Report No. 18801 PAN Panama Poverty Assessment Priorities and Strategies for Poverty Reductioi Volume I Main Report June 28, 1999 Human Development Department Latin America and the Caribbean Region The World Bank Document of the World Bank CURRENCY EQUIVALENTS US1$ = I Balboa FISCAL YEAR January I - December 31 ACCRONYMS AND ABBREVIATIONS BHN National Mortgage Bank CAS Country Assistance Strategy CBN Cost of Basic Needs method for calculating poverty CEFACEI Community and Family Centers for Initial Education CSS Social Security Institute DEC Directorate of Statistics and the Census FDI Institutional Development Fund (Fondo de Desarrollo Institucional) FES Emergency Social Fund GDP Gross Domestic Product GOP Government of Panama HFA Height for age IDDAN National Institute of Aqueducts and Sewers (Instituto de Acueductos y Alcantarillados Nacionales) IDB Inter-American Development Bank ILO International Labor Organization LAC Latin America and the Caribbean LSMS Living Standards Measurement Survey MEF Ministry of Economy and Finance (formerly MIPPE and Hacienda) MIDA Ministry of Agricultural Development MINEDUC Ministry of Education MINSA Ministry of Health MINTRAB Ministry of Labor MIPPE Ministry of Planning and Economic Policy (now Ministry of Economics and Finance) MIVI Ministry of Housing MJMNF Ministry of Youth, Women, Children and Family NFPS Non-financial public sector NGOs Non-Governmental Organizations PARVIS Program for Rapid Support of Social Housing PM Poverty Map PROINLO Local Investmnent Program SCQS Social Capital Qualitative Survey WFA Weight for age WFH Weight for height UNDP United Nations Development Program Vice President: Shahid Javed Burki Country Director: Donna Dowsett-Coirolo Lead Economist: Ian Bannon Sector Leader: Ana-Maria Arriagada Task Manager: Kathy Lindert PANAMA POVERTY ASSESSMENT: PRIORITIES AND STRATEGIES FOR POVERTY REDUCTION TABLE OF CONTENTS VOLUME 1: MAIN REPORT EXECUTIVE SUMMARY INTRODUCTION 1 1. POVERTY, MALNUTRITION AND INEQUALITY IN PANAMA 2 A. MEASURING POVERTY AND WELFARE 2 B. POVERTY IN PANAMA 3 C. MALNUTRITION IN PANAMA 7 D. INEQUALITY IN PANAMA 8 2. UNEQUAL OPPORTUNITIES: THE ASSETS OF THE POOR 10 A. LABOR 10 B. EDUCATION 12 C. HEALTH 15 D. PHYSICAL ASSETS 17 E. BASIC INFRASTRUCTURE 18 F. FINANCIAL ASSETS 20 G. SOCIAL CAPITAL 21 3. THE DETERMINANTS OF POVERTY, MALNUTRITION, AND INEQUALITY 23 A. PROBABILITY OF BEING POOR / CORRELATES OF POVERTY 23 B. DETERMINANTS OF NUTRITIONAL STATUS 25 C. SOURCES OF INEQUALITY 26 4. POVERTY, DUALITY AND DISTORTIONS IN PANAMA'S ECONOMY 27 A. THE LEGACY OF DISTORTIONS AND DUALITY IN PANAMA'S ECONOMY 27 B. RECENT REFORMS: REDUCING DISTORTIONS IN PANAMA'S ECONOMY 29 C. CONSOLIDATING ECONOMIC REFORMS 29 5. POVERTY AND SOCIAL SPENDING 31 A. INEQUITIES AND INEFFICIENCIES IN THE PROVISION OF PUBLIC EDUCATION 31 B. INEQUITIES AND INEFFICIENCIES IN THE PROVISION OF PUBLIC HEALTH CARE 34 C. PROTECTING THE POOR: NUTRITION & SOCIAL ASSISTANCE INTERVENTIONS 36 6. PRINCIPLES, PRIORITIES, AND ACTIONS TO REDUCE POVERTY 39 A. THE MAIN PRINCIPLES OF THE GOVERNMENT'S NEW POVERTY STRATEGY 39 B. PRIORITIES FOR TRANSLATING THESE PRINCIPLES INTO ACTION 39 C. ROAD MAP OF KEY ACTIONS FOR POVERTY REDUCTION 45 REFERENCES OF MAIN REPORT TEXT TABLES: 1.1 DISTRIBUTION AND INCIDENCE OF POVERTY IN 1997, BY GEOGRAPHIC AREA 3 1.2 DEPTH AND SEVERITY OF POVERTY IN 1997, BY GEOGRAPHIC AREA 5 1.3 REGIONAL DISTRIBUTION OF POVERTY IN PANAMA 7 1.4 INTERNATIONAL COMPARISON OF INEQUALITY USING THE GINI INDEX 9 2.1 HEALTH SERVICES BY POVERTY GROUP AND GEOGRAPHIC AREA 17 3.1 DECOMPOSITION OF INEQUALITY IN PANAMA 26 5.1 PUBLIC SPENDING ON EDUCATION 32 5.2 PUBLIC SPENDING ON HEALTH 35 TEXT BOXES: 1. I PAST STUDIES OF POVERTY AND INEQUALITY IN PANAMA 4 1.2 GEOGRAPHY AND ETHNICITY IN THE LSMS 6 2.1 UNEMPLOYMENT IN PANAMA: DEFINITIONS AND OVERALL TRENDS I1 2.2 DISPARITIES IN EDUCATION: KEY GAPS FACING THE POOR IN PANAMA, BY GEOGRAPHIC AREA 13 2.3 THE LANGUAGE POVERTY TRAP: LANGUAGE ABILITY, EDUCATION AND POVERTY AMONG THE ETHNIC INDIGENOUS 13 2.4 ECONOMIC BARRIERS TO EDUCATION 14 2.5 POVERTY AND ACCESS TO TEXTBOOKS IN PANAMA: INADEQUACY OF A KEY INPUT 15 2.6 DISPARITIES IN BASIC INFRASTRUCTURE SERVICES: KEY GAPS FACING THE POOR IN PANAMA, BY GEOGRAPHIC AREA 19 2.7 INCOMPLETE COST RECOVERY FOR WATER AND ELECTRICITY 20 2.8 TRANSPORTATION BOTTLENECKS AND THE URBAN POOR 20 2.9 COMMUNITY PERCEPTIONS OF THE QUALITY OF LIFE: URBAN PESSIMISM & SOCIAL PROBLEMS 22 3.1 DETERMINANTS OF POVERTY, MALNUTRITION, AND INEQUALITY: SUMMARY OF KEY FACTORS 23 3.2 THE CORRELATION BETWEEN BASIC SERVICES AND POVERTY IN PANAMA 24 4.1 PANAMA'S LABOR MARKET: MULTIPLE AND STRINGENT POLICY REGIMES 28 4.2 REDISTRIBUTIVE EFFECTS OF TRADE PROTECTION ON HOUSEHOLDS 28 4.3 KEY ACHIEVEMENTS OF THE ECONOMIC REFORM PROGRAM 29 5.1 INVENTORY OF MAIN GOVERNMENT SOCIAL ASSISTANCE PROGRAMS 37 5.2 DECENTRALIZING SCHOOL FEEDING PROGRAMS IN REMOTE AREAS: THE POTENTIAL OF BLOCK GRANTS 38 6.1 COMMUNITY PRIORITIES BY GEOGRAPHIC AREA 39 6.2 PRIORITY POVERTY GROUPS 40 6.3 THE CHALLENGE OF POVERTY REDUCTION AMONG REMOTE, DISPERSED POPULATIONS 43 6.4 ROAD MAP OF KEY ACTIONS FOR POVERTY REDUCTION 45 TEXT FIGURES: 1.1 GEOGRAPHIC DISTRIBUTION OF POVERTY 4 1.2 ETHNICITY, GEOGRAPHY, AND POVERTY 5 1.3 POVERTY & MALNUTRITION BY GEOGRAPHIC AREA 8 1.4 THE POOREST ARE VERY POOR AND THE RICHEST VERY RICH 8 2.1 UNEMPLOYMENT RATES FOR VARIOUS GROUPS 10 2.2 INFORMAL EARNINGS ARE SUBSTANTIALLY L OWER I1 2.3 POVERTY AND THE LACK OF PROPERTY TITLING 18 2.4 POVERTY AND SOCIAL CAPITAL 21 5.1 PUBLIC SPENDING ON SOCIAL SECTORS IN PANAMA, 1997 31 5.2 WHO BENEFITS FROM EDUCATION SPENDING? 33 5.3 URBAN AREAS RECEIVE A DISPROPORTIONATE SHARE OF EDUCATION SPENDING 33 5.4 DISPARTITIES IN HOUSEHOLD SPENDING EXACERBATE INEQUITIES IN TOTAL SPENDING ON EDUCATION 34 5.5 WHO BENEFITS FROM SCHOOL FEEDING PROGRAMS? 36 VOLUME 2: ANNEXES 1. THE CONSUMPTION AGGREGATE 2. MEASURING POVERTY 3. THE INCOME AGGREGATE 4. STATISTICAL APPENDIX 5. AGRICULTURE AND RURAL POVERTY IN PANAMA, by Tom Wiens, Carlos Sobrado and Kathy Lindert 6. INDIGENOUS POVERTY IN PANAMA, by Renos Vakis and Kathy Lindert 7. URBAN POVERTY IN PANAMA, by Kathy Lindert 8. THE DETERMINANTS OF POVERTY IN PANAMA, by Carlos Sobrado 9. MALNUTRITION IN PANAMA: AN ANALYSIS OF LSMS ANTHROPOMETRIC DATA, by Carlos Sobrado, Kathy Lindert, and Gloria Rivera 10. INEQUALITY IN PANAMA, by Carlos Sobrado I1. POVERTY, VULNERABILITY AND LABOR MARKETS IN PANAMA, by Renos Vakis and Kathy Lindert 12. REDISTRIBUTIVE IMPACT OF AGRICULTURAL TRADE AND PRICING REFORMS, by Rafael Yrarrazaval, Kathy Lindert, Tom Wiens, and Carlos Sobrado 13. IMPROVING THE IMPACT OF PRE-SCHOOLS IN PANAMA, by Kinnon Scott 14. THE RETURNS TO EDUCATION, by Kathy Lindert 15. PATTERNS OF HEALTH CARE USE AND THE INCIDENCE OF PUBLIC HEALTH SPENDING IN PANAMA, by Kinnon Scott and Roberto Gonzalez 16. THE DISTRIBUTIONAL INCIDENCE OF SCHOOL FEEDING PROGRAMS, by Kathy Lindert 17. THE DISTRIBUTIONAL INCIDENCE OF THE MATERNAL-INFANT SUPPLEMENTAL FEEDING PROGRAM, by Kathy Lindert 18. COMMUNITY ORGANIZATION, VALUES AND SOCIAL CAPITAL IN PANAMA, by Maria-Valeria Junho Pena and Hector Lindo Fuentes 19. POVERTY AND FEMALE HEADED HOUSEHOLDS IN PANAMA: SUMMARY OF FINDINGS, by Nobuhiko Fuwa THE POVERTY ASSESSMENT TEAM This report is the outcome of a highly collaborative process by the Government of Panama and the World Bank. The process involved continuous cooperation and numerous missions by World Bank staff to Panama and MIPPE staff to Washington. The World Bank's task team included: Ana-Maria Arriagada (Sector Leader), Kathy Lindert (Task Manager), Kinnon Scott, Carlos Sobrado, Renos Vakis, Maria-Valeria Junho Pena, Hector Lindo-Fuentes, Tom Wiens, Nobuhiko Fuwa, and Rafael Yrarrazaval. Carlos Becerra and Enriqueta Davis provided invaluable inputs in Panama. Judy McGuire, Tilahun Temesgen, Amit Dar, and Saji Thomas also contributed. Lerick Kebeck provided administrative assistance and coordinated the production of the report. Norman Hicks and Jose Sokol were the Peer Reviewers. The team also benefited from the guidance of Ian Bannon (Lead Economist). The task team in the Ministry of Economics and Finance (MEF, formerly the Ministry of Planning and Economic Policy, MIPPE) included: Nuvia Z. de Jarpa (Director of Social Policy), Rosa Elena de De la Cruz (National Coordinator of the LSMS and Poverty Study), Cecilio Gadpaille, Maria Cristina de Pastor, Edith de Kowalczyk, Zuleika de Bustos, Ida de Garcia, Agustin Garcia, Armando Villarreal, Gabriela Montoya, Miguel Achurra, Eudemia Perez, Tatiana Lombardo, Roberto Gonzalez, and Julio Di6gez H. Pedro Contreras, Gloriela de Gord6n and Gloria Rivera of the Ministry of Health also participated in the study. Nuvia Z. de Jarpa and Cecilio Gadpaille also participated in the final revision of the Report. Funding for the study was generously provided by the Government of Panama (GOP), the World Bank (including funding from the Institutional Development Fund (IDF) and the Strategic Compact), the United Nations Development Programme, the Government of Japan, and the Swiss Agency for Development and Cooperation. PANAMA POVERTY ASSESSMENT: Priorities and Strategies for Poverty Reduction EXECUTIVE SUMMARY 1. This report has three main objectives. 1998, is a direct outcome of this on-going effort. The first is to assess the extent and determinants With the technical support of the Bank's Poverty of poverty, malnutrition and inequality in Assessment team, the Govemment also Panama. The second involves examining the constructed a new and improved Poverty Map impact of government spending and policies on that combines data from the LSMS with the the poor in key sectors. The third is to outline National Census to serve as a policy tool for priorities and steps for translating the targeting public spending and interventions. Government's recently approved Poverty Strategy into action. THE PROBLEMS OF POVERTY, 2. This study is primarily based on the first IALNUTRITION AND INEQUALITY... nationally-representative Living Standards 4. Despite Panama's relatively high- Measurement Survey (LSMS), which was income per capita (US$3,080 in 1997), conducted by the Ministry of Planning and poverty remains pervasive. Over one million Economic Policy (MIPPE, now the Ministry of people (37% of the population) live below the Economy and Finance, MEF), with funding poverty line, and of these, over half a million from the Government of Panama, the World (19%) live in extreme poverty. One half of all Bank, the Inter-American Development Bank, Panamanian children are poor. The distribution the Swiss Agency for Development and and magnitude of poverty in Panama varies Cooperation, Japan's Policy and Human significantly by geographic area: Resource Development Fund, and the United Rural poverty bias. Poverty and extreme Nations Development Program. The study also poverty are concentrated in the countryside. draws on results from a Social Capital Rural poverty is higher in relative terms Qualitative Survey, which was designed to (with 65% and 39% of the rurl population complement the LSMS with in-depth living in poverty and extreme poverty, community-level information on perceptions, respectively) and in absolute terms, with social organization, and social capital. over 788,000 rural residents living in 3. The process by which this report was poverty (close to three-quarters of the prepared is as important as the document itself. nation's poor population). It is based on a highly collaborative process Desttution amon the idigenous. between the Government and the World Bank g g designed to build local capacity for, and Poverty in indigenous areas can only be ownership of, the analysis of poverty. The desibedts abysmal er (197,003 process involved continuous cooperation and residen f idenous areasd(97003 numerous missions by Bank staff to Panama and people) fall below the poverty line and 86% MIPPE/MEF staff to Washington. MIPPE/MEF live in extreme poverty. Although has also actively disseminated preliminary indigenous residents represent only 8% of results, regularly publishing the findings in the total population, they account for 19% various bulletins and papers and holding and %oth nion'sp of fertility, numerous seminars and workshops. The process poor.eWith hger te of frily has already served as an important input for ingeno ar the most Ap polcymain ad hedeelpmntofth growing segment of the population. As Bank's masistanc tre forpana the such, Panama's poverty rate will increase in Bank's Assistance Strategy for Panama. The the absence of an aggressive poverty Government's new Poverty Strategy, which was the segy. P overty approved as a Cabinet Resolution in September i poverty are highest among the Ngobe-Bugl6, and the highest incidence among the Ngobe- Panama's largest ethnic indigenous group, Bugle. followed by the Embera-Wounan. Poverty is 7. Panama is among the more unequal lower among the Kuna overall, though it is countries in the world. With a consumption still quite high among those living in Gini of 49 and an income Gini of 60, Panama's indigenous areas. Geography appears to be a more powerful determinant of poverty than inequality ranks among the highest - on par with a more power.iil determinant of poverty than Brazil and just below South Africa, two of the ethnicity, with a higher incidence among world's most unequal countries in the world. .indigenous ara tn those living outsid Panama's poorest are very poor and the richest these areas. are very rich. Although inequality is higher in rural areas, it is more obvious in urban areas, * Urban vulnerability. Although poverty is such as the city of Col6n, where the close not as widespread or deep in urban areas physical juxtaposition of the modern, dynamic, (15% of the urban population), Panama's wealthy sector with poor city slums accentuates cities account for an important share of the the perceived gap between rich and poor. poor (23% or over 232,000 poor urban residents). Close to 40% of the urban poor ... REFLECT UNDERLYING DISPARrTIES IN (over 90,000 people) live in the Panama City ASSETS. - San Miguelito area. Moreover, a 8. The problems of poverty, significant share of city-dwellers live just malnutrition, and inequality in Panama above the poverty line and could be largely reflect disparities in opportunity. The considered vulnerable. distribution of key productive assets - labor, 5. Poverty is a national problem, with human capital, physical assets, financial assets, several key regional pressure points. and social capital - is highly unequal. These Panama's poor are spread across the country. disparities are most prevalent between the poor Although poverty rates are significantly higher and non-poor, but also manifest themselves further away from the capital area, some differently by geographic area. 315,000 poor residents are concentrated in the 9. Labor, the poor's most abundant Provinces of Panama and Col6n. The new asset, accounts for 77% of their total income Poverty Map recently constructed by (69% for the non-poor). Nonetheless, the poor MIPPE/MEF using data from the LSMS and the are constrained in their use of this key asset in a Population Census confirms this tendency: number of ways: poverty rates are highest in San Blas, Darien, Bocas del Toro, Cocle, and Chiriqui, and lowest * High rates of unemployment for the in the Provinces of Panama and Col6n. urban poor. Though poverty and Nonetheless, these latter two provinces account unemployment are not correlated for the for roughly one third of Panama's poor. Within nation as a whole, the poor are twice as each region, poverty rates are highest in the likely to be unemployed as the non-poor in rural, indigenous, and remote areas and lowest in Panama's cities. The difference is even the central urban districts. higher when taking into account seasonal and discouraged job seekers. 6. There is a strong correlation between ad isorged jb sees 6.o Theretis and child lngorration betwena. Unemployment is particularly high for poor poverty and child malnutrition in Panama. urban women and youths. Over 16% of all children under five (close to 50,000) suffer from any form of malnutrition. * Potential underemployment among the About 85% of these are poor. Close to one poor in all areas. The poor work fewer quarter of poor children and one third of extreme total hours than the non-poor in all areas - a poor under five are malnourished, compared to likely sign of underemployment and low 4% among the non-poor. The incidence of productivity. malnutrition mirrors the geographic and ethnic A strong correlation between informal- patterns of poverty, with one-half of all children sector employment and poverty. Close to in indigenous areas suffering from malnutrition three-quarters of the poor work in the ii informal sector (40% of the non-poor). Low are apparent for achievement, coverage, internal levels of household consumption are efficiency and the quality of education: significantly correlated with informal sector employment even after other factors (such as expanding literay and increasing human capital) are taken into account. educational attainment over time, gaps Earnings in the informnal sector are emainal tainmend the gaps significantly lower than those in the formal remain for the poor and the indigenous sector: informal workers earn 60% and 43% (particularly indigenous women). of what those in the private and public * The main gaps in access include the formal sector earn. These differences are indigenous at all levels, and the rural and not explained by differences in human urban poor at the pre-primary and secondary capital, area of residence, or job levels. Moreover, very few students of characteristics. higher education are poor (5%). Key * Indigenous workers face probable wage obstacles to higher enrollment for the poor discrimination. The LSMS shows' that the include (i) the direct costs of schooling differences in the salaries received by (fees, books, etc.) at the primary and indigenous and non-indigenous workers secondary levels; (ii) a lack of "interest" cannot be explained by factors such as among some poor children (particularly poor educannoti experience, type of work, etc. a urban boys) at the secondary level, which education, indigenous work , e In could reflect social pressures as well as addition, indigenous workers have few quality issues in the educational system; and opportunities for employment in the formal q t sector. (iii) a lack of programs at the pre-primary level. 10. Human capital - education and health - is an important complement to labor, v Internal efficiency is also lower among the boosting its productivity and potential for poor and indigenous, who tend to repeat income generation. The LSMS reveals that grades and drop out more frequently than schooling pays off in terms of higher incomes: the non-poor. each year of schooling yields about a 5% * Lower quality education for the poor is increase in hourly earnings. These returns vary evident from the higher share of poor significantly by education level, with primary students without textbooks, a lack of school (which has fairly equitable coverage) bilingual materials and instruction for generating much lower returns than secondary or indigenous primary students, the high share higher education (to which the poor have much of the poor in communities reporting less access). Access to health care also insufficient teachers, and dilapidated school generates productivity gains and contributes facilities in communities with higher directly to well-being. The distribution of human concentrations of the poor. capital assets, however, is highly unequal. 12. Inequities in health status and health 11. Disparities in education are key care also abound. Relatively strong health causes of poverty, malnutrition, and indicators for the nation as a whole mask large inequality in Panama. Education is also a disparities and poor health status among those crucial elevator for the poor to lift themselves living in poorer areas. The poor (particularly the out of poverty. Higher educational attainment indigenous) have lower life expectancy, higher for a household head or his/her companion rates of infant mortality and malnutrition, and significantly reduces the probability of being continue to die from infectious and poor. Mothers' education significantly affects communicable diseases despite Panama's child nutritional status. Disparities in education epidemiological transitioning. The poor and constitute the single most important determinant indigenous have less access to health care, and of inequality, accounting for about 40% of are less likely to seek medical treatment in case Panama's consumption inequality. Inequities of illness than the non-poor. Low access to health services bears a significant link to child 'The analysis is based on salary regressions that do not include malnutrition in Panama. income from self-employment. See Annex 6, Appendix A6.3. 111 13. Physical assets - such as housing and borrowing against these assets for land - also contribute to income-generating emergency income. potential and help households avert risk. The 15. Basic infrastructure services poor commonly use housing and land as a base contribute to higher welfare and productivity. for productive activities and enterprises. Some services, such as potable water and Property also generates rent - through earnings sanitation, contribute directly to overall welfare charged to renters or via the savings from and health status. Others, such as electricity and "imputed" rent. Households can also use telephones, help households use their homes property as collateral for leveraging credit. productively for income generation. The LSMS Emergency income can likewise be generated reveals that access to basic services is highly through sales of property or borrowing against it correlated with a lower probability of being (equity loans). Finally, housing can be used as a tool for extending personal relationships, poor. Inequities in access to such services tooldn foruexnding ernsoal rlapiti. abound in Panama, both between the poor and non-poor and by geographic area (especially 14. The ability of households to use among disperse populations in rural and property as an asset depends largely on the indigenous areas). Key gaps in coverage security of tenure and the flexibility of land include: the indigenous for all services and the and housing markets. In Panama, the rural poor for energy and sanitation services and distribution of housing and land is highly to a lesser extent, potable water. While the unequal, as is access to the titling of available urban poor have much greater access to all types property. of services than their rural counterparts, a lack of sanitation services for an important share of poor *~~~~~~~~ Hosn.Tepo,en olv nmc city-dwellers raises public health concerns. lower quality housing than the non-poor. Moreover, while the majority of the 16. Financial assets - savings and credit - population lacks proof of ownership allow households to smooth their (registered or unregistered titles) for their consumption and invest for future earnings homes, the gap in titling of housing assets is potential. The poor are much less likely to save much worse for the poor. Not owning a than the non-poor; when they do, they tend to house increases a household's probability of put their savings in public institutions (whereas being poor, as does the lack of a registered the non-poor are more likely to use private title. banks). The overall volume of lending to the * Land. The distribution of land is highly poor is much smaller than their contribution to unequal in Panama. The poor, who account the economy. Whereas poor households receive for two-thirds of the rural population, own 3% of total credit, they account for 10% of total one third of land. Among those who own consumption and income in the economy. any land, the Gini coefficient for total land Information constraints, physical distance, lack owned is 77. Disparities in land ownership of formal guarantees, and high costs-per-dollar account for 11% of total consumption borrowed present obstacles to lending to the inequality in Panama. The poor have even poor. less access to titled land: only one third of 17. Social capital2 is one of the assets of all owned agricultural land is fully titled, the poor. The poor and extreme poor account and the non-poor own 84% of it. for a disproportionate share of those living in * Titling and Income Generation. The lack communities with high social capital. The of property titles reduces the ability of the LSMS reveals that the poor tend to associate for poor to obtain credit (since titles are often "public goods" purposes (e.g., in community required as collateral). The lack of associations), whereas the non-poor join guarantees (property titles or other assets) associations that yield higher private gains (e.g., was the main reason poor households were refused credit. Without formal claim, the 2 Social capital - defined as norms, trust, and reciprocity networks poor also lack the possibility of