Nicaragua - Poverty assessment (Vol. 1 of 2) : Main report

Report No. 14038-NI Republic of Nicaragua Poverty Assessment (In Two Volumes) Volume 1: Main Report June 1, 1995 Country Department II Country Operations Division Latin America and the Caribbean Regional Office u Document of the World Bank  CURRENCY EQUIVALENTS Currency Unit = C6rdoba (C$) US$1 = C$6 (November 1993, time of survey) ACRONYMS BANADES National Development Bank CDC Centers for Disease Control CONAGRO National Commission for Agriculture CORNAP State-owned Corporations Holding Company CIDA Canadian International Development Agency DANIDA Danish International Development Agency EEC European Economic Community ENABAS State Agricultural Marketing Company FAO Food and Agriculture Organization FINNIDA Finnish Department of International Development Cooperation FISE Emergency Social Investment Fund FONIF Fondo Nicaraguense de la Niiiez y la Familia GDP Gross Domestic Product GON Government of Nicaragua IDA International Development Association IDB Inter-American Development Bank IMF International Monetary Fund INAA National Water Authority INATEC National Technological Institute INCAP Institute of Nutrition of Central America and Panama INE State Public Utility Company INEC National Institute of Statistics and Census INIFOM Municipal Development Institute INRA Agrarian Reform Institute INSS Social Security Institute (previously INSSBI) INTA Institute for Agricultural Technology KfW German Credit Institute for Reconstruction LSMS Living Standards Measurement Survey MAG Ministry of Agriculture MARENA Ministry for Natural Resources and the Environment MAS Ministry of Social Action MEDE Ministry of the Economy MIS Management Information System MOE Ministry of Education (or MED) MOF Ministry of Finance MOH Ministry of Health (or MINSA) NGO Non-Governmental Organization PAHO Pan American Health Organization PAMIC Micro-enterprise Support Program PAN Nicaraguan Food Program PHC Primary Health Care RUTA Regional Unit for Technical Assistance SIDA Swedish International Development Authority SILAIS Local Systems of Integrated Health Services UNDP United Nations Development Programme UNESCO United Nations Educational, Scientific and Cultural Organization UNFPA United Nations Fund for Population Activities UNICEF United Nations Children's Fund USAID United States Agency for International Development WFP World Food Program WHO World Health Organization  NICARAGUA POVERTY ASSESSMENT TABLE OF CONTENTS Volume I (In two volumes) EXECUTIVE SUMMARY ............... .................. i I. INTRODUCTION ...................................... 1 II. POVERTY PROFILE, 1993 ....... ................... . 2 A. Measurement and Extent of Poverty ...... ............ . 2 (1) Methodological Approach ................ 2 (2) Incidence and Distribution of Poverty ........... ... . 3 (3) The Index of Unmet Basic Needs ........................ 4 (4) Consumption Patterns of the Poor .............. 6 (5) Sources of Income of the Poor ............... 7 B. The Social Conditions of the Poor ............... 8 (1) Characteristics of Poor Households .............. 8 (2) Housing and Social Infrastructure .............. 9 (3) Access to Basic Social Services ......... .. ............. 10 (a) Education Services ............................. 10 (b) Health Services ............................... 12 (c) Child Malnutrition ............................. 14 C. Employment and Poverty ............................... 15 (1) Labor Force Participation .......... .. ................ 15 (2) Educational Attainment ............. ................ 15 (3) Unemployment and Underemployment ....... .. ........... 16 (4) Sector of Employment ........... .. ................ 17 (a) Agricultural Sector Employment .................... 18 (b) Formal and Informal Sector Employment ............... 20 (5) The Structure of Wages ............................. 21 (a) Male Wages ................................. 21 (b) Female Wages ............................... 22 D. The Probability of Being Poor ...... ..................... 23 III. INCOME-EARNING OPPORTUNITIES FOR THE POOR: THE ECONOMIC FRAMEWORK ...... .................... 25 A. The 1990-93 Economic Reform Program ..................... 25 B. The Effects of the Economic Reform Program on the Poor .... ...... 26 (1) The Incentive Framework ........................... 27 (2) Public Expenditures .............................. 28 (3) The Regulatory Environment ......................... 29 (4) Trends in the Incomes of the Poor ...................... 29 C. The Current Economic Agenda ........................... 33  - 2 - IV. AGRICULTURE AND POVERTY REDUCTION ................. 33 A. The Agricultural Sector in the 1990-94 Period .................. 33 (1) The Government's Agricultural Reform Program ............. 33 (2) Production Trends ................................ 35 B. Main Issues and Constraints ............. ................ 37 (1) Inadequate Implementation of Sectoral Policies .............. 38 (2) Land Tenure Issues ............................... 39 (3) Inadequate Rural Financing Schemes ..................... 41 (4) Infrastructure, Information and Marketing Constraints ..... . . . . . 42 V. HUMAN RESOURCES DEVELOPMENT AND POVERTY REDUCTION. 43 A. Government Strategy for Human Resource Development .. ........ . 44 (1) Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 (2) Health ................. ..................... . 45 (3) Nutrition ............. . ....................... 45 (4) Population ............. . ....................... 46 B. Social Services: Key Issues Faced by the Poor . . .46 (1) Low Educational Attainment ..46 (2) Low Access to Health Care Services ..48 (3) Determinants of Malnutrition . . . . . ..... ... . . . . . . . .... . 51 C. Implementation of Government Strategy: Key Constraints ....... ... 53 (1) Accelerated Population Growth ......... .. ............. 53 (2) Incidence of Social Expenditures ........ .. ............. 54 (a) Incidence of Public Expenditures in Education .... ........ 54 (b) Incidence of Public Expenditures in Health .............. 56 VI. THE SOCIAL SAFETY NET .............. ................ 57 A. Main Safety Net Programs .............................. 58 B. Main Issues in the Social Safety Net ........................ 61 VII. POLICY RECOMMENDATIONS ........................ 63 A. Clearly Define a Poverty Alleviation Strategy .................. 63 B. Sustain Macroeconomic Policies .......................... 63 C. Promote Agricultural Growth and Rural Poverty Alleviation ... ...... 64 D. Invest in the Human Capital of the Poor ..................... 66 E. Improve the Social Safety Net ............................ 67 F. Institutionalize Monitoring and Evaluation .................... 68 Bibliography This report was prepared by a team led by Ana-Maria Arriagada (Task Manager). Major contributors included Kinnon Scott and Nancy Gillespie. It has benefitted greatly from the inputs of [an Bannon, Ulrich Lachler, Miguel G6mez and Alberto Valdes. The report draws on background papers by Gustavo Arcia, Linda Baez, Rene Bernier, Jack Fidler, Carola Pessino and Ellen Piwoz (Consultants). Financial support for the study was provided by the Netherlands Consultant Trust Fund. The Lead Economist, Division Chief, and Department Director are Ian Bannon, Donna Dowsett-Coirolo, and Edilberto Segura, respectively.  -3- TEXT TABLES 2.1 Poverty Lines, 1993 ................... 3 2.2 Contribution to Poverty and Head Count Index by Region and Geographic Area, 1993 ......................................... 4 2.3 Poverty Gap and FGT P2 Measures by Region and Geographic Area, 1993 ......................................... 4 2.4 Basic Needs and Poverty Line by Geographic Area, 1993 ...... ................. 5 2.5 Distribution of Total Per Capita Expenditures by Poverty Group and Geographic Area, 1993. 6 2.6 Distribution of Monthly Per Capita Food Expenditures by Poverty Group and Geographic Area, 1993. 7 2.7 Household Income Sources by Poverty Group, and Geographic Area, 1993. 