If ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I * ... . , . _ . ... ... _ ....... ...... _ ... _ _ ........ _. . _ ...... __ _ . _ _. __. _ RECENT WORLD BANK TECHNICAL PAPERS No. 251 Sharma, Rietbergen, Heimo, and Patel, A Strategyfor thie Forest Sector in Sub-Saharan Africa No. 252 The World Bank/FAO/UNIDO/Industry Fertilizer Working Group, World and Regional Supply and Demand Balancesfor Nitrogen, Phosphate, and Potash, 1992/93-1998/99 No. 253 Jensen and Malter, Protected Agriculture: A Global Review No. 254 Frischtak, Governance Capacity and Economic Reform in Developing Countitries No. 255 Mohan, editor, Bibliographiy of Publications: Technical Department, Africa Region, July 1987 to April 1994 No. 256 Campbell, Design and Operation of Smaliholder Irrigation in South Asia No. 258 De Geyndt, Managing thte Quality of Healthi Care in Developing Coulntries No. 259 Chaudry, Reid, and Malik, editors, Civil Service Reform in Latin America and the Caribbean: Proceedings of a Conference No. 260 Humphrey, Payment Systems: Principles, Practice, and Improvemlents No. 261 Lynch, Provisionifor Chihlren witil Sp7ecial Educational Needs in the Asia Region No. 262 Lee and Bobadilla, Healtlh Statisticsfor the Americas No. 263 Le Moigne, Subramanian, Xie, and Giltner, editors, A Guide to the Formulation of Water Resources Strategy No. 264 Miller and Jones, Organic and Compost-Based Growing Mediafor Tree Seedling Nurseries No. 265 Viswanath, Building Partnersihipsfor Poverty Reduction: Thie Participatory Project Planning Approach of tihe Womnen's Enterprise Management Traininig Outreach Program (WEMTOP) No. 266 Hill and Bender, Developing the Regulatory Environment for Competitive Agricultural Markets No. 267 Vald6s and Schaeffer, Surveillance of Agricultural Prices and Trade: A Handbookfor the Dominican Republic No. 268 Vald6s and Schaeffer, Surveillance of Agricultural Prices and Trade: A Handbookfor Colombia No. 269 Scheierling, Overconming Agricultural Pollution of Water: The Challenge of Integrating Agricultural and Environmental Policies in the Europeatn Union No. 270 Banerjee, Reliabilitation of Degraded Forests in Asia No. 271 Ahmed, Technological Development and Pollutiotn Abatement: A Study of How Enterprises Are Finding Alternatives to Clhlorofluorocarbons No. 272 Greaney and Kellaghan, Equity Issues in Public Examinations in Developing Countries No. 273 Grimshaw and Helfer, editors, Vetiver Grassfor Soil and Water Conservation, Land Relhabilitation, and Embankment Stabilization: A Collection of Papers and Newsletters Compiled by the Vetiver Network No. 274 Govindaraj, Murray, and Chellaraj, Health Expenditures in Latin America No. 275 Heggie, Management and Financing of Roads: An Agendafor Reform No. 276 Johnson, Quality Review Schettmesfor Auditors: Their Potentialfor Sub-Saharan Africa No. 277 Convery, Applying Environmental Economiiics in Africa No. 278 Wijetilleke and Karunaratne, Air Quality Management: Considerationsfor Developing Countries No. 279 Anderson and Ahmed, The Casefor Solar Energy Investments No. 280 Rowat, Malik, and Dakolias, Judicial Reform in Latin America and the Caribbean: Proceedings of a World Bank Conference No. 281 Shen and Contreras-Hermosilla, Environtmtental and Economic Issues in Forestry: Selected Case Studies in Asia No. 282 Kim and Benton, Cost-Benefit Analysis of the Onchocerciasis Control Program (OCP) No. 283 Jacobsen, Scobie and Duncan, Statutory Intervetitiotn in Agricultural Marketing: A New Zealand Perspective No. 284 Valdcs and Schaeffer in collaboration with Roldos and Chiara, Surveillance of Agricultural Price and Trade Policies: A Handbookfor Uruguay No. 285 Brehm and Castro, The Marketfor Water Rights in Chile: Major Issues No. 286 Tavoulareas and Charpentier, Clean Coal Technologiesfor Developing Countries No. 287 Gillham, Bell, Arin, Matthews, Rumeur, and Heam, Cotton Production Prospectsfor the Next Decade No. 288 Biggs, Shaw, and Srivastiva, Technological Capabilities and Learninig in African Enterprises No. 289 Dinar, Seidl, Olem, Jorden, Duda, and Johnson, Restoring and Protecting tihe World's Lakes and Reservoirs (List continues on the inside back cover) WORLD BANK TECHNICAL PAPER NO. 328 Social Development and Absolute Poverty in Asia and Latin America Willy De Geyndt The World Bank Washington, D. C. Copyright X 1996 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing July 1996 Technical Papers are published to communicate the results of the Bank's work to the development com- munity with the least possible delay. The typescript of this paper therefore has not been prepared in ac- cordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal sta- tus of any territory or the endorsement or acceptance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publica- tions, which contains an alphabetical title list (with full ordering information) and indexes of subjects, au- thors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'16na, 75116 Paris, France. ISSN: 0253-7494 Willy De Geyndt is principal public health specialist in the Human Resources and Social Development Division of the World Bank's Asia Technical Department. Library of Congress Cataloging-in-Publication Data De Geyndt, Willy. Social development and absolute poverty in Asia and Latin America / Willy De Geyndt. p. cm. - (World Bank technical paper, ISSN 0253-7494 ; no. 328) ISBN 0-8213-3695-9 1. Social indicators-Asia. 2. Social indicators-Latin America. 3. Poverty-Asia. 4. Poverty-Latin America. 5. Asia-Social conditions-Statistics. 6. Latin America-Social conditions- Statistics. 7. Asia-Economic conditions-Statistics. 8. Latin America-Economic conditions-Statistics. I. Title. II. Series. HN652.7.D44 1996 306'.095-dc2O 96-27954 CIP CONTENTS Foreword ...........................................................v Abstract .......................................................... vii Acknowledgments .......................................................... ix I. Definition of Main Concepts ...........................................................1 II. Methodology ............................................................3 Asian and Latin American Countries Selected ............................................................3 Variables Selected and Time Period ............................................................3 Data Collection ...........................................................3 Data Analysis ...........................................................4 III. Univariate Description ...........................................................5 IV. Correlational Analysis ........................................................... 14 Data Presentation ........................................................... 14 Statistical Methodology: Use of Nonparametric Statistics .......................................... 15 Statistical Analysis .......................................................... 17 Correlating Absolute Poverty and Social Development . ........................................... 17 Correlating 1885 with 1990 Absolute Poverty and 1985 with 1990 Social Development .......................................................... 18 Correlating Absolute Poverty and Income ........................................................... 19 Correlating Social Development and Income ........................................................... 19 Interrelating Rankings on Absolute Poverty, Social Development and Income ........ 20 Discussion .......................................................... 22 V. Other Determinants of Absolute Poverty .......................................................... 24 Material Environment .......................................................... 24 Religious/Spiritual Environment ........................................................... 24 Political Environment .......................................................... 25 Culture .......................................................... 26 Social Organization .......................................................... 27 VI. Conclusion .......................................................... 28 Annex I Absolute Poverty in East Asia, South Asia and Latin America ......................................... 29 Table Al: Number and Percentage of Population in Asia with per capita income less than $1/day .......................................................... 30 Table A2: GDP Growth rates, 1980-1990 and GNP Per Capita, 1985 and 1990 ............. 31 Table A3: Life Expectancy at Birth, 1985 and 1990 ........................................................ 32 Table A4: Infant Mortality Rate (IMR), 1985 and 1990 ................................................... 33 iii Table A5: Under-5 Mortality Rate, 1985 and 1990, and Under-5 Mortality Rate by Sex, 1990 ............................................................. 34 Table A6: Maternal Mortality, 1980 and 1988 ............................................................. 35 Table A7: Annual Incidence Rate of Tuberculosis (per 100,000 population), 1990 ........ 36 Table A8: Total Fertility Rate, 1985 and 1990 .............................................................. 37 Table A9: Nutrition: Babies with low birth weight, 1985 and 1990, and Prevalence of under 5 malnutrition, 1990 ............................................................. 38 Table AIO:Adult Illiteracy, 1985 and 1990 ............................................................. 39 Table All :Education: School Enrollment, 1985 and 1990 .............................................. 40 Table A12:School Enrollment: Females per 100 Males, 1985 and 1990 ......................... 41 Table A13:Health Expenditures, 1990 ............................................................. 42 Table A14:Latin America - Government Health Expenditures as % of GDP, 1980-1990 ............................................................. 43 Table Al 5:Social Indicators for the Largest Countries in East Asia, South Asia and Latin America. Summary Table, 1985 ..................................................... 44 Table A16:Social Indicators for the Largest Countries in East Asia, South Asia and Latin America, Summary Table, 1990 ...................................................... 45 iv FOREWORD In 1990 the Asia region was home to 2.8 billion people or 55 percent of the world's population. One third of the Asian population was classified as absolute poor. The ten largest countries in Asia - five in East Asia and five in South Asia- are the focus of the analysis in this paper that examines the relationships between absolute poverty, social development and level of income. These ten countries account for 93 percent of the Asian population. A set of the six largest Latin American countries is included in the analysis to provide a comparative perspective. Social development and absolute poverty are affected by a number of factors: economic growth; level of income; access to health, education, family planning, nutrition, and water and sanitation; environment protection; the maturity and good governance of institutions and the regulatory, political and legal environments. The relative importance of these factors and their interaction are not well known and are difficult to predict. For example, income is widely acknowledged to be a good predictor of absolute poverty rates but at the same time empirical evidence shows that economic growth is not all that matters in reducing poverty. The author relates the incidence of absolute poverty in a sample of sixteen countries to a set of thirteen social development measures and to the level of income using a nonparametric statistic to correlate associations and to test the statistical significance of the correlations. Other variables that are more difficult to measure and that may affect the level of absolute poverty are briefly explored, viz. a country's material environment, its religious/spiritual history, its political structure, its culture and its social organization. Harold W. Messenger Director Asia Technical Department v ABSTRACT A social development perspective is used to seek an answer to the question: why is the incidence of absolute poverty so different among developing countries in East Asia, South Asia and Latin America? "Economic growth is clearly not all that matters to reducing poverty ... and there are other 'non-income' aspects of poverty - such as lack of access to schools or health care ...' This paper relates the incidence of absolute poverty to the degree of presence or absence of thirteen social development indicators for the years 1985 and 