to poverty reduction, there was considerable mobility out of poverty between 2003 and 2008. For instance, of those who were extremely poor in 2003, at least 17 percent were able to climb out of destitution by 2008. However, because of the lack of a comprehensive safety net system, at least 0.7 percent of the non-destitute fell into extreme poverty during the 2003-08 period. Had the country had a more effective safety net system in place, upward mobility alone could have reduced extreme poverty to at least 14 percent. Our more optimistic upper bound estimates indicate that extreme poverty could have been reduced even further to 9 percent if safety net programs were in place to prevent downward mobility. 5.4. Increased urban-to-rural migration appears to be linked to the drop in rural poverty and the rise in urban poverty. As fast growth in the services and industry sectors between 2003 and 2008 have attracted the moderate poor from rural areas, poverty fell in these areas and increased in urban areas. The same effect was not observed for extreme poverty which fell in urban areas and stayed constant in rural areas. In fact, between 2003 and 2008 there was an increase in migration flows from urban to rural areas by the indigenous. Given that the indigenous tend to be extremely poor, this might partially explain the drop in extreme poverty in urban areas. It is also likely that this increased reverse migration was induced by the expansion of the CCT program Red de Oportunidades. 5.5. The prevalence of female headed households has increased by 30 percent between 1997 and 2008. As individuals living in female headed households tend to be poorer than otherwise, policy interventions targeted to women heads of households may need to be developed. For instance, access to publicly funded and/or provided child care may increase the welfare of those living in such households. 5.6. Expanding and enhancing the effectiveness of Panama’s social protection system can be accomplished well within the current fiscal space. In addition to high spending on the social sectors, the country has a large program of untargeted indirect subsidies. It spends approximately 1.2 percent of GDP on subsidies for water, electricity, fuel, and mortgage interest rates. Together these amount to almost US$280 million a year and have been shown to be vastly regressive. If this sum were to be transferred directly to those below the extreme poverty line, this would amount to a per capita transfer of approximately US$1.60 a day, which would completely eradicate extreme poverty, even if almost 40 percent of the transfers leaked to the non-destitute. Panama should build upon its own recent experience with successful safety net programs to enhance the effectiveness of its social spending. Its recent efforts to strengthening its social protection system by adopting direct transfers targeted to the poor have shown promising results. While extreme poverty for the indigenous remains extremely high, targeted conditional cash transfers distributed by the Red de Oportunidades program seem to be having a significant impact. Despite having reached less than a third of its targeted population in 2008, our 64 estimates indicate that the program had a positive impact on indigenous household consumption of approximately nine percent and is likely responsible for approximately 25 percent of the drop in extreme poverty in indigenous areas. Since the indigenous contribute the most to the extreme poverty head count and the extreme poverty gap, once the program is fully implemented, Panama is likely to show substantial progress in malnutrition and extreme poverty reduction in the future. 5.7. The effectiveness of the Red de Oportunidades program in poverty reduction is likely a result of its excellent targeting. While still in its infancy, the program had already reached 28 percent of the extreme poor population by 2008. Moreover, the large majority of the transfers go to the neediest. Almost 60 percent goes to those in the first decile of the income distribution, and almost 90 percent goes to the first two deciles. Its benefit incidence compares very favorably to similar programs in Mexico and Brazil. 5.8. While safety net programs have shown to effectively prevent downward mobility, fostering upward mobility will require enhancing access to quality education in indigenous and rural areas. Education is positively linked with upward mobility and negatively with downward mobility. Our conservative estimates indicate that 36-38 percent of those with either a complete secondary education or a primary degree moved out of (0.4 percent moved into) destitution. For those with no primary schooling, only one fifth moved out and 17 percent moved into extreme poverty. 5.9. Access to education has significantly increased in Panama, but more needs to be done to improve its quality. The heavy investments that Panama undertook to mitigate the disparities in access to education in the past appear to be bearing fruit. Children are entering the school system earlier and staying longer. Most importantly, all of this expansion seems to have happened in rural and in indigenous areas where lack of services had severely restricted pre-primary and secondary education opportunities for the poor. 5.10. While rural adults seem to be catching up with their urban peers in education attainment, the indigenous continue to lag far behind. For rural dwellers born in the 1980s, average years of schooling is now close to 65 percent of the urban average, up from less than 50 percent for those born in the 1930s. But the indigenous born in the 1980s accumulated only 32 percent of the number of years of schooling accumulated by urban adults. 5.11. Efforts to expand access to pre-school programs in the country are also noteworthy. The proportion of children less than six years of age enrolled in pre-school programs has increased substantially between 1997 and 2008. As in the case of secondary school, most of this increase in enrollment has occurred in rural and indigenous areas. As there is mounting evidence that early childhood education and pre-school programs have significant impacts on cognitive development and future student performance, this outcome is likely to positively affect human capital accumulation and poverty reduction in the long run. 5.12. Nevertheless, returns to further investments in education are likely to decrease in Panama if the country fails address issues related to both the supply and demand for education, and to attract investments in high productivity sectors. Returns to education have been decreasing over the past decade. They dropped by almost 16 percent between 1997 and 2008. In 1997, one additional year of schooling yielded in average 10.2 percent more income. These returns have steadily decreased to 9.5 percent in 2003 and 8.6 percent in 2008. 5.13. The drop in returns to education is likely associated to the increase in the share of workers employed in low productivity sectors. Most of the increase in employment between 1997 and 2008 occurred in the construction and service sectors, which are not known for being highly productive. More businesses that demand high productivity workers will be needed if returns to education are to increase in the future. Parallel to its efforts to expand access to secondary schooling, Panama needs to enact policies to attract businesses that demand highly skilled workers. 65 5.14. The expansion of secondary education in rural and indigenous areas may also be linked to the diminishing returns to schooling. As the country has strived to rapidly increase access to education to the poor in rural and indigenous areas, the average quality of education may have suffered. Between 1997 and 2008, secondary enrollment increased by 40 percent in rural areas, and 90 percent in indigenous areas. As a result, enrollment rates among the poor increased by more than 50 percent. The decline of the cognitive ability of students entering the school system may also be contributing to the drop in returns to education. As school access expands to the rural and indigenous areas, where the prevalence of chronic malnutrition is high, the performance of students may deteriorate. Chronic malnutrition, or stunting, is known to be highly associated with low cognitive development and school performance. 