7 2.8 Characteristics of Households by Poverty Group and Geographic Area, 1993 ......................................... 8 2.9 Characteristics of Female Headed Households by Poverty Group and Geographic Area, 1993. 9 2.10 Characteristics of Housing by Poverty Group and Geographic Area, 1993 ........................................ 10 2.11 Illiteracy by Gender, Poverty Group and Geographic Area, 1993 ...... . . . . . . . . . . . . 11 2.12 Mean Years of Schooling by Gender, Age Poverty Group and Geographic Area, 1993 .11 2.13 School Enrollment by Age Group, Education Level, Poverty Group and Geographic Area, 1993 .12 2.14 Public School Enrollment by Level, Poverty Group and Geographic Area, 1993 .12 2.15 Consultation, Coverage and Concentration Rates by Poverty Group, 1993 .13 2.16 Consultation, Coverage and Concentration Rates by Region, 1993 .13 2.17 Malnutrition Among Children under Five by Poverty Group and Geographic Area, 1993 . ....................................... 14 2.18 Labor Force Participation and Unemployment Rates by Sex, 1993 ................. 15 2.19 Average Years of Schooling of the Labor force, 1993 ........................ 16 2.20 Labor Force Participation, Unemployment, Underemployment, and Poverty by Region, 1993 . ...................................... 17 2.21 Underemployment Rates by Region and Geographic Area, 1993 .................. 17 2.22 Distribution of Employed by Economic Sector, Poverty Group and Geographic Area, 1993 .18 2.23 Comparison of Agricultural Households with and Without Agricultural Land, 1993 .19 2.24 Distribution of Crops by Poverty Group and Geographic Area, 1993 ............... 19 2.25 Loans Received by Agricultural Households in Previous 12 Months, 1993 ........................................ 19 2.26 Mean Monthly Earnings in the Formal and Informal Sectors, 1993 ................ 21 2.27 Mean Years of Schooling and Log Hourly Wages by Region and Sex, 1993 .22 2.28 The Probability of Being Poor ....................................... 24 2.29 Probability of Poverty by Number of Children < 13 ....... .................. 24 2.30 Probability of Poverty by Years of Schooling of the Head ...... ................ 25  -4- TEXT TABLES (continued) 4.1 GDP and Agricultural GDP Growth 1998-1994 ............................ 35 4.2 Value of Agricultural Exports and Imports, 1989-1994 ........................ 36 4.3 Area Planted by Type of Crop, 1989-1994 (in '000 Mzs.) ...................... 37 4.4 Estimate of Total Resources for Projects in Rural Areas (1990-2000) ............... 40 4.5 Estimated Resources for Public Sector Rural Credit Programs (1991-96) ............. 41 4.6 Spread Between Farmgate and Wholesale Prices in Nueva Guinea, 1992 ............. 42 4.7 Roads Network (kilometers) ........................................ 43 5.1 Mean Gap Between Expected and Actual School Attainment by Age, Poverty Group and Geographic Area, 1993 ............................... 47 5.2 Predicted Probability of Child School Attendance by Age, 1993 .................. 48 5.3 Immunization Coverage for Children Under 6 by Poverty Group and Geographic Area, 1993 ........................................ 49 5.4 Maternal Health by Poverty Group and Geographic Area, 1993 ..... . . . . . . . . . . . . . 49 5.5 Use of Health Services Among Children by Poverty Group and Geographic Area, 1993 .50 5.6 Incidence and Level of Health Expenditures Among Those Reporting Illness or Accident in Past 30 Days by Poverty Group and Geographic Area, 1993 .51 5.7 Child Malnutrition Indicators by Poverty Group and Region, 1993 ..... . . . . . . . . . . . . 52 5.8 Child Malnutrition Indicators by Age, 1993 ............................... 52 5.9 Distribution of Public Education Expenditures, 1990-94 ....................... 54 5.10 Distribution of Subsidies to Public Pre-primary Education, 1993 ..... . . . . . . . . . . . . . 55 5.11 Distribution of Subsidies to Public Primary Education, 1993 ...... . . . . . . . . . . . . . . 55 5.12 Distribution of Subsidies of Public Secondary Education, 1993 ...... . . . . . . . . . . . . . 55 5.13 Distribution of Subsidies to Public Vocational Training, 1993 .................... 56 5.14 Distribution of Subsidies to Public Higher Education, 1993 ...... . . . . . . . . . . . . . . . 56 5.15 Distribution of Subsidies to Public Health, 1993 ............................ 57 5.16 Distribution of Subsidies to Public Health by Geographical Area, 1993 .............. 57 6.1 Government Safety Net Programs, 1991-1994 ............................. 59 TEXT BOXES 1.1 Data, Analysis, and Policy Making ..................................... 2 1.2 Limitations of the Unmet Basic Needs (NBI) Method ........ . . . . . . . . . . . . . . . . . . 5 1.3 Inequality ..................................................... 6 3.1 Achievements of the Structural Reform Program ............................ 26 4.1 Background: Agriculture in the 1970-89 period ............................ 35 TEXT FIGURES 1.1 Percentage of Malnourished Children by Region and Geographic Area ..... . . . . . . . . . 14 1 Index of Average Salaries plus Benefits in Constant April 1991 Cordobas ..... . . . . . . . 30 2 Index of Average Salaries plus Benefits in Constant April 1991 Cordobas ..... . . . . . . . 30 3 Index of Average Salaries plus Benefits in Constant April 1991 Cordobas ..... . . . . . . . 31 4 Real Producer Price Index (April 1991CS, deflated by RCPI) .................... 31 5 Producer Prices (USS nominal) ...................................... 32 4.1 Basic Grains Production, 1982-1994 (in thousands of QQ) ....... . . . . . . . . . . . . . . . 36  Nicaragua - Priority Poverty Indicators Poverty Lines (1993, cordobas)9' Urban Rural Total Upper Poverty Line (c$214.5) Headcount (% of population) 31.9% 76.1% 50.3% Lower Poverty Line (c$101.3) Headcount (% of population) 7.3% 36.3% 19.4% Memorandum Item GDP p.c. 1992 (1990, US$) 422h' Short-Tenn Income Indicators 1990 1991 1992 1993 1994| CPI inflation rate (year end) 13,490 865.6 3.5 19.5 12.4 Social Indicators 1960 1970 1981 1989 1990 1991 1992 1993 Share of Public Expenditures for Social Services in GDP-' 4.0% 8.7% 6.1% 14% 9.7% 9.2% Gross Primary Enrollments' 80% 99% 95% 98% Under-5 Mortalityg' 210 132 92 78 56 Immunizatiori' 70% Child Malnutritionf' 57% 56% 28% Life ExpectancyE' 53 58 65 67 Total Fertility Rate- 6.8 6.1 5.3 4.6i Maternal Mortalityg' 159 a/ The upper poverty line is the cut-off for the poor; the lower poverty line is the cut-off for the extreme poor. World Bank, Nicaragua. Poverty Assessment, June 1995. b/ Health and education. c/ 1970-90 World Bank, Human Resources in Latin America and the Caribbean - Priorities and Action, January 1993. 1992, World Bank, World Development Report, 1994. d/ Per thousand live births. " % of one-year-olds immunized in 1987, UNDP, Human Development Report, 1990. " children 0-59 months with any form of malnutrition, 1970 & 1981 World Bank, Nicaragua. Review of Social Sector Issues, February 1993. 1993, World Bank, Nicaragua, Poverty Assessment, June 1995. "Ibid. Per 100,000 live births. h/ World Bank, Nicaragua, Country Economic Memorandum, February 1994, Table 2.4 i' Profamilia/USAID/CDC, Nicaragua Fertility Survey 1992-93, 1993.   NICARAGUA POVERTY ASSESSMENT EXECUTIVE SUMMARY 1. This report has two main objectives. The first objective is to assess the extent and distribution of poverty in the country. The second objective is to outline the framework for a sound poverty reduction strategy: a three pronged strategy including accelerated economic growth, particularly the revitalization of the agricultural sector; investment in human resource development of the poor; and a well targeted social safety net. The main findings and recommendations of this report are intended to assist the Government of Nicaragua (GON), IDA, and other donors refine and strengthen the design and targeting of poverty reduction policies and programs in a context of continuing adjustment. A. POVERTY IN NICARAGUA 2. This poverty profile is based on the nationally representative Nicaraguan Living Standards Measurement Survey (LSMS) carried out with IDA support by the National Statistics Office (INEC), with funding from SIDA, UNDP, UNICEF, USAID and the GON, in 1993. Data were collected from 23,135 people living in 4,213 households representing urban and rural areas in each of the country's regions. 