selling or that facilitate mutually beneficial cooperation in a community - is an important asset that can reduce vulnerability and increase opportunities. iv cooperatives). These patterns suggest that the labor - and thus reducing demand for the poor's poor rely on community action (social capital) to most abundant asset - these distortions have compensate for a lack of other assets created by swelled the ranks of the unemployed and public services. Social capital also helps encouraged informality. Moreover, while they communities leverage assistance. Communities benefit those who work in formal sector jobs, the with high social capital report a higher resulting segmentation of the labor market can frequency of assistance from the Government put a heavy toll on informal sector workers by and NGOs. This correlation is observed in all reducing their wages, making it difficult for the three geographic areas, and is particularly strong working poor to grow out of poverty through in rural and indigenous areas. Interventions their own labor. Indeed, the LSMS reveals that should work with communities to build on this the poor in Panama do not benefit from such important avenue for public action. distortions, but may be hurt by them: (i) the majority of the working poor receive wages that SUCH DISPARITIES LARGELY REFLECT A are below the official minimum wage; (ii) they LEGACY OF DISTORTIONS IN PANAMA'S do not receive the "mandated" fringe benefits; ECONOMY... (iii) the majority are employed in the informal 18. The disparities in the distribution of sector, where wages are lower and employment reflect terms less favorable; and (iv) the urban poor are Panama's uniquely dualistic pattern of hurt by high rates of open unemployment and development. Panama's privileged geographic the rural working poor appear to be location and its monetary regime anchored in the underemployed. use of the US dollar as legal tender have fostered 21. The modern services sector is its comparative advantage in services, which juxtaposed with a traditional sector contribute over three quarters of GDP and (primarily agriculture and industry), that has generate two-thirds of employment in Panama. been constrained by policy-induced rigidities These strategic factors have also spurred the and low productivity for decades. Until rapid development of internationally-oriented, recently, a highly distorted trade regime (one of modem, dynamic service enclaves, including the the most protectionist in Latin America) Canal Zone, the Col6n Free Zone, and the combined with a complex web of price controls, International Banking Center. While these had the result of protecting inefficient sectors, enclaves generate large shares of GDP, they distorting resource allocation, generating rents create little employment (3% of the labor force) for certain groups, and raising the cost of basic or fiscal revenue. Moreover, they inject staples. Simulations using the LSMS suggest negative spillovers into the economy due to the that the net redistributive effects of such huge differentials between wages paid in the protection was regressive, effectively taxing the enclaves, particularly the Canal Zone (which is poor and increasing poverty and inequality. subject to the U.S. labor code) and the rest of the 22. The Government has undertaken a economy. number of fundamental reforms since 1994 in 19. Indeed, factor markets have been an attempt to stimulate higher and more segmented by policies that drive up the cost of inclusive growth and reduce poverty and labor relative to capital. Panama's labor inequality. It has dismantled the labyrinth of market is characterized by a multiplicity of trade barriers and price controls, launched a far- policy regimes, with separate regimes for the reaching privatization program, issued anti-trust private sector, the public sector, the Panama legislation, unified fiscal incentives for Canal Commission, and the Export Processing manufacturing firms, and adopted modest but Zones. These regimes have created large wage important reforms in the Labor Code. After differentials between workers in the Panama several years of little growth, the economy Canal Commission, public sector employees, began to respond in 1997, with strong growth and those employed in the rest of the economy. and some decline in unemployment for the first 20. Labor-market interventions not only time in years. Even more important for.the hamper growth, but ao have a direct link to long-term, employment has apparently become poverty. By increasing the relative price of more responsive to changes. in growth. Despite V these reforms, an important policy agenda these biases by developing a number of remains, with several key additional reforms social assistance programs. While several needed to reduce poverty. programs appear to be quite effective at reaching the poor, lack of targeting of some ... AND BIASES AND INEFFICIENCIES IN PUBLIC of the larger programs, inefficiencies in SPENDING. program delivery, and possible duplication 23. Disparities in key assets - notably of functions reduces the effectiveness of human capital - also reflect a lack of these transfers. targeting of social policy and public spending. Despite high spending on the social sectors GOVERNMENTs NEW POVERTY STRATEGY (21 % of GDP overall, 6% for education, and 7% for health in 1997), inefficiencies and inequities 24. The Government's Poverty Strategy have prevented improved outcomes for the poor: and Action Plan3 is a direct outcome of the joint analysis of the LSMS, and several inter- Total public spending on enefiting institutional poverty seminars. This strategy regressive, with the non-poor benefiting complements the Government's economic more than the poor. This inequity largely reform program and is designed to strengthen reflects the large share of public spending the key assets of the poor, taking into account allocated to higher education (close to one- third of the public education budget), aphicierenmein the poverty situation virtually all of which (95%) benefits the and priorites. The main underlying principles non-poor. It also reflects the gaps in f t coverage of the poor at the pre-primary and * Coherence with, and implementation of, the secondary levels. Geographic inequities in economic reform program for sustained public spending on education are also broad-based economic growth; apparent, with biases in favor of urban areas in social spending; and against rural (indigenous and non- indigenous) areas. Furthermnore, the poor * Targeting of resources to the poor; quality of public education stems from . . functional inefficiencies in the delivery of , Decentralization of servitces for improved education, including over-centralized efficiency and quality of interventions; decision-making, weak policy-making and * Increased community participation for planning capacity in the Ministry of improved effectiveness; Education, lack of management information tools, and a disconnect between teacher * A multi-dimensional approach with strong- salaries and performance. intersectoral coordination; and * Inequities in public spending on health * Monitoring of the poverty situation and care also abound. The poorest quintile of implementation of the strategy itself. the population benefits the least from public The strategy also includes an action plan for spending on health care due to low levels of interventions in key areas, including: education, use and access to these services. health, social assistance and nutrition, urban Inefficiencies in the health sector - poverty, and rural poverty. including the fragmentation of the sector by three major providers (the Ministry of TRANSLATING THESE PRINCIPLES INTO Health, the Social Security Institute, and the PRIORITIES AND ACTION private sector), few incentives for efficiency and peformane, andweak plicy-mking, 25. To translate these principles into action, and performance, and weak policy-making, the Government should: financing, and regulation capabilities in the Ministry of Health - have also prevented * Prioritize among poverty groups. Given improved health outcomes among the poor. the distribution of poverty, first priority * Since the early 1990s, the Government has attempted to compensate for some of 3Nuevo Enfoque Etrategico Frente a la Pobreza, Cabinet Resolution No. 134, September 17, 1998. vi should be given to: the rural poor, the transportation, electricity, telephones and indigenous (particularly the Ngobe-Bugle), potable water have been identified by rural poor children and youths, and under- communities as priority; potable water is top nourished children and pregnant and priority in indigenous communities; and lactating women. Second priority should be transportation and sanitation are key for assigned to combating urban poverty. Third urban communities. priority should be given to programs that Decentralize and promote community target the poor elderly, poor child laborers, participation in service delivery to poor informal-sector workers and the poor improve the effectiveness and efficiency of unemployed. poverty interventions. Examples include: * Reallocate public expenditures. The top (i) decentralizing personnel decisions to priority for effective action to reduce regional education boards and expanding poverty should involve reallocating public innovative participatory pre-school expenditures. Given the high level of social programs; (ii) decentralizing food purchases spending, it is unlikely that a large amount in remote areas under school feeding of additional resources will be forthcoming. programs; (iii) expanding use of NGOs and As such, the Government needs to reallocate communities as intermediaries in social existing spending toward areas that benefit programs; and (iv) responding to community the poor, boost cost recovery for services preferences for service delivery. used by the non-poor, and improve efficiency in service delivery. A thorough * Implement key policy reforms to reduce review of public spending should be disparities in assets. While maintaining conducted in 1999 to provide guidance on existing reforms is critical, the agenda for such reallocations. This review should the second-generation of economic reforms emphasize both the functional aspects of the designed to promote growth is large. budget (including the distributional Special efforts should be made to ensure that incidence of spending in key sectors) as well key reforms to reduce disparities in assets, as the management of public expenditures. and hence poverty, are undertaken, C lear candidates for reallocation of including: (i) deepening reforms to the labor education spending include: (i) enforcing code; (ii) expanding property titling (both of higher cost recovery for higher education housing and land), which will also help and shifting fieed resources toward basic improve the poor's access to credit; education; (ii) focusing spending on demand (iii) continuing trade reforms; and side education schemes to reduce economic (iv) reducing distortions in public and barriers faced by poor households to freight transport. increase enrollment by the poor. Although a * Improve targeting mechanisms. The thorough analysis of the incidence of healtl Government should apply the new poverty spending remains to be done, some map (combining data from the LSMS and candidates for reallocation include: the census) to the allocation of expenditures (i) enforcing higher cost recovery for as soon as possible. It should also seek to curative and hospital care develop additional mechanisms for (disproportionately benefit the non-poor) targeting, including means-testing and self- and (ii) shifting resources to cost-effective targeting. primary interventions in poor areas. Spendimary interventi insuooran ard Allocate sufficient resources to monitor Spending on social insurance andpoetanthimlettonfte assbtance should also be streamlined to poverty and the implementation of the ensure a comprehensive, efficient, well- strategy. The Government is developing a argted saftyet An inve ntory of service poverty monitoring system to track living urge safshould be overlayed with te new conditions and provide data for impact covegrW map to uide spending allocations evaluation of interventions. This system po 'l to a'spending allocaps includes LSMS-type surveys to . be amongbac the poor. as broa taetkeyrmps, conducted every three years. The among the poor. In broad termsl5, Government should also seek to develop a vii key set of indicators for monitoring the implementation of the poverty strategy in 1999 and subsequent years. Program- specific questions should also be included in the next LSMS for additional impact and implementation analysis. 26. Some key steps for immediate implementation during the remainder of 1999 should include: * Conducting a thorough review of public expenditure allocations and developing proposals for reallocating expenditures such that they better reach the poor starting with the 2000 budget; * Developing a set of indicators to monitor implementation of the strategy (including key budget categories) and agreeing on an inter-institutional process for reporting on such indicators and implementation. Funds from the on-going IDF grant could be used to contract technical assistance to help MEF staff developing and monitoring such indicators; * Applying the new Poverty Map as a tool for targeting and resource allocation; and * Continuing dissemination efforts for both the Poverty Assessment and the Government's Poverty Strategy in both Government circles and public forums. 27. Areas for further research include: public expenditure analysis (incidence and management), the links between poverty and the environment, a thorough analysis of the impact of existing legislation on labor markets including the informal sector, an analysis of the distributional incidence of social security, impact evaluations of social assistance programs, participatory research on the obstacles to increased school enrollment among indigenous children, and participatory research on poverty, crime and violence. viii PANAMA POVERTY ASSESSMENT: Priorities and Strategies for Poverty Reduction INTRODUCTION 1. Objectives. This report has three main objectives. The first is to assess the extent and determinants of poverty, malnutrition and inequality in Panama. The second involves examining the impact of government policies and spending on the poor in key sectors. The third is to outline priorities and steps for translating the Government's Povdrty Strategy into action. 2. Quantitative and Qualitative Data Sources. This study is primarily based on the first nationally-representative Living Standards Measurement Survey (LSMS),' which was conducted with funding from the Govemment, the World Bank (which also provided technical support), the Inter- American Development Bank, the Swiss Agency for Development and Cooperation, Japan's Policy and Human Resource Development Fund, and the United Nations Development Program. The World Bank's Institutional Development Fund (IDF) also provided funds for building the capacity of staff in the Ministry of Economy and Finance (MEF, formerly known as MIPPE)2 in social policy analysis and formulation. MIPPE carried out the survey in 1997 with active collaboration of an inter-institutional committee consisting of the Ministry of Health, the Ministry of Education, the Ministry of Agriculture, the Directorate of the Census and Statistics, the Social Fund (FES), and MEF/MIPPE. 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? The LSMS community questionnaire, covering 436 communities, involved focus-group discussions to collect quantitative and qualitative information from community members on their priorities, perceptions, social problems, and social capital. 3. The study also draws on results from the joint MIPPE-World Bank Social Capital Qualitative Survey (SCQS), which was designed to complement the LSMS with in-depth community-level information on perceptions, social organization, and social capital. Qualitative focus-group discussions for the SCQS were carried out in 16 communities in urban, rural and indigenous areas in November 1997. 4. Collaborative Process. The process by which this report was prepared is as important as the document itself. The objectives of this process were to build capacity for, and ownership of, the analysis in the Government. It is based on a highly collaborative process between the Government and the World Bank, involving continuous cooperation and numerous missions by Bank staff to Panama and MEF staff to Washington. As a result, MEF staff, as well as representatives of several other line ministries and agencies, have been intimately involved in generating and understanding the results. MEF has also actively disseminated results, regularly publishing the findings in various bulletins and papers and holding numerous seminars, workshops and discussions both within the Government and in public forums. 5. PoHcy Impact of the Poverty Assessment and LSMS. The process has already served as an impoftant input for policy making. First, the Government's "New Strategy for Reducing Poverty,"'4 which was passed as a Cabinet Resolution in September 1998, is a direct outcome of the Poverty 'E amt* NeI ks LSMS _iwsc wm i fieWed * June to Sep_teber 1997. DatA am available for public use on the World Wide Web at: _ MWJlsc.tbd.gcibtft.lhvc7hwJw.wodd _ __Ihns/udMssAmhomchgl In 1qrMy 1_99, the Minutzy of Planing aid Eeomzc Policy (MI nhlrged wit Hacia cei a nw _i called the Ministry of Economics mad Finae. 'The final s mple indudus 4,938 louseholds, of e 650,726 houseoWds atde ndaionl level, d 21,410 individuals of the 2,732,316 individuals in the nation population. 4Nuevo Enfoqw Ewraaigico Freck a la Pobreza, Cabinet Resolution No. 134 of Sepumber 17,1998. Assessment process, the analysis of the LSMS (including community focus-group discussions), and several high-level inter-ministerial poverty seminars. Second, MEF with the FES and technical support from the Bank's Poverty Assessment team constructed an imnproved Poverty Map that combines data from the LSMS with data from the National Census. This new Poverty Map will serve as an important policy tool for targeting public spending and specific programs (including the FES). Third, the LSMS is being used to conduct policy analysis in a number of sectors, including inter-alia a demand-and-supply study of health care services and incidence, and cost-benefit analysis in the education sector. Fourth, the analysis substantially contributed to the development of the Bank's Assistance Strategy for Panama' 6. Road Map for the Report. This report is divided in two volumes. Volume 1 presents the main body of the report and is organized as follows. Chapter 1 examines the extent and distribution of poverty in Panama answering the question Who are the Poor in Panama? It also assesses the extent of malnutrition and inequality. The problems of poverty, malnutrition, and inequality in Panama largely reflect underlying disparities in key assets (labor, human capital, physical assets, financial assets, and social capital), as shown in Chapters 2 and 3. Such disparities can largely be traced to a legacy of distortions in Panama's economy, as discussed in Chapter 4, and biases in public spending, as discussed in Chapter 5. Chapter 6 outlines the main principles of the Government's Poverty Strategy, which is designed to strengthen the key assets of the poor, and recommends priority actions for translating this strategy into action. 7. In addition to providing additional details in technical and statistical appendices, Volume 2 includes a number of in-depth background papers that were prepared during the Poverty Assessment process. These include papers on: Rural Poverty, Indigenous Poverty, Urban Poverty, the Determinants of Poverty, Malnutrition, Inequality, Poverty and Labor Markets, Poverty and Trade Reforms, Pre-School Programs, Health Care Use and the Incidence of Health Care Spending, School and Health Feeding Programs, Social Capital, and Female Headed Households. Chapter 1 - Poverty, Malnutrition and Inequality in Panama 1.1. This chapter identifies the magnitudes of the problems of poverty, malnutrition and inequality in Panama. It starts with a definition and measure of poverty and a description of its extent and distribution. It then examines the magnitude and distribution of malnutrition, which is highly correlated with poverty. Finally, it reviews indicators of inequality in Panama, with geographic comparisons within thecountry and with other nations. A. MEASURING POVERTY AND WELFARE 1.2. Assessing poverty requires some measure of welfare. Typical measures include consumption, income, and basic-needs indicators. For the first time in Panama, this study uses consumption to measure poverty and welfare. Consumption data tend to be more reliable than income data and consumption tends to fluctuate less than income, making it a better indicator of living standards. Consumption is also less subjective than basic-needs indices and is more responsive to changes in economic circumstances. Using data from the LSMS, two poverty lines were calculated using consumption as a measure of welfare: an extreme poverty line and a full poverty line. 1.3. The extreme poverty line is defined as the level of per capita annual consumption required to satisfy the minimum average daily caloric requirement of 2,28O. The annual cost of this minimum caloric requirement yields an extreme poverty line of B./519. Below this level of expenditures, 'World Bank Report No. 18421-PAN (October 1998). 6Estimated by the hzgtituto de Nutricion de Centro Amdricay Panamd (INCAP) and MINSA, this minimum represents a weighted average based on the assumption of moderate activity, taking into account the actual age and gender distribution of the Panamanian population in accordance with official population projections for 1997. 2 individuals cannot maintain the minimum level of caloric consumption even if all resources were allocated to food. 1.4. The full poverty line is defined as the extreme poverty line (the cost of the minimum caloric requirement) plus an allowance for non-food items. This allowance is calculated as the non-food budget share of those individuals with total consumption that is close to the extreme poverty line.7 It is assumed that, since these individuals would barely meet the minimum caloric requirements evenif they spent all of their resources on food, whatever share of total consumption they actually allocate to non-food consumption must constitute true necessities. The analysis found that the non-food share for individuals near the extreme poverty line was 43 percent. This method yields a full poverty line of B./905, below which individuals would be considered poor.8 B. POVERTY IN PANAMA 1.5. Despite Panama's relatively high level of GDP per capita (US$3,080 in 1997), over one million people (37 percent of the population) live below the poverty line and over half a million (19 percent of the population) live in extreme poverty (Table 1.1). Overall poverty is in line with the LAC Regional average (37 percent) but extreme-poverty in Panama is higher (16 percent in LAC). 1.6. The costs of reducing poverty in Panama are not high. Given average consumption levels of the poor, it is estimated that the minimum annual cost to bring all poor Panamanians to the poverty line represents roughly five percent of GDP and ten percent of total government spending9 For extreme poverty it would cost about one percent of GDP and six percent of total government spending. Although these minimum costs exclude the inevitable administrative costs orleakages to the non-poor associated with virtually all poverty-alleviation schemes, the amounts are quite low. To put this in context, this minimum annual cost of eradicating extreme poverty roughly equals the share of public spending on higher education in 1996, virtually none of which went to the poor. Table 1.1 - Distribution and Incidence of Poverty in 1997, By Geographic Area % of Incidence of Poverty Contribution to National National Headcount Indexb (%) Poverty (% of:) Populationa All Poor' Extreme Poor All Poor Extreme Poor Total Panama 100 37.3 18.8 100 100 By Geographic Area Urban 55.6 15.3 3.1 22.7 9.2 All Rurald 44.4 64.9 38.5 77.3 90.8 Rural (non-indigenous) 36.9 58.7 28.7 57.9 56.2 Indigenouse 7.6 95.4 86.4 19.3 34.6 Source: Panamna LSMS 1997. (a) The distribution of the national population (2,732,316) is derived firom the LSMS samnple using the sample weights (expansion factors) from the National Census. It closely reproduces the figures from the National Census. (b) The Headcount Index is the share of the populabion whose total consumption falls below the poverty line. (c) Throughout this study, "all poor" includes the extreme poor. (d) The "all rural" area category includes: rural (non-indigenous) and indigenous areas. (e) Indigenous in this table refers to geographic area not ethnicity (see Box 1.2). 