5.15. In addition to income transfers to the extreme poor, expanded access to adequate prenatal and infant care services will be needed to combat chronic malnutrition in indigenous areas. Preventive prenatal and infant health care are known to protect infants from stunting. Access to prenatal care has increased substantially in urban and rural areas, covering almost all pregnancies. However, in indigenous areas, which contribute the most to stunting, 40 percent of women go through their whole pregnancies without a single visit to a health center. Therefore, while improving the incomes of the indigenous is crucial for enhancing their diets, it may not be enough for tackling chronic malnutrition. Expanded access to preventive prenatal and infant care among the indigenous will also be needed. 5.16. Lack of access to adequate sanitation is also a potential cause of chronic malnutrition in poor areas. While broad access to education and clean water services puts Panama among the countries with highest equality of opportunity in Latin America, the country still lags behind in terms of access to electricity and sanitation. Access to sanitation is a particular area for concern since the HOI of 33 is only about half that of the region overall. The gap between the region and Panama in access to electricity is less stark but only two-thirds of the opportunities for equitable access are available and equitably distributed in Panama. On the positive side, the HOI for school enrollment is almost universal suggesting an area where a level playing field does in fact exist. Panama leads the regional average completing sixth grade on time. It also does well in terms of access to clean water. The Human Opportunity Index (HOI) for completing sixth grade on time and for access to water are both roughly 80. 5.17. In areas in which the country lags behind the region in terms of equality of opportunity, improvements have been mixed. While Panama lags behind the region in providing universal and equitable access to electricity, it is making significant progress to catch up. Its rate of increase of the HOI for access to electricity is roughly 1.3 percentage points per year. In contrast, given how low the HOI for access to sanitation is, the low rate of growth is a sign that the situation is unlikely to improve in the near future, unless significant investments are made in the sector. Therefore, to continue to improve equality of opportunity and returns to education, the country should continue to focus investments in access to electricity and sanitation. POLICY CONSIDERATIONS 5.18. Panama’s slow progress in reducing poverty and extreme poverty is not a consequence of a lack of public resources. The country has grown faster than most countries in Latin America in the last decade, and government revenues have increased substantially. Nevertheless, considerable scope under current spending is available for Panama’s public sector to become more effective in the fight against destitution, malnutrition and inequality of opportunity. Our analysis suggests that the following themes are key elements to successful poverty reduction strategies in the years to come:  Continue to improve the effectiveness of its social spending. Continuing to improve the targeting and effectiveness of spending in education, health and social protection ought to put the country 66 on a virtuous cycle in which social spending relative to GDP would persistently decrease, as growth driven by faster human capital accumulation accelerates, and fiscal requirements for poverty alleviation gradually decline. Example of policy measures toward this goal are: o Improve the efficiency of education spending and the quality of education services. o Improve the efficiency of health spending. o Reorient social assistance spending away from costly untargeted subsidies on electricity, water, cooking gas, and gasoline, and towards programs targeted to the poor like the Red de Oportunidades which has shown to be effective, especially in indigenous communities. In addition, the government should considered employing more effective targeting tools as the ones used in RdO in its new transfer programs Bono 100 a los 70 and Beca Universal. o Strengthening monitoring and evaluation systems in all government institutions in charge of social programs, to facilitate a transparent and easy to monitor use of public resources, and to ensure that the benefits of social programs are received by the targeted groups and have the desired impacts.  Continue to reduce inequality of opportunity, especially between the indigenous and non- indigenous population. Strengthening human capital accumulation and improving the returns to education, with special emphasis on improving the educational, health and nutritional status of the indigenous, should be a key part of enhancing the effectiveness of Panama’s development strategy. Key policy options include: o Expanding access to preventive health services in rural and indigenous areas, with especial attention to enhancing pre-natal and infant care, immunization, nutritional monitoring and mother’s education. o Continuing to expand the supply of pre-primary, primary and secondary education in rural and indigenous areas, while insuring relevance and quality of teaching via community based monitoring and teacher’s incentives. o Continue to expand Red de Oportunidades which delivers targeted conditional cash transfers in order to alleviate liquidity constraints and provide incentives for the poor to attend school and periodically visit basic health service providers, especially in rural and indigenous areas. o Expand access to sanitation in indigenous areas to help reduce the high rates of chronic malnutrition in these areas. 67 REFERENCES Albornoz, Facundo and Marta Menéndez, 2002. “Analyzing Income Mobility and Inequality: The Case of Argentina during the 1990’s� Datt, Gaurav, and Martin Ravallion, 1991. “Growth and Redistribution Components of Changes in Poverty Measures: A Decomposition with Applications to Brazil and India in the 1980s.� Living Standards Measurement Study Working Paper No. 83, World Bank, Washington, DC. Khor, Niny and John Pencavel, 2010. “Evolution of Income Mobility in the People’s Republic of China: 1991–2002� Manila, Philippines: ADB. Lanjouw, Peter, Jill Louto, David McKenzie, 2009. “Using Repeated Cross-Sections to Explore Movements in and out of poverty� Working Paper. Washington, DC: World Bank. Luttmer, Erzo, 2002. “Measuring economic mobility and inequality: Disentangling real events form noisy data� Washington, DC: World Bank, mimeo. Ministerio de Desarrollo Social 2009. “Evaluación de Programas de Asistencia Social y Propuesta de Reordenamiento y Racionalización�.Informe Final. Programa de Protección Social. Red de Oportunidades. Nuñez, Javier I., and Miranda Leslie, 2010. “Intergenerational income mobility in a less- developed, high-inequality context: The case of Chile� The B.E. Journal of Economic Analysis & Policy, 2010: 10(1). Ravallion, Martin, 2007. “Inequality Is Bad for the Poor� In Inequality and Poverty Re- Examined, ed. Stephen Jenkins and John Micklewright. Oxford, UK: Oxford University Press. Ravallion, Martin & Chen, Shaohua, 2003. “ Measuring pro-poor growth� Economics Letters , Elsevier, vol. 78(1), pages 93-99, January. World Development Indicators. 