3. This report sets the first poverty and extreme poverty lines ever estimated for Nicaragua based on a method generally used in other developing countries. This method allows one to make poverty comparisons among population groups, target antipoverty programs, and monitor changes overtime. The poverty line is defined as the level of total per capita monthly expenditures at which an individual obtains the minimum caloric requirement (2,226 cal. per adult). The extreme poverty line is defined as the level of per capita monthly food expenditures required to obtain the daily minimum caloric requirement. This approach leads to a poverty line of US$429 per year, and to an extreme poverty line of US$202 per year. Half of the population, or about 2 million people fall below the poverty line. Forty percent of the poor, or 20 percent of the population, fall below the extreme poverty line and are food poor. That is, they cannot meet the daily minimum caloric requirement even if they were to devote all of their consumption to food. The most important conclusions of the poverty profile for Nicaragua are: (a) rural poverty is much higher and deeper than urban poverty. Three quarter of those living in rural areas are poor compared to 32 percent in urban areas. Rural areas, with 41 percent of the country's population have 63 percent of all the poor and 78 percent of all the extreme poor. The urban poor have average expenditures 11 percent below the poverty line, while the rural poor have average expenditures 37 percent below the poverty line; (b) by far the poorest regions of the country are the Northern (Jinotega and Matagalpa) and the Segovias (Esteli, Madriz and Nueva Segovia) regions. With 23 percent of the total population, they have 46 percent of the country's extreme. In these regions, the rural poor have average expenditures 48 percent below the poverty line. In contrast, Managua, with almost one third of the country's population has only 7 percent of all extremely poor people; (c) poverty is concentrated in agriculture. Over three quarters of the country's poor households derive most of their income from agriculture either as farmers and/or wage laborers. Small farmers engaged in basic grains production are the poorest of all; (d) the key poverty issue in the labor market is not unemployment but underemployment. Overall unemployment is 13.4 percent, but with wide regional variations. The three regions with highest unemployment (Western, Managua and Southern) comprise over 80 percent of the country's unemployed, and have the lowest incidence of extreme poverty. At the same time, the poorer  - ii - regions have the highest underemployment rates showing lower reservation wages, as they are too poor to afford to be unemployed; (e) 70 percent of the employed work in the informal sector. Informal sector employment is not a good predictor of poverty. Monthly earnings in the informal sector are above the poverty line, indicating that it is high dependency ratios that make these earnings inadequate on a per capita basis. Moreover, no evidence of labor market segmentation arising from formal or informal sector employment was found, with workers sorting themselves out into the sector where they receive higher wages according to their individuals skills; (f) female labor force participation rates among the poor are lower than among the non-poor because they face high demand foT household work stemming from a large number of children coupled with general lack of daycare, lack of access to basic services and the ensuing need to collect wood for cooking and fetch drinking water; (g) larger households are poorer, mostly because they have more children under age 12. The predicted probability of a household being poor increases by 10 percent with each additional child; (h) there is no evidence that female headed households are at a greater risk of poverty; (i) poverty and low educational attainment are highly correlated. Over half of the extremely poor adults in rural areas and over a third in urban areas are illiterate. On average, a Nicaraguan has 4.5 years of schooling. The extremely poor have much less schooling: 3 years in urban areas and only 1.6 years in rural areas; (j) today, access to primary school is almost universal for boys and girls. Net enrollment of over 90 percent compares very well with other Central American countries. However, the poor do not complete primary school due to late entrance to school and/or high repetition and low quality of schooling. The system must finance 11 student years to obtain a primary school graduate. Because the poor fail to complete primary education, their enrollment in secondary education is half that of the non-poor. For the same reason, enrollment at the university level is exclusive by the non-poor. Attendance in preschool programs, which helps to develop learning skills among disadvantaged children, largely caters to the urban non-poor; (k) poverty and child malnutrition are highly correlated. Overall, almost one third of children aged 0-5 years old are mnalnourishied. Four out of ten extremely poor children living in rural areas are malnourished. Consistent with the distribution of poverty, almost half of the rural children in the Northern and Segovias regions are malnourished. Besides income, this study found that: [i] a literate mother nourishes her children better, [ii] each additional child in the household increases the probability of malnutrition, [iii] breastfeeding is associated with lower malnutrition among children under 18 months, and [iv] children in households without water and sanitation are twice more likely to be malnourished than children in households with water and sanitation; (1) basic infrastructure services are inadequate, and the poor have less access to these services than the non-poor, particularly in rural areas. Low coverage of water and sanitation services (60 and 32 percent) explain why diarrhea and vector borne diseases are still the main causes of morbidity and mortality in the country. Almost 80 percent of the rural poor do not have water (inside or outside the dwelling). Rural areas have no access to sanitation services (sewer, septic tanks or cesspools), and over half of the extremely poor do not even have latrines. In urban areas, over 70 percent of the poor are not connected to sanitation services, compared to 35 percent of the non- poor. At 70 percent, coverage of electricity is 20 percent higher than estimated by the State Power Company (INE), but the rural poor have lower access (30 percent coverage);  - 111 - (m) the poor, particularly in rural areas, have less access to health care services than the non-poor. Consultation rates show that a non-poor person perceiving illness is three times more likely to obtain care than an extremely poor, and twice as likely to receive at least one consultation. On average, the non-poor receive more consultations per person than the extreme poor. The poorest regions, particularly the Northern region, suffer disproportionately from lower access to health care services. Although immunization rates compare favorably with other Latin American countries, poor children are less likely to receive full immunization, especially in rural areas. Pre- natal and birth care are particularly inadequate for poor and rural women, largely explaining the country's high maternal mortality (159 per 100,000). Over 25 percent of poor women in urban areas and 50 percent in rural areas did not receive any pre-natal care during their last pregnancy. Almost one-third of births were cared for by midwifes (some not trained). Actually, 23 percent of extremely poor women in rural areas gave birth without assistance from health care personnel. B. GOVERNMENT POLICIES AND POVERTY: PROGRESS AND ISSUES 4. Recent Economic Trends and the Poor. Promotion of economic growth, which is the basis for poverty reduction, has been a driving force behind the GON's economic reform program initiated in 1991. The program has been highly successful in stabilizing the economy. However, in the 1990-93 period, reactivation of growth remained elusive as GDP per capita fell by 8 percent. Economic conditions improved markedly in 1994. Real GDP growth is estimated at 3.2 percent, the highest rate since 1983, and inflation declined from 20 percent in 1993 to 12 percent. And most encouraging, 1994 saw strong growth in agriculture and non-traditional agricultural exports. 