1990 and to the average level of income for the same time period. There are three parts to the analysis. First, a univariate statistical description presents and explains each variable used. Second, correlational analysis examines the rank correlation between the incidence of absolute poverty and the degree of social development and income. Nonparametric statistical techniques are used for measuring association and for testing the statistical significance of the correlations. Third, five other forces that may affect the level of absolute poverty are explored, viz. a country's material environment, its religious/spiritual history, its political structure, its culture and its social organization. M. Ravallion: Economic Letters 48 (1995) 411-417. vii ACKNOWLEDGMENTS It would not have been possible at all to write this paper without the help of Ms. Junko Otani who provided invaluable support in collecting and formatting the data and especially in the tedious process of verifying the accuracy of each data item and the reliability of the data sources. A large number of World Bank staff helped verify data and supplied the best available estimates based on their intimate country knowledge. I am endebted to B. Bidani, E. Bos, G. Datt, F. Delannoy, B. Duza, S. Habayeb, A. Losada, J. Martins, R. McColgan-Mohamed, P. Mitchell, J. Otani, M. Over, I. Pathmanathan, T. Rana, M. Ravallion, F. Saadah, A. Saliba and L. Shrestha. Colleagues outside the World Bank also helped in checking data and provided advice in interpreting the information. I thank K. Hill, A. Hyder and H. Mosley from Johns Hopkins University for their kind support as well as Dr. Marian Lief Palley from the University of Delaware for her comments. Advice on the correct use of nonparametric statistics was gratefully received from my former doctoral thesis advisor and mentor Professor Vernon E. Weckwerth of the School of Public Health at the University of Minnesota. I received critical comments from J. Socknat, J. Hammers and P. Musgrove on an early draft and from M. MacDonald on the final draft but hasten to say that all comments were greatly appreciated but that some could not be taken into account. ix I. Definition of Main Concepts The definition of the main concepts in Section I is followed by a description of the research methodology in Section II. The dependent and explanatory variables are described individually in Section III and are correlated in Section IV. Section V explores non-quantitative variables and is followed by the concluding section. The data used in the analysis are contained in sixteen tables in the annex. Social Development is a multidimensional concept including social services, social integration, environment protection, rural development as well as political and institutional development. Absolute Poverty is defined as a per capita income of less than one US dollar a day. For comparative purposes income is calculated on the basis of equivalent and comparable levels of purchasing power. Another frequently used poverty definition is income necessary to maintain a subsistence intake of 2,150 calories a day. A comparison of the results of using these two definitions shows slight upward and downward shifts in the number of absolute poor per country but the orders of magnitude revealed by the two approaches are quite close. Social Services forrn a subset of the broader concept of social development. Access to quality social services has a direct impact on reducing poverty incidence and on improving the lives of disenfranchised and economically marginalized populations. Provision of targeted social services is an effective and necessary complement to an economic growth strategy and to social integration. A social services delivery strategy would include the following components: (a) Health Care: reduced morbidity increases productivity, incomes and the capacity to learn; (b) Education: allows the poor to participate in modem sectors at better wages increasing productivity and incomes; educated women choose lower fertility rates; (c) Population Growth: lower fertility reduces mortality and morbidity and lower population growth makes more available per capita within a given global economic growth; demographic trends change and increase the demands on health services and education; (d) Food Security and Nutrition: adequate intake of calories and proteins increases productivity, learning and well being; I (e) Safe Drinking Water and Sanitation: poor sanitation in dense human settlements contaminates surface waters with human excreta; the cost in infant deaths is high and the cost of high adult morbidity is probably higher especially in lost productive potential; waterbome diseases reduce the ability to absorb nutrients and add to the cost of medical care. Per Capita Income: A country's Gross National Product (GNP) and Gross Domestic Product (GDP) divided by the population. GDP per capita measures the total output of goods and services for final use produced by residents and nonresidents. GNP per capita measures the total domestic and foreign value added claimed by residents; it comprises GDP plus net factor input from abroad. Both income indicators are measures of location, i.e. they represent the mean or the average, and do not provide information about the dispersion around the mean or the distribution of resources among the population. Thus two countries may have the same average income level but may have a vastly different distribution of income among the population. It is also generally acknowledged that perfect cross-country comparability of GNP per capita estimates cannot be achieved because of differences in national accounting and demographic reporting systems and the use of official exchange rates for converting GNP data expressed in different national currencies to a common denomination (conventionally the US dollar). 2 II. Methodology Asian and Latin American Countries Selected The extent of poverty varies among countries, within countries and among population groups. In Asia about 80% of the poor are located in rural areas. Women are overrepresented and suffer more intense poverty than men. This paper divides Asia into two geographic groups of countries with sharply different poverty incidence: East Asian countries selected are China, Indonesia, Malaysia, Philippines and Thailand or 92 percent of the East Asia population; South Asia countries selected are Bangladesh, India, Nepal, Pakistan and Sri Lanka or 96 percent of the South Asia population. Latin American countries selected for inclusion in this analysis are Brazil, Mexico, Argentina, Colombia, Peru and Venezuela or 79 percent of the population of the Latin American continent. The present analysis focuses on differences between countries. However it is known, that as the proportion of the population in poverty declines, the nature of poverty changes and becomes more intractable and more localized (e.g. the southwest of China, Java in Indonesia, center islands and Tagalog in the Philippines, the Northeast in Brazil, the Andean mountain area in Peru). New poverty groups emerge and the needs of the poor change. Poverty incidence analysis and strategies to reduce poverty within countries and among population groups (the displaced, the disenfranchised, the refugees) would require a more targetted and a country-specific approach. Variables Selected and Time Period The dependent variable in this analysis is absolute poverty. Four of the five social services components mentioned above provide fifteen social development measures. The safe drinking water and sanitation component was not included for lack of reliable comparative data for the years and the countries selected. A measure of GNP per capita and a measure of GDP growth rates provide the economic background. Most data are for the years 1985 and 1990 allowing a five-year evolution comparison. In one case (maternal mortality) data for 1980 and 1988 were used. Data Collection Data were derived from a variety of existing sources: World Bank and United Nations published reports, inhouse World Bank background papers, national health, demographic and fertility surveys, and country statistical reports. World Bank task managers working on specific countries were consulted to assist in resolving conflicting data values. Data can be considered valid and reliable, and acceptable for indicating orders of magnitudes and time trends. 3 Data Arkaly Section III contains the univariate statistical analysis of all variables used in the study with data tables being presented in the annex. Section IV presents the results of the correlational analysis using nonparametric statistics on the ordinal data set that is shown in Tables 1-4 on pages 13-16. The statistical methodology used is explained in Section IV.2. Section V deals with qualititative parameters and is not data-based as are the two prior sections. 4 III. Univariate Description This section presents a statistical description of the dependent variable, of the economic explanatory variable and of fifteen social development explanatory variables. Tables A- I through A- 14 in the annex contain the variable-specific numerical data and the matrices in tables A- 15 and A- 16 summarize the data for the years 1985 and 1990. Absolute Poverty (Table A-I). Absolute poverty is measured using the comparative yardstick of per capita income of less than one US dollar per day expressed in equivalent 1985 purchasing power levels or purchasing power parity (PPP). Between the years 1985 and 1990 and among the sixteen countries selected, the proportion of poor decreased most dramatically in Indonesia, Malaysia, Thailand and Colombia; a lesser decrease occurred in the Philippines, India, Nepal, Pakistan and Sri Lanka. In Venezuela the percentage of absolute poor tripled, in Peru it doubled, in Bangladesh, Brazil and China it increased 65, 32 and 22 percent respectively, and it remained unchanged in Mexico. A decrease or increase in the proportion of poor was matched by a corresponding increase or decrease in the number of poor except in the case of India and Pakistan where high fertility rates caused an increase in the number of poor even though the proportion of poor decreased slightly. Because of their predominant population size, the largest country in each region accounts for the largest number of poor, i.e. 88 percent of the South Asian poor are in India, two thirds of the East Asian poor are in China and 44 percent of Latin American poor are in Brazil. GNP per Capita (Table A-2). GNP per capita increased for most countries during the five year period under consideration. The highest increase was registered by Thailand with 77.5% - which also had the highest annual growth rate - followed by Brazil (63%), Bangladesh (40%), India (30%), the Philippines (26%), Sri Lanka (24%) and Mexico (20%). Pakistan's GNP per capita did not change between 1985 and 1990 and two Latin American countries (Colombia and Venezuela) showed a slight decrease. In general, GNP per capita is highest in the Latin American countries and lowest in the South Asian countries with East Asia presenting a mix. An analysis of economic growth rates however offers a different picture as explained in the next paragraph. Economic Growth Rates (Table A-2). The average annual growth rate in the Gross Domestic Product (GDP) serves as a broad economic development measure for the five year period preceding the years in which absolute poverty was measured, i.e. 1980- 85 for 1985 and 1985-90 for 1990. The six Latin American countries have the slower average annual growth rates and are in the bottom half in both periods measured confirming the often quoted "lost decade" statement. Colombia is the strongest performer among them and Peru the weakest. The five South Asian countries situate 5 themselves squarely in the middle with strong economic performances by India and Pakistan. China, Malaysia, Thailand and Indonesia lead their region as well as all countries with average annual GDP growth rates between five and ten percent. Life Expectancy at Birth (Table A-3). A South Asian newborn, except a Sri Lankan, will on average live ten years less than an East Asian or Latin American newborn. Life expectancy at birth is roughly between 60 and 70 years of age in East Asia and in Latin America whereas in South Asia, except Sri Lanka, it is between 50 and 60 years of age. Increases in life expectancy were registered between 1985 and 1990 in all countries in our sample. Female life expectancy follows the expected pattern of being longer than the corresponding male number except