2009. Wasington, D.C.: World Bank Group 68 ANNEX 1.1 POVERTY STATISTICS BY AREAS Table A.1.1.1: Overall Poverty Poverty Headcount Rate Poverty Gap Squared Poverty Gap (P0) (P1) (P2) 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Urban 15.3 20.0 17.7 -2.3 3.9 5.6 4.7 -0.9 1.5 2.3 1.8 -0.5 Rural 58.7 54.0 50.7 -3.3 25.2 20.6 19.7 -0.9 14.0 10.5 10.1 -0.4 Indigenous 95.4 98.4 96.3 -2.1 66.2 68.8 67.0 -1.8 49.2 51.1 50.5 -0.6 Total 37.3 36.8 32.7 -4.1 16.4 15.2 13.4 -1.8 9.7 8.7 7.6 -1.0 Extreme poverty line at 2008 prices Urban 3.1 4.4 3.2 -1.2 0.7 0.9 0.6 -0.3 0.2 0.3 0.2 -0.1 Rural 28.7 22.0 22.2 0.1 10.2 6.6 6.4 -0.2 5.0 2.8 2.6 -0.1 Indigenous 86.3 90.0 84.8 -5.2 47.0 47.9 47.7 -0.2 29.7 29.6 31.4 1.9 Total 18.8 16.6 14.4 -2.2 7.7 6.4 5.6 -0.7 4.2 3.4 3.1 -0.3 Note: Changes shown between years 2003 and 2008 Table A.1.1.2: Distribution of the poor Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Urban 22.8 32.9 34.7 1.8 55.6 60.5 64.3 3.7 Rural 58.0 46.5 44.4 -2.1 36.9 31.7 28.6 -3.1 Indigenous 19.2 20.6 20.9 0.3 7.5 7.7 7.1 -0.6 Total 100.0 100.0 100.0 100.0 100.0 100.0 Extreme poverty line at 2008 prices Urban 9.2 16.0 14.2 -1.8 55.6 60.5 64.3 3.7 Rural 56.3 42.1 44.0 1.9 36.9 31.7 28.6 -3.1 Indigenous 34.5 41.9 41.8 -0.1 7.5 7.7 7.1 -0.6 Total 100.0 100.0 100.0 100.0 100.0 100.0 Note: Changes shown between years 2003 and 2008 69 Table A.1.1.3: Mean Expenditure for Urban, Rural and Indigenous Areas Mean Percentage 1997 2003 2008 change Urban 3,207.9 2,889.0 3,080.3 6.6 Rural 1,305.8 1,421.5 1,505.1 5.9 Indigenous 416.6 373.1 389.5 4.4 Quintiles Lowest quintile 403.5 446.8 479.4 7.3 2 946.1 980.7 1,056.9 7.8 3 1,557.6 1,525.8 1,656.4 8.6 4 2,502.8 2,407.3 2,536.2 5.4 Highest quintile 6,067.4 5,779.9 6,457.0 11.7 Total 2,296.4 2,229.0 2,437.9 9.4 Note: Changes shown between years 2003 and 2008 Table A.1.1.4: Poverty Decomposition Absolute Percentage change change Moderated poverty line at 2008 prices Change in poverty (P0) -4.13 100.00 Total Intra-sectoral effect -2.62 63.52 Population-shift effect -1.54 37.18 Interaction effect 0.03 -0.70 Intra-sectoral effects: Urbana -1.41 34.13 Rural -1.05 25.53 Indigena -0.16 3.86 Extreme poverty line at 2008 prices Change in poverty (P0) -2.18 100.00 Total Intra-sectoral effect -1.08 49.80 Population-shift effect -1.08 49.45 Interaction effect -0.02 0.75 Intra-sectoral effects: Urbana -0.73 33.35 Rural 0.04 -2.02 Indigena -0.40 18.46 70 Table A.1.1.5: Poverty Headcount Index by Education level Urban Rural Indigenous National 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Moderated poverty line at 2008 prices Education level No education 24.8 35.3 32.6 71.5 67.9 65.6 97.5 99.3 98.3 57.8 58.2 54.8 Primary incomplete 24.3 29.0 27.1 66.2 61.5 59.6 96.4 98.9 97.4 52.4 50.7 48.1 Primary complete 22.0 23.9 22.7 59.6 55.8 52.8 91.5 98.8 96.2 46 42.6 40 Secondary incomplete 13.9 19.9 17.9 42.6 42.7 43.1 89.8 98.0 91.1 23.4 27.9 26.4 Secondary complete 7.2 11.3 10.0 26.6 26.5 26.0 71.7 72.9 65.6 11.8 15.3 13.3 Superior 2.2 2.6 2.5 8.8 12.1 11.1 26.6 66.9 45.2 3 3.9 3.6 Total 15.3 20.0 17.7 58.7 54.0 50.7 95.4 98.4 96.3 37.3 36.8 32.7 Extreme poverty line at 2008 prices Education level No education 6.6 10.3 7.1 39.7 33.9 32.7 89.0 92.4 89.8 35.3 33.3 30.2 Primary incomplete 5.1 6.8 6.1 34.5 26.3 28.3 89.0 91.7 85.6 28.4 24.5 24.1 Primary complete 5.2 5.0 4.1 30.3 22.1 23.9 80.4 86.7 82.6 22.6 17.4 17.1 Secondary incomplete 1.9 3.6 2.5 10.3 11.1 14.6 75.3 83.7 68.7 6 7.8 7.5 Secondary complete 0.9 1.2 1.3 4.1 4.7 5.7 43.8 53.3 40.1 2 2.6 2.4 Superior 0.3 0.3 0.1 2.2 1.3 1.7 17.3 31.9 29.6 0.5 0.5 0.4 Total 3.1 4.4 3.2 28.7 22.0 22.2 86.3 90.0 84.8 18.8 16.6 14.4 Note: Changes shown between years 2003 and 2008 71 Table A.1.1.6: Poverty by Household Head's Gender Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Male 39.5 38.6 33.4 -5.2 84.2 79.9 75.3 -4.6 79.6 76.3 73.7 -2.6 Female 28.9 31.2 30.7 -0.5 15.8 20.1 24.7 4.6 20.4 23.7 26.3 2.6 Total 37.3 36.8 32.7 -4.1 100 100 100 0 100 100 100 0 Extreme poverty line at 2008 prices Male 20.3 17.8 15.4 -2.4 85.9 81.8 78.8 -2.9 79.6 76.3 73.7 -2.6 Female 13 12.8 11.6 -1.1 14.1 18.2 21.2 2.9 20.4 23.7 26.3 2.6 Total 18.8 16.6 14.4 -2.2 100 100 100 0 100 100 100 0 Note: Changes shown between years 2003 and 2008 Table A.1.1.7: Headcount Index by Household Head's Gender and Area Urban Rural Indigenous National 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Moderated poverty line at 2008 prices Male 14.7 19.8 16.6 60.4 55.3 49.6 95.1 98.1 96.2 39.5 38.6 33.4 Female 16.9 20.5 20.1 48.8 47.6 55.6 97.6 99.5 96.6 28.9 31.2 30.7 Total 15.3 20.0 17.7 58.7 54.0 50.7 95.4 98.4 96.3 37.3 36.8 32.7 Extreme poverty line at 2008 prices Male 2.6 4.2 2.9 30.0 23.0 22.0 85.1 89.5 84.7 20.3 17.8 15.4 Female 4.7 4.9 3.9 21.0 17.0 22.8 94.5 92.1 85.4 13.0 12.8 11.6 Total 3.1 4.4 3.2 28.7 22.0 22.2 86.3 90.0 84.8 18.8 16.6 14.4 Note: Changes shown between years 2003 and 2008 72 Table A.1.1.8: Poverty by Age Groups Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Age 0-5 52.9 54.1 49.4 -4.7 18.0 19.0 17.2 -1.8 12.7 12.9 11.4 -1.5 6-14 47.7 48.1 46.0 -2.1 25.3 25.0 26.1 1.1 19.8 19.1 18.5 -0.6 15-19 41.0 39.8 36.9 -2.9 10.7 10.3 9.9 -0.4 9.7 9.6 8.8 -0.7 20-24 33.6 33.2 28.0 -5.2 7.8 7.7 7.3 -0.4 8.6 8.5 8.5 0.0 25-29 32.6 33.1 28.3 -4.8 7.0 6.7 6.1 -0.5 8.0 7.4 7.1 -0.3 30-34 33.2 33.1 26.3 -6.8 6.0 6.4 5.8 -0.6 6.7 7.1 7.2 0.1 35-39 28.8 30.8 27.7 -3.0 5.1 5.8 6.1 0.4 6.6 6.9 7.2 0.3 40-44 26.6 25.4 26.8 1.4 3.9 4.3 5.1 0.8 5.4 6.2 6.2 0.1 45-49 25.4 21.3 19.7 -1.6 3.4 2.9 3.4 0.5 5.0 5.0 5.7 0.7 50-54 29.5 26.0 20.9 -5.1 3.4 3.0 3.0 0.0 4.3 4.2 4.7 0.5 55-59 26.7 25.3 19.5 -5.8 2.5 2.3 2.2 -0.1 3.5 3.4 3.7 0.3 60-64 29.6 26.8 20.3 -6.6 2.3 2.1 2.1 0.0 3.0 2.8 3.4 0.5 65+ 26.4 24.1 24.2 0.1 4.8 4.6 5.7 1.1 6.7 6.9 7.6 0.7 Total 37.3 36.8 32.7 -4.1 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Extreme poverty line at 2008 prices Age 0-5 29.4 29.2 24.6 -4.6 19.9 22.9 19.4 -3.5 12.7 12.9 11.4 -1.5 6-14 26.6 23.2 22.1 -1.1 28.0 26.8 28.3 1.5 19.8 19.1 18.5 -0.6 15-19 20.5 18.0 16.9 -1.1 10.6 10.4 10.3 -0.1 9.7 9.6 8.8 -0.7 20-24 15.8 14.3 11.3 -3.0 7.2 7.4 6.7 -0.7 8.6 8.5 8.5 0.0 25-29 13.5 13.3 10.6 -2.7 5.7 5.9 5.2 -0.7 8.0 7.4 7.1 -0.3 30-34 13.9 11.6 9.8 -1.8 5.0 5.0 4.9 -0.1 6.7 7.1 7.2 0.1 35-39 13.8 11.1 11.1 -0.0 4.8 4.6 5.5 0.9 6.6 6.9 7.2 0.3 40-44 13.4 10.3 10.6 0.3 3.9 3.8 4.6 0.7 5.4 6.2 6.2 0.1 45-49 12.8 9.2 9.1 -0.1 3.4 2.8 3.6 0.8 5.0 5.0 5.7 0.7 50-54 13.2 10.6 8.7 -1.9 3.0 2.7 2.8 0.1 4.3 4.2 4.7 0.5 55-59 13.3 9.7 7.3 -2.5 2.4 2.0 1.9 -0.2 3.5 3.4 3.7 0.3 60-64 13.5 10.2 7.9 -2.4 2.1 1.8 1.8 0.1 3.0 2.8 3.4 0.5 65+ 11.1 9.1 9.4 0.2 4.0 3.8 5.0 1.1 6.7 6.9 7.6 0.7 Total 18.8 16.6 14.4 -2.2 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Note: Changes shown between years 2003 and 2008 73 Table A.1.1.9: Poverty by Age Groups. Urban Areas Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Age 0-5 24.3 34.6 31.1 -3.5 17.3 20.2 18.4 -1.9 10.9 11.7 10.4 -1.2 6-14 20.8 28.0 27.1 -0.9 23.7 23.8 25.2 1.4 17.4 17.0 16.4 -0.6 15-19 18.2 21.9 20.0 -1.9 11.4 10.3 9.7 -0.6 9.6 9.4 8.6 -0.8 20-24 13.5 18.9 15.3 -3.6 8.3 8.5 8.0 -0.5 9.4 9.0 9.3 0.3 25-29 14.5 18.9 15.9 -3.0 8.5 7.6 6.8 -0.8 9.0 8.0 7.6 -0.5 30-34 16.9 19.1 16.1 -3.0 8.1 7.4 7.2 -0.2 7.3 7.8 7.9 0.2 35-39 11.4 16.3 15.4 -0.9 5.3 6.0 6.6 0.6 7.1 7.3 7.6 0.3 40-44 7.5 13.2 13.9 0.7 2.8 4.5 5.0 0.6 5.7 6.7 6.4 -0.3 45-49 8.7 9.6 9.2 -0.4 3.1 2.6 3.2 0.5 5.5 5.5 6.1 0.6 50-54 10.3 11.3 9.8 -1.5 3.1 2.4 2.8 0.4 4.6 4.3 5.0 0.7 55-59 7.0 9.6 7.3 -2.3 1.7 1.7 1.7 -0.0 3.7 3.5 4.0 0.6 60-64 9.7 10.6 6.1 -4.4 1.9 1.6 1.2 -0.4 2.9 3.0 3.4 0.4 65+ 10.7 9.9 10.3 0.3 4.8 3.4 4.2 0.9 6.8 6.8 7.3 0.5 Total 15.3 20.0 17.7 -2.3 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Extreme poverty line at 2008 prices Age 0-5 5.7 9.7 6.4 -3.2 19.8 25.7 21.0 -4.7 10.9 11.7 10.4 -1.2 6-14 4.6 6.9 5.2 -1.7 25.8 26.6 26.9 0.3 17.4 17.0 16.4 -0.6 15-19 4.4 5.6 3.6 -2.0 13.5 12.0 9.7 -2.4 9.6 9.4 8.6 -0.8 20-24 2.7 3.6 2.3 -1.3 8.1 7.4 6.6 -0.8 9.4 9.0 9.3 0.3 25-29 2.0 3.2 2.6 -0.6 5.8 5.8 6.1 0.2 9.0 8.0 7.6 -0.5 30-34 3.2 3.1 2.3 -0.7 7.6 5.4 5.8 0.4 7.3 7.8 7.9 0.2 35-39 1.6 1.9 2.6 0.6 3.7 3.2 6.1 2.9 7.1 7.3 7.6 0.3 40-44 1.3 3.0 2.3 -0.7 2.3 4.7 4.6 -0.0 5.7 6.7 6.4 -0.3 45-49 2.0 2.5 1.5 -1.0 3.5 3.1 2.9 -0.2 5.5 5.5 6.1 0.6 50-54 1.5 2.1 1.2 -0.9 2.3 2.1 1.9 -0.1 4.6 4.3 5.0 0.7 55-59 2.2 0.2 1.2 1.0 2.6 0.2 1.5 1.4 3.7 3.5 4.0 0.6 60-64 0.9 2.2 0.6 -1.7 0.8 1.6 0.6 -0.9 2.9 3.0 3.4 0.4 65+ 1.9 1.5 2.8 1.3 4.1 2.3 6.3 4.0 6.8 6.8 7.3 0.5 Total 3.1 4.4 3.2 -1.2 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 74 Table A.1.1.10: Poverty by Age Groups. Rural Areas Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Age 0-5 73.4 68.5 65.3 -3.3 17.4 16.4 14.5 -2.0 13.9 12.9 11.2 -1.7 6-14 67.4 62.7 61.4 -1.3 24.6 24.3 24.7 0.4 21.4 20.9 20.4 -0.5 15-19 62.2 59.0 56.4 -2.6 10.2 10.9 10.2 -0.7 9.6 10.0 9.1 -0.8 20-24 58.2 50.9 49.8 -1.1 7.6 7.4 7.2 -0.2 7.7 7.9 7.4 -0.5 25-29 58.1 53.2 47.5 -5.7 6.9 6.6 6.1 -0.5 6.9 6.7 6.5 -0.2 30-34 52.8 52.5 42.0 -10.5 5.6 6.0 5.1 -0.9 6.2 