5. Thus far, the main welfare impacts of the economic program are: (a) all have gained with the end of hyperinflation; (b) the flow of external aid allowed stabilization to take place with substantial increases in wages and consumption; (c) commerce and services have gained with trade deregulation; (d) some urban workers have benefitted, with dramatic salary gains; (e) agriculture and other exportables show mixed results, depending on product competitiveness under the new relative price regime, and the speed of replacement of dismantled market structures-small scale agriculture, where Nicaragua's poorest are concentrated, has lost, as the elimination of state market structures is only slowly being replaced in rural areas; (f) the poorest consumers have lost, as general subsidies on basic goods have been eliminated and commerce has been freed, allowing vastly greater product availability, but also higher prices; (g) the unemployed and underemployed have lost, as employment guarantees are eliminated and rising real wages have only benefitted the employed. 6. Agriculture and Poverty. The GON's agriculture policy framework complements the overall economic reform program, aiming to reestablish market-based agriculture and increase production in the short-run. However, the supply response of the agricultural sector has been slow. Main reasons have been overall political instability and a fragile macroeconomic situation. The real exchange rate remains overvalued in competitiveness terms. The actual impact of this implicit tax on the performance of the agricultural sector warrants detailed analysis. 7. In addition, there are four sector specific constraints, particularly relevant from a poverty perspective: (a) inadequate implementation of sectoral policies and lack of a strategy for addressing rural poverty. There is a mismatch between policies and what actually gets done. There is little coordination among institutions, which have weak formulation and implementation capacity. No emphasis has been placed on defining policies and programs that are cost-effective while benefiting poor farmers. Excessive fragmentation of donor programs further complicates the situation. There are over 90 separate donor- financed projects (for over US$588 million) with overlapping or inconsistent objectives and activities, or doubtful effectiveness. In general, and probably due to this fragmentation of initiatives, project implementation has been less than satisfactory, with less than one third of available resources spent by December 1994; (b) insecurity of land tenure and inefficient land markets are still a critical constraint. Although the GON is addressing this issue, progress is slow; (c) although there are over 30 donor  - iv - financed programs for about US$68 million to provide credit to small farmers, no viable financing schemes for small and medium size farmers have been developed. Some programs use credit as a means to transfer a subsidy while others make no distinction between resources for social assistance and productive use; and (d) infrastructure, information and marketing constraints. Although Nicaraguan basic grains can compete effectively at the regional level, small producers are unable to benefit because of high costs of information about local and regional markets, poor rural infrastructure (particularly roads), high cost of transportation, and non-tariff barriers (such as complicated and discretionary export requirements). 8. Human Resources Development and Poverty. The GON has developed sound health and education strategies. They focus on improving the delivery of basic services, which benefit most the poor, and include major institutional reforms to improve efficiency in the use of available resources and implementation capacity. The MOH has made remarkable progress decentralizing operations and budgets to the SILAIS, reorganizing manpower, developing pharmaceutical supply and management systems, implementing hospital management programs, and diversifying financial sources. The MOE's municipal decentralization model is promising, in terms of increased accountability at the local level through direct community participation in school management and mobilization of resources. Donor programs in both sectors are well coordinated around GON strategies. 9. Despite significant progress, two major issues remain: (a) accelerated population growth increases demand for health and education services faster than they can be expanded. There is no prospect of seriously increasing average incomes if the population continues to double every 26 years as it will happen if the current fertility pattern of 4.6 children per woman is maintained. In rural areas, total fertility is 6.4 children per woman largely due to inadequate access to family planning services. About 48 percent of married women of reproductive age are using contraception compared to 75 percent in Costa Rica, 58 percent in Panama, and 53 percent in El Salvador; and (b) with the exception of primary education, public social spending is not well targeted to the poor, resulting in inadequate funding for basic services. Nicaragua's public spending in the social sectors is higher than most Central American countries. In 1994, the share of public spending on the social sectors was over 9 percent of GDP and about 40 percent of total public sector spending. However, per student spending in primary education is US$55 per year compared to a Latin American average of US$100. At the same time, higher education, which was shown by this study to serve exclusively the non-poor, receives US$800 per student per year. Compared with other Latin American countries, Nicaragua's higher/primary spending ratio (14.5 times) is the second highest in the region (after Brazil). In 1994, annual public health care spending was US$19 per capita (US$23.2 including donor support), which is higher than that of any other Central American country except Costa Rica, and much higher than is typical for an economy with Nicaragua's GDP per capita. However, as shown by this study, coverage of primary health care services is insufficient, particularly ir rural areas. 10. The Social Safety Net. In the 1991-94 period, the GON has mobilized between 1.0 and 1.2 percent of GDP for safety net programs (rehabilitation of basic infrastructure, integrated rural development programs for war victims, emergency employment, and nutrition). The social safety net has several positive features. First, some programs have a demand-driven approach. Second, and partly as a result of this orientation, it gives local communities an incentive to organize themselves, improves community participation, and fosters the democratization process. Finally, its strongest agency, FISE, is a financing entity; implementation is left to private contractors, local bodies, and NGOs. This approach avoids building up bureaucracies, circumvents inefficiencies in the public sector, and promotes participation of the private sector. 