in four South Asian countries where male life expectancy in 1990 exceeded female life expectancy by one or two years. Infant Mortality (Table A-4). In most countries under consideration infant mortality rates decreased between 1985 and 1990 with drops of more than a quarter in Malaysia, Mexico, Colombia and Peru. Four countries (Bangladesh, India, Nepal and Pakistan) still have rates of more than 90 infant deaths per 1,000 live births. Three countries (Sri Lanka, Colombia and Malaysia) have rates of less than 20. Deaths during the post-neonatal period (1-11 months) account for most deaths in countries with high rates whereas in low rate countries the neonatal deaths (0-28 days) predominate. Sustained economic growth and access to adequate social services reduce postneonatal deaths but neonatal deaths are more difficult to reduce and gains are more costly and are achieved more slowly. At lower levels of IMR, say below 50, and higher levels of life expectancy at birth, say above 60 years, health care services contribute only marginally to improving these indicators and interventions are more costly. For example, intensive neonatal care services can reduce the neonatal mortality rate and thereby reduce IMR but require an infrastructure and highly skilled personnel. Cardiovascular surgical interventions, organ transplants and orthopedic surgery (hip, knee) may prolong life and reduce suffering but require large capital investments in sophisticated equipment and highly trained staff and a regular flow of funds to cover operating costs. Providing health care makes a larger differences with high levels of IMR and low levels of life expectancy especially through reducing vaccine-preventable diseases, improving nutrition, providing safe drinking water and adequate sanitary conditions. Under-5 Mortality (Table A-5). All sixteen countries show a decrease between 1985 and 1990 in the rate of children dying by age five. The U5MR is highest in Bangladesh, India, Nepal and Pakistan and in the same countries the risk of dying by age five is actually higher for females than for males. The expected pattern is one of male children under five having a higher death rate than females. Female babies have a 23 percent lower risk of dying by age five than male babies in industrial market economies. 6 Maternal Mortality (Table A-6). Reporting of deaths occurring during pregnancy and childbirth is notoriously inaccurate and what this table presents are best estimates and orders of magnitude for the years 1980 and 1988. No plausible explanation can account for the large increases shown by Mexico, Argentina and Colombia between 1980 and 1988. Keeping in mind that the maternal mortality ratio in industrial market economies is about ten per 1 00,000 live births one cannot be but extremely concerned with ratios that in some countries are 60 to 80 times higher. Tuberculosis Incidence (Table A-7). This is also a difficult rate to collect and report and Table 6 shows best estimates of the 1988 incidence rate of tuberculosis. Noteworthv is the very wide range from 44 per 100.000 population in Venezuela to 280 in the Philippines. Five countries have a rate of less than 100 and five countries have rates of more than 200. Population Growth (Table A-8). The total fertility rate (TFR) has declined consistentlv in the 1980s for 14 out of the 16 countries as shown in Table 7 for the year-s 1980 and 1988. Women in four South Asiaii countries still average mnore than four children whereas five countries show rates of less than two children on average. T'he TFR is between 3 and 4 in the other seven countries. The poor nearly always have higlher fertility and changes in fertility rates may be an early predictor of where povertv will be concentrated even when the absolute poverty incidence decreases. In regions within countries there may be a strong association between fertility decline and poverty incidence and distribution. Low Birth Weight Babies (Table A-9). An important nutritional indicator is the proportion of children that are born weighing less than 2,500 grams. Only in four of the 15 reporting countries did the percentage of babies with low birth weight decrease from 1985 to 1990, it increased in eight and three showed the same percentage for both years. Babies in the six Latin American countries fare slightly better than babies in East Asia but the percentages in South Asia have remained high and are double the oncs in Latin America. Research indicates that East European countries show an increase in low birth weight babies during the 1980s as the level of poverty increased. Under-5 Malnutrition (Table A-9). Child malnutrition data arc onily available tor 1990 and the prevalence pattern among the twelve reporting countries is similar to thle pattern shown for babies with low birth weight. This confirms an intuitive and statistical relationship between low birth weight babies and malnutrition in children under five. Adult Illiteracv (Table A-10). The total rate and the female rate for adult illiteracy decreased in almost all countries between the years 1985 and 1990. Female illiteracy is higher than total adult illiteracy in all countries except in Venezuela where it is five percentage points lower. Illiteracy affects less than 20 percent of the population of Latin American countries, between 20 and 30 percent of the population in three East Asian 7 countries (but ten percent in the Philippines and seven percent in Thailand). However, illiteracy rates in four South Asian countries are up to fifteen times higher than the lowest rate observed in Argentina (5%). It must be noted that literacy figures must be interpreted with caution as literacy definitions vary greatly from country to country. Primary and Secondarv Education (Table A-l 1). Enrollment rates for primary education are high and at about the same level in East Asia and Latin America for both indicator years 1985 and 1990. Four South Asian countries show markedly lower percentages with Pakistan enrolling less than half of its children. Some small changes can be noted between 1985 and 1990 but they do not show a trend. This is partly due to the ratio used which is a proportion of pupils to the population of school-age children. Some pupils that are included in the numerator may belong to school-age groups that are not included in the denominator, i.e. they may be younger than six or older than eleven. This counting artefact affects less the enrollment percentages in secondary education which show a consistent increase between 1985 and 1990 for all countries except Bangladesh, Mexico and Venezuela. Only six countries enroll more than half of the secondary school-age children and four countries enroll less than one third. This variable documents school enrollment. Improving access to primary and secondary education are important policy goals. The quality of available schooling, especially in countries where enrollment rates are already acceptable, is the single most important factor determining primary school completion2. Female Education (Table A- 12). Six countries are not reporting 1985 or 1990 female enrollment rates in primary and secondary education making the analysis less complete for female education than for the other variables. Eight countries increased female enrollment in primary and secondary education between 1985 and 1990. The ratio of females to males in primary schools in 1990 is well above 90 for all Latin American and East Asian (except China) countries and for Sri Lanka. The other four South Asian countries range from a low of 47 in Nepal to a high of 81 in Bangladesh. Secondary education data show five countries with more than 90 females per 100 males enrolled and two countries with less than 50 (Bangladesh and Pakistan). Malaysia, Sri Lanka and Venezuela enrolled more females than males in secondary schools in 1990. Health Expenditures in 1990 (Table A-13). Generally as countries achieve higher economic wealth they tend to spend more on health. This is also the case here. Per capita expenditures for health care are highest in Latin America and lowest in South Asia mirroring the GNP per capita levels in the regions under analysis. Noteworthy is the very wide range from US$7 to US$312. This wide variation in per capita health expenditures persists even when grouping countries by levels of GNP per capita: the five countries 2 "Investing in Education": Poverty Lines, No 2, World Bank, April 1996. 8 with GNP per capita of more than US$2,000 have the widest range from US$67 (Malaysia) to US$312 (Argentina), and the eight countries with GNP per capita of less than US$1,000 range from US$7 (Nepal) to US$21 (India) still a difference by a factor of three. Total health expenditures as a percentage of GDP ranges from 2.0 percent in Indonesia and the Philippines to 9.6 percent in Argentina. East Asia countries (except Thailand) spent less of their GDP on health than do South Asia and Latin America. Brazil and India rank behind Argentina and spent, respectively, 6.4 and 6.0 percent of their GDP on health in 1990. The countries in this sample also support the general worldwide trend of higher public sector expenditures in richer countries and higher private sector - mainly out-of-pocket - expenditures in poorer countries. OECD countries health spending is financed about 80 percent out of general taxation or wage taxes. In eleven countries private sector expenditures for health surpass public sector expenditures whereas in four countries the reverse is true. The largest difference between public and private sector spending occurs in Thailand where private sector expenditures in 1990 were 3.5 times the public sector expenditures. More recent statistics (1994) indicate that out-of-pocket payments account for about 85 percent of total health care spending in Thailand. Government Spending on Health in Latin America in the 1980s (Table A- 14). Government spending on health in Peru and Venezuela dropped sharply in the second half of the 1980s falling by about half in the case of Venezuela and by as much as two thirds in the case of Peru over a ten-year period. The other four countries experienced dips in government spending on health during the 1980s but finished the decade with higher percentages of GDP allocated to health as compared with the beginning of the 1980 decade. An increase of two thirds in public sector spending on health was registered by Argentina, almost half by Mexico, one third by Colombia and one quarter by Brazil. 9 Table 1: Absolute Poverty in East Asia, South Asia and Latin America. Ranking of the 16 Largest Countries on Social Indicators for 1985. Country % of Life IMR U5MR MMN R* TFR LBW Illiter Prim Sec PriFem Sum of Overall Poor Exp Educ Educ Educ Rankings Rank China 2.5 4 6 6 1 1 1.5 12 1 9.5 12 54 3 Indonesia 13 13 11 12 15 9 8.5 10 3 9.5 9.5 100.5 12 Malaysia 4 5 4.5 1 2 7 5 11 11 6 5 57.5 6 Philippines 12 10 9 9 4 11 12 7 8 2.5 3 75.5 8 Thailand 6 9 7 7 11 2.5 7 2 12 12 9.5 79 9 Bangladesh 8 14.5 16 16 14 14 16 14 15 14 13.5 147 16 India 15 12 13 13 13 13 15 13 13 11 13.5 129.5 13 Co Nepal 14 16 15 15 16 16 8.5 16 14 13 16 145.5 15 Pakistan 7 14.5 14 14 8.5 15 13 15 16 15 15 140 14 Sri Lanka 11 2 1 4 7 2.5 14 5.5 10 4 6.5 56.5 5 Brazil 10 7.5 10 10 10 6 3 9 9 16 9.5 90 11 Mexico 9 6 8 8 6 11 10.5 3 5 5 4 66.5 7 Argentina NA 2 4.5 3 5 4.5 1.5 1 6.5 1 9.5 38.5 1 Colombia 2.5 7.5 3 5 8.5 4.5 10.5 4 4 7 1 55 4 Peru 5 11 12 11 12 11 5 8 2 2.5 6.5 81 10 Venezuela 1 2 2 2 3 8 5 5.5 6.5 8 2 44 2 *1980 Sources: Tables A1, A3 - A13 Table 2: Absolute Poverty in East Asia, South Asia and Latin America. Ranking of the 16 Largest Countries on Social Indicators for 1990. Country % of Life IMR U5MtR NIMR* TB TFR LBW Ilite Prim Sec PriFem per cap % GDP Sum of Overall Poor Exp Educ Educ Educ H Exp Health Rankings Rank China 4 4.5 8 8 6 8 1 2.5 12 1 7 12 14 10 94 9 Indonesia 8 12 12 12 13 13 9 13 11 3 8.5 9.5 12.5 15.5 144 12 Malaysia 1 4.5 1 1 1 4.5 7 6.5 10 11 4 4.5 6 14 75 5 Philippines 11 10 9 9 4 16 11 14 3 5.5 2 6.5 11 15.5 116.5 10 Thailand 2 8.5 6 6 2 11 2.5 12 2 14 13 4.5 5 5 91.5 7 Bangladesh 10 15.5 15 15 15 13 14 16 14.5 15 16 13 15.5 12 189.5 16 India 15 13 13 13 12 13 13 15 13 10 10 14 9 3 151 13 Nepal 14 15.5 14 14 14 9.5 16 10.5 16 13 14 16 15.5 5 173 14.5 Pakistan 5 14 16 16 16 7 15 2.5 14.5 16 15 15 12.5 11 170.5 14.5 Sri Lanka 6 1.5 2 2 5 9.5 2.5 2.5 4.5 9 1 9.5 10 9 68 3 Brazil 13 8.5 11 10 7.5 3 6 8.5 9 8 11 8.5 2 2 95 8 Mexico 9 4.5 7 7 10.5 5 11 10.5 6.5 4 5 6.5 3 8 88.5 6 Argentina NA 1.5 5 4.5 7.5 2 4.5 1 1 5.5 8.5 1 1 1 44 1 Colombia 3 7 3 3 5 4.5 4.5 6.5 6.5 7 6 3 7 4 67 2 Peru 12 11 10 11 9 15 11 8.5 8 2 3 8.5 8 13 118 11 Venezuela 7 4.5 4 4.5 3 1 8 2.5 4.5 12 12 2 4 7 69 4 *1988 Sourc: Tables Al, A3 - A13 Table 3: Absolute Poverty in East Asia, South Asia and Latin America. GNP per Capita and Rankings, 1985 - 1990. 