6.2 6.2 -0.0 35-39 48.1 49.1 45.9 -3.2 5.0 5.9 6.3 0.4 6.1 6.5 7.0 0.5 40-44 48.6 43.2 44.8 1.6 4.5 4.6 5.6 1.0 5.4 5.8 6.3 0.6 45-49 44.6 37.1 34.9 -2.2 3.5 3.1 3.6 0.5 4.6 4.6 5.2 0.7 50-54 51.6 44.3 39.0 -5.3 3.6 3.5 3.4 -0.2 4.0 4.3 4.4 0.1 55-59 51.6 42.2 42.8 0.6 3.0 2.7 2.9 0.2 3.4 3.5 3.4 -0.1 60-64 50.5 51.8 41.2 -10.6 2.8 2.7 2.9 0.2 3.2 2.8 3.6 0.8 65+ 43.3 38.8 40.9 2.1 5.5 5.8 7.6 1.7 7.5 8.1 9.4 1.3 Total 58.7 54.0 50.7 -3.3 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Extreme poverty line at 2008 prices Age 0-5 39.8 36.2 33.4 -2.8 19.2 21.2 16.9 -4.3 13.9 12.9 11.2 -1.7 6-14 36.3 26.3 28.9 2.6 27.1 25.0 26.5 1.6 21.4 20.9 20.4 -0.5 15-19 29.1 23.9 26.5 2.6 9.7 10.8 10.9 0.1 9.6 10.0 9.1 -0.8 20-24 26.8 21.7 21.3 -0.4 7.1 7.7 7.1 -0.7 7.7 7.9 7.4 -0.5 25-29 24.8 22.4 17.5 -4.9 6.0 6.8 5.1 -1.6 6.9 6.7 6.5 -0.2 30-34 22.1 15.4 16.9 1.5 4.8 4.3 4.7 0.4 6.2 6.2 6.2 -0.0 35-39 21.8 15.9 19.0 3.1 4.7 4.7 6.0 1.3 6.1 6.5 7.0 0.5 40-44 23.9 16.1 17.3 1.2 4.5 4.2 4.9 0.7 5.4 5.8 6.3 0.6 45-49 21.3 12.7 16.4 3.7 3.4 2.6 3.9 1.3 4.6 4.6 5.2 0.7 50-54 22.0 16.1 17.2 1.2 3.1 3.1 3.4 0.3 4.0 4.3 4.4 0.1 55-59 23.5 14.3 13.6 -0.7 2.8 2.2 2.1 -0.2 3.4 3.5 3.4 -0.1 60-64 23.7 18.0 14.3 -3.7 2.7 2.3 2.3 0.0 3.2 2.8 3.6 0.8 65+ 19.1 13.5 14.7 1.2 5.0 5.0 6.2 1.2 7.5 8.1 9.4 1.3 Total 28.7 22.0 22.2 0.1 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Note: Changes shown between years 2003 and 2008 75 Table A.1.1.11: Poverty by Age Groups. Indigenous Areas Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Age 0-5 96.6 99.3 98.2 -1.1 20.7 23.2 21.0 -2.1 20.4 22.9 20.6 -2.3 6-14 95.8 99.0 97.5 -1.5 29.1 28.7 30.5 1.8 29.0 28.5 30.1 1.6 15-19 97.3 98.6 96.2 -2.4 11.2 9.1 9.9 0.7 11.0 9.1 9.9 0.8 20-24 94.3 98.6 96.1 -2.6 7.6 6.9 6.2 -0.7 7.7 6.9 6.3 -0.7 25-29 92.7 97.8 94.4 -3.4 5.5 5.4 5.2 -0.2 5.7 5.4 5.3 -0.1 30-34 88.6 97.7 94.9 -2.8 4.8 5.7 4.9 -0.7 5.1 5.7 5.0 -0.7 35-39 92.7 98.2 95.4 -2.8 5.1 5.1 4.9 -0.1 5.3 5.1 5.0 -0.1 40-44 95.9 93.7 93.5 -0.2 3.2 3.1 4.1 1.0 3.2 3.3 4.3 1.0 45-49 96.2 93.6 95.1 1.5 3.6 2.8 3.4 0.6 3.6 2.9 3.4 0.5 50-54 93.6 96.4 94.9 -1.4 3.2 2.4 2.5 0.0 3.3 2.5 2.5 -0.0 55-59 100.0 97.9 95.1 -2.8 1.9 2.6 1.7 -1.0 1.8 2.6 1.7 -1.0 60-64 93.7 100.0 91.3 -8.7 1.6 1.4 1.8 0.4 1.6 1.4 1.9 0.5 65+ 100.0 98.4 91.8 -6.6 2.4 3.6 4.0 0.4 2.3 3.6 4.1 0.6 Total 95.4 98.4 96.3 -2.1 100 100 100 100 100 100 100 100 Extreme poverty line at 2008 prices Age 0-5 88.4 92.2 88.6 -3.6 20.9 23.5 21.6 -2.0 20.4 22.9 20.6 -2.3 6-14 89.1 91.0 86.6 -4.4 30.0 28.9 30.8 1.9 29.0 28.5 30.1 1.6 15-19 88.0 92.4 85.0 -7.4 11.2 9.4 9.9 0.5 11.0 9.1 9.9 0.8 20-24 80.1 90.5 85.0 -5.5 7.2 7.0 6.3 -0.7 7.7 6.9 6.3 -0.7 25-29 79.8 85.4 79.8 -5.6 5.3 5.2 5.0 -0.2 5.7 5.4 5.3 -0.1 30-34 77.4 86.6 82.3 -4.3 4.6 5.5 4.8 -0.6 5.1 5.7 5.0 -0.7 35-39 88.9 90.9 83.5 -7.4 5.4 5.2 4.9 -0.2 5.3 5.1 5.0 -0.1 40-44 86.9 85.5 82.0 -3.5 3.2 3.1 4.1 1.0 3.2 3.3 4.3 1.0 45-49 83.6 88.4 86.6 -1.7 3.4 2.8 3.5 0.6 3.6 2.9 3.4 0.5 50-54 80.3 86.7 84.8 -1.9 3.1 2.4 2.5 0.1 3.3 2.5 2.5 -0.0 55-59 90.2 84.6 87.4 2.7 1.9 2.5 1.7 -0.8 1.8 2.6 1.7 -1.0 60-64 83.9 84.4 79.2 -5.2 1.6 1.3 1.7 0.4 1.6 1.4 1.9 0.5 65+ 85.7 82.9 65.5 -17.4 2.3 3.3 3.2 -0.1 2.3 3.6 4.1 0.6 Total 86.3 90.0 84.8 -5.2 100 100 100 0.0 100 100 100 0.0 Note: Changes shown between years 2003 and 2008 76 Table A.1.1.12: Poverty by Demographic Composition Poverty Headcount Rate Distribution of the Poor Distribution of Population 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices No Children 21.5 19.7 18.8 -1.0 26.9 25.2 29.8 4.6 46.7 47.2 52.0 4.8 1 34.2 34.7 32.3 -2.4 24.6 25.7 27.2 1.5 26.9 27.3 27.5 0.2 2 58.8 59.9 60.3 0.4 24.8 23.1 24.4 1.3 15.8 14.2 13.2 -1.0 3 or more 82.8 84.5 83.5 -1.1 23.6 25.9 18.6 -7.3 10.7 11.3 7.3 -4.0 children Household size 1 11.9 9.0 6.1 -3.0 0.9 0.7 0.7 0.0 2.8 2.8 3.7 1.0 2 11.4 10.2 10.7 0.5 1.9 2.2 3.1 0.9 6.4 8.0 9.4 1.4 3 18.0 15.0 13.2 -1.9 5.9 5.5 6.0 0.5 12.2 13.5 14.8 1.3 4 21.6 19.8 19.1 -0.7 10.4 10.6 12.6 2.0 18.0 19.7 21.5 1.9 5 27.9 29.3 32.0 2.7 13.9 14.0 18.0 4.1 18.5 17.5 18.5 0.9 6 41.1 48.9 45.0 -3.9 15.3 16.3 15.9 -0.3 13.9 12.3 11.6 -0.7 7 or more 68.2 71.1 69.8 -1.2 51.7 50.8 43.7 -7.1 28.3 26.3 20.5 -5.9 Total 37.3 36.8 32.7 -4.1 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Extreme poverty line at 2008 prices No children 8.1 5.4 6.3 1.0 20.1 15.2 22.8 7.5 46.7 47.2 52.0 4.8 1 15.3 10.3 11.2 0.9 21.8 16.9 21.3 4.4 26.9 27.3 27.5 0.2 2 30.2 30.6 28.5 -2.1 25.3 26.3 26.2 -0.1 15.8 14.2 13.2 -1.0 3 or more 57.9 61.2 58.8 -2.4 32.8 41.6 29.8 -11.8 10.7 11.3 7.3 -4.0 children Household size 1 3.4 2.0 1.9 -0.1 0.5 0.3 0.5 0.2 2.8 2.8 3.7 1.0 2 4.4 2.6 3.1 0.6 1.5 1.2 2.0 0.8 6.4 8.0 9.4 1.4 3 5.5 2.6 4.1 1.5 3.5 2.1 4.2 2.1 12.2 13.5 14.8 1.3 4 7.4 5.0 4.8 -0.2 7.1 6.0 7.2 1.2 18.0 19.7 21.5 1.9 5 11.4 8.7 10.3 1.6 11.2 9.2 13.2 4.0 18.5 17.5 18.5 0.9 6 17.2 17.5 15.9 -1.6 12.7 12.9 12.7 -0.2 13.9 12.3 11.6 -0.7 7 or more 42.2 43.0 42.4 -0.7 63.4 68.2 60.1 -8.1 28.3 26.3 20.5 -5.9 Total 18.8 16.6 14.4 -2.2 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Note: Changes shown between years 2003 and 2008 77 Table A.1.1.13: Other measures of poverty Sen Index Sen-Shorrocks-Thon Index Watts Index 1997 2003 2008 change 1997 2003 2008 change 1997 2003 2008 change Moderated poverty line at 2008 prices Area Urban 0.086 0.116 0.103 -0.013 0.078 0.112 0.094 -0.018 0.051 0.074 0.060 -0.014 Rural 0.390 0.336 0.323 -0.013 0.502 0.411 0.392 -0.019 0.390 0.302 0.288 -0.014 Indigenous 0.779 0.791 0.790 -0.002 1.316 1.368 1.333 -0.035 1.294 1.349 1.370 0.020 Total 0.266 0.254 0.226 -0.027 0.328 0.304 0.268 -0.036 0.270 0.245 0.218 -0.026 Extreme poverty line at 2008 prices Area Urban 0.017 0.024 0.017 -0.007 0.014 0.018 0.012 -0.007 0.008 0.012 0.007 -0.004 Rural 0.179 0.126 0.129 0.002 0.204 0.132 0.128 -0.004 0.146 0.089 0.085 -0.003 Indigenous 0.628 0.626 0.628 0.002 0.936 0.954 0.951 -0.004 0.785 0.799 0.850 0.050 Total 0.131 0.112 0.098 -0.013 0.153 0.127 0.112 -0.015 0.118 0.097 0.089 -0.007 Note: Changes shown between years 2003 and 2008 78 ANNEX 1.2. POVERTY STATISTICS BY PROVINCES Table A.1.2.1: Poverty by Provinces Poverty Headcount Distribution of the Distribution of Rate Poor Population 1997 2003 2008 Change 1997 2003 2008 change 1997 2003 2008 Change Moderated poverty line at 2008 prices Bocas del Toro 69.7 69.8 64.7 -5.1 8.0 6.5 6.6 0.1 4.3 3.4 3.3 -0.1 Coclé 55.4 57.1 51.6 -5.5 10.6 10.9 10.9 0.0 7.2 7.0 6.9 -0.1 Colón 43.4 42.7 26.9 -15.9 9.5 8.3 5.9 -2.4 8.2 7.1 7.1 0.0 Chiriquí 52.4 35.3 29.3 -6.0 22.3 13.0 10.8 -2.2 15.9 13.6 12.1 -1.5 Darién 75.3 71.8 60.2 -11.7 4.7 3.1 2.5 -0.7 2.3 1.6 1.3 -0.3 Herrera 39.7 29.9 33.7 3.8 4.1 2.7 3.4 0.8 3.8 3.3 3.3 0.0 Los Santos 22.7 27.0 33.8 6.8 1.8 2.3 2.8 0.5 2.9 3.1 2.7 -0.4 Panamá 19.7 20.3 19.2 -1.1 24.9 26.0 29.8 3.8 47.2 47.3 50.9 3.5 Veraguas 63.8 53.7 52.4 -1.3 14.1 10.7 10.7 0.0 8.2 7.3 6.7 -0.6 Comarca Kuna 99.1 91.9 -7.2 4.4 3.1 -1.3 1.6 1.1 -0.5 Yala Comarca Emberá 95.9 87.4 -8.5 0.9 0.8 -0.1 0.3 0.3 -0.1 Comarca Ngöbe 98.1 98.0 -0.1 11.3 12.8 1.5 4.2 4.3 0.0 Bugle Total 37.3 36.8 32.7 -4.1 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Extreme poverty line at 2008 prices Bocas del Toro 51.7 51.0 43.2 -7.8 11.7 10.5 10.0 -0.5 4.3 3.4 3.3 -0.1 Coclé 32.5 27.7 23.0 -4.7 12.4 11.7 11.0 -0.7 7.2 7.0 6.9 -0.1 Colón 25.0 10.3 7.8 -2.5 10.9 4.4 3.9 -0.5 8.2 7.1 7.1 0.0 Chiriquí 27.9 10.8 11.3 0.5 23.5 8.8 9.4 0.6 15.9 13.6 12.1 -1.5 Darién 50.0 37.4 31.6 -5.8 6.2 3.6 2.9 -0.7 2.3 1.6 1.3 -0.3 Herrera 13.6 4.5 13.1 8.6 2.8 0.9 3.0 2.1 3.8 3.3 3.3 0.0 Los Santos 4.4 7.6 6.3 -1.2 0.7 1.4 1.2 -0.2 2.9 3.1 2.7 -0.4 Panamá 5.8 5.4 4.1 -1.3 14.7 15.5 14.5 -1.0 47.2 47.3 50.9 3.5 Veraguas 39.2 20.0 22.8 2.8 17.2 8.8 10.6 1.7 8.2 7.3 6.7 -0.6 Comarca Kuna 92.3 68.8 -23.5 9.1 5.3 -3.8 1.6 1.1 -0.5 Yala Comarca Emberá 93.6 57.0 -36.6 1.9 1.1 -0.8 0.3 0.3 -0.1 Comarca Ngöbe 91.6 91.7 0.1 23.4 27.1 3.7 4.2 4.3 0.0 Bugle Total 18.8 16.6 14.4 -2.2 100.0 100.0 100.0 0.0 100.0 100.0 100.0 0.0 Note: Changes shown between years 2003 and 2008 79 Table A.1.2.2: Mean per Capita Expenditure by Provinces Mean Percentage 1997 2003 2008 change Bocas del Toro 1,001.5 1,109.1 1,414.4 27.5 Coclé 1,484.3 1,342.3 1,561.9 16.4 Colón 1,642.3 1,885.9 2,233.1 18.4 Chiriquí 1,671.5 1,902.4 2,182.1 14.7 Darién 