11. However, a number of issues exist: (a) the safety net remains fragmented because it lacks a coherent framework for short-term interventions. It does not have clear priorities on what are the problems it should address nor what populations to target. The MAS, which was created in 1993 to improve social sector coordination and to monitor social policy and programs, has become an executing agency managing over 15 programs, many of which belong to agricultural sector agencies. This study  - v - found that transfers from donor and government programs are mostly received by non-poor households. This is not surprising since different agencies and programs have assorted definitions of "poor" and use different targeting criteria; (b) institutional arrangements are inefficient. The safety net includes an excessive number of programs and overlaps with line agencies. The MAS alone implements about 11 programs for war victims, and three employment programs. Different agencies finance similar types of project (FISE, MAS, INIFOM), which requires coordination arrangements between safety net programs and line ministries and sectoral agencies further complicating management; (c) information systems in implementing agencies and programs are inadequate. As a result, there is no data on actual targeting criteria, program costs and unit costs, or effectiveness of different interventions; and (d) there are gaps in the safety net, with insufficient attention to water and sanitation, and short-term nutritional assistance to pregnant and lactating mothers, children under 5 years of age, pre-school and school-age children. C. RECOMMENDATIONS 12. Clearly Define a Poverty Alleviation Strategy. The GON should clearly spell out a poverty alleviation strategy and express it in a policy document. Such document should specify priorities, programs and measures to be pursued over the short- and medium-term, including a defined and articulated implementation plan with institutional responsibilities. This strategy should have three main elements. First, it should sustain a stable environment to enable rapid economic growth. International experience shows that countries with sustained economic growth and increased efficiency of resource use have been successful in significantly reducing poverty. Policies that absorb and make productive use of rural labor will expand the employment and income-earning opportunities for the majority of the Nicaraguan poor. Second, the strategy should focus on the provision of adequate basic social services for the poor, especially quality primary education, primary health care, family planning, and nutrition. Third, a well-targeted and coherent social safety net is needed to protect the most vulnerable groups. Such strategy document should be the central instrument to coordinate donor support. There is a need for improved donor coordination around a single Government strategy, particularly regarding agricultural sector and safety net programs. 13. Sustain Macroeconomic Policies. Sustaining and deepening the economic reform program is essential to maintain macroeconomic stability and promote economic growth. Nicaragua's economic transformation is far from complete. The GON must increasingly focus on strengthening the institutional structures that are needed for markets to function efficiently so as to generate growth and reduce poverty. This will involve rebuilding the state apparatus to support the private sector and to assist the poor. The public sector must increasingly focus on providing essential infrastructure and strengthening the human capital base for equitable growth. Policy and incentive reforms must be designed to increase the responsiveness and flexibility of product and factor markets to price signals. Given Nicaragua's economic structure, constraints imposed by the difficult property rights problem, considerable political uncertainty and a difficult external environment, over the medium term the revival of growth must be based on the agricultural sector. Strong agricultural growth would not only provide a powerful stimulus to the rest of the economy and contribute to export diversification, but would also have a marked positive impact on the incomes of the majority of the poor. 14. GON's reform program aims to further downsize the public sector to increase public savings, restore a safer foreign reserve level and permit expansion of credit to the private sector. From a poverty perspective, the key measures of the economic program would encourage broadly based and non- distortionary economic growth, which would absorb and make productive use of the poor's most abundant resource-labor. Provision of a stable business climate is essential for increased investment, but economic reforms must be complemented by political and social stability. Elimination of remaining tax, tariff, and other policies that distort the relative prices of capital, labor and land, would also encourage a labor absorbing pattern of growth. Public sector reform is also crucial to the poor, first, to create a small and responsive public sector capable of facing declining foreign aid without fiscal crises. Second, a smaller public sector will also permit fiscal savings that can be deployed to: (a) alleviate key  - vi - bottlenecks identified in this report as constraining poverty reducing growth such as lack of rural infrastructure, poor quality of primary education, and inadequate access to basic health care; and (b) expand efficient and well-targeted safety net programs until sustained growth is achieved. 15. Promote Agricultural Growth and Rural Poverty Alleviation. To this end, the GON should, first, streamline programs and strengthen sectoral implementation capacity. The Government should focus only on those areas where the development of a public sector role is appropriate. As a first step, the GON should evaluate all on-going rural programs, including a comprehensive review of their institutional arrangements. Second, the GON should select a limited number of interventions and programs to be supported both by the Government and donors alike. A key challenge will be to ensure that selected programs are delivered in a coherent and coordinated fashion. To this end, each sectoral institution should develop an operational strategy, including an institutional strengthening action plan, with the National Commission for Agriculture (CONAGRO) ensuring coherence of the strategies as a whole. MAS' agricultural programs, if maintained, should be returned to the sector. Second, land titling and dispute resolution should proceed as a main priority. The GON has committed to resolve all outstanding title disputes and complete land titling by June 1996. Adequate registration of properties is a critical element for the operation of a land market. As the legal titles of many small farms that were not involved in the agrarian reform process have not been properly updated, assistance for small farmers to update title registration must be also a priority. Third, the GON should emphasize policies and programs that are cost-effective while benefiting small farmers. Major areas of support for small farmers include: (a) Developing a viable rural financing system for small and medium size farmers. Existing credit programs should be reformulated to ensure they follow a sustainable and nondistortionary approach. A key lesson from successful international small rural credit schemes is that access to reliable credit for the poor at market rates is a more important factor in contributing to their income growth than efforts to subsidize interest rates, which inevitably lead to reduced access through credit rationing. Within the GON, the PAMIC should be the institution responsible for regulating and assisting organizations providing non-conventional credit for these producers. Finally, the GON should phase out BANADES by transferring its activities to the private sector. This would be consistent with ongoing financial sector reforms, and clearly indicate to the public the Government's commitment to end direct state intervention in the financial sector; (b) Removing information, marketing and infrastructure constraints. To this end, improving the efficiency of marketing channels, especially for basic grains, should be a priority. Key actions should include timely provision and wide dissemination of price and market information to farmers. This task should be at the