1.7. Poverty appears to have fallen over time. Comparing estimates of poverty over time in Panama is risky due to large differences in sampling and methodology (Box 1.1 below). Estimates using income data from the 1983 Socioeconomic Survey calculate poverty at 39 percent of the population and extreme poverty at 20 percent. Adjusting these figures to include indigenous and remote population°o yields rough estimates of 46% of the population living in poverty and 28% living in extreme poverty in ' In other words, individuals whose total consumption is more or lss (+/-10 percent) equal to the extreme povaty line (Annex 2). Trhis is known as the Cost of Basic Needs (CBN) method. ' An advantage of the above method for calculating the extreme and full poverty lines ova the cost of a pre-detennined basket of food and non- food items is that it does not impose assumptions about the consumption prefernces and patems of the population. These patterns are o,served - representing te true consumption patterns of the poor in Panama - rafher tban imposed. 9 As a percent of spending by the non-financial public sector (NFPS). "' Which currently represent 23 percent of the poor population in Panama and 36 percent of the extreme poor population. This adjustment also takes into consideration the fact that indigenous and remote populations represent about nine percent of the total population. 3 1983, suggesting a fall in poverty of roughly nine percentage points over the past 14 years. Trends in social indicators support this apparent improvement in living standards: life expectancy has risen (from 70 years in 1980 to 74 in 1997), infant mortality is down (from 31 per thousand live births in 1980 to 22 in 1997), and educational indicators, such as literacy and secondary school enrollment, have improved." Box 1.1 - Past Studies of Poverty and Inequality in Panama Concern with poverty increased in Panama during the 1970s and 1980s, and a number of studies of poverty and inequality were conducted during that time (many of which were commissioned by the Government). A number of agencies (including MIPPE, ECLAC, and ILO) have also tracked poverty on an annual basis using household labor surveys since the late 1980s. Inferring conclusions regarding the trends in poverty and inequality from these studies is risky due to large differences in data coverage (none were nationally representative), methodology, and the definition of welfare variables (most used incomplete measures of income) and poverty lines (most set the extreme poverty line equal the cost of a basic food basket in Panama City and doubled it to obtain the full poverty line). Some Past Findings on Poverty (poverty figures reported on an individual basis). * Molina (1982) estimated poverty at 39% in 1970 and 37% in 1980 using Census data on incomes. * Sahota (1990) estimated extreme poverty at 20% in 1980 using Census data on incomes * Sahota (1990) estimated extreme poverty at 20% and full poverty at 39% using a fairly complete measure of income from the 1983 Socioeconomic Survey (incomplete sampling of indigenous and remote areas). * MIPPE (1996) tracked poverty using data on cash incomes from the SIAL/ILO Household Survey database (excludes indigenous areas) on an annual basis from 1991-1995. These estimates suggest a fall in poverty from 55% in 1991 to 47% in 1995 and in extreme poverty from 27% in 1991 to 22% in 1995. This drop mainly comes from a reduction in the share of urban residents living in poverty (from 47% in 1991 to 35% in 1995); rural poverty remained fairly stable during that period (at 71% in 1991 and 72% in 1995). Some Past Findings on Income Inequality (all Ginis calculated on a per person basis). * McLure (1974) calculated a Gini coefficient of 55.7 using wage data from the 1969 Household Survey * Sahota (1972) calculated a Gini coefficient of 59.7 for 1970 using wage and non-wage income data from the Population Census and Agricultural Census combined with data from the Household Survey for 1969. * The DEC (1975) used income data from the 1971 Income Survey to calculate a Gini coefficient of 56.3. * Sahota (1990) derived a complete measure of income using data from the 1983 Socioeconomic Survey to calculate Gini coefficients of 55 for the nation as a whole (excluding indigenous and remote areas), 47 for urban areas and 56 for rural areas. 1.8. One-half of all Panamanian children live in poverty. Due to higher fertility rates among the poor, a large share of children live in poverty. In fact, 53 percent of children under five (over 160,000) and 48 percent of all children under 18 (over 500,000) live below the poverty line. Close to one third live in extreme poverty. In contrast, 27 percent of senior citizens over age 60 live in poverty, with 12 percent in extreme poverty. This lower share of the elderly in poverty (compared with 37 percent of the overall national population) suggests lower life expectancy among the poor than the average population. 1.9. Contrary to conventional wisdom, poverty is not higher among households headed by women than those headed by men. Overall, female-headed households are not over-represented among the poor in Panama, regardless of the method used to define household headship.12 Some exceptions do emerge, however: (i) in urban areas, the poverty rate among female-headed households with unmarried partners (unida) is significantly higher than households headed by males with common-law partners; and (ii) in indigenous areas, poverty is more prevalent ,1.1 among households headed by widows thn by widowers. These exceptions represent a very small Geow"mgDI S O lPauoV percentage of total households however, and as (LSMS1997) such, female headship is not likely to be a useful UAW proxy for targeting poverty interventions. 1.10. Poverty and extreme poverty are concentrated in the countryside (Figure 1.1). A disproportionate share of the poor and extreme poor live in rual areas in comparison with the "Hicks and Pecters (Septmber 1998). 12 Definitions of headship include self-eported, ecnomic and demographic (Annex 19). 4 share of rural residents in the national population. Three-quarters of the poor and 91 percent of the extreme poor live in the countiyside. Two thirds of all rural residents fall below the full poverty line and close to 40 percent live in extreme poverty. Table 1.2 - Depth and Severity of Poverty in 1997, By Geographic Area % of Pop. below # Individuals Average Total Minimum Cost Depth Severity Index3 Pov. Line I below Pov. Line Consumption (B./)a (nn. B .b Indexc Full Povert, Line (B./905), inc uding those below the extreme poverty line Total 37.3 _ 1,020,158 506 406 16.4 9.7 By Area: Urban 15.3 232,006 674 53 3.9 1.5 All Rural: 64.9 788,152 457 353 32.1 20.0 Rural Non-Ind. 58.7 591,148 517 229 25.2 14.0 Indi2enoust 95.4 197,003 276 124 66.1 49.2 Extreme Xoverty Line (B./519) Total 18.8 514,839 306 109 7.7 4.2 By Area: Urban 3.1 47,283 406 5 0.7 0.2 All Rural: 38.5 467,557 297 104 16.5 9.2 Rural Non-Ind. 28.7 289,236 334 53 10.2 5.0 Indigenouse 86.4 178,321 236 50 47.0 29.7 Source: Panama LSMS 1997. (a) Represents total annual per capita consumption for those below the poverty line. Figures have been adjusted to national prices using a price index to take into account spatial variation in prices. (b) Represents the minimum cost to bring the total per capita consumption of all individuals below the poverty line up to the line. It is the absolute poverty gap. (c) The Poverty Depth Index (PI) represents the amount needed to bring all poor individuals up to the poverty line (full or extreme), expressed as a percent of the poverty line taking into account the share of the poor population in the national population. (d) The Poverty Severity Index (P2) is a derivation of the Poverty Depth Index that takes into account the distribution of total consumption among the poor. In other words, it measures degree of inequality among the population below the poverty line. (e) Indigenous in this table refers to geographic area not ethnicity (see Box 1.2). 1.11. Poverty in indigenous areas can only be described as abysmal. Over 95 percent of residents in indigenous areas (close to 200,000 people) fall below the poverty line. Some 86 percent live in extreme poverty. Poverty is also deepest and most severe in indigenous areas (Table 1.2). On average, the indigenous poor would have to more than double their annual consumption to reach the poverty line. Although residents of indigenous areas represent only eight percent of the national population, one fifth of the poor population and one third of the extreme poor live in indigenous areas. With higher rates of fertility, indigenous areas are the most rapidly growing segment of the population (5.4 percent per year). As such, in the absence of an F 1.2 aggressive poverty reduction strategy Figure for indigenous areas, the share of the Bhnicity, Geography, and Poverty: poor in the national population will 100% % below poverty line by ethnic group (LSMS 1997) increase. 92% 80% 96% 80% 84% 65% 1. 12. While poverty is high among 60- 53% all ethnic indigenous groups, there 60% is substantial variation in the 40% 32% prevalence of poverty between 20% them.13 Using language indicators of 20% * ethnicity (as opposed to geographic, 0%- _ see Box 1.2), some 84 percent of , O O m @ co 0 0 ethnic indigenous people live below c @ 5.2 -* 7.0 *4 5.1 !! 3.3 (d) Current Coverage: (c) Pre-Primary * 68% !! 26% !! 46% !! 23% !! 14% Primary /95% /90% / 92% / 91% !! 83% Secondary / 82% !! 37% !! 60% !! 37% ! 16% Higher /31% *:*3% *:7% * 2% *: 0% Internal Efficiency & Quality: Repetition Rate - Primary / 4% !! 9% !! 10% *> 8% !! 12% Repetition Rate - Secondary *> 8% * 9% !!11% * 8% 4. 4% No Books-Primary 4% ! 29% *: 7% ! 24% !! 61% No Books - Secondary. / 5% *> 12% *> 9% + 12% !! 19% Source: Panama LSMS 1997. See Annex 4 for detailed figures. (a) Percent over aged 9 who can read and write. (b) Average years of education completed for those aged 12 and older. (c) Net enrollment rates. (d) Large gender disparities with lower rates of literacy and educational attainment for indigenous women. / Represents fairly comprehensive coverage, not a key priority. *> Represents incomplete coverage. H! Represents very low coverage and key priority. Box 2.3 - The Language Poverty Trap: Language Ability, Education and Poverty among the Ethnic Indigenous Spanish-speaking ability appears to be an important determinant of educational outcomes and even poverty status: * One fifth of the ethnic indigenous population in Panama does not speak Spanish (monolingual indigenous speakers). * Literacy is lowest among the indigenous who do not speak Spanish: only 20% of monolingual indigenous speakers read and write (as compared with 79% and 98% of bilingual and monolingual Spanish-speakers of indigenous ethnic origin) * Educational attainment is also lower. Monolingual indigenous speakers average less than one year of schooling. + Among the indigenous, poverty is highest among households that are headed by monolingual indigenous speakers. Virtually all households headed by monolingual indigenous speakers live in extreme poverty. * An inability to speak Spanish is also associated with lower enrollment for the current generation: just one-half of monolingual indigenous speaking children aged 6-11 are currently enrolled in primary school. + Though Spanish is clearly the dominant language for schooling in Panama (even among the indigenous), a few indigenous students (15019) do report speaking indigenous languages in class at the primary level. Bilingual education at the primary level should be promoted where necessary to help monolingual indigenous children make the transition into the Spanish speaking world - and eventually escaping poverty. Source: Panama LSMS 1997. See Annex 6 for details. Gaps in Current Coverage 2.7. There are a number of gaps in current educational coverage, primarily among the poor and indigenous, which can be summarized as follows: * Despite efforts to boost national pre-school enrollment, the increase has not been well-targeted and there are serious gaps for the poor and indigenous. Of those who do attend, the poor are highly reliant on public facilities while a third of the non-poor attend private schools. * Despite widespread primary school coverage for both boys and girls, even among the poor, there are serious gaps for indigenous children. The net enrollment rate is highest among the Kuna and the Embera-Wounan and lowest among the Ngobe-Bugl. Almost all poor primary students and the majority of the non-poor enroll in public facilities. 13 * There are large gaps in secondary school enrollment among rural, poor and indigenous children."3 More girls enroll than boys, and most students enroll in public facilities. * Access to higher education is extremely unequal in Panamna, with virtually none of the poor attending. Some 95 percent of students in higher education are not poor and half of those enrolled belong to the highest quintile of the population.14 These disparities reflect inequities in public education, since over three-quarters of all students of higher education attend public facilities. Enrollment rates in higher education are much higher for women than men. 2.8. Demand-side factors, more than the availability of schools, appear to be the key obstacles to increased enrollment by the poor and indigenous. "Lack of money" was the single most common reason given for not enrolling in primary and secondary school (accounting for roughly half of absentees)." This suggests that the direct costs (fees, materials, etc.) of attending school are prohibitively high for the poor and indigenous. Indeed, the costs to households of public schooling (particularly informal fees and the cost of school supplies) are quite high in Panama (Box 2.4). Indirect costs - such as foregone earnings from work or domestic duties - are not frequently cited as barriers to enrollment. "Lack of interest" is the second main impediment to enrollment for the poor, particularly at the secondary level (accounting for one fifth of absentees).'6 Disinterest is particularly high among poor urban boys (accounting for 40 percent of poor male non-enrollees), although none cited work as an obstacle. This suggests a rather large cohort of poor, young urban males opting neither to work nor to attend school, who are at risk for both crime and continued poverty. Only four percent of poor non-enrollees at the secondary level and 18 percent at the primary level cited supply-side factors - such as distance to schools, inadequate class space, and lack of facilities - as key obstacles.'7 Box 2.4 - Economic Barriers to Education The direct costs of schooling in Panama are high, and could constitute a barrier to enrollment among the poor. Households allocate 6.2% of their total consumption to education. Even the average total annual cost of attending public school is high: B.A109 per student for primary and B./253 for secondary. To put this in context, these direct costs of schooling represent 12% and 28% of the full poverty line respectively. Private school is over nine times more costly at the primary level and four times more costly for secondary school. Formal fees for public schools constitute only a small share of the cost. In fact, informal fees (covering fundraisers, field trips, extra-curricular activities, etc.) are much higher than formal fees. The largest single cost of attending school appears to be supplies (uniforms, materials), followed by transport and books. Higher spending by the non-poor - particularly for key educational inputs - could signal differences in the quality of education being received by the poor and non-poor. There is substantial variation in costs by poverty group and geographic area, with much higher spending among the non-poor and in urban areas. These differences arise primarily in spending on school supplies and books; as discussed below, an important share of poor and indigenous students do not possess textbooks. Spending on informnal fees - which contribute to educational and extra-curricular activities - is also higher among the non-poor and in urban areas. Source : Panama LSMS, see Annex 4 for details. Disparities in Internal Efficiency and the Quality of Education 2.9. Poor children tend to repeat grades more than non-poor children at the primary level. Twelve percent and nine percent of extremely poor and poor primary school students respectively report repeating their current grade, as compared with only four percent of non-poor children. Roughly double the share of indigenous primary students report repeating their current grade than those in urban and rural 13 In fact, national secondary school enrollment lags below what would be expected given Panama's level of per capita income. See Hicks and Peeters (September 17, 1998). 4 With an average total annual per capita consumption of B./4,8 12, or 5.3 times the full poverty line. 15 Figures from the LSMS household questionnaire. These results were confirmed in the LSMS community questionnaire, in which economic factors were given as the main reason children did not attend school. 16 Additional research (perhaps via focus group discussions) would be worthwhile to better pinpoint the factors underlying this disinterest. It could reflect social factors (parents who didn't attend school, lack of family support, peer pressure, etc.) or a lack of relevance of the curriculum. Lack of family support was cited as a key factor (after economic reasons) for children not attending school in the LSMS community questionnaire. " Lack of facilities is a much more serious constraint at the pre-primary level, though demand-side factors (such as poverty status, mother's education, etc.) do play a role in determining pre-primary enrollment. See Annex 13. 14 areas. Dropout rates are also higher among indigenous primary-school students.18 These disparities in internal efficiency could arise due to differences in the quality of education or demand-side constraints. They are less apparent at the secondary level, where very few poor students enroll. 2.10. The quality of education received by poor and indigenous children lags that received by the non-poor. There are two ways to assess educational quality: the quality of learning (via standard achievement tests) and the quality of inputs (materials, books, teachers, etc.). The LSMS suggests that the poor and indigenous lack access to key quality-related inputs. First, a large share of poor and indigenous children do not have textbooks (Box 2.5). Second, lower household spending on school supplies among the poor and indigenous suggests lower use of educational materials in their learning process (see Box 2.4). Third, a lack of bilingual materials and instruction at the primary level apparently creates an obstacle to better attainment among indigenous children (see Box 2.3). Fourth, poor and indigenous children are much more likely to belong to communities reporting an insufficient number of teachers. Finally, poor and indigenous children are much more likely to live in communities with dilapidated school facilities (infrastructure, classrooms, desks, etc.). Box 2.5 - Poverty and Access to Textbooks in Panama: Inadequacy of a Key Quality Input * Lack of books among the poor. Close to 30% of poor primary and 12% of poor secondary students do not have textbooks, as compared with only 5% of the non-poor (at both levels). "Lack of money" is the main reason cited by the poor for not possessing books (over half of those without books). One-third indicated that "they weren't asked" for books at school, suggesting a complete neglect of book use in these schools and a symptom of poverty (teachers may not ask students for books if they know they cannot afford them). + Lack of books among indigenous children. Close to 60% of all primary and one fifth of secondary students in indigenous areas do not have books (in any language). "Lack of money" is the main reason (70% of those without books). * Sources of books. The majority of poor and indigenous children who do have books purchase or borrow them. Very few receive free books from the Govemment or schools, although existing donations do appear to be well targeted: 9%/o and 2% of poor primary and secondary students respectively received free books, as compared with only 1% of the non-poor at both levels; 33% and 7% of pimary and secondary indigenous students respectively received free books. Source: Panama LSMS 1997 Household Questionnaire. C. HEALTH 2.11. Disparities in both health status and access to health care services are prevalent between the poor and non-poor in Panama. Good health contributes directly to individual well-being. It is also an important complement to labor in terms of its contribution to productivity. 2.12. Relatively good health indicators for the nation as a whole mask large differences and poor health status among the poor and indigenous. Although life expectancy overall is high for Latin America (74 years at birth), it is as low as 63 in the poorest regions in Panama!9 Likewise, in the poorest 40 districts, infant mortality rates are 40-50 per 1000 live births (on par with low-income countries), despite a national average of 19 per 1000. Intestinal diseases, malnutrition, and respiratory diseases still account for a significant share of deaths in provinces such as Bocas del Toro and the Comarca of San Blas; tuberculosis is also high in Bocas del Toro; and malaria is common in Bocas del Toro, Darien and Veraguas, all of which have a high concentration of poor and indigenous people?0 Moreover, as discussed in Chapter 1, malnutrition is more prevalent among the poor and indigenous than among the non-poor. Other social concerns frequently reported among the poor include violence, alcoholism, and teenage pregnancy (over half of pregnant teenagers are poor and one quarter of all pregnancies among the poor occur during the teenage years). 2.13. Available data suggest that the poor and indigenous have less access to health care. Access is determined both by the physical availability of services (as measured by the distance or time needed to " Of indigenous residents aged 13-17 who enrolled at some point, half did not complete primary school. This compares with only four percent and 13 percent of urban and rural youths respectively. " MINSA (September 1996). Health Statistics. Overall life expectancy in Panama is actually slightly higher than what would be expected given per capita incomes. Hicks and Peeters (September 1988). 20 MINSA (September 1996). Health Statistics. 15 reach the nearest facilities) and the ability of households to pay for such services. Though these two constraints call for very different policy responses, their effect is similar for the household: an inability to afford either to travel to distant services or to pay for nearby services (or both); Using the World Health Organization definition of access as living within one hour of a health care facilities (physical availability of services), roughly 20 percent of the population lacks access. LSMS data suggest that this gap is worse among the poor, particularly the indigenous. Among indigenous residents whodid seek health care when ill, the average time taken to reach health facilities was 52 minutes (Table 2.1). For those who didnot seek treatment, both cost and distance were key obstacles (over half of those in the poorest quintile and a quarter of those in the second quintile did not seek care when ill due to cost and/or distance). 2.14. The poor and indigenous are less likely to seek medical treatment in case of illness than the non-poor.21 Among those reporting illness or an accident, 45 percent of the poor consulted a medical professional, as compared with 64 percent of the non-poor. The share of those seeking treatment is even lower among the indigenous (Table 2.1). 2.15. The poor who do seek treatment are more likely to use public services - such as public clinics, hospitals and healtl centers - than their non-poor counterparts, though a substantial share of the non-poor do use public services (Table 2.1). The indigenous likewise rely substantially on public facilities (mainly centers and sub-centers), but also self-treat at home. The non-poor are more likely to visit doctors than the poor and indigenous, who also rely on nurses and relatives for treatment. 2.16. The low quality of health services also appears to be a problem facing the poor, particularly in rural and indigenous areas. At present, there are few incentives in the public health system to ensure quality of care. With limited access to services, poor consumers have little choice of provider. Health services are not adequately supervised and there are no appropriate accreditation and quality control systems in place. Though the LSMS did not collect many indicators of quality, the community survey does suggest that medical inputs are less available to the poor, who are more likely to live in communities reporting longer waiting times, unavailability of medicines, and sufficient medical equipment. Lack of medicines appears to be particularly problematic in rural areas, as does medical equipment in indigenous areas. Poor quality of care was the primary reason for not seeking care for three percent of those in the poorest quintile who were ill but did not seek treatment. 