1985 1990 Country USD Rank USD Rank China 310 13 370 13 Indonesia 530 10 570 10 Malaysia 2,000 4 2,320 5 Philippines 580 9 730 9 Thailand 800 8 1,420 6 Bangladesh 150 16 210 15 India 270 14 350 14 Nepal 160 15 170 16 Pakistan 380 11.5 380 12 SriLanka 380 11.5 470 11 Brazil 1,640 5 2,680 1 Mexico 2,080 3 2,490 3 Argentina 2,130 2 2,370 4 Colombia 1,320 6 1,260 7 Peru 1,010 7 1,160 8 Venezuela 3,080 1 2,560 2 Source: Table A2 12 Table 4: Absolute Poverty in East Asia, South Asia and Latin America. Annual GDP Growth Rates and Rankings, 1980 - 1990. 1980- 1985 1985- 1990 Country % Rank % Rank China 9.8 1 7.7 2 Indonesia 5 6 5.9 5 Malaysia 5.4 4 6.9 3 Philippines -1. 1 14 4.8 8 Thailand 5.6 3 10.1 1 Bangladesh 4.9 7 3.7 10 India 5.2 5 6.2 4 Nepal 4.2 9 5.1 7 Pakistan 6.4 2 5.8 6 SriLanka 4.7 8 2.8 11 Brazil 0.9 12 1.9 13 Mexico 1.2 11 1.6 14 Argentina -1.2 15.5 -0.3 15 Colombia 2.2 10 4.4 9 Peru -0.9 13 -2.7 16 Venezuela -1.2 15.5 2 12 Source: Table A2 13 IV. Correlational Analysis Data Presentation Annex Tables A-15 for 1985 and A-16 for 1990 summarize the data from Tables A-I through A-14 in a matrix format. The sixteen countries are listed on the vertical axis in three geographical groupings (East Asia, South Asia, Latin America). The first data column on the horizontal axis is the inicidence of absolute poverty expressed in percentages followed by a measure of income and respectively twelve and fifteen measures of social development for the years 1985 and 1990. Three social development measures available for 1990 ( malnutrition. per capita health expenditures. and percentage of GDP for healtlh) could not be obtained for 1985 for most counltries. Table I for 1985 and Table 2 for 1')90 on1 the previous pages contain the same information as T'ables A-15 and A-16 but expressed in a different level of measurement. i.e. an ordinal or ranking scale. Two measures have been deleted in T ables I and 2 (secondary female education and under-5 malnutritioni) because they accounted for most of the missinig values. Little information is lost because female education is still captuled in the variable "female primary school enrollmiienit" and under-5 malnutrition is statistically related to the measure of "low birth weight babies". The matrix in Table I therefore has ten social measures instead of twelve and has only five missing rankings (3.1%) out of a maximum 160 rankings. Similarly the Table 2 matrix for 1990 has thirteen social measures instead of fifteen withi only four missing values (1.9%) out of a maximum of 208 rankings.3 The missing rankinigs were assignied the mean value of 8. Table 3 "GNP per Capita" presents an indicator of average economic wealth of each country for the two years of interest. Data are converted to an ordinal scale and countries are ranked from I to 16 assigning I to the highest level of income. Table 4 "GDP Growtlh Rates" contains the actual average annual growtl rates for each country as well as the same data converted to an ordinal scale. Percentages are ranked from I to 16 with I indicating the highest growth rate. 3There are four ways of handling missing values: (a) eliminate the measures with missing data; (b) assign the mean value to the missing value; (c) rerank using the proportional events method; and (d) use alternate methods to estimate the missing value. (Personal communication with Professor V. E. Weckwerth. LJniversitv of Minnesota). To solve the missing value issue here, (d) was tried first but without satisfactory results, then (a) and (b) were used to solve the missing value issue. (c) was also applied and it gave the same results as (b). 14 Statistical Methodology: Use of Nonparametric Statistics The data in Tables I to 4 have been ranked for the sixteen countries from 1 to 16 with I indicating a more positive ranking. There are two reasons for using an ordinal scale and ranking tests: (a) Using ordinal scaling allows the use of nonparametric ranking statistics and the calculationi of correlation coefficients based on rankings. Most of the measurements made by behavioral scientists culminate in ordinal scales. Socioeconomic status. for example, constitutes an ordinal scale. Parametric tests as \\ell as ranikinig tests make the assumption that scores are drawn from an 1idrdCri Vil0 CcOntilnUOUS distributioni. liowever parametric statistical tests, which LISc nmean.s and standard deviations, ought not to be used with data in an ordinal 4 scale1t : andb (b) Regression analysis is a statistical technique preferred by economists because it cani estimate the coefficient of each variable and test the statistical significance of individual coefficients. A sample of only sixteen countries and of fifteen variables limits the use of regression analysis. Increasing the sample size would sharply reduce the number of indicators for lack of comparable data to as few as M\O. i.e., life expectancy at birth (LE) and the infant mortality rate (IMR) which ale estimated for most developing countries. Larger countries usually have more data and by selecting only the largest countries in each region it was possible to increase the numiiber of variables thereby trading off fewer countries against more v ariables. 4 S. Siegel: Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, 1956, p.26. 15 Econometric studies use a larger number of countries but only two or three indicators of health status or social development. The typical measures used are life expectancy at birth and infant mortality rate. At times these two variables are augmented with the under five mortality rate, the perinatal mortality rate or public sector expenditures on health.5 These macro indicators purportedly measure health status of the populations of the countries under review. The statistical methodology is seducingly elegant with precise regression estimates, loglinear specifications to estimate the various effects, simple regression equations and statistically convincing correlation coefficients. These studies usually acknowledge that other explanatory variables (access to health care, access to schools, clean drinking water, nutrition, etc) should be taken into consideration and/or lament the shortcomings of a regression approach and/or recommend a muitisectoral modeling framework. The data analysis in this paper uses a nonparametric measure of correlation to determine whether two sets of ranks are related. The data set consists of sixteen countries and fifteen measures of social development. It uses a statistical test - the Kendall rank correlation coefficient (Kendall's Tau) - to measure the degree of association or correlation between the two sets of ranks and to determine the statistical significance of the observed association. The null hypothesis states that the variables are unrelated in the population. The sampling distribution of Tau is practically indistinguishable from the normal distribution and therefore the significance of Tau can be tested using the normal curve table. 5Anand and Ravallion use a sample of 86 developing countries and life expectancy (LE) at birth and the infant mortality rate (IMR) as health status measures (S. Anand and M. Ravallion: "Human Development in Poor Countries: On the Role of Private Incomes and Public Services"; Journal of Economic Perspectives, 7:1, 1993, pp 133-150); Bidani and Ravallion use 35 developing countries and three aggregate health indicators: LE, IMR and the perinatal mortality rate (B. Bidani and M. Ravallion: "Decomposing Social Indicators Using Distributional Data"; Policy Research Working Paper 1487, World Bank, July 1995); Carrin and Politi use 57 developing countries and three health status variables,viz. LE, IMR and the under-five mortality rate (G. Carrin and C. Politi: "Exploring the Health Impact of Economic Growth, Poverty Reduction and Public Health Expenditure"; "Macroeconomics, Health and Development" Series Technical Paper No 18, WHO, Geneva, March 1996.). 16 Statistical Analysis Null Hypotheses Tested Ho l The incidence of absolute poverty in 1985 is not associated with 1985 measures of social development. Reject (p = .0027) Ho2 The incidence of absolute poverty in 1990 is not associated with 1990 measures of social development. Reject (p = .0207) Ho3 The 1985 ranking on absolute poverty is not associated with the 1990 ranking on absolute poverty. Reject (p = .0038) Ho4 The 1985 ranking on social development indicators is not associated with the 1990 ranking on social development indicators. Reject (p = .00003) Ho5 The incidence of absolute poverty in 1985 is not associated with per capita income in 1985. Reject (p = .0239) Ho6 The incidence of absolute poverty in 1990 is not associated with per capita income in 1990. Accept (p = .2005) Ho7 Social development rankings in 1985 are not associated with per capita income rankings in 1985. Reject (p = .0013) Ho8 Social development rankings in 1990 are not associated with per capita income rankings in 1990. Reject (p = .0008) Correlating Absolute Poverty and Social Development. Each country's ranks on the 1985 ten social development measures and on the 1990 twelve social development measures were summed and the sums of ranks were ordered as shown in the last column of tables 1 and 2. There are two sets of ranks to be tested: (a) a ranking of fifteen6 countries on the incidence of absolute poverty; and (b) a ranking of the presence of a set of social development measures. These two rankings are the first and last columns of Tables 1 and 2. Calculation of the degree of correlation between these two sets of ranks results in a rank correlation coefficient (Tau) of 0.5359 for 1985 and 0.3924 for 1990. The p values are .0027 for the 1985 data and .0207 for 6 Data on absolute poverty for Argentina are not available. 17 1990 data. We therefore reject the null hypothesis of no association at the .05 level and conclude that for both years the two measures are associated but with a stronger association for 1985 than for 1990. Stated in non-statistical language, a country's ranking on the scale of absolute poverty is associated with its ranking on the scale of social development. The higher the incidence of absolute poverty the lower the score on measures of health, nutrition, education, and population growth. Poor health and nutrition, high fertility rates, low school enrollment rates, especially of girls, and low life expectancy are indicators that are correlated with a higher percentage of the total population living in absolute poverty. The statistical association between absolute poverty and social development holds in the aggregate for all countries but some individual countries deviate. For example, in 1990 five of the six countries with the highest incidence of absolute poverty also had the lowest ranking on social development (Bangladesh, India, Nepal, Philippines and Peru). This is in line with the statistical finding. Brazil is the sixth country with a high incidence of absolute poor but for social indicators it ranks in the middle. Pakistan shows the opposite pattem: it is ranked fifth on absolute poverty but 14th on social development. Brazil's GNP per capita is seven times that of Pakistan and in 1990 it spent 6.4 percent of its GDP or US$172 on health; Pakistan on the other hand spent only 3.4 percent of its GDP or US$13 on health. Yet Pakistan has relatively fewer poor than Brazil. Why is the association less robust in 1990 than in 1985? The two additional measures used in 1990 are per capita health care expenditures and percent of GDP spent on health. Eliminating these two measures and using the same ten measures for 1990 as for 1985 results in a much higher correlation coefficient of .5619 with a p value of 0.0018 or as strong an association between poverty ranking and social development in 1990 as in 1985. Correlating 1885 with 1990 Absolute Poverty and 1985 with 1990 Social Development Is the 1985 ranking of the fifteen countries on the incidence of absolute poverty associated with the 1990 ranking of the same variable? And is the 1985 ranking of the sixteen countries on social development associated with the corresponding 1990 ranking? Kendall's rank correlation coefficient Tau was calculated with the following results: (a) the 1985 - 1990 comparison on incidence of absolute poverty yielded a Tau value of 0.493 9 and a p-value of .003 8; and (b) the comparison for the rankings of social development yielded a Tau value of 0.8204 and a p value of .00003. In each case we reject the null hypothesis that the variables are unrelated and conclude that the two variables are associated. The correlation is particularly strong for the social development rankings. 