core of MAG's work plan. Regarding infrastructure, the GON should: [i] assess the needs for rehabilitation, upgrading and construction of rural infrastructure (particularly rural roads) by region and determine their costs; [ii] select among alternative interventions, those highest return investments to give them priority in the Public Investment Program. The generation of temporary jobs would be an additional benefit of such investments; and [iii] assess options for transferring to the private sector ENABAS' remaining grain treatment and storage facilities. (c) Strengthening agricultural research, training and extension services to improve productivity and create marketable surpluses among small farmers. These services should be designed with a participatory approach and evaluate constraints of small farmers throughout the productive cycle. Extension services should be complemented by training on small farm management techniques; 16. Invest in the Human Capital of the Poor. Although the GON has started implementation of sound health and education strategies, three major recommendations arise from this study. First, family planning should be a national priority. This study found that large family size resulting in high dependency ratios is a major determinant of poverty in Nicaragua. Moreover, this study also found that the higher the number of children in a household, the higher the probability that children under 5 will  - vii - be malnourished, which has long-term physical and negative cognitive consequences. Finally, the study also found that women's labor force participation, which could help increase family income, is strongly restricted by the number of small children in the household. Nicaragua is one of the few Latin American countries that still has high fertility and has not succeeded in extending family planning services to all those who need them. Therefore, the need is to strengthen broad based effective programs at the MOH and private sector in support of temporary methods of contraception for the more than 160,000 women, mostly rural and poor, needing protection against unwanted fertility. 17. Second, the GON should target better public social expenditures. If the GON is to implement its primary education strategy and improve quality and internal efficiency of primary education, more resources are needed at this level. To this end, the generous subsidization of higher education must be corrected. Improving targeting of pre-primary and secondary education resources is also needed. Because their purpose is to compensate children from disadvantaged socio-economic backgrounds, publicly financed pre-primary programs should be exclusively targeted to the extreme poor. In the case of secondary education, where public spending is regressive, there is ample room for cost-recovery. In the health sector, improving access for the poor will require strengthened efforts of the GON to: (a) further reallocate public expenditures toward primary health care services. An important finding of this study is that the provision of preventive services, such as pre-natal and birth care is a good mechanism to reach the poor both in urban and rural areas; (b) increasing cost-recovery measures for the non-poor who are major users of public health services, particularly in urban areas. 18. Third, addressing child malnutrition should be a priority. To this end, the GON should refine its nutrition strategy and 5 year work plan to emphasize: (a) MOH nutrition interventions such as nutrition education, surveillance and monitoring, micronutrients, promotion of breastfeeding and adequate weaning practices; and (b) initiatives for fortification of foods with micronutrients. Additional initiatives on targeted nutrition programs and water and sanitation should be undertaken by the social safety net. 19. Improve the Social Safety Net. To this end, the GON should first evaluate all safety net programs, particularly those of FISE and MAS, including their institutional setups. Second, based on this evaluation and the strategies of sectoral agencies, the GON should clearly define the role and interventions to be handled by the social safety net vis-a-vis line agencies. As a third step, the GON should develop a safety net strategy which should: (a) select a small number of priority programs emphasizing rural areas; (b) develop and apply a consistent targeting criteria based on poverty. The poverty profile provides a good basis to start developing an adequate targeting mechanism with a well- defined target population. It shows the geographic and regional location of the poor. Also, it identifies a number of factors at the household level that are strongly associated with poverty. Finally, municipal level information gathered by MAS and INIFOM could provide valuable inputs; (c) transfer to the line agencies those programs that belong in sectoral agencies; (d) if specific assistance is still needed for war- affected populations, consolidate their management in one program with a single administration unit; (e) eliminate duplication of programs; (f) establish adequate information systems including, inter alia, target population, actual coverage, total costs, administrative costs, and unit costs; and (g) define appropriate inter-institutional coordination mechanisms between safety net programs and line agencies. Finally, the GON should revise the role of the MAS. The MAS should not be another executing agency. Its role should be one of coordination and monitoring of the implementation of the poverty alleviation strategy. 20. Based on the poverty profile and experience to date, the GON's safety net should: (a) continue to support FISE, its most robust program to date; (b) strongly emphasize the expansion of water and sanitation, particularly in rural areas; (c) evaluate existing nutrition programs, and select a limited but well-targeted number of interventions for support both by the GON and donor agencies. As shown by this study, the highest priority should be assigned to children and women in rural areas and to the Segovias and Northern regions; (d) as most of the country's unemployed are in the urban areas of Managua, the Southern and Western regions, and increased demand for labor may not be forthcoming in the short-term, employment generation through a public works scheme may be advisable. However,  - viii - such program should concentrate in the above mentioned regions and not compete with other employment: it should pay wages below the minimum wage. 21. Institutionalize Monitoring and Evaluation. The GON is well positioned to institutionalize the experience of the first LSMS to measure poverty conditions on a periodic basis, and to develop a monitoring system which can provide policy makers with timely information to evaluate the poverty alleviation strategy. Under the FISE II project, IDA will provide technical assistance to the GON for the second round of the LSMS and to analyze its results. MEDE would oversee its implementation. Because of the extent of rural poverty, the second survey would include a special module on agricultural activities. To develop its institutional capacity to effectively monitor poverty and evaluate the implementation of its poverty alleviation strategy, the GON should organize a high level technical unit, responsible for monitoring and evaluating the Government's poverty alleviation strategy. This unit would require staff with adequate analytical skills, capable of performing tasks going from conception to data analysis. Institutional affiliation of such unit should be consistent with the Public Sector Reform Program.  