21 Data on health consultations should be treated with caution as they are self-reported (cross-country evidence suggests that the non-poor are more likely to report illnesses than the poor) and the poor tend to self-treat more than the non-poor. 16 Table 2.1 - Health Services by Poverty Group and Geographic Area |Extreme All I Non- I l R l I Poor I Poor Poor Urban Rural Indig. Total Health Visits, Sickness During Past Month Consultancy Ratea 0.7 1 0.8 1.2 1.2 1 1.0 0.6 1.1 Coverage Rateb 0.4 0.5 0.6 0.6 0.5 0.3 0.6 Concentration Ratec 1.9 11.8 1.9 1.9 11.8 2.0 1.9 Place of Treatment Public Facility 81% 83% 62% 62% 77% 81% 68% Private Facility 19% 17% 38% 38% 23% 19% 32% Public Hospital/Clinic 27% 37% 42% 47% 33% 20% 41% Health Center/Sub-Center 55% 46% 20% 15% 44% 61% 27% Private Hospital/Clinic/Physician 7% 9% 32% 32% 17% 2% 26% Pharmacist/Other 3% 4% 4% 5% 3% 1 % 4% At home 9% 4% 2% 2% 3% 16% 3% Distance to Treatment (Average minutes to place of treatment, excluding treatment at home) Average minutes (for those who sought treatment) | 37 T 32 27 24 35 52 29 Private (Household) Spending on Health Services % of sick paying something (B./> 0) 58% i 61% 1 72% 1 71% 67% 47% T 68% Source: LSMS 1997. (a) Consultancy rate = number of visits/ number of sick; in other words, the number of visits on average for an individual reporting illness. (b) Coverage rate = number of people with at least one visit/ number of sick; in other words, the percent of people reporting illness who had at least one medical visit (c) Concentration rate = number of visits/ number of people with at least one visit. D. PHYSICAL ASSETS 2.17. The distribution of physical assets - such as housing and land - is highly unequal. Such assets contribute to income-generating potential by providing a base for productive activities. Property also generates rent - through earnings charged to renters or via the savings from imputed rent, which would otherwise be spent by the household. Households can also use property as collateral for leveraging credit. Emergency income can likewise be generated through sales of property or borrowing against it (equity loans). Finally, housing can be used as a tool for extending personal relationships, building trust, and generating social capital. The ability of households to use property as an asset depends largely on their quality, the security of tenure, and the flexibility of land and housing markets. 2.18. The poor tend to live in much lower-quality housing than the non-poor. Although there are significant cultural differences among the poor in Panama, the LSMS suggests many similarities in their housing conditions. A typical poor household lives in an individual house (65 percent) or hut (choza, rancho; 28 percent). Their homes tend to be quite crowded (averaging three members per room, as compared with 1.2 among the non-poor), particularly among the indigenous (over four members per room). The materials used to construct these homes are less durable than for the housing of the non-poor; the average estimated value of housing of the non-poor is more than double that of the poor. 17 2.19. The majority of the Fi population lacks titling for their g homes, but the gap is worse for the Poverty and the Lack of Property Titling: poor. Most households in Panama own % of property owners w ithout docunmentation (registered or their homes, including the poor 0 unregistered tites or deeds; LSMS 1997) (76 percent overall, 82 percent of poor 0 100% households). However, formal proof of ' 80% - 64% ownership - titles or deeds (either cm X 60% 50% 51% registered or unregistered) - is rare, a 40%0- particularly among the poor 20%- (Figure 2.3). 3I 0% -_ _ _ ,_ l_ lll ( 2 All home Pbor Non- All land Pbor Non- 2.20. The distribution of land is ow ners home Pbor ow ners land Pbor highly unequal, particularly untitled owners home owners land land. The poor, who account for two owners owners thirds of the rural population, own less than one third of agricultural land and 16 percent of titled land. Of those with land, non-poor households own 3.6 times as much land as poor households (with average holdings of 36 and 10 hectares respectively), and nine times more titled land. The Gini coefficient for total land owned was 77 in 1997.22 2.21. Land - particularly untitled land - may be a constraint for the poor. Half the poor and 70 percent of the non-poor in rural areas are landless. Among the poor, those with small- and medium-sized plots have the lowest incomes - even lower than the landless. Only those with larger farms (greater than 15 hectares) have incomes that approach those of the landless.23 2.22. The lack of property titles reduces the ability of the poor to obtain credit (since titles are often required as collateral). In fact, as discussed in para. 2.30, lack of guarantees (property titles or other assets) was the main reason poor households were refused credit. Without formal claim, the poor also cannot sell or borrow against this asset for emergency income. Titling contributes significantly to land values (adding an average of B./456 per hectare taking into account other differences in land value), lending promise to the use of a titling program as part of an anti-poverty strategy. 2.23. The poor also have fewer durable goods than the non-poor. On average, the total value of the durable goods owned by the non-poor is 16 times higher than the value of those owned by the poor. In addition to the overall store of wealth that can be used for emergency income, the poor lack the conveniences and productive opportunities that these assets can help generate. E. BASIC INFRASTRUCTURE 2.24. The poor - particularly the rural poor and the indigenous - also lack access to many basic services (Box 2.6). Households without basic services (water, sanitation, electricity, telephones) are more vulnerable than those who command these services. Potable water and sanitation services contribute directly to overall welfare and health status. Electricity and telephones help households use their homes productively for income generation. 22 Source: Panama LSMS 1997. A previous national estimate put the Gini coefficient for land at 87 in 1990. These compare with 84 in Guatemala (1979), 81 in Costa Rica (1973) and El Salvador (1971), 78 in Honduras (1974), 56 in Indonesia (1973), 53 in the Philippines (1980), and 35 in Korea (1980). Source: Deininger (1997). 23 This does not take into account differences in land quality. While the LSMS does not include information on many indicators of land quality (climate, soil type, slope, etc.), it does suggest that the poor have very little access to irrigated land (see Annex 5). Other indicators could be included in the next LSMS through the use of GIs data collection techniques. 18 Box 2.6 - Disparities in Basic Infrastructure Services: Key Gaps Facing the Poor in Panama, by Geographic Area Non-Poor All Poor- Urban Poor Rural Poor Indigenous Poor Piped Water: (a) Access (house, yard, community) V 95% 4* 72% ' 92% * 72% !! 42% (d) o/w % with uninterrupted service *: 70% *: 46% * 65% * 45% ! 23% (d) Sanitation Services: 4 .. 4. 4.> ! Fonnal connection (sewer/septic) 70% 15% 44% 6% 8% Latrine/pit 28% 66% 48% 80% 32% No facility 2% 20% 8% 14% 60% Garbage Collection: _ 73% 4 20% * 77% = 6% * 0% Energy: Electric lighting in home (b) / 92% * 45% v 92% * 38% O 7% Gas/electricity for cooking (c) 1 93% * 45% _/ 91% !! 38% !! 8% (e) Source: Panama LSMS 1997. See Annex 4 for detailed figures. (a) Other sources include wells, river, other; uninterrupted service refers to receipt of water 30 days/month, 24 hours/day. (b) Other sources include candles, gas, fuel, nothing. (c) Other sources include firewood, other (d) 57% of indigenous poor obtain water from rivers/lakes; only 3% treat their unpiped water. (e) 92% of indigenous poor use firewood. I Represents fairly comprehensive coverage, not a key priority. C*> Represents incomplete coverage. '! Represents very low coverage and key priority. 2.25. Although the majority of the poor have access to piped water, a large share of indigenous residents lack access, and service quality problems abound. Three-quarters of the poor receive piped water.24 Close to 60 percent of indigenous residents, however, lack any form of connection to piped water; the remainder collect water from rivers, lakes, rain or other sources25 The lack of access to potable water in indigenous areas is a likely cause of the high levels of diarrhea and hence malnutrition among indigenous children. Although the Government is undertaking a number of projects to extend piped water to rural and indigenous communities, anecdotal evidence suggests problems with overlapping efforts and a lack of coordination in service delivery and approach. Of those who do receive piped water, a large share of the poor in all areas report problems with service quality (irregular flow, inadequate water pressure, contamination and pollution problems, and inadequate maintenance). Higher efficiency and cost recovery in the provision of water services could help finance improvements in quality and coverage of services (Box 2.7). 2.26. There are large disparities in access to sanitation services by area and by poverty group. Overall, 55 percent of the Panamanian population has connections to the sewer system or a septic tank. Access, however, varies greatly by poverty group, with formal connections serving the majority of the non-poor but very few poor households. Instead, most poor households have pit toilets or latrines, while one-fifth have no formal toilet system. Differences are also apparent by geographic area, though the pattern of inequities in service delivery remains. In urban areas, half of poor households rely on pit toilets or latrines, raising public health concerns. Sanitation connections are rare in the countryside, particularly among the poor. While the majority of poor rural (non-indigenous) households have pit toilets or latrines, only a third of indigenous residents do, with the remainder having no formal toilet access. With respect to garbage collection services, differences are largely geographic, with most of the urban population covered (though qualitative data suggest that services are sporadic), and few rural and indigenous residents receiving the services. 24 The majority of the poor teceive piped water on their property (en elpatio), but not in their homes. 25 The lack of access to piped water is particularly widespread among the Ngobe-Bugle. 19 Box 2.7 - Incomplete Cost Recovery for Water and Electricity Incomplete cost recovery for basic services leads to waste, and lower quantity and quality of service provision. Water. Close to one-fifth of Panamanian households with water connections do not pay anything for their water consumption.26 This figure is higher for the poor (27%) than the non-poor (15%), and is particularly high among the urban poor (40%/0). For those who do pay something, the cost is minimal: the poor pay less than B./8.4 per person per year (2% of their average total consumption) while the non-poor pay B.3 5 per person per year (1% of their average total consumption). Electricity. Some 13% of households with connections to the electricity grid do not pay anything for their service. Illegal connections (tapping into public lighting for example) appear to be particularly widespread among poor city dwellers: 40% of poor urban households with connections did not pay for their consumption. Source: Panama LSMS 1997. 2.27. While virtually all urban residents have access to electric lighting and gas for cooking, coverage is extremely limited in rural and indigenous areas, particularly among the poor. The majority of the poor in rural and indigenous areas rely on kerosene lighting and firewood for cooking. The widespread use of firewood for cooking in these areas raises serious concerns regarding deforestation27 in already marginal areas and respiratory diseases. Low cost recovery and efficiency reduces the funds available for expanding electric service coverage (Box 2.7). 2.28. Virtually no poor households have telephone connections. Only 14 percent of all households in Panama report owning telephone lines and only two percent of the poor have a line. Box 2.8 - Transportation Bottlenecks and the Urban Poor Urban communities cited transportation - inadequate public services, deteriorating roads, and traffic congestion - as a top priority. The urban poor are largely reliant on public transportation. Half of all poor urban residents (71% in Panama City and San Miguelito) rely on public transportation to commute to work. The main form of public transportation is buses, which are run by private operators. Relatively low fixed prices and a complex system of licensing for entry and route assignments have dampened competition in service provision and greatly reduced the quality of services. Commuting times are lengthy: on average, it takes the poor close to one hour to commute to work using public transportation, thereby reducing time available for productive activities. The Government is currently exploring ways to reduce distortions in public transport, improve the quality of service provision, and strengthen the regulatory framework for vehicle standards, inspection, and pollution control. Source: Panama LSMS 1997. See Annex 7 for details. F. FINANcIAL ASSETS 2.29. The non-poor are much more likely to save than the poor. Savings28 can be an important asset for the poor, for consumption smoothing and investment purposes. Only eleven percent of the poor maintained savings, however, compared with close to half of the non-poor. Geographic differences are also notable, with half of urban households maintaining savings as compared with one fifth and one tenth of rural and indigenous households respectively. The poor tend to put their savings in public institutions, while the non-poor are more likely to use private banks. 2.30. The overall volume of lending to the poor is much smaller than their contribution to the economy. Borrowing and making purchases on credit are another important way for households to smooth their consumption and invest for future earnings potential. Whereas poor households receive three percent of total credit, they account for ten percent of total consumption and income in the economy. Information constraints, physical distance, lack of guarantees, and high costs-per-dollar borrowed could present obstacles to lending to the poor. Thus: 26 This compares with Ecuador, for example, where 74 percent who are connected pay nothing for their watelservices. Source: World Bank, Ecuador Poverty Report, Volume 1, June 30, 1995. 27 Most deforestation in Panama, however, is not caused by households, but by forestry and livestock enterprises. 28 The Panama LSMS questionnaire did not include questions on household food stocks, a form of in-kind savings that can be quickly liquidated for cash. Hence a complete picture of savings cannot be assembled. In addition, data collected by the LSMS on amounts saved per month and total savings are not consistent and the figures cannot be reconciled. Use of these figures is not recommended. Future LSMS questionnaires should try to correct for these errors. 20 * A smaller share of the poor solicited credit than non-poor households. Although the approval rate is fairly similar (slightly higher among the non-poor), the expectation of rejection was an important deterrent for soliciting credit among the poor. Other factors that discouraged the poor from soliciting credit include risk and excessive paperwork. * Lack of guarantees (titles or other assets) is the main reason that the poor were refused credit in all areas. The size of land holdings and the possession of land titles are the two main factors influencing access to, and volume of, agricultural lending in rural areas. * Physical distance to lenders affects the sources of loans obtained by the poor. Physical distance affects the ability of the borrower to access lenders and the ability of the lender to assess the borrower's credit risk and supervise repayment. The sources of credit accessed by the rural poor, for example, are on average only three kilometers away, compared with ten kilometers for the non-poor. Formal credit sources (private banks or producers associations) are much further from borrowers than government programs, cooperatives, moneylenders, NGOs and family. Thus the more distant, formal sources seek collateral to substitute for knowledge of the borrowers, and thereby exclude the poor. * The poor borrow a greater amount for consumption purposes, and the size of loans they obtain is much smaller than those obtained by the non-poor. Three quarters of borrowing by the poor is for consumption purposes, as compared with about half for the non-poor. Average annual borrowing by the poor is B./730 as compared with about B./4,600 for the non-poor. * Poor borrowers pay a higher average annual interest rate for farm lending. Effective annual interest rates vary greatly with source, term, collateral, and riskiness of the borrower. In rural areas, poor borrowers pay a median rate of 24 percent versus 12 percent for non-poor borrowers. * Repayment varies significantly by lender, with a high delinquency rate for government programs. Among those who repaid something, median repayments were only 12 and 17 percent for government banks and programs respectively, but 100 percent for all other sources of cash loans. G. SOCIAL CAPITAL 2.31. Social capital is one of the assets Figure 2.4 - Poverty and Social Capital of the poor.2 Social capital - defined as nor the rust, andciprocapity netwok tat Percent of individuals in each quintle living in norms, trust, and reciprocity networkg thatcnTkeswRloardhgScilapl facilitate mutually beneficial cooperation in a community - is an important asset that 40% 0300% -X-Low SCl can reduce vulnerability and increase ° X20% - -o-- :HighSCl opportunities. The poor and the extreme 7 20x poor account for a disproportionate share 0% of those living in communities with high 1 2 3 4 5 social capital.30 Among those in the Populabon Quintles poorest quintile, it is more than double the share living in communities with low social capital (Figure 2.4). The reverse is true for the highest quintile. 2.32. The poor tend to associate for "public goods" purposes, whereas the non-poor join associations which yield higher private gains. The poor tend to participate more in community committees and associations than the non-poor, whereas wealthier households tend to participate in cooperatives. The poor are close to two times more likely to join local committees and four times more 29 The LSMS included a number of questions pertaining to social capital in both the household survey and the community focus group interniews. In addition, to gather in-depth information on social capital, focus group discussions were conducted in 16 communities as part of a complementary Social Capital Qualitative Survey (SCQS). 30 Though this could be explained by cultural differences, with over-representation of the indigenous (who have higher social capital) in the lower quintiles, social capital does appear to be stronger among the poor even within geographic areas (within urban areas for example). 21 likely to participate in local associations than the non-poor. In contrast, wealthier households were close to three times more likely to join cooperatives than the poor. These patterns suggest that the poor substitute community action for a lack of public services. 2.33. Social capital helps communities to leverage assistance. Communities with high social capital report a higher frequency of assistance from the Government and NGOs. Overall, they are close to four times more likely to receive Government assistance and two times more likely to receive NGO assistance than those with low social capital. This correlation is observed in all three geographic areas, and is particularly strong in rural and indigenous areas. 2.34. The SCQS and the LSMS reveal a sharp contrast between urban, rural and indigenous communities in terms of the prevalence of community organizations and social capital.. Some 75 percent and 80 percent of rural and indigenous communities in the LSMS sample affirm having some sort of community organization, as compared with just half of urban communities in the sample. Indigenous households also participate in organizations much more than their urban and rural counterparts (40 percent of indigenous households report participating in some type of organization, as compared with only 28 and 30 percent of urban and rural households respectively). Not surprisingly, indigenous and rural communities also have higher social capital than urban communities,31 reflecting more numerous and active networks of associations in these communities. The SCQS found that horizontal connections are stronger among communities with higher social capital, and consequently among indigenous communities. These horizontal connections manifest themselves either through different organizations in the same community cooperating with each other, or through the establishment of links with groups in neighboring communities. Box 2.9 - Community Perceptions of the Quality of Life: Urban Pessimism & Social Problems * A Paradox of Perceptions? Community perceptions of changes in overall well-being contradict the profile of poverty as measured by annual consumption. Despite having higher average total consumption and a lower incidence of poverty, urban communities have more pessimistic perceptions of recent changes in their overall living standards than their "poorer" rural and indigenous counterparts. In contrast, urban communities were more optimistic when asked about specific changes in living conditions (basic services and housing). Clearly non-economic factors are causing the apparent paradox between economic conditions and community perceptions. * Social Problems as Priorities. Social issues - particularly crime and safety but also drugs, alcohol addiction, gangs, and youth problems - were the most frequently cited priority among urban communities. Over 8O0/o reported problems with theft, drugs and alcohol, and close to half reported problems with gangs and prostitution. Official statistics suggest that crime and violence are rising, and that most crime (particularly violent crime) is concentrated in urban areas, particularly Panama City and Col6n. * Poverty, Crime and Violence. Although the LSMS does not reveal a strong correlation with poverty, crime and violence can exact a high cost on the poor, who are more vulnerable to shocks, in a number of ways: * Direct costs: Stolen property, lost productivity, lost life, destruction of public infrastructure, etc. * Prevention costs: lnvesttients in metal doors, window bars, etc. Common in the low-income barriadas surrounding Panama City. * Lost employment opportunities: Deterred investors in high-crime areas; reduced safety for commuting, etc. * Erosion of social capital: reduced trust in communities and between households limits the use of this important asset. See Sollis et al. (1995) for a discussion of crime and violence in Panama's barriadas and Moser (1996) on impact of crime and violence on the poor. 3' As measured by a Social Capital Index (SCI), see Annex 18. 22 Chapter 3 - The Determinants of Poverty, Malnutrition, and Inequality 3.1. Chapter 2 exarnines the disparities in key assets between the poor and non-poor and by geographic area. This chapter analyzes the relative importance of these and other determinants of poverty, malnutrition, and inequality in a multi-variate setting. It summarizes the key results for each problem, highlighting key priorities for policy interventions. Two key conclusions emerge from the multi-variate analysis of all three problems (Box 3.1): * The correlates of poverty, malnutrition, and inequality overlap substantially - and most involve crucial policy levers. * Disparities in key assets - labor, education, basic services, and physical and financial assets - are indeed strongly correlated with poverty, malnutrition, and inequality. Box 3.1- Determinants of Poverty, Malnutrition and Inequality: Summary of Key Factors Poverty Nutritional Status Inequality Labor V Agriculture & informal sector . Sector of employment, earnings ______________ dependency ratio Education / Household head/companion V Mother's V Household head/companion, all members Basic / Depending on geographic area / Health services, vaccinations V Telephones, electricity, Services Sanitation (covers water) Physical V Housing quality & tenancy, V Durable goods, titled land V Housing tenancy, titled land Assets titled land holdings holdings holdings Financial V Credit v Credit Assets Location V Geographic area V Geographic area V Geographic area Fertility, V Household size V Fertility, pregnancies, number of V Household size Household children under 5 Structure Other V Social Capital (some types) V Economic situation (food V Prices and incentives Factors consumption) V Language V Diarrhea (correlation with water) V Exclusive breastfeeding v Age, gender of child_ A. PROBABILITY OF BEING POOR / CORRELATES OF POVERTY 3.2. This section examines the correlates of poverty in a multi-variate setting. The analysis is useful, first, to verify the relative role of the various factors in determining poverty status, and second, to assess the potential impact that policy-induced changes in these factors are likely to have on the probability of being poor, holding all other factors constant. The probability of a household being poor was examined for the nation as a whole, as well as for urban, rural and indigenous areas.' 