971.4 1,031.2 1,200.2 16.4 Herrera 1,957.4 2,219.0 2,099.7 -5.4 Los Santos 2,238.3 2,332.1 2,243.1 -3.8 Panamá 3,152.2 2,982.4 3,134.1 5.1 Veraguas 1,177.4 1,480.5 1,473.7 -0.5 Comarca Kuna Yala 360.3 616.4 71.1 Comarca Emberá 495.9 717.2 44.6 Comarca Ngöbe Bugle 348.0 295.2 -15.2 QUINTILES Lowest quintile 403.5 446.8 479.4 7.3 2 946.1 980.7 1,056.9 7.8 3 1,557.6 1,525.8 1,656.4 8.6 4 2,502.8 2,407.3 2,536.2 5.4 Highest quintile 6,067.4 5,779.9 6,457.0 11.7 Total 2,296.4 2,229.0 2,437.9 9.4 Note: Changes shown between years 2003 and 2008 80 ANNEX 1.3 SELECTED INEQUALITY STATISTICS Table A.1.3.1: Gini coefficient by areas 1997 2003 2008 Total Panama 0.49 0.47 0.48 Urban areas 0.41 0.42 0.44 Rural areas 0.45 0.44 0.46 Rural-Non indigenous 0.41 0.39 0.41 Indigenous areas 0.40 0.35 0.41 Table A.1.3.2: Gini coefficient by provinces 2008 2003 1997 Prov. de Bocas del Toro 0.5572 0.49904 0.52124 Prov. de Coclé 0.42411 0.40188 0.47252 Prov. de Colón 0.37239 0.4059 0.41263 Prov. de Chiriquí 0.42023 0.3812 0.49653 Prov. de Darién 0.39979 0.37317 0.46937 Prov. de Herrera 0.41445 0.37862 0.41469 Prov. de Los Santos 0.42608 0.39436 0.33694 Prov. de Panamá 0.45017 0.43282 0.43935 Prov. de Veraguas 0.41438 0.41077 0.44933 Total Comarcas 0.43025 0.34025 Comarca de Kuna Yala 0.30723 Comarca Emberá 0.34031 Comarca Ngobe Bugle 0.41732 81 Table A.1.3.3: Decomposition of inequality by regions 1997 2003 2008 GE(0) GE(1) GE(2) GE(0) GE(1) GE(2) GE(0) GE(1) GE(2) Total 44.0 41.8 61.1 40.2 38.7 54.6 42.3 42.1 67.3 Between-group inequality 7.7 7.1 6.7 11.0 8.0 6.8 10.0 6.9 5.7 Between as a share of total 17.5 16.9 11.0 27.3 20.7 12.4 23.6 16.4 8.4 Within-group inequality 36.3 34.8 54.3 29.2 30.7 47.9 32.3 35.2 61.7 Bocas del Toro 50.3 48.9 76.4 43.0 45.0 72.2 55.8 59.0 110.3 Coclé 40.9 38.8 53.8 26.6 28.2 38.2 30.4 32.2 47.6 Colón 31.4 28.0 32.6 28.0 28.7 38.8 24.1 23.0 27.3 Chiriquí 46.4 46.2 79.1 24.6 25.1 33.3 30.2 36.4 72.8 Darién 38.2 37.2 47.8 23.2 24.3 32.1 27.5 27.8 37.9 Herrera 29.5 29.0 36.9 23.9 24.2 30.9 30.0 30.6 47.7 Los Santos 19.1 19.4 25.1 26.2 27.0 35.9 29.6 32.4 46.4 Panamá 33.4 33.5 46.2 32.3 32.4 43.7 34.2 36.5 55.9 Veraguas 35.5 35.3 49.7 28.5 31.0 47.4 29.0 29.7 39.9 Comarca Kuna Yala 13.4 13.8 17.9 15.2 17.0 23.4 Comarca Emberá 11.0 15.8 30.0 19.5 21.5 30.6 Comarca Ngöbe Bugle 22.8 26.9 55.9 30.3 31.1 45.7 Table A.1.3.4: Decomposition of inequality by areas 1997 2003 2008 GE(0) GE(1) GE(2) GE(0) GE(1) GE(2) GE(0) GE(1) GE(2) Total 44.0 41.8 61.1 40.2 38.7 54.6 42.3 42.1 67.3 Urban 28.7 30.1 41.8 29.7 30.7 41.7 31.6 34.6 54.2 Rural 29.5 30.0 44.4 25.7 26.6 36.7 28.2 29.8 44.5 Indigenous 26.8 29.6 44.4 21.9 24.1 44.7 29.7 29.1 40.0 Within-group inequality 28.9 30.0 50.7 27.8 29.8 47.2 30.5 33.7 60.5 Between-group inequality 15.1 11.8 10.3 12.4 8.9 7.4 11.8 8.4 6.8 Between as a share of total 34.3 28.2 16.9 30.8 23.1 13.6 27.9 19.9 10.2 82 Table A.1.3.5: Growth and redistribution decomposition of poverty changes Change in incidence of poverty 2003 2008 Actual change Growth Redistribution Interaction Moderated poverty line at 2008 prices Total 36.83 32.70 -4.13 -3.89 -0.80 0.56 Area Urbana 19.99 17.67 -2.33 -2.80 -0.12 0.59 Rural 53.97 50.65 -3.32 -3.23 -0.53 0.43 Indigena 98.37 96.31 -2.06 -0.49 -1.59 0.02 Extreme poverty line at 2008 prices Total 16.61 14.43 -2.18 -1.77 -0.09 -0.31 Area Urbana 4.39 3.20 -1.20 -0.68 -0.40 -0.11 Rural 22.04 22.18 0.14 -1.79 2.86 -0.93 Indigena 89.99 84.79 -5.20 -0.49 -4.24 -0.47 83 ANNEX 1.4. GROWTH INCIDENCE CURVES Figure A.1.4.1: Growth Incidence Curve by area, Panama (1997-2008) in Balboas of 2008 Total Urban Total (years 1997 and 2008) Urbana 3 Growth-incidence 95% confidence bounds 3 Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 1 1 Annual growth rate % Annual growth rate % -1 -1 -3 -3 -5 -5 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Rural Indigenous Rural Indigena 3 Growth-incidence 95% confidence bounds 3 Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 1 1 Annual growth rate % Annual growth rate % -1 -1 -3 -3 -5 -5 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Source: World Bank staff calculations based on ENV 1997, 2003 and 2008 data 84 Figure A.1.4.2: Growth Incidence Curve by area, Panama (1997-2003) in Balboas of 2008 Total Urban 7 Total (years 1997 and 2003) 7 Urbana Growth-incidence 95% confidence bounds Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 4 4 Annual growth rate % Annual growth rate % 1 1 -2 -2 -5 -5 -8 -8 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Rural Indigenous 7 Rural 7 Indigena Growth-incidence 95% confidence bounds Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 4 4 Annual growth rate % Annual growth rate % 1 1 -2 -2 -5 -5 -8 -8 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Source: World Bank staff calculations based on ENV 1997, 2003 data 85 Figure A.1.4.3: Growth Incidence Curve by area, Panama (2003-2008) in Balboas of 2008 Total Urban 6 Total (years 2003 and 2008) 6 Urbana Growth-incidence 95% confidence bounds Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 3 3 Annual growth rate % Annual growth rate % 0 0 -3 -3 -6 -6 -9 -9 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Rural Indigenous 6 Rural 6 Indigena Growth-incidence 95% confidence bounds Growth-incidence 95% confidence bounds Growth in mean Mean growth rate Growth in mean Mean growth rate 3 3 Annual growth rate % Annual growth rate % 0 0 -3 -3 -6 -6 -9 -9 1 10 20 30 40 50 60 70 80 90 100 1 10 20 30 40 50 60 70 80 90 100 Expenditure percentiles Expenditure percentiles Source: World Bank staff calculations based on ENV 2003 and 2008 data 86 Figure A.1.4.4: Lorenz Curve by Areas Total Urbana 1 1 2003, Gini=46.9 2003, Gini=42.1 2008, Gini=47.9 2008, Gini=43.55 Line of equality Line of equality .8 .8 .6 .6 Lorenz curve .4 .4 .2 .2 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Cumulative population proportion Cumulative population proportion Rural Indigenous 1 1 2003, Gini=39 2003, Gini=34.93 2008, Gini=40.79 2008, Gini=40.77 Line of equality Line of equality .8 .8 .6 .6 .4 .4 .2 .2 0 0 0 .2 .4 .6 .8 1 0 .2 .4 .6 .8 1 Cumulative population proportion Cumulative population proportion 87 ANNEX 2.1 BASIC DESCRIPTIVE STATISTICS Table A.2.1.1 Characteristics by poverty status 2008: Moderate Poverty Non-poor Poor Difference SE N Male 0.749 0.796 0.047*** (0.014) 4,380 Year of Birth 1,967 1,968 1.519*** (0.265) 4,380 Years of Education 10.532 5.758 -4.774*** (0.128) 4,360 Mother of head completed primary 0.485 0.386 -0.099*** (0.016) 4,380 Mother of head completed secondary 0.175 0.035 -0.139*** (0.009) 4,380 Mother of head completed post-secondary 0.053 0.005 -0.048*** (0.004) 4,380 Mother tongue is indigenous 0.021 0.297 0.276*** (0.013) 4,380 Mother tongue is other 0.023 0.000 -0.023*** (0.003) 4,380 Agriculture 0.111 0.497 0.385*** (0.015) 4,348 Mining 0.003 0.005 0.002 (0.002) 4,348 Manufacturing 0.071 0.052 -0.020** (0.008) 4,348 Electricity, gas, water supply 0.008 0.002 -0.006*** (0.002) 4,348 Construction 0.107 0.084 -0.023** (0.010) 4,348 Transport, communications 0.088 0.020 -0.068*** (0.006) 4,348 Financial and professional services 0.064 0.019 -0.046*** (0.006) 4,348 Public administration and defense 0.081 0.029 -0.052*** (0.007) 4,348 Education, Health 0.134 0.065 -0.069*** (0.009) 4,348 Domestic Service 0.024 0.028 0.004 (0.005) 4,348 Missing or inactive 0.086 0.112 0.025** (0.010) 4,348 Note: Means for poor and non-poor are weighted. The difference and standard error are obtained from a regression of poverty status (binary) on the characteristic. Robust standard errors reported. 