NICARAGUA POVERTY ASSESSMENT I. INTRODUCTION 1. Starting in 1991, Nicaragua has undergone a major stabilization and adjustment process that has been highly successful in stabilizing the economy. Reactivation of growth, however, has remained elusive: GDP per capita fell by 8 percent between 1990 and 1993. The social impact of this process was strongly cushioned by a large influx of foreign assistance that allowed a 12 percent expansion of private consumption per capita in the same period'. Reviving growth is a pre-condition for poverty alleviation. To this end, the Government of Nicaragua (GON) faces a major economic challenge: to maintain macroeconomic stability in the face of declining foreign aid2, while extending and deepening structural adjustment to improve the conditions for resumption of private sector led growth. A significant reduction in private consumption will be needed to increase domestic savings and release resources for investment. Political support to sustain these policies is critical. Widespread chronic poverty and socio-political instability pose another major challenge for the GON: to protect the poorest and most vulnerable groups in the short run and to better the provision of basic social services. 2. Meeting these policy challenges requires a sharper focus on priorities for poverty reduction. Thus far, lack of up-to-date and reliable data to quantify, place, and characterize the poor has prevented the GON from designing an adequate poverty alleviation strategy, and donor resources from focusing on key poverty issues and/or target groups. This report has two main objectives. First, to systematically assess the prevalence and distribution of poverty in the country. How many poor are there? Who are the poor? Where do they live? What are their sources of income? Second, to outline the framework for a sound poverty reduction strategy: a three pronged strategy including accelerated economic growth, particularly the revitalization of the agricultural sector; investment in human resource development of the poor; and a well targeted social safety net. The main findings and recommendations are intended to assist GON, IDA, and other donors refine and strengthen the design and targeting of poverty reduction policies in a context of continuing adjustment3. 3. The report is divided in two volumes. The first volume presents the main body of the report, and is organized as follows. Chapter II presents a summary of the poverty profile for Nicaragua constructed on the basis of the 1993 Living Standards Measurement Survey (LSMS) (Box 1.1). Chapter III analyzes the 1990-93 economic reform program, its likely impact on the poor, and the challenges ahead for enhancing poverty-reducing growth. The linkages between the agricultural sector and poverty reduction are examined in Chapter IV. Chapter V focuses on human resources development issues including health, education, nutrition, and population. Chapter VI reviews GON's social safety net programs and main issues. Finally, Chapter VII presents main recommendations. The second volume of the report contains all the tables of the poverty profile including methodological and technical details; and a series of technical notes and tables supporting each of the chapters in the main report. 1. During 1990-92, gross disbursements of grants and loans averaged about US$600 million per year, equivalent to about US$150 per capita. 2. In 1993, gross disbursements of grants and loans fell to US$390 million, 35 percent below the 1990-92 average, a drop equivalent to 12 percent of GDP. 3. Other IDA economic and sector work has addressed several issues relevant for poverty analysis. The Public Sector Expenditure Review (Report No.10785-NI, 1992) addressed the key issues in public sector resource allocation and institutional capacity. The Social Sector Review (Report No. 1067 1-NI, 1993) examined trends, programs and key issues in the social sector, including the level of public expenditures, its appropriateness and composition within each sector, and main institutional constraints to effectively contribute to poverty reduction. It covered the areas of population, water and sanitation, nutrition, health and education. A Country Economic Memorandum (Report No. 12066-NI, 1994) evaluated economic developments since 1990, and identified barriers to private sector growth, including the agricultural sector and the labor markets.  -2 - Box 1.1: Nicaragua - Data, Analysis, and Policy Making In Nicaragua, the war and fimited capacity affected data collection and the development of a national statistical system. The Living Standards MeasurementSurvey (LSMS) initiative. was launchedby the GON with IDA support to: (a) address the generalized lack of up-to-date and reliable information on the welfare level of the population; and (b) set the basis for the development of the country's statistical capacity building. The survey was: designed to address a number of important policy questions, including, but not limited to, poverty levels and characteristics of the poor, levels of child malnutrition, access to, and use of, health and education services, labor participition, levels of unemployment, and informal sector employment. The survey was implemented in 1993 withC a sample of 4,213 households (23,135 individuals) representative for urban and rural areas at the regional level. (See details on the sample and LSMS methodology in Volume 1, Poverty Profile.): The executing agency was the National Institute of Statistics' and Census of Nicaragua (INEC) at the Ministry of the Economy (MEDE), and was co-financed by the United Nations Development Programme (UNDP)V Children's Defense Fund (UNICEF), the Swedish International Development Authority (SIDA), the United States Agency for International Development (USAID) and GON. In addition to its use in this study, the LSMS data has already fueled a number of operational and policy- oriented analyses. For example, the Nicaraguan Social Investment Fund (FISE) is using the LSMS to evaluate the suitability of its geographic targeting mechanism and criteria for selection of individual beneficiaries. The GON has used the LSMS data for: designing new consumer price index (MOF), updating patterns of private consumption for national accounts (Cental Bank), updating basic consumption basket (MOF and Ministry of Labor), estimating unmet basic needs (Ministry of Social Action (MAS). Finally, the LSMS exercise assisted the GON. to make progress toward the implementation of a national population census (the last one was in 1972) by updating census maps and aiding mobilization of donor financing. The census commenced to be fielded in April 1995. 1!. POVERTY PROFILE, 1993 4. This chapter identifies the characteristics of the poor, where they live, and depicts their economic and social conditions. It starts with a definition and measure of poverty and a description of its extent and distribution (Section A). Section B characterizes the social conditions of the poor, including access to and use of basic infrastructure and social services. Section C examines the labor market conditions of the poor. Finally, Section D assesses the relative role of different factors determining the probability of being poor. A. Measurement and Extent of Poverty (1) Methodological Approach 5. In this study, poverty is measured with the poverty line method4. Prior to this study, no poverty line had ever been constructed for Nicaragua. The poverty line is defined as the level of total Per capita monthly expenditures at which an individual obtains the minimum daily caloric requirement (2,226 cal. per adult).5 This method has two advantages over using the cost of a set food basket as the basis for the poverty line. First, it takes into account non-food items in which households have expenditures regardless their income level (such as housing, clothing, transportation). Second, it requires no assumptions about consumption preferences and patterns in the population. This is particularly important in Nicaragua, where the existing food basket is based on an income and expenditure survey done in the mid 1970s. 4. See detailed information and discussion of conceptual issues involved in constructing the poverty line in Volume 11, Poverty Profile, Appendix 3. 5. Note that this is not the cost of calories but a level of per capita expenditures sufficient to obtain the required caloric level.  