3.3. It is important to note the limitations of this analysis at the outset. First and foremost, the analysis does not capture the dynamic impact of certain causes of poverty over time. Most notably, the impact of changes in economic growth - most certainly a key determinant of poverty - cannot be assessed using this static, cross-section model. Second, the analysis is limited by the variables available at the household level from the 1997 LSMS. Other factors -- such as social conditions like crime and violence, or physical conditions such as variations in climate or access to markets -- could not be included due to a lack of data at this level. Finally, though theory holds that many of the variables included in the analysis do indeed contribute to (cause) poverty (or poverty reduction), the statistical relationships should be interpreted as correlates and not as determinants since causality can run both ways for some variables. 'See Annex 8 for a more detailed description of the methodology and results. Annexes 5 and 6 include detailed models for rural areas and for the ethnic indigenous (as opposed to geographic areas). 23 3.4. Poverty is clearly associated with lower levels or constrained use of key assets, including labor, education, physical assets, basic services, credit and social capital. Geographic location and household size are also found to be important correlates of poverty. Key findings are summarized below. * Informal-sector and agricultural employment are correlated with poverty. Households deriving most of their income from agriculture or the informal sector have a higher probability of being poor due to lower earnings (and hence productivity) in these sectors. * Education can serve as an important elevator for escaping poverty.2 The higher the educational attaimnent of the household head or his/her companion, the lower the household's probability of being poor. Overall, if a household head or companion has completed primary education, the household is close to ten percent less likely to be poor than if he/she had no education. Even a few years of secondary education further reduce the probability of being poor. * Low quality housing and lack of housing titles are strongly correlated with poverty, though the direction of causality is not clear. Poverty itself is a cause of makeshift housing, but low quality housing can also limit the ability of households to use their homes as a productive asset - as a location of independent businesses for example (particularly when accompanied by a lack of basic infrastructure services as discussed below). Not owning a house is also associated with a higher probability of being poor, as does the lack of a registered housing title in urban and rural areas. * Small land holdings are associated with poverty, even if they are titled. Households with titled land holdings greater than 15 hectares are less likely to be poor. Land holdings of less than 15 hectares does not lower the probability of being poor. * Access to basic services significantly reduces the probability of being poor for the nation as a whole (Box 3.2). The correlation between basic services and poverty depends largely on the existing coverage of these services. In urban areas, water and electricity are not significant determinants of poverty status because most city-dwellers (including the poor) are covered by these services. Likewise, in indigenous areas, lack of sanitation services and formal water supply do not register as significant correlates of poverty because almost no indigenous households have access to these services. Box 3.2 - The Correlation Between Basic Services and Poverty in Panama Nation Urban Areas Rural Areas Indigenous Areas Water Supply X __X Sanitation Services x v X Electricity . X _ _ _ Telephone Services _ v_ _ w' Indicates significant correlation between the lack of access to the service and a higher probability of being poor. XIndicates no significant correlation between service and probability of being poor. * Limited access to credit is strongly correlated with being poor, though the direction of causality is unclear. Low incomes and assets could prevent the poor from obtaining credit; likewise, low access to credit could hamper the poor's ability to use their labor and other assets productively. * Membership in certain types of organizations reduces households' probability of being poor. Households that belong to a cooperative - which generates "private gains" to members - are less likely to be poor, particularly in urban areas. In indigenous areas, participation in community organizations (such as committees, associations,juntas, and consejos) is associated with higher total consumption, which is consistent with the findings discussed above. * Even after accounting for disparities in assets, geographic location is a major factor in explaining poverty in Panama, particularly for indigenous poverty. Households in indigenous and 2 The analysis uses the maximum educational attainment of the household head or his/her companion. Since the educational attainment of these members (adults) precedes their current economic status, it could validly be considered as having a causative influence on poverty status (whereas the educational levels of young dependents in the household may be low because poverty prevents them from affording an education). 24 remote areas are 25 percent more likely to be poor than urban households, even after other differences are taken into account. Finally, living in rural (non-indigenous, non-remote) areas increases the probability of being poor by nine percent. * Larger households tend to be poorer. Overall, each additional child under five increases the probability of being poor by 16 percentage points. The correlation between fertility and poverty is particularly strong in rural areas, underscoring the importance of access to family planning services. B. DETERMINANTS OF NUTRITIONAL STATUS 3.5. Malnutrition is strongly correlated with poverty in Panama, as discussed in Chapter 1. Not surprisingly, the determinants of nutritional status largely coincide with the correlates of poverty. The key determinants of nutritional status are summarized below. * More education is associated with better nutritional status, holding consumption and other factors constant. Mothers' education significantly affects nutritional status: children whose mothers have some secondary education or more are less likely to be malnourished than those whose mothers have no education. * Access to health services improves nutritional status. Closer proximity to health facilities and timely receipt of vaccines are correlated with improved nutritional status. Children with recent bouts of diarrhea are significantly more likely to be underweight, underscoring the importance of access to water and sanitation services, as well as the availability of oral rehydration therapy and deworming interventions. * The possession of physical assets affects nutritional status. Children living in households with few durable goods - a proxy for wealth - are more likely to be malnourished. * Paradoxically, children of farmers are more likely to be malnourished. Children living in households with agricultural land holdings (even titled) of 1-15 hectares are more likely to be malnourished than those living in households within little or no land. * Geographic disparities are a significant determinant of nutritional status. Children in indigenous areas are significantly more likely to be stunted than those in urban areas. * Older children are more likely to be malnourished - which highlights the vulnerability of children during the weaning period. Boys are also more likely to be malnourished, which is a common result in an absence of discriminatory practices. - Competition with other children under five lowers nutritional status. The more members under five, the worse the nutritional status of each child. Similarly, a higher number of pregnancies reduces the nutritional status of each child. This could relate to health problems associated with short birth spacing (for both mother and child), shorter periods of exclusive breastfeeding for each child, and competition for food and child care among siblings. These findings highlight the need for increased availability of family planning services and outreach. * Exclusive breastfeeding strongly improves nutritional status. The benefits of breastfeeding (stronger immune systems, safe source of nutrients, etc.) are well known. The determinants of nutritional status were ascertained using several multi-variate linear regression models on z-scores expressed in percentage terms (continuous variables) for height-for-age (indicator of chronic malnutrition or stunting) for children under five and for weight-for-age for children under five. Other variables that were tested but were not significant include an index of housing quality, lack of sanitation services, rooms-per-person, lack of electricity, cooking with firewood, lack of piped water (with and without treatment), garbage dump in yard, gender and marital status of household head, childcare providers, health status of mother (smoking, illness), etc. Some of these variables may indeed have an impact on nutritional status but di'd not prove significant due to correlation with other variables (e.g., water with diarrhea and indigenous areas). See Annex 9 for details. 25 C. SOURCES OF INEQUALY 3.6. In Panama, the sources of inequality overlap substantially with the determinants of poverty. As with poverty, disparities in assets - labor, education, housing, land, basic services, and credit - explain a large share of inequality (Table 3.1). Other sources of inequality include geographic location and household size and composition.4 Two-thirds of Panama's consumption inequality can be explained by disparities in labor, education, land, certain services, geography and household sizer Education constitutes the single most important determinant of inequality, reflecting the large disparities in access to education (particularly secondary and higher education) and the large wage differentials associated with the various levels of educational attainment. Table 3.1- Decomposition of Inequality in Panama________ ___ Group Sub-Group % Inequality of Explained by: I Labor # of workers per household member 14% 2 Formal vs. Informal Earnings (largest source in HH) 13% 3 Public/Private/Canal Earnings (largest source in HH) 7% 4 Years of Education All household members 13% 5 Maximum for HH Head or Companion 41% 6 Household Head 39% 7 Housing Ownership With registered title vs. no title vs. doesn't own house 15% 8 Hectares of .Owned (6 groups) 11% 9 agricultural land 'Worked (6 groups) a 14% 10 ____ With title (6 groups)' 5% 11 Access to No telephone line (vs. possesses line) 35% 12 Basic Services No electricity connection (vs. has connection) 22% 13 No sanitation services (vs. has services) b 33% 14 Credit More than B./100 in annual credit obtained 6% 15 Geography & Four Areas: Urban, Rural, Indigenous, Remote 29% 16 Language Indigenous Areas vs. Non-Indigenous 10% 17 Indigenous Language vs. Non-Indigenous Language 8% 18 Gender Male/Female 0% 19 Household Size # members from 0-23 years of age 26% 20 # members from 24-59 years of age 2% 21 # members aged 59+ 0% Total: % of inequality explained by factors 2, 5, 9, 13, 15, 19 61% Based on a decomposition of Entropy Measures of Inequality. Figures presented for the E-2 measure of inequality. See Annex 10 for detailed methodology and results.a\ Divided by sizes of 0-0.01, 0.01-1, 1-3, 3-10, 10-30 and more than 30 hectares. bSanitation services include connection to sewer system or septic tank. This variable also reflects access to water. 3.7. The long-term nature of many of these variables is significant: as international experience has shown, it is very difficult to reduce inequality. Most of these variables are somewhat structural (geography), difficult for Governments to change (land distribution, sectoral employment opportunities), or affect consumption over various generations (e.g., education and fertility).6 3.8. One measure that can reduce inequality in the short term is the reduction of trade barriers and price controls. As discussed in Chapter 4, simulations suggest that reductions in trade and price distortions for agricultural products are likely to reduce inequality: complete trade liberalization would reduce the consumption Gini by 0.6 points. Although small, such reductions are significant, particularly in light of the tendency for Gini coefficients to show little variation over time. 4The analysis reveals that gender (the ratio of males to females and the sex of the household) is not an important personal attribute in explaining inequality in Panama (or poverty, as discussed above). The unimportance of gender consistent with intenational evidence for developing countries (see Ferreira and Litchfield, 1997). It is important to note, however, that inequality within households is not capdtur by LSMS data. ' It is important to note that, with the exception of this "joint" analysis (analyzing the different variables together), the decompositon of inequality using each individual variable is not additive. See Annex 10 for additional explanations. 6 Even Chile, which has aggressively pursued structural policies to promote growth, and reduce poverty and inequaity for some time (trade policies, education reform and targeted public investments, etc.), has only reduced the Gini index of inequality by 1.5 percentage points in seven years. See Ferreira and Litchfield (January 1997) and World Bank (June 10, 1997). 26 Chapter 4 - Poverty, Duality and Distortions in Panama's Economy 4.1. Chapters 2 and 3 establish the links between disparities in the distribution of assets and poverty, malnutrition, and inequality. These disparities in economic opportunity reflect Panama's uniquely dualistic pattern of development, which has in turn been spawned by a legacy of policy distortions adopted largely in the 1970s. This chapter examines these distortions and their likely impact on poverty and inequality. It also reviews recent reforms that have sought to correct these distortions and highlights actions remaining on the reform agenda that are key for reducing poverty and inequality. A. THE LEGACY OF DiSTORTIONS AND DUALITY IN PANAMA'S ECONOMY 4.2. With a GDP per capita of US$3,080 (1997) and low inflation, the Panamanian economy is among the more prosperous and stable in the region. Its privileged geographic location and its monetary regime anchored in the use of the U.S. dollar as legal tender have fostered a comparative advantage in services, which contribute over three quarters of GDP and generate two-thirds of employment in Panama. Such services are becoming even more competitive with increased privatization (e.g., in ports services). 4.3. Panama's Modern Service "Enclaves" as a Source of Duality. These strategic factors have also spurred the rapid development of internationally-oriented, modern, dynamic service "enclaves," including the Canal Zone, the Col6n Free Zone, and the International Banking Center. While these enclaves generate an important share of GDP (eight percent for the Col6n Free Zone, seven percent for the Canal Zone) they create little employment (three percent of the labor force) or fiscal revenue that could be used for investment in other regions of the country. Moreover, they inject negative spillovers into the economy: the huge wage differentials between those paid in the enclaves, particularly in the Canal Zone where the U.S. minimum wage is enforced, and the rest of the Panamanian economy sharply raise wage expectations relative to the opportunity cost of labor. 4.4. Factor Market Segmentation. Indeed, factor markets have been segmented by policies that drive up the cost of labor and cheapen the cost of capital. Panama's labor market is characterized by a multiplicity of labor regimes (Box 4.1), which have created large wage differentials between workers in the Panama Canal Commission, public sector employees, and those employed in the rest of the economy. 4.5. These interventions not only hamper growth, but also have a direct link to poverty. Although poverty is measured by consumption rather than income (Chapter 1), income is a key determinant of consumption levels and the poor derive most of their income from labor earnings. By increasing the relative price of labor - and thus reducing demand for the poor's most abundant asset - these distortions swelled the ranks of the unemployed and encouraged informality. Indeed, unemployment rose sharply and remained fairly unresponsive to changes in growth following the introduction of the Labor Code in 1972.1 Moreover, while they benefit those who are lucky enough to find formal sector jobs, the resulting segmentation of the labor market can put a heavy toll on informal sector workers by reducing their wages, making it difficult for the working poor to grow out of poverty through their own labor. Indeed, the LSMS reveals that the poor in Panama do not benefit from the generous provisions mandated by labor legislation - and they may even be hurt by the distortions they create. A number of aspects support this conclusion: (i) the majority of the working poor receive wages that are below the lowest official minimum wage;2 (ii) they do not receive the fringe benefits mandated by the Labor Code (such as the "13'b month" salary bonuses); (iii) the majority of the working poor are employed in the informal sector, wherewages are lower and employment terms less favorable; and (iv) the working poor throughout the country appear to be underemployed and the urban poor are hurt by high rates of open unemployment. 'Following the introduction of the Labor Code in 1972, unemployment climbed from an average of 6.5% in the 1960s to an average 7.3b% in the 1970s and 11.9% in the 1980s. (Source: MIPPE/DPS). Moreover, Gonzalez (1998) shows that unemployment became less responsive (elastic) to changes in economic growth after the Labor Code was instituted. 2 Close to two thirds of extremely poor paid workers and over half of all paid workers living below the poverty line receive less than the lowest hourly minimum wage (as compared with only 17 percent of the non-poor). 27 Box 4.1- Panama's Labor Market: Multiple and Stringent Policy Regimes A multiplicity of policy regimes govern Panama's Labor Market: * The generous Labor Code, which was introduced in 1972 and modified in 1995, is the principle regulatory mechanism for the private sector. The Labor Code and associated legislation still distorts the private labor market in a number of ways: * A complex structure of minimum wages (by region and sector) and prohibition on wage reductions under any circumstances. * Short probationary period after which strong severance restrictions apply (the 1995 reform maintained the probationary period but removed the automatic conversion of workers to permanent status at its end) * Severe limits and penalties for dismissal. * Numerous fringe benefits including, inter alia, an extra "13'h month" salary bonus, length-of-service bonuses, and overtime bonuses. * Some 100 special labor laws govern public sector employment, with even more generous provisions for state workers and higher wage schedules than those of the Labor Code. * U.S. labor laws cover workers in the Panama Canal Commission (PCC) and this regime is being maintained as assets are transferred to Panamanian ownership. Wages paid by the PCC are far higher than those paid in the rest of the economy (even higher than those in the public sector). * A separate - and less restrictive - legislative regime governs employment in the Export Processing Zones. See Annex 11 for details. 4.6. The Traditional Sector: Further Distortions. The modem services sector is juxtaposed with a domestically-oriented traditional sector (primarily agriculture and industry), that has been constrained by policy-induced rigidities and low productivity for decades. Until recently, a highly distortedtrade regime (one of the most protective in Latin America), combined with a complex web of price controls in the domestically-oriented sectors, had the result of protecting inefficient sectors, distorting resource allocation, generating rents for certain groups, and raising the cost of the basic consumption basket. Simulations using the LSMS suggest that the net redistributive effect of such protection was highly regressive, effectively taxing the poor and increasing poverty and inequality (Box 4.2). Box 4.2 - Redistributive Effects of Trade Protection on Households Until recently, Panama maintained one of the most highly protected trade and pricing regimes in Latin America. These interventions had a direct redistributive effect on the economy, and thus a potential direct impact on poverty and inequality. The redistributive impact of complete trade and price liberalization for basic food items was simulated using household data from the LSMS. Food is important from a poverty perspective because food consumption represents 59% of the total budgets of the poor, and because agriculture generates 29% of the total incomes of the poor and 41% of the incomes of the rural poor. The net impact of complete trade liberalization on poverty and inequality is expected to be positive. Simulations suggest that: * Overall poverty would fall close to two percentage points from the 1997 level (from 37.3% to 35.6%). * Inequality would fall (with a reduction in the consumption Gini of 0.6 percentage points). * Poor households would incur a net gain of B./26 per capita or 5% of total consumption (as compared with B.157 per capita or 2% of total consumption for the non-poor). * Although some producers would lose in net terms (even taking into account their gains as consumers from the fall in the cost of basic staples), net losers represent only 2% of the population and less than one third are poor. * Despite the fact that very few poor producers (under 15,000 people in total) would lose in net terms from an elimination of trade barriers, a significant share of these poor net losers are concentrated in the Provinces of Darien and Veraguas. Geographically-targeted compensating measures might thus be warranted to protect poor producers in these regions from the localized adverse effects of possible future trade liberalization measures. See Annex 12 for details on tariff rates and the methodology used in the simulations. 4.7. Effects of Distortions and Duality. The highly protectionist and distortionary policies ultimately stifled growth, and generated a large external debt. They also exacerbated poverty and engendered a highly unequal distribution of income. In the 1980s, these distortions and the confrontation with the U.S. plunged the economy into severe crisis. By end-1989, GDP had fallen by 16 percent, mismanagement and external sanctions had demonetized the economy, and unemployment had climbed to 20 percent (using the national definition). 28 B. RECENT REFORMS: REDUCING DISTORTIONS IN PANAMA'S ECONOMY 4.8. Since 1990, Panama has gone a long way toward achieving political and economic recovery. In the first several years, growth was based mainly on a boom in construction and other non-tradables, as well as arrears accumulation. By 1994, GDP reached pre-crisis levels, but, in the absence of fundamental structural changes, growth fell back to a modest two percent. 4.9. In an attempt to stimulate more inclusive growth and reduce poverty and inequality, the Government has undertaken a number of fundamental reforms since 1994. It has dismantled a labyrinth of trade barriers and price controls, launched a far-reaching privatization program, issued anti-trust and consumer protection legislation, unified fiscal incentives for manufacturing firms, and adopted modest but important reforns to the Labor Code (Box 4.3). 