18 A country's ranking on absolute poverty and on social development in 1985 is highly associated with its ranking on the same measures in 1990. No significant change took place in the five year period between 1985 and 1990. This may in part be due to the fact that a five year period is too short a time span to observe differences in macro indicators. Correlating Absolute Poverty and Income Is a country's rank on absolute poverty associated with its ranking on GNP per capita? Table 3 presents income data expressed as GNP per capita for the sixteen countries and their corresponding ranks. Calculating Kendall's rank correlation coefficient Tau yielded a different result for 1985 than for 1990. For 1985 the p value is .0239 and therefore the null hypothesis of no association is rejected (the Tau value is 0.38 1). For 1990 however the p value is only .2005 (the Tau value is 0.1619) causing the acceptance of the null hypothesis of no association indicating that the incidence of absolute poverty in a country in 1990 is not associated with its level of income. The ranking on the incidence of absolute poverty in 1985 is moderately correlated with the 1985 GNP per capita ranking but the 1990 ranking on the incidence of absolute poverty is not correlated with the 1990 GNP per capita ranking. A preliminary explanation - subject to further research and empirical evidence - is that the income distribution worsened between 1985 and 1990. In 1985, the ranking of four countries on absolute poverty was five or more ranks apart from their ranking on income and in 1990 this was the case for seven countries. These four or seven countries had a very different ranking on the poverty indicator than on the income measure. The direction of the difference varies as well: three Latin American countries (Mexico, Venezuela, Brazil) tend to have a high GNP per capita but also a high percentage of poor whereas four Asian countries (China, Pakistan, Sri Lanka, Bangladesh) show the reverse with a lower income level but also a lower incidence of absolute poverty. Brazil and China lie at the extreme ends of this spectrum. Brazil ranks first on income but thirteenth on absolute poverty incidence whereas China ranks thirteenth on income but fourth on absolute poverty incidence suggesting a more egalitarian income distribution in China than in Brazil (25 percent of income or consumption accrues to the top ten percent of China's population whereas the comparable figure for Brazil is 51 percent7). Correlating Social Development and Income Is a country' rank on social development associated with its ranking on GNP per capita? The Tau rank correlation coefficient yields a p value of 0.0013 and of 0.0008 for 1985 and 1990 respectively thus rejecting the null hypothesis of no association. 7World Development Report 1995; Table 30:Income Distribution and PPP Estimates of GNP, pp 220-2 1; The World Bank, Washington D.C. 19 A country's ranking on the scale of social development is associated with its ranking on the scale of GNP per capita. This may occur because the level of expenditures on social services reflects a country's level of income. Latin American countries situate themselves in the higher half of the GNP per capita ranking and also (except Peru) in the higher half for social development indicators. East Asian countries stradle the middle on both measures and South Asian countries fall in the lower half for both measures. Thus the ranking of all countries on the two measures is quite close except for Sri Lanka and Brazil which are the outliers. Sri Lanka ranks high (third place) on social development but only eleventh on income or a difference of eight ranks. Brazil shows the opposite: it has the highest per capita income but falls right in the middle on the scale for social development or a difference of seven ranks. Thus the six times higher income level of Brazil compared to Sri Lanlca does not translate in a better social development ranking. Given an identical distribution of income the country with a lower average income would have a higher incidence of absolute poverty and a lower ranking on social services. This is clearly not the case for Brazil and Sri Lanka suggesting an inequitable income distribution in Brazil (39 percent of income or consumption accrues to the top twenty percent of the Sri Lanka population; the comparable percentage for Brazil is 688). Conversely, the level of social development may also be a reflection of income distribution: higher relative expenditures for social services and relative fewer absolute poor may indicate a higher degree of equity in the distribution of income. Central and Eastern European (CEE) countries provide an interesting contrast. The national wealth and health status of these countries is broadly similar to that of Latin America and the Caribbean9. A recent study examined health status trends in ten CEE countries and documented the following reverse development trends: the wealthier the society, the less healthy its population, particularly for its males. Moreover, it concluded that despite declining infant mortality, life expectancy at birth in the former socialist 10 economies decreases as per capita income rises Interrelating Rankings on Absolute Poverty, Social Development and Income Table 5 on the next page juxtaposes the three 1990 sets of ranks that were analyzed two at a time in the previous sections. Three South Asian countries fall clearly in the bottom third on all three measurements. Interestingly most other countries show 8WDR 1995, op.cit. 9A. S. Preker and R. G. A. Feachem: "Market Mechanisms and the Health Sector in Central and Eastern Europe"; World Bank Technical Paper No 293, World Bank, Washington D.C., December 1995, p.6. 10 0. Adeyi, A. S. Preker, E. Goldstein and G. Chellaraj: "Reverse Development Trends: The Transition in Central and Eastern Europe"; World Bank, Washington D. C., mimeograph, March 1996. 20 different and opposing patterns. Four Latin American countries (Brazil, Mexico, Peru, Venezuela) and one East Asian country (Philippines) have their best rank on income, their second best on social development followed by their position on the absolute poverty scale. For these five countries a higher income does not translate in better outcomes on social services and reducing absolute poverty suggesting that the level of income may be less important than its distribution. Indeed China with a more equitable income distribution presents the opposite picture: it has its worst rank on income, a better rank on social development and a very good rank for number of absolute poor. Analogous to China - but less dramatic in its differences between ranks - are three countries (Sri Lanka, Colombia, Nepal) that have a lower score for income than for social services and for the number of absolute poor. An income distribution analysis - to be treated with caution - of the percentage share of income or consumption that flows to the top twenty percent of the populations of the fifteen countries in our sample (not including Argentina for lack of data) reveals the following pattern: (i) the percentage for five South Asian countries and two East Asian countries (China and Indonesia) ranges from 39 to 42; (ii) for the three other East Asian countries and for four Latin American countries the range is 48 (Philippines) to 56 (Colombia and Mexico); and (iii) the outlier is Brazil with 68 percent ' WDR 1995, op.cit. 21 Table 5: Ranking on Absolute Poverty, Social Development and GNP per Capita, 1990 Country % of Poor Social Dev. GNP per Capita China 4 9 13 Indonesia 8 12 10 Malaysia 1 5 5 Philippines 11 1 0 9 Thailand 2 7 6 Bangladeslh 10 16 15 Inidia 15 13 14 Nepal 14 14.5 16 Pakistan 5 14.5 12 Sri Lanka 6 3 1 1 Brazil 13 8 Mexico 9 6 3 Argenitina N/A 1 4 Colombia 3 2 7 Peru 12 1 1 8 Venezuela 7 4 2 Discussioin A country's ranking on a scale measuring the incidence of absolute poverty is statistically correlated with its ranking on a set of social indicators measuring social development but less so with its ranking on income. Thus, a country with a high number of absolute poor usually has social services that are less well developed but a lower level of income does not necessarily mean that a country has a higher incidence of absolute poverty. In 1990 Brazil ranks highest in terms of GNP per capity (US$2,680) but over one third of its population was classified as absolute poor. A higher level of income does not necessarily translate into fewer poor or a better availability of social services. Would consistent economic GDP growth - as shown in Table 4 - eventually reduce the absolute poverty level as well as increase the availability of health care, nutrition, family planning and education? In the long run this may very likely be the case as shown by Thailand, Malaysia and Colombia which have a core of absolute poor of less than ten percent and relatively well developed social services. Indeed "no country has been successful in 22 continually improving its social indicators, feeding its people, and providing health care and education for its children, without sustained economic growth. Equally, we know that the countries which have pursued the most successful economic policies have also ,,2 invested heavily in their people' Absolute poverty measures typically respond quite elastically to aggregate economic growth but "it would not be correct to say that growth always benefits the poor, or that none of the poor lose from pro-growth policy reforms ... The gains to poor people from such a growth process will tend to be lower the higher the extent of initial inequality"'3. A rising tide lifting all the boats has drowned many people. Targeted social interventiotns may achieve quicker results in the short run in alleviating poverty. Targeting metliods include geographic targeting (poverty regions), programmatic targeting (e.g. tuberculosis control, food supplements) or economic targeting (e.g. income transfers). Direct compensatory social development interventions may benefit the poor more directly and more quickly than longer term non-poverty focused aggregate economic growth especially in countries with large income distribution inequalities. "Countries which give priority to basic human capabilities in schooling, health and nutrition not only directly enhance well-being, but are also more likely to see improving income distributions and higher average incomes over the longer term"'4. Of the three countries in the study sample that rank lowest for female primary education in 1990 (India, Nepal, Pakistan), two also have the largest incidence of absolute poverty (India, Nepal). The single most effective policy action that can be taken in Sub-Saharan Africa today to reduce fertility and child mortality is to raise female education. Each additional year of female schooling results in a 0.6 to 0.9 percent reduction in child mortality during the first two years of life'5. A statistical analysis of economic and social variables and their interdependence can only partially explain the differences among countries in levels of absolute poverty. Other forces must be at work that cannot be captured in a classical quantitative socioeconomic analysis. The next section turns to other potential influences. These armchair musings do not lead to any conclusions. Their purpose is to broaden and elucidate our thinking and understanding of the observed inter-country differences. 12 C. Koch-Weser: "Children in Development: The Key Challenges Ahead", The Carter Center, Atlanta, Georgia; April 9, 1996. '1 M. Bruno, M. Ravallion and L. Squire: "Equity and Growth in Developing Countries"; Policy Research Working Paper 1563, World Bank, January 1996, pp. 21-22. ibid, p. 23. 15 M. Ainsworth: The Socio-Economic Determinants of Fertility in Sub-Saharan Africa. The World Bank. 1994. 