88 Table A.2.1.2 Characteristics by poverty status 2008: Extreme Poverty Non-poor Poor Difference SE N Male 0.754 0.824 0.070*** (0.017) 4,380 Year of Birth 1,967 1,968 1.010*** (0.355) 4,380 Years of Education 9.828 4.275 -5.553*** (0.154) 4,360 Mother of head completed primary 0.488 0.237 -0.251*** (0.020) 4,380 Mother of head completed secondary 0.150 0.018 -0.133*** (0.008) 4,380 Mother of head completed post-secondary 0.044 0.000 -0.044*** (0.003) 4,380 Mother tongue is indigenous 0.047 0.482 0.435*** (0.021) 4,380 Mother tongue is other 0.019 0.000 -0.019*** (0.002) 4,380 Agriculture 0.154 0.709 0.554*** (0.020) 4,348 Mining 0.004 0.004 -0.000 (0.003) 4,348 Manufacturing 0.071 0.030 -0.041*** (0.008) 4,348 Electricity, gas, water supply 0.006 0.002 -0.005** (0.002) 4,348 Construction 0.108 0.052 -0.056*** (0.011) 4,348 Transport, communications 0.078 0.002 -0.076*** (0.005) 4,348 Financial and professional services 0.057 0.007 -0.050*** (0.005) 4,348 Public administration and defense 0.075 0.005 -0.069*** (0.005) 4,348 Education, Health 0.125 0.037 -0.088*** (0.010) 4,348 Domestic Service 0.027 0.011 -0.016*** (0.005) 4,348 Missing or inactive 0.092 0.110 0.018 (0.014) 4,348 Note: Means for poor and non-poor are weighted. The difference and standard error are obtained from a regression of poverty status (binary) on the characteristic. Robust standard errors reported. 89 ANNEX 2.2 CONSUMPTION MODELS Table A.2.2.1 Consumption Models for year 2008 Log (Consumption) (2008 prices) [1] [2] Rural (1=Yes) -0.220*** (0.031) Indigenous (1=Yes) -0.916*** (0.083) Male (1 = Yes) 0.053* 0.025 (0.031) (0.031) Year of Birth -1.764** -1.253* (0.766) (0.752) Year of Birth Squared 0.000** 0.000* (0.000) (0.000) Years of Education of head of household 0.013 0.013 (0.011) (0.011) Years of Education of head of household squared 0.003*** 0.002*** (0.001) (0.001) Head's mother has completed primary education 0.087*** 0.065** (0.029) (0.028) Head's mother has completed secondary education 0.220*** 0.161*** (0.043) (0.040) Head's mother has completed post-secondary education 0.576*** 0.425*** (0.073) (0.065) Mother tongue: Indigenous -0.384*** -0.574*** (0.053) (0.063) Mother tongue: English or other 0.548*** 0.476*** (0.071) (0.094) Agriculture -0.367*** -0.238*** (0.049) (0.050) Mining -0.135 0.032 (0.211) (0.244) Manufacturing -0.166*** -0.141*** (0.050) (0.049) Electricity, gas, water supply 0.217 0.342* (0.197) (0.183) Construction -0.110** -0.122*** (0.047) (0.045) Transport, communications 0.121** 0.166*** 90 Table A.2.2.1 Consumption Models for year 2008 Log (Consumption) (2008 prices) [1] [2] (i) (0.058) (0.058) Financial and professional services -0.012 -0.053 (0.053) (0.054) Public administration and defense -0.176*** -0.149*** (0.049) (0.047) Education, Health -0.196*** -0.142*** (0.044) (0.044) Domestic service -0.243*** -0.239*** (0.074) (0.081) Missing or inactive -0.192*** -0.179*** (0.045) (0.047) Constant 1,752.072** 1,247.390* (753.800) (739.303) Indicator variables for place of residence five years ago N Y Number of observations 4,333 4,333 R2 0.524 0.629 Adjusted R2 0.522 0.588 Number of clusters 813 813 note: *** p<0.01, ** p<0.05, * p<0.1. Dependent variable is log consumption in 2008 prices. Column 2 controls for a number of indicator variables derived from retrospective questions on whether the respondent currently resides in the same location as five years ago (the time of the 2003 survey). 91 ANNEX 3.1 EDUCATION STATISTICS Table A.3.1.1 Descriptive Statistics by Area Total National Urban Rural Indigenous 1997 2003 2008 1997 2003 2008 1997 2003 2008 1997 2003 2008 Average years of education Pop>18 years 8.4 8.9 9.4 10.0 10.4 10.9 5.3 6.4 7.2 3.1 3.6 3.5 Literacy rate 90.6 91.9 93.9 96.7 96.6 97.9 87.0 88.8 91.7 60.0 63.6 64.8 Higher educational level achieved Pop>18 years Completed primary education 76.5 80.4 83.2 87.9 88.7 91.0 61.8 66.4 72.5 29.5 32.2 32.1 Completed secondary education 78.0 77.9 78.9 47.4 47.5 53.4 14.1 19.0 21.5 4.5 5.3 5.2 Completed tertiary education 7.9 9.9 13.2 11.8 12.8 17.9 2.0 3.2 4.3 0.5 0.4 0.9 Average years of schooling of household head 7.9 8.4 8.9 10.0 10.0 10.6 5.3 5.9 6.3 3.1 3.5 3.1 Average years of schooling by gender Male 6.1 6.4 6.9 7.7 7.6 8.3 5.5 5.0 5.4 2.6 2.9 2.9 Female 6.5 6.8 7.4 8.1 8.0 8.8 4.7 5.4 5.9 2.0 2.1 2.3 School assistance by age 6-11 95.3 94.9 97.7 98.2 97.3 99.4 95.2 95.5 98.5 83.7 82.0 89.4 12-17 78.0 83.3 85.8 90.4 90.4 90.8 68.1 79.6 84.2 53.9 54.2 64.9 18-22 35.3 40.0 35.5 46.8 47.7 40.1 20.6 29.1 32.1 5.9 12.4 13.9 92 Table A.3.1.2 School attendance ratios and out-of-school by level, according to background characteristics, [1997] Post- Primary Secondary secondary Proportion Proportion Gross Net Gross Net Gross of out-of- of out-of- Attendance Attendance Attendance Attendance Attendance school school Total 107.79 96.74 2.46 73.87 80.11 0.32 24.07 Gender Boys 108.76 96.88 2.59 71.00 77.55 0.41 20.34 Girls 106.75 96.59 2.33 76.90 82.70 0.23 27.67 Area of residence Urban 103.95 95.81 2.92 96.63 89.57 0.27 34.57 Rural 111.44 97.67 2.00 50.30 66.68 0.40 7.81 Residence and gender Urban - Boys 102.02 95.36 3.84 98.40 89.29 0.55 32.00 Urban - Girls 106.03 96.29 1.94 94.97 89.83 0.00 36.83 Rural - Boys 115.31 98.43 1.31 45.63 61.75 0.23 4.47 Rural - Girls 107.43 96.88 2.71 55.84 71.97 0.58 11.59 Household wealth Quintile 1 113.91 98.42 1.45 22.32 38.42 1.04 1.12 Quintile 2 109.39 97.55 2.45 69.29 77.98 0.00 5.76 Quintile 3 107.44 96.85 2.37 92.01 86.43 0.21 15.10 Quintile 4 102.10 95.43 2.89 102.42 94.04 0.25 34.75 Quintile 5 97.63 92.89 4.38 104.01 94.09 0.31 59.52 Household wealth and gender 93 Quintile 1 - Boys 116.15 98.73 1.03 21.52 34.55 0.58 0.75 Quintile 2 - Boys 113.64 98.17 1.83 63.58 72.73 0.00 4.52 Quintile 3 - Boys 109.55 97.53 1.58 86.41 83.31 0.42 11.39 Quintile 4 - Boys 99.79 95.21 3.86 104.45 95.08 0.47 28.93 Quintile 5 - Boys 91.92 91.15 7.85 109.86 94.56 0.67 61.79 Quintile 1 - Girls 111.48 98.09 1.91 23.31 42.81 1.56 1.59 Quintile 2 - Girls 105.34 96.96 3.04 75.38 83.10 0.00 7.15 Quintile 3 - Girls 105.21 96.13 3.19 97.72 89.47 0.00 19.05 Quintile 4 - Girls 105.01 95.70 1.66 100.17 92.90 0.00 38.96 Quintile 5 - Girls 103.18 94.51 1.11 99.35 93.69 0.00 57.66 Gender of the household head Male 107.07 96.97 2.21 71.92 79.61 0.41 23.94 Female 111.31 95.57 3.68 81.16 81.93 0.00 24.46 Education of the household head No education 113.58 98.00 1.79 45.30 62.35 0.44 8.17 Incomplete primary 105.95 96.12 2.84 91.87 86.07 0.20 22.53 Primary 110.49 97.81 1.97 70.22 78.93 0.21 17.22 Incomplete secondary 97.37 94.90 3.31 98.33 93.77 1.77 55.92 Secondary 101.22 95.24 3.37 104.43 92.58 0.00 31.83 Some higher 96.50 93.74 3.57 108.25 94.75 0.00 68.91 Geographic areas Urban 103.95 95.81 2.92 96.63 89.57 0.27 34.57 Rural 110.64 97.34 2.25 57.85 73.33 0.26 9.29 Indigenous 114.30 98.98 1.02 20.01 32.88 1.12 0.66 94 Table A.3.1.3 Percentage of the population that has ever attended school, by age according to background characteristics, [1997] Age Age 6 Age 7 Age 8 Age 9 Age 10 Age 11 Age 12 Age 13 Age 14 Age 15 Age 16 Age 17 Total 95.61 46.69 13.77 9.85 7.30 6.44 4.92 4.67 3.14 2.68 1.91 3.12 Gender Boys 96.19 44.75 13.09 10.45 6.80 7.91 4.70 4.87 2.36 1.71 2.07 3.47 Girls 95.04 48.84 14.50 9.18 7.88 5.06 5.17 4.48 3.99 3.55 1.72 2.71 Area of residence Urban 96.34 42.48 7.61 3.63 4.13 1.44 2.19 1.50 0.90 1.17 0.78 1.45 Rural 94.97 50.98 19.33 16.06 10.40 11.19 7.20 8.34 5.27 4.33 3.14 5.05 Residence and gender Urban - Boys 98.00 40.54 10.41 3.53 4.18 1.09 1.62 0.00 0.85 1.68 0.00 2.83 Urban - Girls 94.60 44.61 4.58 3.73 4.07 1.78 2.72 2.64 0.97 0.80 1.57 0.00 Rural - Boys 94.52 49.02 15.56 17.55 9.28 14.66 6.93 9.25 3.99 1.74 4.04 4.13 Rural - Girls 95.40 53.18 23.26 14.46 11.71 8.04 7.53 7.20 6.48 7.18 1.91 6.24 Household wealth Quintile 1 95.99 57.55 30.85 18.10 11.73 14.45 9.34 14.40 10.14 6.49 5.76 8.70 Quintile 2 96.38 49.58 8.12 13.65 9.00 5.17 5.32 4.21 0.92 3.16 0.00 2.43 Quintile 3 93.60 38.36 6.16 2.90 3.96 3.09 2.85 0.00 0.00 1.47 2.12 2.83 Quintile 4 95.81 42.09 7.17 0.89 4.90 0.00 1.91 2.64 1.11 0.87 0.98 1.00 Quintile 5 96.07 35.17 1.94 4.67 4.83 4.89 0.00 1.80 0.00 0.00 0.00 0.00 Household wealth and gender Quintile 1 - Boys 94.74 55.47 28.17 20.20 9.66 16.73 9.67 16.93 7.94 4.83 5.13 8.45 Quintile 2 - Boys 97.49 46.18 9.92 14.24 7.03 10.23 4.20 4.10 1.85 0.00 0.00 2.28 95 Quintile 3 - Boys 94.63 38.45 9.29 3.84 4.13 1.82 3.47 0.00 0.00 0.00 4.19 5.29 Quintile 4 - Boys 100.00 38.33 1.94 1.44 6.28 0.00 0.00 0.00 0.00 1.74 0.00 0.00 Quintile 5 - Boys 96.15 34.80 3.78 4.06 5.90 5.78 0.00 0.00 0.00 0.00 0.00 0.00 Quintile 1 - Girls 97.65 59.83 33.43 15.91 13.89 12.05 8.96 11.32 12.18 8.74 6.75 9.05 Quintile 2 - Girls 95.73 53.45 5.77 13.08 11.50 1.62 6.46 4.33 0.00 6.13 0.00 2.60 Quintile 3 - Girls 92.28 38.25 3.49 2.01 3.80 4.66 2.11 0.00 0.00 2.69 0.00 0.00 Quintile 4 - Girls 92.09 47.28 13.73 0.00 2.58 0.00 3.89 5.15 