4.10. After several years of little growth, the economy began to respond in 1997, despite a series of climatic and economic shocks.3 Real GDP rose from 2.9 percent in 1994 to 4.7 percent in 1997. Unemployment started to fall for the first time in years, declining from over 14 percent in 1994 to 13.2 percent in 1997 (using the national definition). Even more important for the long-term, employment has apparently become more responsive to changes in growth since the introduction of modest, but important, reforms to the Labor Code in 1995 4 Box 4.3 - Key Achievements of the Economic Reform Program Liberalizing the Trade Regime. Reduced tariffs on many commodities, transformed specific tariffs into ad valorem rates, eliminated non-tariff barriers to imports, removed price controls. Transformed most highly distorted trade regime in LAC to one of the most open. Promoting Domestic Competition. Passed the Competition Law, which removed remaining price controls and established anti-trust and consumer protection framework. Reforming the Labor Code. Modest reforms to the Labor Code, including: (i) curbing the intervention of government in employer-employee relations; (ii) giving employers more flexibility to reassign employees within firms; and (iii) easing the automatic conversion of employees to permanent employment status after the three-month probationary period. Reforming and Privatizing Public Enterprises. (i) Privatizing the state telecommunications institute (INTEL); (ii) reforming pricing principles and passing law to permnit private generation of electricity; (iii) privatizing service delivery functions of state port authority (APN); (iv) opening traditionally state-dominated activities in the financing, construction, operation, and maintenance of major highways. Unifying Fiscal Incentives in Manufacturing Sector. Passing Law 28/1995 "Ley de Universalizaci6n a los Incentivos Tributarios" to equalize fiscal incentives for all manufacturing firns irrespective of their size. Sources: Republic of Panana: Desarrolo Social con Eficiencia Economica 1997-1999, Cabinet Resolution No. 222 of September 19, 1997 World Bank (1998): Economic Reform Loan (ERL), Implementation Completion Report. C. CONSOLIDATING ECONOMIC REFORMS 4.11. Though progress has been made in reducing some distortions and disparities in economic opportunity, an important economic reform agenda remains for Panama to sustain growth and reduce unemployment.5 With timely implementation of these reforms, growth in real GDP is expected to climb from four percent in 1998 to five percent by 2001.6 Assuming a constant distribution of consumption, these expected growth rates could reduce poverty by up to 3.3 percentage points by the year 2001 (from 37.3 percent in 1997 to 34.0 percent in 2001). Extreme poverty, which tends to be more stable, is projected to fall by 2.0 percentage points over that same period (from 18.8 percent to 16.8 percent). 3Notably, the El Niflo climatic phenomenon, a downturn in Canal operations (due to decreased grain imports by China), and declining U.S. Defense Department demand. 'Gonzalez (1998). s As discussed in detail in the Country Assistance Strategy for Panama, World Bank (October 29, 1998). 6 See Country Assistance Strategy for Panama for details on base case reform and growth projections. 29 4.12. In addition to fostering growth, which is critical for poverty reduction, four areas also have the potential to weaken disparities in key assets and economic opportunity and are crucial for reducing poverty and inequality: * Deepening reforms to the labor code and related legislation to ease restrictions on dismissal and lower the relative cost of labor. Such distortions continue to hurt the poor by encouraging unemployment, underemployment, and informality. Further reforms to reduce these distortions would allow the poor more flexible use of their most abundant asset. * Expanding the land tenure regime, by modernizing land administration services and launching a massive national land titling program focusing on peri-urban and rural areas. This program should fill the gaps in land titling that currently prevent the poor from leveraging credit and buying and selling property. It would also have a likely impact on inequality, since disparities in titled property assets explain a significant share of inequality in Panama (see Chapter 3). * Continuing trade reforms. Though tariffs have come down substantially, further reductions are still needed (particularly for rice and milk). Simulations using the LSMS suggest that further reductions in tariffs would indeed reduce poverty and inequality (Box 4.2). Although some producers would lose in net terms from lower prices (even taking into account their gains as consumers), these net losers represent a very small share of the population and the majority are non-poor. * Reducing distortions in public and freight transport. The Government should actively pursue reforms to reduce distortions in the provision of public transport, which is largely used by the urban poor. Barriers to competition in freight transport would also reduce the cost of living in general. 30 Chapter 5 - Poverty and Social Spending 5.1. Disparities in human Figure5.1 capital reflect a lack of targeting g of social policy and public resources. Despite high public Public Spending on Social Sectos in Panama, 1997 spending on the social sectors, 60%0l inefficiencies and inequities have 50% 46% %ofSocialSpending: prevented improved outcomes for 5 the poor (Figure 5.1).' Although 40/0o the Government has attempted to 26% 290/o 30% compensate for some of these 21% biases against the poor by 20% 9% developing a number of social 10% s 5% assistance programs, the impact of 0%__ these programs could be improved through systematic targeting and v c improvements in program delivery. Given the clear links between , disparities in human capital and poverty, inequality and malnutrition established Chapters 2 and 3, this chapter highlights the crucial need for a reallocation of public spending on the social sectors toward areas that benefit the poor. A. INEQUITIES AND INEFFICIENCIES IN THE PROVISION OF PUBLIC EDUCATION 5.2. With close to six percent of GDP allocated to public spending on education, Panama ranks among the highest in Latin America.2 Moreover, education has absorbed an increasing share of resources, rising from 4.7 percent of GDP in 1992 to 5.6 percent in 1997 and from nine percent of total public spending in 1992 to twelve percent in 1997 (Table 5.1). The LSMS reveals that household spending on education accounts for about three percent of GDP, yielding a total of nine percent of GDP (public and private spending) allocated to education in 1997. 5.3. While the overall level of public expenditures on education appears to be adequate, its distribution is inequitable. Indeed, disparities in educational outcomes largely reflect biases in public spending: * Overall, public spending on education is regressive, with the non-poor benefiting more in absolute terms. Whereas the poor represent 43 percent of the school-aged population aged 5-24, they only receive 28 percent of total public spending on education (Figure 5.2). In relative terms, however, subsidies are more important to the poor, representing 35 percent of the poor's average per capita total consumption as compared with eleven percent for the non-poor. 'A lion's share (93 percent) of spending on labor and social welfare is absorbed by the Social Security Fund's Disability, Old Age, and Life Insurance Fund (invalidez, vejez, y muerte under the Caja de Seguridad Social, CSS). The remainder goes to employment and training, worker- employer relations, sick and matemity leave benefits, a CSS fund for professional risks, welfare and social assistance, and administration and regulation. The majority (83 percent) of housing spending goes to investments in public housing channeled through the Ministry of Housing (MIVI), the CSS, and the National Mortgage Bank (BHN). Social Investment Programs include spending by the Ministry of the Presidency on the Social Fund (FES) and programs supported by the Office of the First Lady, and on MIPPE/MEF's Local Investment Program (PROINLO). 2 Other top spenders include Mexico (5.8%/0), Venezuela (5.1%), and Costa Rica (4.70%). Like Panama, these countries not only allocate a high share of GDP to public spending on education, but such spending is high in on a per capita basis in absolute purchasing-power-parity (PPP) terms. The average share of GDP allocated to public spending on education in developed countries is about 5.3 percent. Sources: World Bank and UNICEF EdStats database; World Bank World Development Reports; MIPPE and MINEDUC. 31 * Public spending on pre-primary and secondary education is regressive in absolute terms (Figure 5.2). In relative terms, however, subsidies at these levels are more important to the poor than the non-poor (as a share of the poor's relatively lower average total consumption per capita). * Public subsidies to higher education are extremely regressive, in both absolute and relative terms. Virtually all public spending (94 percent) on higher education benefits the non-poor. Over forty percent of subsidies on public higher education accrues to the top quintile of the population. Moreover, public spending on higher education is regressive in relative terms, representing over five percent of the average per capita total consumption of the non-poor, compared with only three percent for the poor (despite the poor's having relatively lower levels of total consumption). Hence, with a lower rate of cost recovery at this level,3 the large govermnent subsidies to higher education truly benefit the upper echelons of Panamanian society. Table 5.1 - Public Spending on Education 1992 1993 1994 1995 1996 1997(P) Total Spending on Education (B./OOO) (a) 313,668 338,143 364,377 373,022 431,859 490,355 % Recurrent 95% 87% 92% 91% 91% n.a. % Investment 5% 13% 8% 9% 9% n.a. Total Spending on Education as Share of: GDP 4.7% 5.4% 4.7% 4.7% 5.2% 5.6% Total Government Spending (b) 9.2% 12.8% 12.0% 11.0% 11.5% 12.2% Total Gov't Spending on Social Sectors (c) 25.1% 26.6% 23.90/o 21.6%/e 23.9% 26.5% Education Spending by Level (% of total) (d) Primary and Pre-Primary 40% 30% 33% 34% 29% 29% Secondary 25% 37% 30% 29%' 34% 33% Higher 29% 27% 29% 30% 29% 30% Other 6% 7% 7% 7% 7% 8% Total Subsidies Per Student (B./) (d) Primary and Pre-Primary 358 326 342 351 342 374 Secondary 476 831 641 604 805 872 Higher 1,623 1,676 1,722 1,683 1,747 2,142 Sources: MIPPE/MEF: Direcci6n de Presupuesto de laNaci6n; MINEDUC; Contralorsiade laRepublica. 1997 figures are preliminary. a\ Data on education expenditures exclude spending on education investmnents by FES and PROINLO (demand-driven local investment programs) as well as the School Lunch Program supported by FES. b\ As share of Non-Financial Public Sector Spending._elncludes Government spending on Education, Health, Labor and Social Security, and Housing. Excludes spending on FES, programs supported by the First Lady, PROINLO, and the "Proyectos de Desarrollo Social" (Circuitales), some or all of which could be considered as "social" spending. _d\ Includes administrative spending * Only at the primary level are subsidies progressive in both absolute and relative terms. While the poor represent 49 percent of primary-school aged children, they receive 56 percent of total government spending on primary education. In relative terms, primary school subsidies are ten times more important for the poor, representing 18 percent of the poor's per capita consumption as compared with 1.6 percent for the non-poor. * Geographic inequities in public spending on education are also apparent, with biases in favor of urban areas and against indigenous and non-indigenous rural areas. While indigenous residents (virtually all of whom are poor) represent ten percent of the school-aged population (5-24 years), they only receive five percent of total government spending on education (Figure 5.3). Rural residents also receive a disproportionately low amount of education subsidies, despite the fact that the majority of the poor are concentrated in the countryside. In contrast, whereas about half of the population aged 5-24 lives in urban areas, city-dwellers receive two thirds of public spending on education. As with biases against the poor, geographic inequities in public spending arise at the preschool, secondary, and higher education levels. Cost recovery is lowest for higher education: subsidies account for 84 percent of the total cost of public higher education, as compared with 78 percent for primary and secondary school 32 5.4. Despite these patterns, the share of total spending allocated to primary education has fallen substantially in recent years. The share allocated to basic education4 now equals that accorded to higher education - which almost exclusively benefits the non-poor (Table 5.1). Figure 5.2 Who Benefits from Public Spending on Education? % of Subsidies to Poor, Non-Poor by Level (LSMS 1997) 100% 90% 94% 80% 70% 72% 70%- 59% 56% 57% 60%- 50% 410/ 44% 430/ o All Pbor 40% - 30 28 n Non-Poor 30%- 20%- 10% 0% Pre-Prinary PrirTry Secondary Higher Total Sub. % Pbp. 5-24 Figure 5.3 Urban Areas Receive a Disproportionate Share of Public Spending on Education: % of Subsidies, by Geographic Areas and Level (LSMS 1997) 100% 8 90% 86% 80% _ 70% 61% 65% 60% 0;/4 % 5 i % Urbano 50% ~ ~ ~ ~ ~ 1 38 -5 % 62% 40% 36% 30%42 38% * Rural 40/o - 30% 1 Indigenous 30% 20% -13% 14%10 10% - %3% 0% 0% II III Pre-Prinary Rimary Secondary Figher Total Sub. % Pop. 5-24 5.5. Disparities in household spending on education exacerbate inequities in the distribution of total resources devoted to education. The non-poor spend substantially more than the poor on education (Figure 5.4). As discussed in Box 2.4 in Chapter 2, these differences arise primarily in The Government does not systematically disaggregate spending allocations to primary and pre-primary (basic) education. Available data for 1993, however, indicate that primary school absorbs 96 percent of the allocation to basic education. Source: World Bank ReportNo. 13701-PAN "Panama: Issues in Basic Education," March 1995. 33 spending on school supplies, books, and informal fees. Such disparities probably add to inequities in the quality of education received. Figure 5.4 Disparities in Household Spending Exacerbate Inequities in Total Spending on Education (LSMS 1997) 100% 89a 90% 8%78% 80% 72% 70% 57% 60% m AlI Poor 50% 43% Non-Poor 40% 28% 22% 30% 20% 11% 0%6 % of public % of household % of total % Pop. 5-24 spending spending spending 5.6. Functional inefficiencies in education management have also prevented improved educational outcomes among the poor. The poor quality of public education also stems from inefficiencies in the delivery of education, including: (i) highly centralized decision-making and lack of parent and community participation; (ii) weak policy-making and planning capacity in the Ministry of Education; (iii) the lack of a management information system and standardized evaluation system as decision-making tools; and (iv) a disconnect between the pay and performance of teachers and administrators. While such inefficiencies are not limited to the poor, the resulting quality of the education received by the poor lags that of the non-poor (as discussed above). 5.7. Although the Government has recently started to address some of these deficiencies, the major challenge is to remove existing biases in public spending. The Government's ten-year Education Strategy (1997-2006) outlines four major policy objectives: (i) equity of access; (ii) quality improvements; (iii) institutional efficiency with an emphasis on decentralization of administrative and educational decisions; and (iv) expanded community and family participation. The key challenge for translating these principles into action and for reducing disparities in human capital is to reallocate public spending on education such that a larger share benefits the poor. Cost recovery for higher education, which exclusively benefits the non-poor, should be increased in order to liberate resources that could be allocated toward basic education, which is a priority for helping the poor grow out of poverty. B. INEQUITIES AND INEFFICIENCIES IN THE PROVISION OF PUBLIC HEALTH CARE SERVICES 5.8. With over seven percent of GDP and 15 percent of government expenditures allocated to public spending on health in 1996, Panama ranks among the top spenders in Latin America.5 Moreover, health has absorbed an increasing share of resources, with public spending, rising from six percent of GDP and twelve percent of government expenditures in 1992. Despite this large investment, health outcomes and performance indicators are lower than expected, particularly among the poor and indigenous (see Chapter 2). Inequities in public spending and inefficiencies in the sector largely explain this apparent paradox. 5 Other high spenders include Argentina (10.6 percent of GDP), Costa Rica (8.5 percent), and Colombia (7.4 percent). Source: Govindaraj et. al. (March 1995). 34 Table 5.2 - Public Spending on Health 1992 1993 1994 1995 1996 Total Spending on Health (B./000) 406,709 442,724 475,859 547,090 584,510 % Recurrent 91% 89% 88% 81% 79% % Investment 9% 11% 12% 19% 21% Total Spending on Health as Share of: GDP 6.1% 6.1% 6.2% 6.9% 7.1% Total Government Spending 11.9% 14.6% 15.7% 16.1% 15.6% Total Gov't Spending on Social Sectors 32.5% 30.3% 31.2% 31.7% 32.3% Health Spending by Type (% of total) Maternal-Infant Health 28% 28% 28% 26% 25% Adult Health 34% 34% 32% 30% 29% Environmental Health 16% 15% 17% 13% 12% Construction of Health Centers 6% 8% 8% 17% 20% Administration and Regulation 16% 15% 16% 15% 14% Health Spending by Agency (% of total) Ministry of Health (MINSA) 33% 36% 31% 29% 30% Social Security Fund (CSS) 51% 50% 1 53% 59% 59% Water and Sanitation Institute (IDAAN) 3% 3% 3% 3% 3% Metropolitan Waste Management (DIMA) 13% 11% 13% 10% 9% Source: MIPPE/MEF: Direccion de Presupuesto de la Nacl6n. 5.9. While the overall level of public resources allocated to health care appears to be adequate, disparities in health outcomes reflect biases in public spending. Extensive data from the LSMS on use patterns, poverty status and geographic location support the following conclusions (see Annex 17):6 * The poorest and richest quintiles benefit the least from public spending on health care. For those in the top quintile, this is due to reliance on private health care. For the poorest, however, this bias reflects low levels of health care use in general, regardless of the type of care. - Health care spending on curative care is regressive, while spending on maternity-related care is progressive. However, a lack of institutional data on the costs of different types of health care prevents an assessment of the overall incidence of public spending. * Hospital-based care is used more by the non-poor and those in urban areas and thus spending on hospital care is likely to be regressive. Poorer groups tend to use health centers, posts and sub- centers more, and hence spending on these services is likely to be progressive. 5.10. Inefficiencies in the health sector have also prevented better health outcomes among the poor. Such inefficiencies arise largely due to the existence of a fragmented health system comprised of three major players: the Ministry of Health (MINSA), the Social Security Fund (CSS) and the private sector. Disintegration in the financing and provision of health services among these three players promotes duplication and under-utilization of services, negates economies of scale, and creates unnecessary increases in the cost of services. In addition, MINSA has weak policy-making, financing, and regulatory capabilities, deteriorated infrastructure, and obsolete management and information systems. There is also an excessive reliance on physicians in instances in which nurses could more cheaply provide services. Decision-making and management are overly-centralized, which combined with high levels of bureaucracy, has promoted a culture of poor management, lack of accountability for results and performance, and an inability to respond to local needs in a tirnely fashion. There is no link between payment transfers to providers and performance, with little incentive to improve efficiency and effectiveness. Moreover, health service delivery and referral systems are disorganized and lack appropriate regulation and financial incentives to manage demand for health services. While these issues are not limited to the poor, it is the poor who suffer the results in terms of less access to quality services and worse health outcomes. 6 A complete analysis of the incidence of public spending on health was not possible due to a lack of data on the value of health care services and tne share of this value that was covered by government subsidies (versus the share covered by household payments). The patterns presented here largely reflect the incidence of the use of public services. See Annex 17 for details. 35 5.11. While the Government is undertaking a number of steps to improve efficiency in the health sector, it should also seek to correct biases in public spending. MINSA and CSS have launched a health sector reform program to be piloted in the San Miguelito region that seeks to improve efficiency by (i) establishing a strategic partnership between the two institutions; (ii) separating the financing and provision of services; (iii) increasing private-sector participation in service provision; (iv) strengthening the policy-making, regulatory and management capabilities of MINSA; (v) decentralizing services to the regional level and increasing the administrative and financial autonomy of service providers; and (vi) increasing community participation in health affairs and health care. Increasing efficiency is only half the story, however. The Government should also seek to improve equity of access to health care services by: (i) improving the targeting of public spending using the new poverty map developed by MIPPE/MEF; and (ii) expanding the coverage of cost-effective primary health interventions in poor rural, indigenous, and urban areas. A thorough review of public spending on health services, combined with data from the LSMS, would facilitate the identification of specific actions needed for such a reallocation. C. PROTECTING THE POOR: NUTRITION & SOCIAL ASSISTANCE INTERVENTIONS 5.12. To compensate for some of the biases against the poor, the Government has mobilized considerable resources for social assistance programs. An estimated one percent of GDP is currently allocated to social programs, most of which were initiated since the early 1990s (see Box 5.1). 5.13. These programs rightly address key poverty issues and several appear quite effective at reaching the poor. While Panama does not have a comprehensive social safety net, its existing assistance programs do seek to reduce key disparities, including: Figure 5.5 (NOTE: all poor includes extreme poor) - Removing economic barriers Who Benefits from School Feeding Programs? to increased school Share of benefits as conpared with share in population, enrollment by the poor. by Poverty Group (LSMS 1997) Panama's compensatory 100% __ education programs - several 10% 74% school feeding schemes and a 0 53% 47% 55 53% 47% new cash transfer program - 0% 28 26% 30'm rm Extreme Ror seek to remove economic _ All Poor barriers to education by 20% _ *Non-Pbor providing assistance needed to 0% generate incentives for % of Snack % of Lunch % of Pbp. increased enrollment, Benefits Benefits Aged 3-11 attendance, and performance. While an evaluation of the educational impact of these programs is currently underway, there is evidence that some are quite effective at reaching the poor.7 In particular, the LSMS suggests that the FES School Lunch Program is extremely well targeted, reaching a disproportionate share of poor and extreme poor children (Figure 5.5). * Responding to community requests for basic and social infrastructure. Disparities in access to basic services are closely linked to poverty, inequality and malnutrition. The Government currently operates a number of demand-driven local investment programs - including the FES and PROINLO - designed to respond to community needs to fill gaps in basic service provision. The FES and the World Bank are currently designing an evaluation of the impact of school feeding programs on school attendance and performance. 