23 V. Other Determinants of Absolute Poverty Are there other - non-quantitative factors - that may help explain the large differences in the level of absolute poverty observed among the sixteen countries? Are there common elements that would allow a differential diagnosis and consequently permit a prescriptive set of recommendations to alleviate the observed levels of poverty? There is no longer anymore a "quick fix" solution. Pasteur provided an example of a linear relationship when he demonstrated convincingly the correlation between bacteria and disease. His germ theory of disease specified that specific diseases are caused by specific microbes. Unfortunately no specific etiology exists that sets forth a set of criteria that prove conclusively that absolute poverty is caused by a specific "disease". Causes of absolute poverty are deeply rooted in historical, societal, religious, economic, environmental and cultural contexts. A better insight may be obtained by a holistic approach which analyzes absolute poverty in terms of interrelationships among exogenous determinants of poverty. Five building blocks are posited and examined. Material Environment Countries differ widely in their geographical make-up and in population size. Geographic location determines the degree of abundance or scarcity of natural resources and of raw materials, and contributes to a country's burden of disease. Some raw materials are more valuable on the world market than others, e.g. oil versus bananas. Population density and population distribution put pressure on land. Arable land is a very small percentage of total land mass in China and Egypt. Regions with the highest percentages of absolute poor are usually located in difficult material environments. Competition for land is intense especially in high birth rate countries and is often exacerbated by a continual subdividing of plots for inheritance reasons. A relationship can be hypothesized between geographic location, population density and distribution and the incidence of absolute poverty. Material environment variables are measurable and should be included in any future research on poverty. Religious/Spiritual Environment Confucianism was a philosophy of social organization and provided Chinese society with strict conventions of social etiquette. It reinforced China's imperial tradition of respect for authority and hierarchy. Buddhism - the dominant spiritual tradition in most parts of Asia - is psychological and a doctrine of psychotherapy. Buddha was concerned with the sufferings and frustrations of human beings. Hinduism is mythological and ritualistic with numerous sects and cults and the worship of countless gods and goddesses. The spiritual tradition of Hinduism mirrors the geographical, racial, 24 linguistic and cultural complexities of India'6. To what extent have these religions influenced or determined political, social and cultural life in Asia? To what extent can they help explain the large variations in absolute poverty among the countries in East and South Asia? The Judeo-Christian and Islamic tradition adheres to the belief in a male God as the source of ultimate power ruling the world from above by imposing his divine law on it. It has encouraged the view of man as dominating nature and woman and the world. It has supported the scientific method and its rational, analytic thinking. It underpinned an expansionist behavior to discover, conquer, invade and impose its culture and religion. The Arab-Muslim empire spread around the Mediterranean and into the Iberian peninsula between the eighth and the thirteenth centuries fueled by the religious duty to spread Islam. The Spanish empire in the sixteenth and seventeenth centuries conquered most of the Americas and the Philippines imposing its language and religion. The British empire expanded in the eighteenth and nineteenth centuries and at its apogee in 1897 contained 372 million people in eleven million square miles which was 91 times the area of Great Britain.'7 In contrast China in the fifteenth century with a navy unparalleled in numbers, in skills and in technology did not conquer, crusade or establish trading posts. Before the seafaring voyages of Henry the navigator, Vasco da Gama, Columbus, Magellan or Vespucci, the Chinese had already sailed down the East Coast of Africa. Yet after these grand expeditions China withdrew into its own borders'8. Is current Western material wealth and technological and cultural world dominance a product of earlier expansionist behavior supported by its religions? Political Environment Are democracy and democratic pluralism necessary and/or sufficient conditions to reduce the level of absolute poverty? The concept of democracy is often equated with its least common denominator, i.e. holding free elections in a multiparty system. Yet holding elections is only one aspect of a democratic environment. More important aspects for the people include freedom of speech, of association, of religious preferences, of ownership and protection of personal property. These are freedoms usually associated with a democratic way of life but are not guaranteed to result from multiparty elections. In practice, democracy becomes a multifaceted concept interpreted and adapted by countries and by the ruling elites to accommodate internal and external pressures. Examples of managed democracy abound: democracy with one strong party (Mexico), democracy with one strong party and one strong leader (Malaysia, Indonesia, Philippines in the 1960s to 1980s), democracy with implicit support from the military (Thailand, South Korea, Philippines, Peru, Indonesia), and democratic pluralism (Colombia, Brazil, India, Pakistan, Bangladesh, Sri Lanka, Argentina since 1983). South Asian democracies 16F. Capra: The Tao of Physics. Bantam Books. 1984. 17J. Morris: Pax Brittanica. Harcourt, Brace and World Inc, NY. 1968, page 42. ] Daniel J. Boorstin: The Discoverers. Vintage Books. 1983, pp 186-94. 25 have been troubled by political assassinations of their democratically elected top leaders. Non-democratic authoritarian and military regimes include China (with a deified strong leader from 1949 to 1976), South Korea in the 60s to the 80s, Brazil in the 1 960s and 70s, Argentina in the 1970s, Peru in the 1970s and 80s, and Chile in the 1970s and 1980s. Do centrally planned economies need a strong hand to guide them towards market reforms? Would democracy unleash powerful centrifugal forces as witnessed in the implosion of the former USSR? Do Eastern Europe and Central Asian paradigmatic changes increase or decrease the number of poor in the short run or the long term? There is no demonstrable direct correlation between social and economic development and democracy and there is no conclusive evidence that a particular form of government or political system is any more likely to promote sustained economic growth and to reduce poverty. Rapid economic growth can take place under authoritarian governments, witness South Korea in the 60s and 70s, Singapore, Taiwan, China, and Indonesia, as well as under democratic governments, witness Japan, India, Pakistan, Malaysia and more recently the Philippines. Economic stagnation can occur under authoritarian governments, witness Vietnam, Myanmar and Cambodia, as well as under democratic governments, witness present day Sri Lanka. The evidence is not overwhelming that democracy and democratic pluralism is necessary for achieving social and economic development. There are many examples of dictators ruining economies and exploiting their people but there are also examples of corrupt and inept democratic governments. Democracies at times have an unfortunate tendency to produce crude populism inimical to good governance and to avoid problems instead of focusing on achieving long term goals. Freedom of association and participation is part of democracy and institutional pluralism. It empowers people and groups. However it is not always conducive to political stability particularly in countries where people tend to organize themselves along ethnic and tribal lines. Respect of human rights is closely associated with democracy and freedom of association and participation. In many cultures it is a non negotiable right and the cornerstone of public policy. Yet one must be willing to understand - not necessarily accept - that in some Asian countries the Confucian influence pervades policy decisions, and that the Confucian system was mainly concerned with maintaining social order even at the expense of the individual rights of the members of the group. Could there be a relationship between incidence of absolute poverty and the rights of the group and of the individual? Culture Anthropologists have documented a wide variety of traditions, ritualistic practices, mores, beliefs and value systems among the many peoples of the world. Some codified patterns of behavior have been shown to be inimical to health - even destructive of health - and to obstruct formal educational efforts. Cultural taboos and customs relating to, inter alia, child bearing, breast feeding, child nutrition, food distribution 26 within a family, girls education exacerbate the disease burden and affect the productivity of a population and the level of absolute poverty. Some codified patterns of behavior, on the other hand, promote health, e.g. post-partum abstinence and long-breast feeding practices, and polyandry practiced in resource poor settings is argued to be an adaptive mechanism to keep fertility to sustainable levels. Social Organization Two aspects of social organization may affect levels of poverty and of social and economic development: how society is organized at the macro and the micro level and how society is governed. Family kinships, tribal or ethnic ties, and how they affect cooperative behavior and solidarity have been studied extensively by sociologists and anthropologists. Economists - and development specialists in general - have been paying more attention lately to how a country is governed or the concept of governance. Governance is "the manner in which power is exercised in the management of a country's economic and social resources for development"'9. Three aspects of governance are identified: (i) the form of political regime; (ii) the process by which authority is exercised in the management of a country's economic and social resources for development; and (iii) the capacity of govemments through their public sector administration to design, formulate, and implement policies and discharge their functions. Good governance involves holding competent public officials accountable for their actions and creating an environment that enables policy dialogue and the formulation of policy. Good governance is good order, that is, the existence of a set of rules, known in advance, actually in force and applied (the notions of predictability and transparency), and the existence of legitimized institutions that implement, enforce, and amend the rules according to known procedures (the notion of a transparent judicial system and a legal framework). Applying these criteria of good governance, would a relationship exist between well governed countries or societies and the degree of absolute poverty and their level of social development? 9 World Bank. Governance and Development. Washington D.C., 1992, p. 52. 27 VI. Conclusion The question was raised at the outset why the incidence of absolute poverty is so different among developing countries in East Asia, South Asia and Latin America and then the relationships between absolute poverty, social development and income were examined. Social development and absolute poverty are affected by a number of likely factors: economic growth, level of income, access to the multiple components of social services (health, education, family planning, nutrition, water and sanitation), social integration, environment protection, rural development, the maturity of institutions and the regulatory, political and legal environments. The relative importance of these factors and their interaction, however, are not known with any degree of precision and are arguably difficult to predict. Income is widely acknowledged to be good predictor of absolute poverty rates but at the same time leading thinkers believe that economic growth is not all that matters in reducing poverty. At lower income levels countries would be expected to have more poverty than at higher income levels. However of critical importance is the distribution of this income among the population. Over one third of Brazil's population was absolute poor in 1990 at a time when it had the highest level of income among the sixteen largest developing countries. Income distribution may also affect the level of development of social services and of social development in its broader sense. In 1990 China was ranked quite low (thirteenth) for income level but was ranked quite high (third) on the relative number of absolute poor and found itself in the middle (ninth) on the measure of social development suggesting that - in terms of absolute poverty and social development - a lower but more egalitarian income distribution may compensate for a lower level of income. Conversely, the provision of social services may in itself be a form of income distribution. Targeted social interventions reduce absolute poverty more quickly than long-term economic growth. The impact of government policies, however, may show a great deal of variation as decision making and implementing strategies are modified by factors operating in different socio-politico-cultural-religious contexts and in resource-rich or resource-poor environments. Human behavior also varies widely in response to the same stimuli because external stimuli interact with individual and ideosyncratic organisms causing different behaviors. Similarly levels of absolute poverty may vary widely among and within countries and respond differentially to apparently similar stimuli. A deeper understanding of the non measurable variables of culture, history, and religion may lead to developing or adjusting existing policies more appropriately and to target strategic interventions that may reduce absolute poverty more effectively. 