2.68 0.00 2.16 2.08 Quintile 5 - Girls 95.99 35.47 0.00 5.51 3.75 4.23 0.00 3.03 0.00 0.00 0.00 0.00 Gender of the household head Male 95.96 47.44 13.41 10.54 7.65 6.18 5.29 4.81 2.89 2.81 2.28 2.67 Female 93.87 43.19 15.55 6.41 5.52 7.75 3.51 4.17 4.23 2.24 0.56 4.81 Education of the household head No education 97.20 49.68 25.03 16.65 11.73 14.53 8.81 7.02 7.64 4.50 4.00 5.97 Incomplete primary 97.14 44.74 7.69 4.75 5.72 3.07 3.64 2.69 0.66 1.41 0.00 3.59 Primary 93.69 54.68 13.06 9.97 5.49 3.41 3.27 4.68 0.00 2.76 2.85 1.94 Incomplete secondary 100.00 49.99 10.07 0.00 2.62 6.58 0.00 0.00 0.00 0.00 0.00 0.00 Secondary 89.04 34.56 6.16 6.76 0.00 0.00 0.00 3.29 1.85 2.02 0.00 0.00 Some higher 96.46 34.24 0.00 3.59 10.65 0.00 5.25 4.27 0.00 0.00 0.00 0.00 Geographic areas Urban 96.34 42.48 7.61 3.63 4.13 1.44 2.19 1.50 0.90 1.17 0.78 1.45 Rural 94.55 48.39 12.37 14.68 6.82 8.63 4.68 6.77 1.50 1.94 1.63 2.28 Indigenous 96.34 59.65 41.25 21.91 22.47 22.49 14.83 17.21 18.49 13.97 10.11 17.33 96 Table A.3.1.4 Proportion of 20-29 years olds that completed each grade, according to background characteristics, [1997] 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Total 87.78 84.79 78.45 69.58 66.82 62.35 43.20 38.68 33.29 Gender Boys 87.55 84.60 76.90 66.93 64.31 59.96 39.07 35.38 31.01 Girls 88.00 84.97 79.92 72.11 69.20 64.63 47.13 41.81 35.45 Area of residence Urban 94.20 91.49 84.55 74.75 71.77 66.17 57.77 51.50 43.79 Rural 77.53 74.10 68.70 61.33 58.91 56.25 19.96 18.22 16.53 Residence and gender Urban - Boys 93.98 91.39 82.91 71.27 68.35 62.43 53.76 48.35 41.68 Urban - Girls 94.40 91.58 86.02 77.84 74.80 69.49 61.33 54.30 45.66 Rural - Boys 78.17 74.71 68.14 60.59 58.42 56.35 17.67 16.49 15.46 Rural - Girls 76.84 73.45 69.30 62.12 59.43 56.15 22.38 20.06 17.66 Household wealth Quintile 1 62.18 59.04 54.34 50.49 49.19 48.06 5.87 5.37 5.06 Quintile 2 86.51 80.99 70.65 58.27 55.35 51.05 21.96 20.09 19.00 Quintile 3 92.70 89.18 79.92 69.03 65.86 58.42 40.24 36.96 33.48 Quintile 4 95.03 92.82 86.94 76.77 72.29 67.29 59.24 51.72 44.47 Quintile 5 95.94 95.22 93.86 87.88 86.30 82.66 78.03 69.78 56.49 Household wealth and gender Quintile 1 - Boys 68.80 66.09 60.33 57.14 55.59 54.52 7.71 7.38 6.79 Quintile 2 - Boys 84.27 79.40 66.95 53.30 50.98 46.43 18.07 16.58 15.74 Quintile 3 - Boys 89.73 85.57 75.88 63.04 61.02 54.56 35.28 32.03 29.66 97 Quintile 4 - Boys 95.31 93.33 86.04 74.03 68.68 64.63 54.41 51.19 44.99 Quintile 5 - Boys 97.31 96.64 94.25 87.65 85.93 80.94 78.47 68.36 56.37 Quintile 1 - Girls 54.75 51.15 47.62 43.04 42.02 40.82 3.79 3.12 3.12 Quintile 2 - Girls 88.87 82.66 74.55 63.52 59.96 55.92 26.07 23.78 22.43 Quintile 3 - Girls 95.74 92.87 84.06 75.17 70.82 62.37 45.32 42.00 37.39 Quintile 4 - Girls 94.80 92.39 87.71 79.10 75.35 69.56 63.36 52.16 44.03 Quintile 5 - Girls 94.85 94.10 93.55 88.06 86.59 84.03 77.68 70.91 56.59 Gender of the household head Male 86.83 83.83 77.66 68.77 65.87 61.56 40.60 36.71 31.30 Female 91.04 88.08 81.17 72.39 70.09 65.09 52.16 45.47 40.15 Education of the household head No education 68.03 64.62 59.70 52.43 50.42 47.86 21.00 19.13 17.20 Incomplete primary 95.60 91.10 77.62 62.09 56.60 48.87 40.17 35.98 30.46 Primary 92.41 89.45 83.42 76.18 74.59 70.97 31.94 28.93 25.66 Incomplete secondary 98.91 98.44 96.61 93.65 91.70 88.66 83.25 65.79 49.61 Secondary 97.66 95.84 93.14 84.05 81.37 75.80 72.04 67.98 60.88 Some higher 97.22 95.91 94.06 89.56 87.89 84.36 78.47 69.68 58.70 Geographic areas Urban 94.20 91.49 84.55 74.75 71.77 66.17 57.77 51.50 43.79 Rural 84.23 80.80 75.13 67.17 64.51 61.55 22.57 20.55 18.59 Indigenous 41.81 38.35 34.43 30.21 29.07 27.99 6.03 5.78 5.53 98 Table A.3.1.5 Inequality in years of schooling, [1997] Gini coefficient Theil index 15-19 20-24 25-29 15+ 15-19 20-24 25-29 15+ Total 39.22 37.60 39.30 48.19 17.09 13.76 16.10 17.84 Gender Boys 39.74 38.60 39.61 48.76 16.98 14.65 16.41 18.31 Girls 38.50 36.53 38.81 47.58 17.05 12.85 15.61 17.35 Area of residence Urban 37.94 32.26 34.56 41.20 20.01 13.83 15.47 18.50 Rural 39.31 43.33 43.53 55.49 12.87 12.25 14.63 13.79 Residence and gender Urban - Boys 39.26 33.69 35.13 41.93 21.04 15.24 15.95 19.45 Urban - Girls 36.41 30.83 33.87 40.55 18.80 12.55 14.88 17.67 Rural - Boys 38.40 42.61 43.10 54.23 12.21 12.32 14.78 13.28 Rural - Girls 40.27 43.99 43.89 56.83 13.69 12.13 14.34 14.28 Household wealth Quintile 1 42.91 50.30 49.17 61.25 9.13 9.09 9.75 8.95 Quintile 2 39.91 41.45 42.14 52.34 17.66 17.18 19.17 16.55 Quintile 3 37.30 34.86 37.61 46.06 20.66 15.66 17.19 17.91 Quintile 4 35.98 31.49 31.86 41.58 19.31 13.17 14.02 17.35 Quintile 5 32.34 23.63 26.68 35.09 15.49 7.62 9.15 14.50 Household wealth and gender Quintile 1 - Boys 40.91 44.29 44.79 57.42 8.58 8.98 9.66 8.68 Quintile 2 - Boys 40.55 42.59 44.35 52.17 17.45 16.80 20.08 15.95 99 Quintile 3 - Boys 39.59 38.53 39.78 47.02 21.81 17.41 17.44 18.68 Quintile 4 - Boys 34.97 32.39 31.61 42.03 18.54 15.33 13.72 17.77 Quintile 5 - Boys 32.52 22.22 26.38 35.39 16.91 7.44 10.06 14.79 Quintile 1 - Girls 45.67 57.83 53.16 65.80 9.94 9.21 9.56 9.34 Quintile 2 - Girls 38.97 40.02 39.38 52.32 17.76 17.46 17.83 16.94 Quintile 3 - Girls 34.84 30.69 35.25 45.07 19.47 13.85 16.68 17.14 Quintile 4 - Girls 35.72 30.38 31.99 41.11 18.81 11.53 14.26 16.93 Quintile 5 - Girls 31.99 24.59 26.65 34.81 14.29 7.76 8.21 14.25 Gender of the household head Male 39.74 38.32 40.05 48.45 17.08 13.68 16.23 17.56 Female 37.47 34.99 36.18 47.15 17.09 13.82 15.48 18.85 Education of the household head No education 45.04 51.37 52.67 70.85 13.52 14.85 15.34 15.83 Incomplete primary 36.12 37.02 39.95 41.61 19.19 18.47 22.73 23.92 Primary 37.50 32.90 31.24 30.36 18.57 12.24 13.03 10.61 Incomplete secondary 35.24 19.93 19.91 24.24 18.97 6.13 6.55 8.49 Secondary 31.96 25.01 22.24 24.22 16.30 10.96 8.91 10.71 Some higher 31.57 23.43 25.96 25.40 15.99 7.60 10.64 10.97 Geographic areas Urban 37.94 32.26 34.56 41.20 20.01 13.83 15.47 18.50 Rural 33.95 38.16 38.66 52.14 12.92 11.99 14.40 13.58 Indigenous 61.82 67.97 72.41 76.26 11.94 13.87 17.30 15.82 100 Table A.3.1.6 School attendance ratios and out-of-school by level, according to background characteristics, [2008] Post- Primary Secondary secondary Proportion Proportion Gross Net Gross Net Gross of out-of- of out-of- Attendance Attendance Attendance Attendance Attendance school school Total 107.31 97.94 1.43 83.68 85.45 0.00 26.21 Gender Boys 108.51 97.98 1.16 80.18 83.30 0.00 22.66 Girls 105.96 97.89 1.73 87.41 87.66 0.00 29.77 Area of residence Urban 104.72 97.91 1.27 93.26 89.91 0.00 31.57 Rural 110.75 97.98 1.63 70.37 78.30 0.00 14.32 Residence and gender Urban - Boys 104.23 98.03 0.77 90.83 90.14 0.00 27.75 Urban - Girls 105.30 97.76 1.86 95.76 89.69 0.00 35.43 Rural - Boys 114.46 97.90 1.71 66.01 72.92 0.00 11.13 Rural - Girls 106.80 98.06 1.55 75.27 84.21 0.00 17.43 Household wealth Quintile 1 113.11 97.48 2.02 55.04 66.54 0.00 4.90 Quintile 2 107.01 98.65 1.04 85.13 87.09 0.00 12.31 Quintile 3 105.08 97.72 1.62 91.65 89.83 0.00 22.21 Quintile 4 104.06 98.06 1.34 99.01 93.16 0.00 35.48 Quintile 5 100.47 97.80 0.40 109.01 95.20 0.00 51.60 Household wealth and gender 101 Quintile 1 - Boys 118.17 97.53 1.57 53.10 60.77 0.00 5.04 Quintile 2 - Boys 105.17 98.88 0.70 82.62 87.20 0.00 8.01 Quintile 3 - Boys 105.90 97.86 1.16 85.37 89.48 0.00 16.68 Quintile 4 - Boys 104.95 97.58 1.87 94.10 91.11 0.00 28.72 Quintile 5 - Boys 101.23 97.88 0.00 109.63 95.09 0.00 48.67 Quintile 1 - Girls 108.07 97.42 2.46 57.26 73.43 0.00 4.75 Quintile 2 - Girls 109.29 98.37 1.45 87.41 86.99 0.00 16.45 Quintile 3 - Girls 104.25 97.57 2.08 99.32 90.22 0.00 27.94 Quintile 4 - Girls 102.95 98.66 0.69 104.26 95.25 0.00 41.44 Quintile 5 - Girls 99.48 97.70 0.93 108.39 95.31 0.00 55.00 Gender of the household head Male 106.17 97.79 1.57 82.89 85.54 0.00 25.78 Female 110.59 98.36 1.02 85.57 85.25 0.00 27.27 Education of the household head No education 115.75 98.20 1.39 59.19 68.95 0.00 9.36 Incomplete primary 104.49 98.03 1.28 93.59 88.86 0.00 17.56 Primary 106.64 97.42 1.95 80.63 85.10 0.00 22.54 Incomplete secondary 103.21 97.47 1.79 99.94 94.28 0.00 55.44 Secondary 103.56 98.81 0.51 96.63 93.18 0.00 30.33 Some higher 101.71 96.99 1.95 103.48 94.45 0.00 57.57 Geographic areas Urban 104.72 97.91 