36 Box 5.1 - INVENTORY OF MAIN GOVERNMENT SOCIAL ASSISTANCE PROGRAMS Prograrn Explicit Targeting of Poor? Key Issues for Improved Poverty Effectiveness COMPENSATORY EDUCATION PROGRAMS MINEDUC School Feeding Program X None. Explicitly universal * Targeting Early-moming snack (milk with enriched cookies or by law (all school children). # Improve efficiency in an enriched cereal-based crema). Average annual cost delivery to remote areas per student: B./23; for milk: B./35; for the cookies; * Improve cost- B./9; for the crema: B./12. effectiveness, particularly for milk New MINEDUC School Lunch Program XNone. + Targeting To be provided to children in extended-day programs. * Reduce overlap with FES New program. program FES School Feeding Program " Targeted according to * Targeted expansion Lunch (rice, beans, and oil). Implementation with nutrition & poverty map * Reduce inefficiencies in community participation. Average annual cost per criteria delivery to remote areas student: B./21. Cash School Transfer (Balboa per day) Program v To be targeted according to * Targeting New programn to transfer B./I per student for each day new poverty map and * Efficient administration of school attendance (being piloted in Bard) enrollment I dropout indices & monitoring OTHER NUTRITIOIN & FEEDING PROGRAMS ________ MINSA Supplementary Feeding Program v Targeted using * Evaluate impact on poor Monthly food rations accompanied by micronutrient anthropometric indicators of and efficiency of supplements via health system to malnourished malnutrition and nutritional implementation; expand children under 5 and underweight pregnant and risk; program also self-targeted program lactating women. Clear eligibility and exit criteria due to bland taste of meal (anthropometric screening). Other Nutrition Interventions # Most are targeted by need * Evaluate impact and (MINSA/PATRONATO DE NUTRICION) (e.g., micronutrient efficiency of Other nutrition interventions include micronutrient deficiencies) implementation supplements in areas with deficiencies, deworming interventions, school gardens, etc. HOUSING ASSISTANCE PROGRAM MIVI - PARVIS Program v Targeted to families with * Evaluate ability to reach Home-improvement grants of up to B./1,500 to low- incomes less than B./300 per poorest income families. Beneficiaries can't have benefited month (means testing); priority * Evaluation needed; from other housing assistance or have owned plot for given to those with less than questions should be longer than three years. Also provides basic B./250 per month. included in LSMS 2000 infrastructure (water, sanitation, latrines) and land- tenure and legalization services for participants. SPECIAL GROUP ASSISTANCE PROGRAMS _____ FES Grupos Vulnerables Programs e Various criteria: geographic * Ensure targeting, New programs. NGO-executed programs funded by targeting, self-targeting of especially for women and FES for special vulnerable groups, including: street street children; elderly not youth programs children, poor elderly, poor women, poor and at-risk covered by CSS or other social * Progratn impact youths, Afro-Panamanians, and the indigenous. security programs; etc. evaluations needed LOCAL SOCIAL INVESTMENT PROGRAMS __ FES Infrastructure & Credit Programs 0 XMany - but not all - FES * Design new criteria for Demand-driven social and economic infrastructure investments are targeted using targeting with new investment projects and a new program to support geographic poverty criteria poverty map micro-enterprises. It also funds a number of regional * Expand share of FES development programs (e.g., FIDA-Ngobe Program) spending subject to _ targeting criteria MIPPEIMEF - PROINLO XNone. Explicitly universal + Targeting Demand-driven community investments. with equal amounts to each * Evaluate possible correginmiento functional overlap with FES 37 * Responding to the specific needs of vulnerable groups. The Government also provides tailored assistance to specific vulnerable groups through a number of smaller programs, including housing assistance to the low-income families, and NGO-executed social programs for street children, poor elderly, poor women and youths, indigenous groups and Afro-Panamanians. MINSA also operates a supplemental feeding and micronutrient program for malnourished children and mothers, which appears to be quite well targeted.8 5.14. Nonetheless, a number of weaknesses reduce the effectiveness of these transfers, including: * A lack of targeting. Some of Panama's larger assistance programs - notably MINEDUC's school snack program (see Figure 5.5)9 and MIPPE/MEF's local investment program (PROINLO) - are explicitly universal. Others, such as the FES, are only partially targeted to the poor. The considerable amount of resources devoted to these programs would have a much larger impact on poverty reduction if they were concentrated on the poor. * Inefficiencies in program delivery. Inefficiencies in the delivery of transfers also reduce the cost- effectiveness of these safety-net programs. The MINEDUC school snack program, for example, includes unnecessarily costly inputs, most notably milk (which is backed by the milk processing and packaging lobby)." Cost-effectiveness would be improved with the removal of milk from the program (to be replaced by the cheaper cereal-based crema or the fortified cookies). Both the MINEDUC snack program and the FES lunch program also suffer from inefficiencies in the physical delivery of food to remote areas (see Box 5.2), which could be remedied through a decentralization of food procurement in these areas. * Possible duplication of functions. A possible duplication of functions of the various social investment programs (between the FES and PROINLO and potentially with the line ministries) could reduce their effectiveness. Box 5.2 - Decentralizing School Feeding Programs in Remote Areas: the Potential of Block Grants Both the MINEDUC Snack Program and the FES Lunch Program suffer inefficiencies in the physical delivery of food to remote areas. Before the start of the next school year, both programs need to design more cost-effective mechanisms for transferring resources to these communities (which have high incidences of poverty and malnutrition). Some options: * Government delivery is costly. In some instances, the food has had to be delivered via helicopter (at a cost of B./800 per hour per helicopter!), horses, or boats. * Community pick-up is costly and ineffective. One option that has been tried involves requiring communities to pick up the food themselves at satellite delivery points (the cost remains, but is borne by the communities). This has not been effective, however, and there are numerous reports of the food being left to waste. * Block grants are a promising alternative. An alternative involves the decentralization of food purchases to the communities themselves. The Government could transfer funds as block grants directly to the communities, who would then purchase the food locally (according to pre-advised nutritional criteria consistent with local food supplies). A cost- benefit analysis suggests that, with proper monitoring and mechanisms for accountability, this would be by far the most cost-effective option. Poor children receive over three quarters of MINSA food rations (with the remaining quarter going to the non-poor). Coverage of the program is limited, however, reaching 14% of indigenous children and 8% of all poor children (compaed with 3%ofthe non-poor). See Annex 17. 9 Panamanian law requires that MINEDUC's snack program be expanded to reach all children. In fiat, it is slightly self-targeted to the poor due to the fact that wealthier children voluntarily attend private - rather than public - schools (and the snacks are not provided at private schools). '° While targeting criteria are applied to some of the investments supported by the FES (notably ftose funded by intemational donors such as the World Bank), the remainder of FES investment spending is not explicitly targeted according to poverty criteria. " In terms of the cost per thousand calories transferred, milk is 1.6 times more costly than enriched cookies and 2.4 times more costdy than the enriched crema. Moreover, the common prevalence of milk allergies (particularly among the indigenous) makes the use of ilk less aractive. 38 Chapter 6 - Principles, Priorities, and Actions to Reduce Poverty 6.1. This section outlines the main principles of the Government's Poverty Strategy,' and recommends priority actions for translating this strategy into action. A. THE MAIN PRINCIPLES OF THE GOVERNMENT'S NEW POVERTY STRATEGY 6.2. The Government's Poverty Strategy is founded on sound basic principles. It complement's the Economic Reform Program (Chapter 4) and is designed to strengthen the key assets of the poor, taking into account geographic differences in poverty and priorities. It largely coincides with the priorities expressed by communities themselves (Box 6.1). The main underlying principles for this strategy include: * Coherence with, and implementation of, the economic reform program for sustained broad-based economic growth; Improved efficiency in, social spending; * Targeting of resources to the poor; * Decentralization for improved efficiency and quality of interventions; * Increased community participation for improved effectiveness; * A multi-dimensional approach with strong-intersectoral coordination; and * Monitoring both of the poverty situation and of the implementation of the strategy itself. The strategy also includes an action plan for interventions in key areas, including: education, health, social assistance and nutrition, urban poverty, and rural poverty. Box 6.1 - Community Priorities by Geographic Area Rural Areas Indigenous Areas Urban Areas Transport, roads Education Safety, crime Electricity, telephones Potable water Transportation, roads Potable water Alcohol, drugs Employment, labor Health care Transport, roads Sanitation services Employment, labor Health care Potable water Education Housing Social & community centers Housing Safety, crime Housing, titling Social & community centers Employment / labor Alcohol, drugs Safety, crime Sanitation services Health care Land, titling Electricity, telephones Gangs, youth problems Sanitation services Social & community centers Lighting, telephones Alcohol, drugs Garbage collection Education source: Panama LSMS 1997 community Qu .onnaire; commumty pnornties raniked m order o trequency. B. PRIORITIES FOR TRANSLATING THESE PRINCIPLES INTO ACTION 6.3. In addition to sustaining and deepening the economic reform program as a vehicle for promoting growth, several priority actions are needed to translate the above principles into results, including: (i) prioritizing among poverty groups; (ii) reallocating public spending as the top priority action; (iii) decentralizing and promoting participation for effectiveness of service delivery; (iv) implementing specific policy reforms to reduce disparities in assets; (v) explicitly using and developing targeting 'Nuevo EnfoqW Es,&ategkco Frenr a la Pobreza' Cabinet Resolution No. 134, September 17, 1998. 39 mechanisms; and (vi) allocating resources for monitoring the poverty situation and the implementation of the Poverty Strategy. Prioritizing Among Key Poor and Vulnerable Groups 6.4. The Government should prioritize among poverty groups according to the prevalence, depth and severity of poverty (Box 6.2). First priority should be given to the rural poor, the indigenous, and poor and malnourished children. Second priority should be assigned to combating urban poverty. Finally, third priority should be given to programs that target the elderly poor, child laborers, poor informal-sector workers and the poor unemployed. Box 6.2 - Priority Poverty Groups (.- = high priority; *- = medium priority; . = priority according to scope & severity of poverty) GEOGRAPHIC GRouPs: PRiORrrY * The poor in rural areas. A high share of the rural population is poor. The new Poverty Map (see below) as* would help identify specific districts with high concentrations of rural poor. * All residents of indigenous areas. Geographic targeting to indigenous areas would be administratively *5 simple: virtually all residents of indigenous areas are poor; most live in extreme poverty. * The poor in urban areas. In terms of their absolute number, the urban poor are an important group. A *o significant share of city dwellers also live just above the poverty line and could be considered vulnerable. Urban poverty is less extreme, severe, or deep, however. ETHNIC GRouPs: PRIORITY * The Ngobe-Bugl6. The analysis reveals that Panama's largest ethnic indigenous group has the highest rate of *. poverty in the country. Most Ngobe-Bugld are concentrated in indigenous areas in Chiriquf and Bocas del Toro, making geographic targeting feasible (see above). They tend to live in dispersed clusters of 6-8 households, however, making, service provision difficult * Other indigenous groups. Poverty rates are also high among other indigenous groups, but primarily for those Soo who live within the indigenous areas. DEMOGRAPHC & ECONOMIC GROuPS: PRioRrry * Poor children and youths. One half of Panamanian children under 18 live in poverty. One quarter of poor *. children under five years old are malnourished. The developmental status of children renders them extremely vulnerable to the risks of living in an impoverished environment. Youth is the point in the life cycle when physical, cognitive and psycho-social development occurs at its most accelerated pace and is most susceptible to abnormal development from poverty conditions. Poverty is particularly costly to individuals, society, and future generations when it occurs at these stages of life. * Undernourished children and pregnant & lactating women. The poverty strategy also identifies these groups as priority for interventions. With strong correlation between poverty and malnutrition, they are largely captured by other priority group classifications. There are some specific interventions needed for these groups, however (see Part 3). * Poor elderly. Though a smaller share of the elderly are poor (29%), demographic changes indicate that these will become an important group in the future. There are very few programs in Panama at present to assist poor elderly citizens who fall outside the social security system. * Poor child laborers, informal-sector workers, and unemployed (particularly women and youths). The poverty strategy also identifies these groups as priority for interventions. Indeed, poverty is correlated with child labor, informal work, and unemployment (for urban women and youths). Though they are probably already captured by the above priority group classifications, they may have specific needs. Main Priority Action: Reallocating Public Spending 6.5. Reallocating public expenditures is the top priority for effective action to reduce poverty. Given the high level of social spending in Panama, it is unlikely that a large amount of additional resources will be forthcoming. As such, the Government urgently needs to reallocate existing spending toward areas that benefit the poor, boost cost recovery for services used by the non-poor, and improve efficiency in service delivery. Explicit use of targeting mechanisms (see below) in program and service delivery plans would also help ensure that more resources are channeled to the poor. A thorough review of public spending should, be conducted in 1999 to provide additional guidance on such reallocations. This review should emphasize both the functional aspects of the budget (including the distributional incidence of spending in the main sectors) as well as the administration and management of public 40 expenditures. The LSMS analysis already provides insights into some of the main reallocations that are needed to reduce disparities in assets. 6.6. Public spending on education must be reallocated for improved effectiveness in combating poverty. The analysis clearly shows that disparities in education are the key causes of poverty, malnutrition and inequality in Panama. Investing in the education of the poor is also the principal lever for helping the poor lift themselves out of poverty in the long run. Clear candidates for reallocation in the education sector include: Higher cost recovery for higher education (benefits the non-poor) would liberate resources that could be allocated toward basic education (priority for the poor and indigenous). The annual cost of completing coverage of primary schooling for all children aged 6-11 who are not currently enrolled is expected to be quite small: about B./12 million in total (0.14 percent of GDP), including B./9 million to the Government and B./3 million to households (in the form of fees, informal fees, transport costs, books, and materials).2 The annual cost of completing coverage of secondary education for all children aged 12-17 not currently enrolled would be larger: about B./89 million in total (1% of GDP), including B./69 million in public spending and B./20 million in contribulons from households. Completing secondary school coverage of only the poor would cost about three quarters of this amount. Nonetheless, given the substantially higher rates of return to secondary school education - and the corresponding impact on reducing poverty and inequality - such investments are warranted. . In many areas demand-side factors (economic constraints) pose larger obstacles to increased enrollment by the poor than supply-side factors (availability of facilities). As such, spending should be focused on compensatory education interventions (e.g., targeted school feeding and/or attendance- based cash transfer programs) and the provision of textbooks and school materials (largest cost items for attending school; benefits for quality of education). Investments in instruments such as the TELEBASICA program would also reduce constraints to increased school access in remote rural areas. 6.7. The Government should complement its strong efforts to boost efficiency in the health sector with a reallocation of public spending designed to improve equity. A thorough review of public spending on health services that combines use information from the LSMS with unit cost data on specific services should be undertaken to identify specific actions needed for such a reallocation. Though unit cost data are not currently available, the analysis of the LSMS does highlight some candidates for reallocated public spending: * In terms of the types of services provided, higher cost recovery should be sought for curative care (disproportionately benefits the non-poor) and resources should be shifted toward the expansion of cost-effective primary health interventions in poor, rural, and indigenous areas (using the poverty map, see below), with an emphasis on maternal-infant health, oral rehydration, potable water in rural and indigenous areas, prevention and early detection of TB, family planning, and health education. * In terms of facilities, higher cost recovery (and private rather than public provision) should be sought for hospital-based care (disproportionately benefits the non-poor and urban residents), and public spending on inputs (including medicines), investment and rehabilitation should focus on health centers, sub-centers, and posts (used primarily by the poor). 6.8. Spending on social security and assistance programs should be streamlined to ensure a comprehensive, efficient, and well-targeted social safety net. Although existing assistance programs rightly address key poverty issues (compensatory education, local investment, nutrition, and tailored- interventions for specific vulnerable groups), a number of actions should be taken to streamline spending on these programs and ensure a more comprehensive and efficient safety net: 2 These figures are baed on averae per student costs of schooling at the primary and secondary levels (Source: MIPPE) and the number of children in each age cohort that au not currently enrolled in school (Source: LSMS 1997). These do not include the incremental investment cost associated with the tnsion of tde school day. 41 * Explicitly target spending on all social assistance programs, including existing school feeding and community investment programs, reducing or eliminating allocations to areas with a low incidence of poverty and shifting these resources to poorer areas (see targeting discussion below); * Reduce duplication in spending on local investment programs by combining these programs or focusing them on specific areas (sectoral or geographic); * Develop mechanisms to protect annual budget allocations for programs and interventions (such as the targeted programs supported by the FES, primary education, etc.) with a proven record for reaching the poor. This does not include "all social spending" or "all social investment" and excludes the more political programs such as the "social investment" spending under the "Circuitales" legislators fund. * Use resources generated from better targeting of above programs to expand existing pilot programs that are tailored to specific vulnerable groups if evaluations show them to be successful (e.g., those being supported by the FES and the new Ministry of Youth, Women, Children and Family, MJMNF); allocate resources to develop new social programs using similar model to tackle social problems such as crime, violence, and alcoholism, which were cited by urban and indigenous communities as priorities in the LSMS; and * Conduct a thorough review of the distributional incidence of public spending on social security programs. 6.9. Public spending allocations for basic infrastructure should be allocated according to gaps and poverty criteria. The precise mix of infrastructure needs differs significantly by geographic area. In broad terms, communities in LSMS focus-group discussions voiced the following priorities: potable water in indigenous areas; transportation, electricity, telephones, and water in rural areas; and transportation and sanitation services in urban areas. These priorities are completely consistent with the analysis of disparities in assets using quantitative data from the LSMS (Chapter 2). The upcoming Census 2000 will provide detailed information on existing gaps and coverage in infrastructure services. A mapping of these gaps should then be overlayed with the new poverty map to facilitate the allocation of public spending according to poverty and service gaps (giving priority to corregimientos and districts that have a high prevalence of poverty and service gaps). Improving the Effectiveness and Efficiency of Service Delivery 6.10. Decentralization of, and community participation in, service delivery should improve the effectiveness of poverty-reduction interventions. The Government's Poverty Strategy emphasizes decentralization of the administration and implementation of various functions and services to the local level in order to improve accountability, quality, and efficiency of service delivery. The strategy also highlights the importance of community participation as a way to boost local ownership and client satisfaction, mobilize additional resources, improve accountability among service providers, and build social capital. Some examples include: * Decentralizing personnel decisions to regional education boards and expanding innovative participatory informal pre-school programs (CEFACEIs, Madre-a-Madre); * Decentralizing food purchases under the school feeding programs by transfering block grants to remote communities for local food procurement; * Expanding use of NGOs and communities as intermediaries for the design, development and execution of social programs tailored to the specific needs of vulnerable groups (e.g., those being supported by the FES and MJ4MNF); and * Greater reliance on community expression of preferences for investments, both for services provided by line ministries and for demand-driven local investment funds (such as the FES). 