28 Annex I Absolute Poverty in East Asia, South Asia and Latin America 29 Table Al Number and Percentage of Population in Asia with per capita income less than $1/day (1) Nbr of Poor (millions) % of Poor Total Pop Country (millions) --__1 1985 1990 1985 1990 1990 China 115.8 153.0 11.1 13.5 1133.7 Indonesia 63.7 39.0 39.3 21.7 178.2 Malaysia 2.0 0.8 12.7 4.3 17.9 Philippines 19.0 17.0 34.8 28.4 61.5 Thailand 8.1 2.7 15.6 4.8 55.8 Five Countries 208.6 212.5 15.7 14.6 1447.1 (92%) East Asia 247.6 232.0 15.7 14.7 1577.0 Bangladesh 16.9 29.7 16.8 27.8 106.7 India 558.5 590.0 73.0 69.4 849.5 Nepal 7.7 7.6 46.5 39.8 18.9 Pakistan 15.1 16.6 15.7 14.8 112.4 Sri Lanka 4.8 3.5 30.5 20.5 17.0 Five countrics 603.0 647.4 60.6 58.5 1104.5 (96%) .South Asia 698.0 673.0 60.8 58.6 1148.0 Brazil 36.2 53.1 26.7 35.3 150.4 Mexico 17.4 19.5 22.1 22.6 86.2 Argentina NA NA NA NA 32.3 Colombia 3.2 2.2 11.1 6.8 32.3 Peru 2.8 6.7 15.2 31.0 21.7 Venezuela 1.3 4.1 7.3 20.6 19.7 Six countries 60.9 85.6 NA NA 342.6 (79%) Latin Amec 100.0 120.4 23.1 27.8 433.0 NA: Not available (1) Income is calculated on the basis of 1985 purchasing power parity (PPP) levels. Sources: Shaohua Chen, Gaurav Datt and Martin Ravallion: "Statistical Addendum to 'Is Poverty Increasing in the Developing World?' " ; Policy Research Department, World Bank; and World Development Reports 1987 and 1992 for 1985 and 1990 population data. 30 Table A2 GDP Growth rates, 1980 - 1990 and GNP Per Capita, 1985 and 1990. Country GDP Growth Rates GNP Per Capita 1980- 1985 1985-1990 1985 1990 % % USD USD China 9.8 7.7 310 370 Indonesia 5.0 5.9 530 570 Malaysia 5.4 6.9 2,000 2,320 Philippines -1.1 4.8 580 730 Thailand 5.6 10.1 800 1,420 Bangladesh 4.9 3.7 150 210 India 5.2 6.2 270 350 Nepal 4.2 5.1 160 170 Pakistan 6.4 5.8 380 380 Sri Lanka 4.7 2.8 380 470 Brazil 0.9 1.9 1,640 2,680 Mexico 1.2 1.6 2,080 2,490 Argentina -1.2 -0.3 2,130 2,370 Colombia 2.2 4.4 1,320 1,260 Peru -0.9 -2.7 1,010 1,160 Venezuela -1.2 2.0 3,080 2,560 Source: Trends in Developing Economies 1993. The World Bank, Washington D.C., 1993 for GDP Growth Rates; and World Development Rcports 1987 and 1992 for GNP Per Capita 1985 and 1990 data. 31 Table A3 Life Expectancy at Birth*, 1985 and 1990. Life expectancy at birth (years) Country Total Female Male 1985 1990 1985 1990 1985 1990 China 69 70 70 71 68 69 Indonesia 55 62 57 64 53 60 Malaysia 68 70 70 72 66 68 Philippines 63 64 65 66 61 62 Thailand 64 66 66 68 62 63 Bangladesh 51 52 51 51 50 52 India 56 59 56 58 57 60 Nepal 47 52 46 51 47 53 Pakistan 51 56 50 55 52 56 SriLanka 70 71 72 73 68 69 Brazil 65 66 67 69 62 63 Mexico 67 70 69 73 64 66 Argentina 70 71 74 75 67 68 Colombia 65 69 67 72 63 66 Peru 59 63 60 65 57 61 Venezuela 70 70 73 73 66 67 *Life expectancy at birth is the number of years a newborn infant would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life. Sources: World Development Reports 1987 and 1992 for 1985 and 1990 life expectancy data 32 Table A4 Infant Mortality Rate (IMIR)*, 1985 and 1990. Country 1985 1990 China 35 35 Indonesia 70 71 Malaysia 28 (1) 16 (1) Philippines 50 48 Thailand 37 30 Bangladesh 116 96 India 98 91 Nepal 107 95 Pakistan 99 98 Sri Lanka 25 18 Brazil 63 56 Mexico 44 31 Argentina 28 23 Colombia 27 19 Peru 72 52 Venezuela 26 22 *Infant mortality rate is the number of infants who die before reaching one year of age, per thousand live births in a given year. (1) Source: World Development Reports 1987 and 1992 for 1985 and 1990 data. Source: Kenneth Hill and Ageo Yazbeck: "Trends in Child Mortality, 1960-90: Estimates for 84 Developing Countries"; World Development Report 1993: Investing in Health, Background Paper Number 6. 33 Table A5 Under-5 Mortality Rate*, 1985 and 1990, and Under-5 Mortality Rate by Sex**, 1990. Country U5MR U5MR 1990 1985 1990 Female Male China 43 43 29 40 Indonesia 110 111 75 40 Malaysia 22 (1) 17 (2) 17 22 Philippines 66 62 45 57 Thailand 45 36 28 38 Bangladesh 178 137 160 142 India 142 127 121 116 Nepal 154 135 183 175 Pakistan 144 141 151 145 Sri Lanka 32 22 21 26 Brazil 79 69 62 75 Mexico 56 38 41 51 Argentina 31 26 30 40 Colombia 35 23 40 49 Peru 101 73 78 93 Venezuela 30 26 36 45 *Under-5 mortality rate is calculated by: Number of deaths under 5 years of age in the given y x 1000 Total number of children under 5 years of age at the middle of the year. ** "Under good nutritional and health conditions and in times of peace, male children under 5 have a higher death rate than females." "In industrial market economies, female babies have a 23 percent lower risk of dying by age 5 than male babies; the risk of dying by age 5 is actually higher for females than for males in some lower-income economies." (World Development Report 1992, Technical notes). (1) Estimate for the period 1985-89. (2) Estimate for the period 1990-94. Sources: Under-5 Mortality Rate, 1985 - 1990 is from Kenneth Hill and Ageo Yazbeck: "Trends in Child Mortality. 1960-90: Estimates for 84 Developing Countries"; World Development Report 1993: Investing in Health, Background Paper Number 6. Under-5 Mortality Rate by Sex, 1990 data is from World Development Report 1992. 34 Table A6 Maternal Mortality*, 1980 and 1988. Country 1980 1988 China 44 115 Indonesia 800 450 Malaysia 59 26 Philippines 80 74 Thailand 270 37 Bangladesh 600 600 India 500 437(1) Nepal (2) 850 515 Pakistan NA 833 Sri Lanka 119 80 Brazil 150 140 Mexico 92 200 Argentina 85 140 Colombia 130 200 Peru 310 165 Venezuela 65 55 NA: Not available *Maternal mortality refers to the number of deaths that occur during childbirth per 100,000 live births. (1) The data source for India is the National Family Health Survey for 1992-93. (2) Nepal data is from the National Fertility and Health Survey report. Sources: World Development Reports 1992 and 1993 for 1980 and 1988 data. 35 Table A7 Annual Incidence Rate of Tuberculosis (per 100,000 population), 1990. Country 1990 China 166 Indonesia 220 Malaysia 67 Philippines 280 Thailand 173 Bangladesh 220 India 220 Nepal 167 Pakistan 150 Sri Lanka 167 Brazil 56 Mexico 110 Argentina 50 Colombia 67 Peru 250 Venezuela 44 Source: World Development Report 1993 based on WHO's Tuberculosis Programme. 36 Table A8 Total Fertility Rate*, 1985 and 1990. Country 1985 1990 China 2.3 2.5 Indonesia 4.1 3.1 Malaysia 3.7 3.8 Philippines 4.3 3.7 Thailand 3.2 2.5 Bangladesh 5.7 4.6 India 4.5 4.0 Nepal 6.3 5.7 Pakistan 6.1 5.8 Sri Lanka 3.2 2.4 Brazil 3.6 3.2 Mexico 4.3 3.3 Argentina 3.3 2.8 Columbia 3.3 2.7 Peru 4.3 3.8 Venezuela 3.9 3.6 *Total fertility rate represents the number of children that would be born to a woman if she were to live to the end of her childbearing years and bear children at each age in accordance with prevailing age-specific fertility rates. Sources: World Development Reports 1987 and 1992 for 1985 and 1990 data as complied from the UN Population Division, The UN Statistical Office, and country statistical offices. 37 Table A9 Nutrition: Babies with low birth weight*, 1985 and 1990, and Prevalence of under 5 malnutrition* 1990. Babies with low birth weight (%) Prevalence of under 5 Country _ malnutrition (%) 1985 1990 1990 China 9 NA Indonesia 14 14 14 Malaysia 9 I ) 24 Philippines I8 1 5 19 Thailand 12 i. 26 Bangladesh 1 50 60 India 30 33I NA Nepal N A NA NA Pakistan 2 - 7 Sri Lanka 28 2545 Brazil 8 11 i Mexico 1 12 14 Argentina 6 8 N A Colombia 1 10 12 Peru 9 11 3 Venezuela 9 9 5 NA: Not available *Babies with low birth weight are childrcn borii weighling less th:in 2'50) grains. **Child malnutrition measures the percentage of childreni uinder fiN c %% iti a dcliciCIc\ or Iii c\ccss ol- nutrients that interfere vith their hcalthaniliid genIetic polentiil for grow tii. Mtds ol ;ISSr ' a . t the most commonly used arc the flollowing: less than Xtt percclut of the slanidard x\ cI lit for agce: Iess tdhanll minus two standard deviation fromii the 50th percentile of lhe wei-ht lor agc rel'.rcnce popld:tioin. anid the Gomez scale of malnutrilioni. Sources: World Developienlit Reports (WDR) I 989 wuid 1994 for 1985 anid I "B() tabies \ ith low hiilh weight" data: and WDR MI93 for 1990 'Prevalence of malnutrition (under 5)- data. 38 Table AIO Adult Illiteracy*, 1985 and 1990. Country Total Female 1985 1990 1985 1990 China 31 27 45 38 Indonesia 26 23 35 32 Malaysia 27 22 34 30 Philippines 14 10 15 11 Thailand 9 7 12 10 Bangladesh 67 65 78 78 India 57 52 71 66 Nepal 74 74 88 87 Pakistan 70 65 81 79 Sri Lanka 13 12 17 17 Brazil 22 19 24 20 Mexico 10 13 12 15 Argentina 5 5 5 5 Colombia 12 13 13 14 Peru 15 15 22 21 Venezuela 13 12 15 10 *Adult illiteracy is the proportion of the population over the age of fifteen who cannot, with understanding, read and write a short, simple statement on their everday life. Sources: World Development Reports 1987 and 1992 for 1985 and 1990 data. 39 Table All Education: School Enrollment*, 1985 and 1990. Country Primary Secondary 1985 1990 1985 1990 China 124 135 39 48 Indonesia 118 117 39 45 Malaysia 99 93 53 56 Philippines 106 111 65 73 Thailand 97 85 30 32 Bangladesh 60 73 18 17 India 92 97 35 44 Nepal 79 86 25 30 Pakistan 47 37 17 22 Sri Lanka 103 107 63 74 Brazil * 104 108 16 39 Mexico 115 112 55 53 Argentina 108 111 70 NA Colombia 117 110 50 52 Peru 122 126 65 70 Venezuela 108 92 45 35 NA: Not available *School enrollment data are estimates of children of all ages erlled in school. Figures are expressed as the ratio of pupils to the population of school-age children. Many countries consider primary school age to be 6 to II years and secondary school age to be 12 to 17 years. Sources: World Development Reports 1987 and 1992 for 1985 and 1990 data. 40 Table A12 School Enrollment: Females per 100 Males, 1985 and 1990. Country Primary Secondary 1985 1990 1985 199O China 81 86 67 72 Indonesia 92 93 73 82 Malaysia 94 95 98 104 Philippines 96 94 100 NA Thailand NA 95 NA 97 Bangladesh 67 81 38 49 India 67 71 52 55 Nepal 41 47 30 NA Pakistan 47 52 34 41 Sri Lanka 93 93 108 105 Brazil * Mexico 95 94 86 92 Argentina NA 103 NA NA Colombia 100 98 100 100 Peru 93 NA 88 NA Venezuela 96 99 119 137 NA: Not available *Brazil does not collect education statistics by gender. Sources: World Development Reports 1987 and 1992 for 1985 and 1990 data. 41 Table A13 Health Expenditures*, 1990. GNP per Total health Health expenditures as a percentage of GDP Country capital I expenditure per capita (US$)WS)_ Total Public Sector Private Sector China 370 11 3.5 2.1 1.4 Indonesia 570 12 2.0 0.7 1.3 Malaysia 2320 67 3.0 1.3 1.7 Philippines 730 14 2.0 1.0 1.0 Thailand 1420 73 5.0 1.1 3.9 Bangladesh 210 7 3.2 1.4 1.8 India 350 21 6.0 1.3 4.7 Nepal 170 7 4.5 2.2 2.3 Pakistan 380 12 3.4 1.8 1.6 SriLanka 470 18 3.7 1.8 1.9 Brazil 2680 222 6.4 2.8 3.6 Mexico 2490 112 3.9 1.6 2.4 Argentina 2370 312 9.6 5.9 3.7 Colombia 1260 65 5.1 3.0 2.2 Peru 1160 59 3.1 1.1 2.0 Venezuela 2560 102 4.2 2.0 2.2 *Health expenditures include outlays for prevention, promotion, rehabilitation, and care; population activities; nutrition activities; program food aid; and emergency aid specifically for health. It does not include water and sanitation. **Gross National Product (GNP) = Gross Domestic Product (GDP) + net factor income from abroad. Note: GNP per capita and Total health expenditureper capita are expressed in official exchange rate U.S. dollars. ces: World Development Report (WDR) 1992 for GNP per capita data; WDR 1993 for health expenditures data of Asian countries; and Haeduck Lee and Jose-Luis Bobadilla: "Health statistics for the Americas"; World Bank, Nov. 1994, for updated health expenditures data of Latin American countries. 