1.27 93.26 89.91 0.00 31.57 Rural 109.17 98.05 1.53 80.28 85.08 0.00 16.81 Indigenous 114.95 97.76 1.94 37.36 48.57 0.00 2.88 102 Table A.3.1.7 Percentage of the population that has ever attended school, by age according to background characteristics, [2008] edad en años Age 6 Age 7 Age 8 Age 9 Age 10 Age 11 Age 12 Age 13 Age 14 Age 15 Age 16 Age 17 Total 95.67 32.60 8.15 6.15 4.75 3.77 1.27 2.22 1.28 2.12 1.18 1.78 Gender Boys 95.33 28.33 7.35 5.31 5.07 3.06 0.29 2.71 1.08 2.37 1.33 1.98 Girls 96.01 36.88 9.20 6.93 4.32 4.59 2.14 1.76 1.52 1.84 1.00 1.53 Area of residence Urban 94.64 24.56 8.08 5.86 5.53 3.54 1.36 1.07 0.57 0.00 0.16 0.94 Rural 97.19 43.00 8.24 6.56 3.84 4.11 1.14 3.67 2.18 5.28 2.64 3.07 Residence and gender Urban - Boys 94.28 18.29 5.85 5.15 5.18 2.10 0.00 2.00 0.00 0.00 0.30 1.64 Urban - Girls 95.06 30.89 11.25 6.52 6.04 5.26 2.41 0.20 1.25 0.00 0.00 0.00 Rural - Boys 97.06 41.38 9.32 5.53 4.93 4.51 0.63 3.60 2.45 5.39 2.83 2.54 Rural - Girls 97.31 44.62 6.95 7.49 2.43 3.67 1.70 3.74 1.86 5.14 2.42 3.67 Household wealth Quintile 1 97.20 47.16 12.74 13.41 6.98 6.09 2.54 5.32 3.38 5.62 3.38 5.96 Quintile 2 98.05 33.44 9.98 5.28 2.98 2.35 2.14 2.15 0.00 1.66 0.45 0.00 Quintile 3 91.85 19.11 5.45 1.61 5.26 2.48 0.00 0.98 0.00 1.29 0.00 1.13 Quintile 4 97.46 28.55 3.79 0.00 3.49 4.44 0.00 0.00 0.00 0.00 0.00 0.00 Quintile 5 90.56 20.05 1.30 5.38 2.93 1.99 0.00 0.00 2.47 0.00 2.59 0.00 Household wealth and gender Quintile 1 - Boys 95.36 45.56 12.86 8.73 8.37 5.38 0.75 4.21 3.46 5.66 4.49 6.25 Quintile 2 - Boys 98.35 35.63 10.50 9.11 2.81 1.90 0.22 4.92 0.00 1.08 0.92 0.00 103 Quintile 3 - Boys 91.92 12.59 3.63 3.33 7.36 2.83 0.00 2.02 0.00 2.52 0.00 1.68 Quintile 4 - Boys 98.20 15.75 0.00 0.00 3.26 2.77 0.00 0.00 0.00 0.00 0.00 0.00 Quintile 5 - Boys 89.31 8.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.43 0.00 Quintile 1 - Girls 98.73 48.83 12.60 17.29 5.39 6.87 4.47 6.36 3.27 5.56 2.18 5.62 Quintile 2 - Girls 97.70 31.21 8.83 1.80 3.22 2.86 3.42 0.00 0.00 2.43 0.00 0.00 Quintile 3 - Girls 91.80 25.26 7.27 0.00 2.30 2.12 0.00 0.00 0.00 0.00 0.00 0.00 Quintile 4 - Girls 96.41 40.21 8.44 0.00 3.83 6.93 0.00 0.00 0.00 0.00 0.00 0.00 Quintile 5 - Girls 92.21 33.36 3.15 11.46 7.64 4.51 0.00 0.00 4.28 0.00 3.87 0.00 Gender of the household head Male 96.13 33.36 7.40 6.90 5.45 4.51 1.41 2.80 1.83 2.46 1.61 2.15 Female 94.42 30.39 10.21 3.70 2.83 1.50 0.93 0.90 0.00 1.30 0.00 0.89 Education of the household head No education 97.59 49.80 9.23 10.77 7.29 6.20 1.89 5.83 3.47 4.52 4.06 5.98 Incomplete primary 95.14 28.64 12.85 6.25 3.11 3.43 2.68 2.33 1.39 0.00 0.00 0.00 Primary 97.22 34.67 6.19 4.28 4.53 0.91 0.80 1.18 0.00 3.26 0.35 1.89 Incomplete secondary 93.62 21.45 6.90 10.81 0.00 3.04 0.00 0.00 0.00 0.00 0.00 0.00 Secondary 94.02 21.38 4.12 1.97 3.68 6.36 0.00 0.00 0.00 1.77 0.61 0.00 Some higher 91.07 19.16 2.12 2.76 6.91 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Geographic areas Urban 94.64 24.56 8.08 5.86 5.53 3.54 1.36 1.07 0.57 0.00 0.16 0.94 Rural 96.91 38.15 4.66 2.70 2.37 0.85 0.07 1.28 0.00 3.95 0.72 0.21 Indigenous 97.79 57.45 16.62 17.03 7.81 14.45 4.03 10.46 9.28 11.09 9.81 14.03 104 Table A.3.1.8 Proportion of 20-29 years olds that completed each grade, according to background characteristics, [2008] 1 year 2 years 3 years 4 years 5 years 6 years 7 years 8 years 9 years Total 92.50 89.72 84.88 74.84 71.46 65.84 54.62 49.44 43.88 Gender Boys 92.75 89.02 84.14 72.67 69.74 63.96 51.31 45.91 40.92 Girls 92.24 90.42 85.63 77.03 73.18 67.74 57.96 53.00 46.87 Area of residence Urban 96.69 94.30 89.82 79.62 75.97 70.06 64.02 57.71 51.00 Rural 83.01 79.36 73.74 64.06 61.25 56.33 33.38 30.76 27.81 Residence and gender Urban - Boys 96.85 93.65 89.06 77.25 73.90 67.85 61.04 54.11 47.78 Urban - Girls 96.53 94.96 90.58 82.00 78.06 72.29 67.03 61.34 54.25 Rural - Boys 83.45 78.54 73.01 62.30 60.33 55.17 29.28 27.35 25.40 Rural - Girls 82.57 80.19 74.47 65.83 62.18 57.49 37.51 34.19 30.23 Household wealth Quintile 1 72.16 67.51 61.46 49.80 47.31 42.84 14.02 12.97 12.51 Quintile 2 91.94 88.33 80.13 68.06 64.01 55.12 38.11 35.48 31.00 Quintile 3 96.59 92.92 87.70 75.25 69.06 61.55 54.80 49.60 44.95 Quintile 4 98.37 96.32 92.02 85.38 82.63 78.36 72.64 65.68 58.19 Quintile 5 98.54 97.96 96.81 88.77 87.21 84.15 81.64 72.89 63.69 Household wealth and gender Quintile 1 - Boys 76.28 71.16 66.36 53.49 50.53 45.76 15.56 14.80 13.90 Quintile 2 - Boys 91.05 85.75 78.16 64.98 61.75 53.44 33.40 31.44 28.41 Quintile 3 - Boys 96.18 90.52 83.50 70.82 65.06 57.48 49.38 44.82 41.49 105 Quintile 4 - Boys 97.63 95.06 90.40 82.30 79.75 73.18 66.84 59.75 52.41 Quintile 5 - Boys 98.38 97.36 96.06 85.05 84.24 81.83 78.02 67.55 58.95 Quintile 1 - Girls 67.93 63.77 56.43 46.02 44.01 39.84 12.45 11.09 11.09 Quintile 2 - Girls 92.80 90.82 82.04 71.04 66.20 56.75 42.66 39.38 33.51 Quintile 3 - Girls 96.96 95.06 91.47 79.22 72.66 65.19 59.66 53.88 48.06 Quintile 4 - Girls 99.08 97.53 93.57 88.34 85.39 83.34 78.21 71.39 63.74 Quintile 5 - Girls 98.72 98.68 97.70 93.22 90.77 86.94 85.97 79.29 69.38 Gender of the household head Male 91.94 89.27 84.33 73.88 70.50 65.18 52.44 47.26 42.54 Female 93.96 90.90 86.33 77.38 73.97 67.59 60.36 55.17 47.41 Education of the household head No education 69.19 66.28 60.17 54.59 52.42 49.36 30.27 27.71 25.68 Incomplete primary 96.77 91.98 81.96 60.32 54.92 44.62 39.19 36.23 32.54 Primary 95.26 91.46 87.70 76.61 73.91 68.31 41.17 35.77 30.90 Incomplete secondary 99.22 98.85 96.68 93.28 89.38 86.24 85.78 68.91 54.86 Secondary 99.08 97.73 96.08 90.63 86.81 81.81 79.42 75.24 68.74 Some higher 99.49 99.49 99.04 97.79 96.55 94.55 93.63 86.22 76.85 Geographic areas Urban 96.69 94.30 89.82 79.62 75.97 70.06 64.02 57.71 51.00 Rural 90.78 86.91 81.04 70.38 67.38 61.87 38.56 35.61 32.13 Indigenous 45.50 42.91 38.49 33.54 31.69 29.59 8.42 7.35 6.93 106 Table A.3.1.9 Inequality in years of schooling, [2008] Gini coefficient Theil index 15-19 20-24 25-29 15+ 15-19 20-24 25-29 15+ Total 37.46 32.87 37.47 43.81 18.99 13.91 16.15 17.85 Gender Boys 38.40 33.07 38.49 44.24 19.46 14.64 17.28 18.42 Girls 36.26 32.57 36.23 43.28 18.35 13.10 14.85 17.18 Area of residence Urban 36.09 29.79 31.96 38.42 19.97 13.25 14.77 17.59 Rural 38.30 38.80 47.58 52.06 16.00 14.70 18.42 15.94 Residence and gender Urban - Boys 37.59 30.54 32.50 38.97 21.44 14.20 15.52 18.52 Urban - Girls 34.25 28.90 31.17 37.79 18.19 12.19 13.87 16.61 Rural - Boys 38.58 37.87 49.01 51.63 15.29 14.95 20.48 15.88 Rural - Girls 37.89 39.54 45.85 52.47 16.81 14.41 16.08 15.96 Household wealth Quintile 1 42.05 45.94 54.44 58.02 15.65 16.78 14.59 13.69 Quintile 2 35.05 34.76 40.39 45.83 17.92 16.02 18.25 17.78 Quintile 3 36.74 31.48 34.35 42.24 21.22 15.17 16.97 18.40 Quintile 4 34.39 24.78 27.26 37.67 19.56 10.13 12.49 16.36 Quintile 5 32.64 24.17 24.40 30.82 16.56 9.21 9.57 12.76 Household wealth and gender Quintile 1 - Boys 42.55 41.87 52.29 56.45 14.54 15.65 15.56 13.37 Quintile 2 - Boys 34.78 35.89 42.45 46.59 17.87 17.28 19.62 18.66 107 Quintile 3 - Boys 39.58 33.89 36.25 43.27 23.55 17.07 19.61 19.26 Quintile 4 - Boys 34.31 27.43 27.45 38.37 19.75 12.20 12.16 16.95 Quintile 5 - Boys 33.27 24.03 28.07 32.09 16.88 9.37 12.24 13.96 Quintile 1 - Girls 41.38 49.83 56.73 59.68 16.90 18.01 13.31 14.05 Quintile 2 - Girls 34.89 33.62 38.15 45.09 17.54 14.82 16.89 16.94 Quintile 3 - Girls 32.53 29.08 32.51 41.08 18.08 13.38 14.65 17.41 Quintile 4 - Girls 34.18 21.89 26.75 36.84 19.13 7.96 12.69 15.65 Quintile 5 - Girls 31.88 24.10 19.05 29.37 16.17 8.92 5.99 11.44 Gender of the household head Male 37.85 33.76 37.53 44.10 18.90 14.33 15.87 17.77 Female 36.46 30.47 37.26 42.96 19.19 12.74 16.94 18.03 Education of the household head No education 42.55 45.92 60.26 72.21 16.05 15.96 16.58 17.97 Incomplete primary 36.83 36.68 41.52 41.93 20.85 19.61 25.45 25.67 Primary 34.95 30.73 35.26 29.87 17.80 13.55 15.85 11.56 Incomplete secondary 33.10 21.41 18.66 23.40 18.73 7.21 6.21 8.48 Secondary 34.90 21.25 18.25 21.74 19.82 8.35 7.83 9.41 Some higher 32.42 19.95 15.42 20.80 17.01 6.20 4.48 7.44 Geographic areas Urban 36.09 29.79 31.96 38.42 19.97 13.25 14.77 17.59 Rural 34.54 33.79 41.61 48.14 16.09 14.58 18.23 15.77 Indigenous 51.33 60.12 76.37 73.27 15.25 12.94 19.82 16.23 108