42 Box 6.3 - The Challenge of Poverty Reduction Among Remote, Dispersed Populations A key challenge facing the Government of Panama in its fight against poverty is the development of remote areas with highly dispersed populations. Panama has over 6,000 dispersed villages of less than 50 inhabitants and over 4,300 hamlets of less than 2,000 people. The LSMS indicates that close to two percent of the population lives in non-indigenous "remote areas" (areas de dificil acceso), many of which are only accessible by air, boat, or by horseback. Many indigenous areas are likewise remote, with no road access and dispersed hamlets of 6-8 households each. A lack of Spanish-speaking ability can be a further barrier distancing these groups from the rest of the economy. There is no "magic bullet" in the fight against poverty in remote areas. The costs of delivering basic physical and social services to dispersed populations is extremely high on a per person basis. In an era of tight fiscal budgets, these costs mean even fewer total poor people who can be reached by government assistance, and hence a smaller impact of public spending on poverty reduction. Productive activities, whether small-scale manufacturing, agro-processing or agriculture, are hampered by high transaction costs and little access to markets. Fragile environmental and ecological conditions further constrain the development of these areas. Experiences around the world has shown that ultimately, a common solution and natural response is migration to larger local towns, regional centers or major metropolitan areas, either by entire families or individual members who then send remittances back. Nonetheless, there are some actions that can help improve the living standards of these dispersed populations, including: . providing basic services in local towns; * expanding the use of innovative approaches to service delivery that are well-adapted to reach dispersed populations, such as (i) the contracting of NGOs and community groups already active in these areas to provide services (an existing example is the contracting of NGOs in remote areas of Darimn (e.g., Pro- Niflos del Darien) to deliver food supplements and other nutrition interventions to school-aged children); (ii) the decentralization of service delivery, social investment decisions, and project implementation via municipalities or regional FES offices; and (iii) the use of TELEBASICA and mobile health personnel to provide education and health services to dispersed populations (which would raise their human capital and expand their employment potential even if they do migrate); * reducing transaction costs in and out of these areas to promote private sector activity and improve market access by improving road access or boat service connections (albeit taking into account environmental and ecological considerations) and expanding telecommunications; * providing technical assistance that is targeted to the specific needs of these populations with respect to agricultural production, marketing, and financial services via the use of travelling extension agents and the media; * supporting integrated regional or area development initiatives; and . continuing the economic reform program to promote overall economic growth - particularly in labor- intensive centers - to help mitigate the employment effects of on-going migration of these populations. Key Policy Reforms to Reduce Disparities in Assets 6.11. Implementing key policy reforms to promote broad-based economic growth and reduce duality and disparities in assets. The agenda for the second-generation of economic reforms designed to promote economic growth is large, encompassing reforms to improve state efficiency, foster competition, improve governance, modernize basic infrastructure, and manage the canal and develop the inter-oceanic region (See Chapter 4). The Government also needs to prevent against backsliding against reforms that have already been implemented since the beginning of the 1990s. Special efforts should be made to ensure that key reforms to reduce disparities in assets, and hence poverty, are undertaken, including: * Deepening reforms to the labor code and related legislation to ease restrictions on dismissal and lower the relative cost of labor, the poor's most abundant asset. A thorough study of labor policies and their impact on the labor market would help identify the specific reforms that should be undertaken. * Expanding coverage of land titling, by modernizing land administration services and launching a massive national land titling programn focusing on peri-urban and rural areas. Expanding titling would 43 greatly reduce disparities in land and housing, which are closely associated with poverty and inequality. It is also one of the most effective ways of allowing the poor access to credit - which can help in income generation - since a lack of collateral is the primary barrier to obtaining credit. * Continuing trade reforms and promoting competition in domestic markets. Simulations using the LSMS show that reforms to liberalize trade and domestic markets would indeed reduce poverty and inequality in Panama. Though tariffs have come down substantially, further reductions are still needed, particularly for rice and milk. Moreover, the introduction of new distortions - including non- tariff barriers (such as phytosanitary permits) - should be avoided. * Reducing distortions in public and freight transport. Although efficient public transport is a top priority for the urban poor, particularly as a complement to labor assets, subsidies and price distortions and a conmplex system of licensing have reduced quality and competition in service provision. Competition is also limited by barriers to entry and exit in freight transport, reducing efficiency and increasing the cost of living in Panama. The Government should actively pursue reforms to reduce these distortions. Explicit Targeting of Programs and Interventions to the Poor is Crucial 6.12. Targeting resources to the poor is a crucial tool for improving the impact and cost- effectiveness of public spending and interventions. Targeting improves cost-effectiveness and impact by concentrating public resources on those who need them the most. Geographic targeting (via the poverty map) is the main tool being used for targeting in Panama. It entails a number of advantages, most notably administrative ease: communities can be selected ex ante for interventions based on their concentrations of poor people. The current effort to improve the poverty map by combining data from the Census and LSMS will be useful in promoting this approach? In some areas (e.g., large urban cities), however, districts and corregimientos are quite diverse in their poverty profiles and sample limitations prevent creating poverty mapping tools at more disaggregated community or neighborhood levels. As such, some programs (e.g., transfer programs such as maternal-infant food supplements or the PARVIS housing subsidy) should attempt to further fine-tune targeting by combining the poverty map with additional targeting mechanisms. Individual indicators, such as income (means-testing) or anthropometric assessments can be useful (though administratively burdensome) to further fine-tune targeting in such instances. Self-targeting can also be used for the delivery of services clearly used more by the poor than the non-poor (e.g., targeting benefits via health posts, sub-centers and centers which are used disproportionately by the poor). Sufficient Resources for Monitoring and Evaluation 6.13. Monitoring of both the poor and the strategy is necessary, and adequate resources should be made available for this task. The Govermment's strategy includes two key monitoring components: Monitoring the poverty situation. For the first component, the Government is developing a poverty monitoring system to track living conditions and provide data for the evaluation of the impact of interventions. The system will build on the 1997 LSMS by executing such surveys every three years. Two are planned for the remainder of the strategy period: one in 2000 and one in 2003. In addition, the Govermnent is exploring the possibility of incorporating key supplements into the annual household survey system (managed by the Directorate of Statistics and Census, DEC) to generate ongoing data for monitoring poverty. Finally, the upcoming Census 2000 (which is quite extensive in its breadth of topic coverage in Panama) will provide additional information for the monitoring of poverty. 3 A repeat of this exercise using data from the upcoming Census 2000 and LSMS 2000 will be nessay to updac dis nq. 44 * Monitoring and evaluating the poverty strategy. The Government is also adopting a number of actions to monitor the, implementation of the poverty strategy, including: detailing actions for the strategy period in an action plan, specifying actions for each year (as was done for 1999), and developing monitoring indicators that will be tracked by each institution. Program-specific questions should be included in the upcoming LSMSs to facilitate such monitoring. While all ministries have responsibility for ensuring the implementation of the strategy, MIPPE/MEF has been charged with coordinating monitoring of both the poverty situation and the implementation of the strategy. Resources should also be made available for evaluating the poverty impact of specific interventions. C. ROAD MAP OF KEY ACTIONS FOR POVERTY REDUCTION 6.14. Box 6.4 outlines key constraints and actions in the main stratgic areas for poverty reduction as well as areas for further research. The general timing of the implementation of these actions is also indicated. Some key steps for immediate implementation during the remainder of 1999 include: * Conducting a thorough review of public expenditure allocations (see para. 6.5 above) and developing proposals for reallocating expenditures such that they better reach the poor for the 2000 budget; * Developing a set of indicators to monitor implementation of the strategy (including key budget categories) and agreeing on an inter-ministerial process for reporting on such indicators and implementation (see para 6.13 above). Funds from the on-going IDF grant could be used to contract consultants to build the capacity of MIPPE/MEF staff for developing and monitoring such indicators; * Utilizing the new Poverty Map prepared by MIPPE/MEF/FES as a tool for targeting and resource allocation; and * Continuing dissemination efforts for both the Poverty Assessment and the Govermment's new Poverty Strategy in both Government circles and public forums. Box 6.4 - ROAD MAP OF KEY ACTIONS FOR POVERTY REDUCTION CONSTRAINTS RECOMMENDATIONS Key Issues Priority Key Actions Perioed _ DEEPENING ECONOMIC REFORMS: Priority for poor in ail areas Economic growth is essential for poverty .o Key reforms: Act: o-going reduction. It is also the most effective way * Deepening reforms to the labor code Imp: ST, MT to reduce unemployment (more effective * Modernizing the land tenure regime (also than direct labor interventions). Removing crucial for improving access to credit) key distortions will also reduce disparities in * Continuing trade reforms assets, such as labor, property, and credit. * Reducing distortions in tasport sector ... = top pnority; .. = medium pnornty; . = prionty; I'M = pove -y map; S l = one year perioa; btr = I-J years; Ll MO tmol a 3 yCOM;Act period for implementation of actions; Imp = period for impact on poverty 45 Box 6A Cont'd - EDUCATION: Priority for poor in all areas, large gaps for all indigenous CONSTRAINTS RECOMMENDATIONS ___ ___ Key Issues Priority Key Actions Period Disparities in access. Main gaps: Reallocate spending by: Act: S l, Mtl * Pre-primary: all poor, especially rural and ... . Increasing cost recovery in higher education Imp: LT indigenous (raising fees) to free up resources * Primary: indigenous ... . Increasing the share of spending allocated to basic . Secondary: all poor, esp. rural & indigenous and secondary educationinpoor areas (via PM) . Higher: all poor (but less of priority) using freed resources from above Direct costs, other demand-side factors are key Ease demand constraints through: Act: ST constraints. Availability of schools more of a see * Targeted compensatory education programs Imp: ST, LT problem at pre-primary level. so . Targeted scholarships for the poor (especially rural and indigenous poor) for primary, secondary education ... Ease supply constraints for pre-primary schools through Act: ST, MT targeted expansion of informal programs (CEFACEIs, Imp: LT Madre-a-Madre); ease supply constraints in rural and remote areas at other levels through expansion and use of TELEBASICA program Low quality schooling (iacx or DOOtS, materiats, Improve quality by providing inputs to schools in Act: S 1, M l bilingual materials & instruction, insufficient poorest corregimienlos (using PM): Imp: LT teachers, dilapidated infrastructure) ... * Providing textbooks at primary, secondary level free of charge on a loan basis ... * Providing school materials including bilingual materials at primary level ... . Training bilingual teachers so . Improving facilities on a demand-driven basis with priority to poor areas (via PM) Functional ineffiiencies Improve efficiency of public spending by: Act: St, MI . Decentralizing administration and some spending Imp: LT decisions (e.g., materials); implementing Law 28/1997 . Building planning capacity in MINEDUC . Introducing MIS, evaluation system .0 .. . Improving incentive system for teachers by linking pay to performance HEALTH CARE: Priority for poor, especially in indigenous areas Low access to neali care among poor, esp. .. . Evaluate t,actors limiting access (cost, availability Act: Sl, MIl indigenous; Traditional infectious and of facilities, etc.) Imp: MT for all communicable diseases: intestinal diseases, ... . Expand coverage of poor of cost-effective inter - malnutrition, respiratory diseases, TB, malaria; ventions, with emphasis on matemal-infant health, Social public health issues: teenage pregnancy, oral rehydration, potable water in rural and high fertility, violence & alcoholism indigenous areas, prevention and early detection of TB, family planning, health education, mobile health personnel to reach dispersed populations, etc. . Train bilingual, indigenous health personnel Low quality nealin care bor poor (imirtea ... lo improve quality and efticiency: Act: MIl competition among providers; lack of supervision . Establish a strategic partnership between MINSA Imp: MT & incentives; lack of medicines, medical and the CSS equipment * Separate financing and provision of services * Increase private-sector participation in the provision of services * Strengthen policy-making, regulation and management capabilities in MINSA . Decentralize services, increase community participation Biases in public spending tn ravor ot non-poor to. la improve equity in pubtic spending: (especially middle classes) * Collect data on cost by type and level of senice Act: ST for thorough incidence analysis Imp: MT * Target spending on new investments and Act: ST, MT rehabilitation according to poor using PM Imp: MT * Reorient spending in favor of primary care via Act: ST, MT .________ clinics and away fromn tertiary care Imp: MT Weaknesses m the health services network ... * Development and strengthenmg of the health Act: C(P, PM services network by improving infrastructure, Imp: PM sanitary technology and human resource capacity, and management capacity; implementing new models for financing and management of services; developing and implementing supervision, monitoring and evaluation mechanisms t..=top pnority; ..medium pnority;. pnori;PM pove ry map; S I= one year period; MI = I-i years; Lt = more han 3 years; Act peo for implementation of actions; Imp = period for impact on poverty 46 Box 6.4 Cont'd - SOCIAL ASSISTANCE & NUTRITION: Priority for poor in all areas, eap. indigenous CONSTRAINTS RECOMMENDATIONS Key Issues Priority Key Actions Period Compensatory Education Programs: (School feeding, cash transfers) * Targeting to poor is crucial *-o Improve/maintain targeting to poor using Act: ST, MT nutritional and PM criteria as follows: Imp: ST, MT * Modify law to introduce targeting of & LT MINEDUC feeding programs; graduate schools with low levels of malnutrition * Maintain targeting of FES lunch program during expansion * Target new cash transfer (Balboa) program * Inefficiencies in implementation *-. Reduce inefficiencies as follows: Act: ST, MT * Decentralize food procurement in remote Imp: ST, MT areas under school feeding programs & LT * Replace milk with more cost-effective foods under MINEDUC snack program * Review international experiences with school-based cash transfers for design of Balboa cash-transfer programn Various transfer and nutrition programs: _ _ * Evaluate impact and targeting of programs Act: ST, MT MINSA food supplements, various nutrition (with specific questions in LSMS 2000); Imp: LT interventions, PARVIS housing subsidy, introduce explicit targeting mechanisms in new FES Grupos Vulnerables programs all new programs (e.g., FES programs) Social Investment Programs FES Infrastructure & Credit Programs ec. * Design new criteria for targeting with new Act: ST Key issues: targeting to poor poverty map; Imp: MT * Expand share of FES spending subject to Act: MT targeting criteria Imp: MT MIPPE/MEF's PROINLO so. * Introduce targeting to poorer corregimientos Act: MT Key issues: targeting to poor using new PM Imp: MT * * Review possible functional overlap with FES Act: MT Imp: LT BASIC INFRASTRUCTURE General: Investments should be targeted to the poor by overlapping PM with inventory of existing infrastructure services Priorities should be demand-driven based on community participation Key issues for poor in rural areas: Access to transport, roads *. * Construction, rehabilitation and maintenance Act: MT, of rural access roads Imp: MT Access to electricity i * Istallation of rural electrification for all Access to potable water where deficient * * Coordinated multi-agency effort to improve access to potable water Key issues in indigenous areas: Access to potable water cc. * Coordinated multi-agency effort to improve Act: MT, access to potable water Imp: MT Access to latrines * * Support community/NGO efforts to promote for all construction of latrines Access to electricity, gas (excessive use of * * Promote use of gas over firewood for firewood for cooking) cooking Key issues for poor in urban areas: Efficiency of public transport services *. * Reforms to promote entry and competition in Act: MT, public bussing; rehabilitation and Imp: MT maintenance of urban streets; improvement for all of traffic management Access to sanitation & garbage services *o * Expand access using PM Quality of water services . * Improve quality of service delivery Access to & maintenance of public lighting * * Improve access & maintenance of public lighting with community participation = top priority;.. = medium pnority; . = prion WM poverty map; s I one year period; Mi = 1-3 years; L1 = moreman 3 yars;A period for implementation of actions; Imp = period for impact on poverty 47 Box 6.4 Cont'd - LAND AND PROPERTY TITLING CONSTRAINTS RECOMMENDATIONS Key Issues Priority Key Actions Period In urban areas, poor lack titles to their so* Identify and rectify inefficiencies in titling and Act: ST, MT, homes. Lack of titles also hampers ability registry system; adopt outreach campaign to LT to leverage credit for income generation. inform poor of procedures and opportunities; Imp: LT expand titling and registry giving priority to corregimientos with high concentrations of poor (via PM) In rural areas, poor lack titles to their land. *e Improve efficiency of land cadastre, titling, and Act: ST, MT, Lack of titles also hampers ability to registry system; complete and evaluate land-titling LT leverage credit for income generation. pilot project in Veraguas; expand and refine pilot Imp: LT approach to nationwide land titling campaign, giving priority to corregimientos with high concentrations of poor (via PM) INCOME GENERATION & FINANCIAL SERVICES Poor lack access to savings and credit ... * Adopt actions to extend titling of land and Act: MT, LT services, which are crucial for consumption property to poor (above), which can be used Imp: LT smoothing and income generation as guarantees for credit to improve credit- worthiness of poor *oo * Evaluate in detail constraints on demand and Act: ST supply side for providing better access to Imp: MT financial services among poor; tailor solutions to key constraints * Adopt innovative, market-friendly Act: MT approaches to providing financial services Imp: MT (e.g., solidarity groups) *so * Remove any existing interest-rate subsidies Act: ST, MT Imp: MT The poor rural communities do not have * * Provide TA for the agricultural production Act: MT, LT access to appropriate technology to enable and marketing of traditional and non- Imp: MT, LT them to diversify production and improve traditional products, using cost-effective productivity. methodologies. BUILDING SOCIAL CAPITAL Social capital helps communities reduce ... * Work with community groups in determining Act: ST, ,LT vulnerability, increase opportunities and local investment needs Imp: MT, LT leverage assistance. While the poor tend to . * Provide funds to communities for direct live in communities with higher social investment in social capital in response to capital, these ties are stronger in indigenous their requests (e.g., via community centers, areas and weaker in urban areas. _ sports areas, etc.) .= top prionty; .. = medium priority;. priority; PM p overty map; SI = one year penod; MT= 1-3 years; Lr = more han 3 years; Act period for implementation of actions; Imp =period for impact on poverty. Box 6.4 Cont'd - AREAS FOR FURTHER RESEARCH * Public expenditure analysis, including the distributional incidence and efficiency of public expenditures in key sectors and the management of public expenditures * More thorough analysis of the impact of existing labor legislation on the labor market in terms of overall segmentation, unemployment, and the informal sector; leading to specific reform proposals * Research to explore the links between poverty and the environment * Analysis of the distributional incidence of the social security system and its impact on the poor * Evaluation of the impact of social assistance programs, including: (i) School feeding programs (educational impact); (ii) MINSA supplementary feeding program; (iii) MIVI's PARVIS housing subsidy program; (iv) FES investments, social programs, and credit scheme; (v) MIPPE/MEF's PROlNLO program. * Participatory research via focus groups and other qualitative tools to further examine obstacles to increased school enrollment and attendance by indigenous children * Participatory research via focus groups and other instruments on crime and violence, possible solutions, and links to poverty 48 References for Main Report1 Deininger, Klaus and Lyn Squire. November 1997. "New Ways of Looking at Old Issues: Inequality and Growth." The World Bank. Draft. Fenreira, F.H.G. y Litchfield, J.A. 1997. "Poverty and Income Distribution in a High-Growth Economy Chile: 1987-1995" Capitulo 2 "Income Distribution and Poverty: A Statistical Overview". Fundacion Istmeina de Estudios Econ6micos y Sociales. November 1994. Empleo, Pobreza y Economia Informal. FIEES Series No. 3. Gonzalez, Jose Antonio. 1998. Labor Market Flexibility in 13 Latin American Countries and the United States: Stylized Facts about Structural Relationships Between Output and Employment- Unemployment-Wages. The World Bank. Government of Panama. September 19, 1997. Desarrollo Social Con Eficiencia Econ6mica 1997- 1999. Cabinet Resolution No. 222. Government of Panama. September 17, 1998. Nuevo Enfoque Estrategico Frente a la Pobreza. Cabinet Resolution No. 134. Hicks, Norman and Pia Peeters. September 17, 1998. "Poverty in Latin America and the Caribbean: A Survey." The World Bank. Draft. Maloney, William F. February 1998. "Are LDC Labor Markets Dualistic?" The World Bank. Draft. Maloney, William F. October 26, 1998. "Self-Employment and Labor Turnover in LDCs: Cross- Country Evidence." The World Bank. Draft. Maloney, William F. and Tom Krebs. April 1998. "Informality, Distribution, Rigidity, and Labor Market Institutions in Latin America: An Applied Efficiency Wage Model." The World Bank. Draft. McLure, Charles E. 1974. The distribution of income and tax incidence in Panama. Public Finance Quarterly Vol. 2 No. 2. Ministerio de Planificaci6n y Politica Econ6mica and RUTA SociaL April 1997. El Gasto Pjiblico Social en Panama: Tendencias y Prioridades. Ministerio de Planificaci6n y Politica Econ6mica. December 1993. Panamd: Niveles de Satisfaccion de las necesidades basicas. Mapa de Pobreza. Documento de Trabajo No. 44. Ministerio de Planificacion y Politica Econ6mica. 1996. Perfil e Indicadores de Pobreza en Panamd. Data calculated by SIAL/OIT. Ministerio de Planificaci6n y Politica Econ6mica, Direcci6n de Politicas Sociales. March 1998. Panamn: Informe Social 1997. 'Additional references included in Annexes. Ministerio de Planificaci6n y Politica Econ6mica, Direcci6n de Politicas Sociales. December 1997. Programa de Desarrollo Social: Periodo 1994-1999. Ministerio de Planificaci6n y Politica Econ6mica, Direcci6n de Politicas Sociales and Fondo de Poblaci6n de las Naciones Unidas. 1997. Panama: Pasos Hacia Un Futuro Mejor: Politicas con Enfoque de Genero, Poblaci6n y Pobreza. Ministerio de Salud. September 1996. Estadisticas de Salud. Molina, S., Sergio. December 1982. La Pobreza: Descripci6n y andlisis de politicas para superarla. Revista de la CEPAL 18. Moser, Caroline O.N. 1996. Confronting Crisis: A Summary of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities. The World Bank. ESSD Series No. 7. Ravallion, Martin and Shaohua Chen. May 1997. "What Can New Survey Data Tell Us about Recent Changes in Distribution and Poverty?" The World Bank Economic Review, Vol. 11 No. 2. Sahota, Gian S. 1990. Poverty Theory and Policy: A Study of Panama. Johns Hopkins University Press. UNICEF and Ministerio de Planificaci6n y Politica Econ6mica. July 1997. Panamd: La Niiiez y la Mujer en la Encrucijada del Ano 2000. Second Edition. UNICEF and Ministerio de Trabajo y Bienestar Social. July 1997. Desarrollo Humano en Panamd: Trabajo Infantil y Educaci6n. The World Bank. October 29, 1998. Country Assistance Strategy of the World Bank Group for the Republic of Panama. Report No. 18421-PAN. The World Bank. June 1, 1995. Republic of Nicaragua: Poverty Assessment. Report No. 14038-NI. The World Bank. June 30, 1995. Ecuador Poverty Report. Volume 2. The World Bank. July 20, 1995. Panama: A Dual Economy in Transition. Report No. 13977-PAN. The World Bank. March 17, 1995. Panama: Issues in Basic Education. Report No. 13701-PAN. The World Bank. June 10, 1997. Poverty and Income Distribution in a High-Growth Economy: Chile: 1987-1995. Report No. 16377-CH. World Health Organization. 1997. Global Database on Child Growth and Malnutrition. Yrarrazaval, Rafael E. October 1997. "Panama: Politica Comercial y de Precios de Productos Agropecuarios 1992-1997." The World Bank. Draft.