42 Table A14 Latin America - Government Health Expenditures as % of GDP, 1980-1990. Country 1980 1983 1987 1990 United States 3.9 4.4 4.7 5.2 Argentina 1.5 1.5 1.4 2.5 Brazil 1.2 1.7 1.3 1.6 Colombia 2.1 2.2 2.0 2.9 Mexico 2.1 2.1 2.3 3.1 Peru 2.9 NA 2.9 1.1 Venezuela 2.3 2.0 1.6 1.2 NA: Not available Sorce: Ramesh Govindaraj, Christopher J.L. Murray, and Gnanaraj Chellaraj: "Health Expenditures in Latin America"; World Bank, November, 1994. 43 Table Al 5: Social Indicators for the Largest Countries in East Asia, South Asia and Latin America. Summary Table, 1985 (1) Country % of Life IMR U5MR MMR* TFR LBW Illiter Prim Sec PriFem Poor Exp (%) Educ Educ Educ China 11.1 69 35 43 44 2.3 6 31 124 39 81 Indonesia 39.3 55 70 110 800 4.1 14 26 118 39 92 Malaysia 12.7 68 28 22 59 3.7 9 27 99 53 94 Philippines 34.8 63 50 66 80 4.3 18 14 106 65 96 Thailand 15.6 64 37 45 270 3.2 12 9 97 30 NA Bangladesh 16.8 51 116 178 600 5.7 31 67 60 18 67 India 73 56 98 142 500 4.5 30 57 92 35 67 Nepal 46.5 47 107 154 850 6.3 NA 74 79 25 41 Pakistan 15.7 51 99 144 NA 6.1 25 70 47 17 47 SriLanka 30.5 70 25 32 119 3.2 28 13 103 63 93 Brazil 26.7 65 63 79 150 3.6 8 22 104 16 NA Mexico 22.1 67 44 56 92 4.3 15 10 115 55 95 Argentina NA 70 28 31 85 3.3 6 5 108 70 NA Colombia 11.1 65 27 35 130 3.3 15 12 117 50 100 Peru 15.2 59 72 101 310 4.3 9 15 122 65 93 Venezuela 7.3 70 26 30 65 3.9 9 13 108 45 96 *1980 (1) See Tables Al - A13 for detailed information. Table A16: Social Indicators for the Largest Countries in East Asia, South Asia and Latin America. Summary Table, 1990 (1) Country % of Life IMR U5MR MMR* TB TFR LBW Iliter Prim Sec PriFem per cap % GDP Poor Exp (%) Educ Educ Educ H Exp Health China 13.5 70 35 43 115 166 2.5 9 27 135 48 86 11 3.5 Indonesia 21.7 62 71 111 450 220 3.1 14 23 117 45 93 12 2 Malaysia 4.3 70 16 17 26 67 3.8 10 22 93 56 95 67 3 Philippines 28.4 64 48 62 74 280 3.7 15 10 111 73 94 14 2 Thailand 4.8 66 30 36 37 173 2.5 13 7 85 32 95 73 5 Bangladesh 27.8 52 96 137 600 220 4.6 50 65 73 17 81 7 3.2 a India 69.4 59 91 127 437 220 4 33 52 97 44 71 21 6 Nepal 39.8 52 95 135 515 167 5.7 NA 74 86 30 47 7 4.5 Pakistan 14.8 56 98 141 833 150 5.8 25 65 37 22 52 12 3.4 Sri Lanka 20.5 71 18 22 80 167 2.4 25 12 107 74 93 18 3.7 Brazil 35.3 66 56 69 140 56 3.2 11 19 108 39 NA 222 6.4 Mexico 22.6 70 31 38 200 110 3.3 12 13 112 53 94 112 3.9 Argentina NA 71 23 26 140 50 2.8 8 5 111 NA 103 312 9.6 Colombia 6.8 69 19 23 200 67 2.7 10 13 110 52 98 65 5.1 Peru 31 63 52 73 165 250 3.8 11 15 126 70 NA 59 3.1 Venezuela 20.6 70 22 26 55 44 3.6 9 12 92 35 99 102 4.2 *1988 (1) See Tables Al - A13 for detailed information. DistWibutors ot World CZECH REPUBUC IRAEL PAKRSTAN SPAIN Bank Publications National Information Yozmot Literature Ltd. MirzaBookAgency Mundi-Prensafibros, S.A. Pr.sandaud#s=vary Center P.O. Box 56055 65, Shahrah-e-Quaid-e- Castello 37 timantylocormny. prodejna, Konviktska 5 Tel Aviv 61560 Azam 28001 Madrid Consuyour /rcaIdWor CS -11357 Prague 1 R.O.Ylntemational RO. Box No. 729 Mundi-Presa Barcelona t&vac0anoraer. P.O. Box 13056 Lahore 54000 Conseil de Cent, 391 DENMARK TelAviv 61130 08009 Barcelona ALBANIA SamfundsLitteratur Oxford University Press Adrion Ltd. Rosenoems All 11 PalesfinianAuthority/ 5 Bangalore Town SRIILANKA, Pedal Rexhepi Str. DK-1970 Fredenksberg C Middle East Sharae Faisal THE MALDIN Pall. 9, Shk. 1,Ap. 4 Index Information Services RO. Box 13033 Lake House Bookshop Tirana EGYPT,ARABREPUBLIC PO.B. 19502 Jerusalem Karachi-75350 P.O. Box 244 OF 100, Sir ChittampalamA. ARGENTINA AlAhram fTALY PERU Gardiner Mawatha Oficina del Libro Al Galaa Street Licosa Cornmissionaria Editorial Desarrolbo SA Colombo 2 Intemacional Cairo Sansoni SPA Apartado 3824 Av. Cordoba 1877 Via Duca Di Calabria, 1/1 Lima 1 SWEDEN 1120 Buenos Aires The Middle East Observer Casella Postale 552 Fritzes Customer Service 41, Sheri Street, Cairo 50125 Firenze PHFLPPES Regeringsgaton 12 AUSTRALA, FtJI, PAPUA Intemational Booksource S-106 47 NEWGUtNA,SOLOMON FINLAND JAMAICA Center, Inc. Stockholm ISLANDS,VANUATU,AND Akateeminen Kirjakauppa Ian Randle Publishers Ltd. Suite 720, Cityland 10 WESTERNSAMOA P.O. Box 23 206 Old Hope Road, Condominium Tower 2 Wennergren-Williams AB D.A. Inormiation Services FIN-00371 Helsinki Kingston 6 H.V dela Costa, corner P. O. Box 1305 648 Whitehorse Road Valero St., Makati, Metro S-171 25 Mitcham 3132, Victoria FRANCE JAPAN Manila Solna World Bank Publicabons Eastem Book Service AUSTRIA 66, avenue d'l6na Hongo 3-Chome, POLAND SWItZERLAND Gerold and Co. 75116 Paris Bunkyo-ku 113 Intemational Publishing Librairie Payot Graben 31 Tokyo Service Service Instutionnel A-1011 Wien GERMANY Ul. Piekna 31137 C6tes-de-Montbenon 30 UNO-Veriag KENYA 00-577 Warzawa 1002 Lausanne BANGLADESH Poppelsdorier Allee 55 Africa Book Service (E.A.) Micro Industries 53115 Bonn Ltd. PORTUGAL Van Diermen Editions DevelopmentAssistance Quaran House Livraria Portugal Techniques Society (MIDAS) GREECE Mtangano Street Rua Do Carmo 70-74 Ch. de Lacuez 41 House 5, Road 16 Papasotiriou S.A. P.O. Box 45245, Nairobi 1200 Lisbon CH1807 Blonay Ohanmondi R/Area 35, Stournara Str., 106 82 Dhaka 1209 Athens KOREA, REPUBLIC OF ROMANtA TANZANIA Daejon Trading Co. Ltd. Compani De Librarii Oxford University Press BELGIUM HONG KONG, MACAO P.O. Box 34 Bucuresti S.A. Maktaba Street Jean De Lannoy Asia 2000 Ltd. Yeoeida Str. Lipscani no. 26 P.O. Box 5299 Av. du Roi 202 Sales & Circulation Seoul sector 3 Dar es Salaam 1060 Brussels Department Bucharest Seabird House, MALAYSIA THAILAND BRAZIL unit 1101-02 University of Malaya RUSSIANFEDERATION Central Books Distrbution Publicacoes Tecnicas 22-28 Wyndham Street, Cooperative Bookshop, lsdatelstvo 306 Silom Road Intemacionais Ltda. Central, Hong Kong Limited, P.O. Box 1127 9a, Kolpachniy Pereulok Bangkok Rua Peixoto Gomide, 209 Jalan Pantai Baru Moscow 101831 01409 So Paulo, SP HUNGARY 59700 Kuala Lumpur TRINIDAD &TOBAGO, Foundation for Market SAUDI ARABIA, QATAR JAMAICA CANADA Economy MEXICO Jarir Book Store Systermtics Studies Unit Renouf Publishing Co. Dombovari Ut 17-19 INFOTEC P.O. Box 3196 #9 Waets Street Ltd. H-1117 Budapest Apartado Postal 22-860 Riyadh 11471 Curepe 1294 Algoma Road 14060 Tlalpan Trinidad Ottawa, Ontario K1B 3W8 INDIA Mexico D.F. SINGAPORE, TAIWAN, West Indies Allied Publishers Ltd. MYANMAR,BRUNEI CHINA 751 Mount R NETHERLANDS Asahgate Publishing Asia UGANDA China Financial & Madras -600002 De Lindeboomn/nOr- Pacific Pte. Ltd. GustroLtd. Economic Publishing Publikaties 41 Kallang Pudding Road Madhvani Building House NDONESIA P.O. Box 202 #04-03 P.O. Box 9997 8, Da Fo Si Dong Jie, Pt. Indira Limited 7480 AE Haaksbergen Golden Wheel Building Plot 16/4 Jinja Rd. Beijing Jalan Borobudur 20 Singapore 349316 Kampala PO. Box 181 NEWZEALAND COLOMBIA Jakarta 10320 EBSCO NZLtd. SLOVAKREPUBLIC UNITED KINGDOM Infoenlace LUda. Private Mail Bag 99914 Slovart G.T.G. Ltd. Microinfo Ltd. Aparlado Aereo 34270 IRAN New Market Krupinska 4 PO. Box3 BogotaD.E. Kowkab Publishers Auckland P.O. Box 152 Afton, PO. Box 19575-511 852 99 Bratislava 5 Hampshire GU34 2PG COTED'IVOIRE Tehran NIGERIA England Centre d'Edition et de University Press Limited SOUTH AFRICA, DiflusionAiricaines Ketab Sara Co. Publishers Three Crowns Building BOTSWANA ZAMBIA (CEDA) Khaled EslamboliAve., Jericho For single titles: University Bookshop 04 B.R 541 6th St. Private Mail Bag 50 Oxford University Press Great East Road Campus Abidjan 04 Plateau Kusheh Delatrooz No. 8 Ibadan Southern Africa P.O. Box 32379 Tehran P.O. Box 1141 Lusaka CYPRUS NORWAY Cape Town 8000 Center of Applied IRELAND Narvesen Information ZIMBABWE Research Govemment Supplies Center For subscription orders: Longman Zimbabwe Cyrus College Agency Book Department Intemational Subscription (Pte.) Ltd. 6,Diogenes Street, OPfigan.tSeathair .O. Box 6125 Etterstad Sev. Tourle Road, Ardbennie Engomi 4-5 Harcourt Road N-0602 Oslo 6 PO. Box 41095 P.O. Box ST;25 PO. Box 2006 Dublin 2 Craighall Southerton Nicosia Johannesburg 2024 Harare RECENT WORLD BANK TECHNICAL PAPERS (continued) No. 290 Weijenberg, Dagg, Kampen Kalunda, Mailu, Ketema, Navarro, and Abdi Noor, Strengthening National Agricultual Research Systems in Eastern and Central Africa: A Frameworkfor Action No. 291 Valdes and Schaeffer in collaboration with Errazuriz and Francisco, Surveillance of Agricultural Price and Trade Policies: A Handbook for Chiile No. 292 Gorriz, Subramanian, and Simas, Irrigation Management Transfer in Mexico: Process and Progress No. 293 Preker and Feachem, Market Mechanisms and the Health Sector in Central and Eastern Europe No. 294 Valdes and Schaeffer in collaboration with Sturzenegger and Bebczuk, Surveillance of Agricultural Price and Trade Policies: A Handbookfor Argentina No. 295 Pohl, Jedrzejczak, and Anderson, Creating Capital Markets in Central and Eastern Europe No. 296 Stassen, Smnall-Scale Biomiass Gasifiersfor Heat and Power: A Global Review No. 297 Bulatao, Key Indicatorsfor Family Planning Projects No. 298 Odaga and Heneveld, Girls and Scihools in Sub-Saharant Africa: From Analysis to Action No. 299 Tamale, Jones, and Pswarayi-Riddihough, Technlologies Related to Participatory Forestry in Tropical and Subtropical Coutitries No. 300 Oram and de Haan, Technologiesfor Rainfed Agriculture in Mediterranean Climates: A Review of World Bank Experiences No. 301 Edited by Mohan, Bibliography of Publications: Technical Department, Africa Region, July 1987 to April 1995 No. 302 Baldry, Calamari, and Yameogo, Environmtienital Impact Assessment of Settlement and Development in the Upper Leraba Basin No. 303 Heneveld and Craig, Scihools Count: World Bank Project Designs and tihe Quality of Primary Education in Sub-Saharan Africa No. 304 Foley, Photovoltaic Applications in Rural Areas of the Developing World No. 305 Johnson, Education and Training of Accountants in Sub-Saharan Anglophone Africa No. 306 Muir and Saba, Improving State Enterprise Performance: The Role of Internal and External Incentives No. 307 Narayan, Toward Participatory Research No. 308 Adamson and others, Energy Use, Air Pollution, and Environmental Policy in Krakow: Can Economic Incentives Really Help? No. 309 The World Bank/FOA/UNIDO/Industry Fertilizer Working Group, World and Regional Supply and Demand Balances for Nitrogen, Phosphate, and Potash, 1993/94-1999/2000 No. 310 Edited by Elder and Cooley, Sustainable Settlement and Development of the Onchocerciasis Control Programme Area: Proceedings of a Ministerial Meeting No. 311 Webster, Riopelle and Chidzero, World Bank Lendingfor Small Enterprises 1989-1993 No. 312 Benoit, Project Finance at the World Bank: An Overview of Policies and Instruments No. 313 Kapur, Airport Infrastructure: The Emerging Role of the Private Sector No. 314 Valdes, Schaefferin collaboration with Ramos, Surveillance of Agricultural Price and Trade Policies: A Handbookfor Ecuador No. 316 Schware and Kimberley, Informatiotn Techinology and National Trade Facilitation: Making the Most of Global Trade No. 317 Schware and Kimberley, Information Techinology and National Trade Facilitation: Guide to Best Practice No. 318 Taylor, Boukambou, Dahniya, Ouayogode, Ayling, Abdi Noor, and Toure, Strengthening National Agricultural Research Systems in the Humid and Sub-humid Zones of West and Central Africa: A Frameworkfor Action No. 320 Srivastava, Lambert and Vietmeyer, Medicinal Plants: An Expanding Role in Development No. 321 Srivastava, Smith, and Fomo, Biodiversity and Agriculture: Iniplicationsfor Conservation and Development No. 322 Charles M. Peters, The Ecology and Management of Non-Timber Forest Resources No. 323 Edited by Dominique Pannier, Corporate Governance of Public Enterprises in Transitional Economies No. 324 Cabraal, Cosgrove-Davies, and Schaeffer, Best Practicesfor Photovoltaic Household Electrification Programs No. 325 Bacon, Besant-Jones, and Heidarian, Estimating Construction Costs and Schedules: Experience with Power Generation Projects in Developing Countries I L ^ - x - L - - - ^ ^ B i F _ ; * ^ Il'ltw l I _t_ w c ' | - ' J - ML - & i | - [ T x t 3 1 s s _s , s ' iL L.t S i L i . 1 1eiz5c .s | L s; ^ ierzl i i .> z i. - L.i _^ E! .rx-ad At !s L Ls >11 . .__ Q h AS s L -"INY:I§cCZ* C -,^itu5 t1w, -- - -t m t - Ei 1 tl:wlm ,, 411.1111 4r - i i ' 2s z ] F ' - :s1er ;d . h .: , w R rrl > r7 n ls Tll * * ! rr o . sT z li _ i A i | ¢T ll | 1 1i l.i nl I5li l, 1 ,, , , ,, ,C;& - I a^> Es1|" 1 S z 1|S jk li ii., *i i 51iCI A A [ L X,, S i i il il i ! _ ' t _ n ! 11 ] _