6. The extreme poverty line is defined as the level of per capita monthly food expenditures required to obtain the daily minimum caloric requirement. Below this level of expenditures, individuals cannot maintain the needed level of caloric consumption, even if total income is spent on food. 7. Clearly it is important, a priori, to take into account regional price differentials. Unfortunately, accurate data on regional prices is not available. Using unit price data for food from the household interview data a regional price index was created. No non-food index could be created so both food and non-food items were adjusted by the regional food price index. The effects of this adjustment on national poverty estimates and regional rankings are minor, and hence will not have a significant effect on the findings6. Therefore, the remainder of the analysis does not include regional price adjustments. (2) Incidence and Distribution of Poverty 8. Half of the population, or about 2 million people fall below the poverty line. Forty percent of the poor, or 20 percent of the population (0.79 million people) fall below the extreme poverty line and are food poor. That is, they cannot meet the daily minimum caloric requirement even if they were to devote all of their consumption to food.7 The magnitude and distribution of poverty are described using three poverty indices8: (1) the head count index, which measures the incidence of poverty or the proportion of the population below the poverty line; (ii) the poverty gap index, which measures the depth of poverty by indicating the gap between the poverty line and the expenditures of the average poor; and (iii) the FGT P2 index, which measures the severity of poverty by weighting the poverty gap by the population at each level. Table 2.1: Poverty Lines, 1993 Current Cordobas US Dollars9 Percent of population Poverty Lines | (30 days) (annual) below each line Poverty Line 214.47 428.94 50.3% Extreme Poverty Line ] 101.32 [ 202.64 19.4% Note: Poverty lines are for per capita expenditures 9. Poverty and extreme poverty are overwhelningly rural. Three-quarters of all rural inhabitants live in poverty as opposed to 32 percent of urban dwellers (Table 2.2, panel [b]). Although only 41 percent of the total population lives in rural areas, 63 percent of all poor and 78 percent of all extremely poor people live in rural areas (Table 2.2, Panel [a]). 10. Poverty incidence varies by region. The Northern (Jinotega and Matagalpa) and Segovias (Esteli, Madriz, and Nueva Segovia) regions have 46 percent of all the extreme poor in the country although they represent only 23 percent of the country's population (Table 2.2, Panel [a]). In contrast, Managua with almost one third of the total population has only 7 percent of the country's extremely poor people (Table 2.2, Panel [a]). 6. See Volume II, Poverty Profile, Table 3. 7. The term "poor" includes the extremely poor unless specific differentiation is made between the "extremely poor' and "poor" categories. Population estimates are based on population projections of INEC based largely on the last census (1972), which may not be accurate for obvious reasons. 8. See Volume 11, Poverty Profile, Appendix 7 for a detailed discussion of these poverty measures. 9. The exchange rate at the time of the survey (November 1993) was at US$1 =C$6.  - 4 - 11. Poverty is deeper and more severe in rural areas. The gap between the poverty line and the average expenditures of the urban poor is 11 percent while it is 37 percent for the rural poor (Table 2.3, Panel [a]). In the poorest Northern and Segovias regions, the average rural poor spends about 48 percent less than the poverty line. Table 2.3, Panel [b] shows not only that there are more poor in rural areas, but also that their poverty is both deeper and more severe than in urban areas. Table 2.2: Contribution to Poverty and Head Count Index by Region and Geographic Area, 1993 (Percent of individuals) Region Panel [a] Panel [b] Contrib. to Nat'l. Poverty Head Count Index Share of Share of Share of Urban Rural Total Nat'l All Poor Extrem. Popul.__ _____________ Poor E. Poor Poor E. Poor Poor E. Poor Poor Nafl ~AllPoor Eotrem E T _ _ __ PO __ Segovias 9.7 14.4 19.1 21.5 59.3 52.0 87.6 38.2 74.8 Western 17.3 16.4 17.0 7.5 30.2 34.8 71.5 19.0 47.6 Managua 27.7 16.5 7.3 4.2 26.4 9.6 49.0 5.1 29.9 Southern 16.5 14.5 12.7 5.6 26.3 28.3 70.0 14.9 44.2 Central 6.6 8.8 9.6 13.1 48.7 31.6 78.0 24.6 67.0 Northern 13.2 18.7 26.6 9.1 37.9 52.2 85.9 39.0 71.2 Atlantic 9.0 10.8 9.1 8.0 35.4 30.3 82.9 19.6 60.6 National 100.0 100.0 100.0 7.3 31.9 36.3 76.1 19.4 50.3 Urban 58.3 37.0 22.0 7.3 31.9 Rural 41.7 63.0 78.0 36.3 76.1 Table 2.3: Poverty Gap and FGT P2 Measures by Region and Geographic Area, 1993 (%) Panel [a] Panel [b] Region Poverty Gap Measure FGT P2 Measure l____________ j Urban [ Rural Total Urban I Rural | Total Segovias 25.2 48.2 37.8 13.8 30.5 23.0 Western 11.1 34.6 21.0 5.3 21.0 11.9 Managua 7.9 18.4 9.5 3.4 9.0 4.2 Southern 8.0 29.9 16.9 3.4 16.4 8.7 Central 18.8 36.6 29.9 9.3 20.5 12.3 Northern 13.0 47.5 36.9 5.9 30.0 22.7 Atlantic 12.1 35.9 24.5 6.0 18.8 12.7 National 10.9 37.1 21.8 5.1 21.9 12.1 (3) The Index of Unmet Basic Needs 12. In 1983 and 1990, the GON used an index of unmet basic needs (NBI) to measure poverty. Based on results of a household survey (ESDENIC-85), the NBI defined a household to be poor according to the following indicators: (i) inadequate access to water or sanitation facilities if no toilet or latrine, or no access to public water source; (ii) crowding if four or more people slept in the same bedroom; (iii) low educational levels if at least one child aged 7-14 did not attend school; and (iv) high dependency ratio, if two or more individuals depended on a head of household with less than primary school. Households with one unmet basic need were deemed poor, those with two or more, extremely poor.  -5- Box 1.2: Limitations of the Unmet Basic Needs (NBI) Method The NBl can be an important complement to an expenditure measure as it helps to pinpoint those individuals neither able to meet their minimum caloric requirement nor obtaining basic services: the chronic poor [MAS 1994a]. However, using the NBI approach alone can lead to problems with targeting of social policy. First, the NBI measures, to a large extent, prior governrent investrnents in social services and infrastructure. Thus any person living in areas with limited or non-existent public investment will be considered poor, regardless of his or her current income levels, which actually determine his or her capacity to purchase goods and services. In Nicaragua, which suffered from over a decade of low investments, this leads to an overestimation of poverty. A second drawback is that the NBI approach preselects the key correlates of poverty. The definition of what constitutes basic needs is deternined a priori by policy makers or the statistical office. To the extent that there are omissions in basic needs and/or other critical components of poverty are overlooked, social policy and targeting will be less effective and, hence, more costly. Finally, there is no agreement on what constitutes basic needs, nor how much of a basic service should be considered adequate. As a change in definitions will have a significant impact on the results of using NBI, this tends to undermine its validity when used on its own. For example, in the case of Nicaragua, while the 1983 NBI study included the quality of housing in the index of basic needs, the 1990 and 1994 studies did not. Also, the definition of education and who should be in school has changed between NBI studies [MAS 1994b]. J 13. Based on the LSMS-93, the Ministry of Social Action (MAS) utilized the NBI method to determine current poverty levels in Nicaragua to aid the GON to target poverty reduction programs. According to MAS, 74.8 percent of Nicaraguan households are poor, and 43.6 percent of the poor are extremely poor [MAS 1994b]. However, the comparison of the NBI with the poverty line indicator used in this report shows potential problems with just using the NBI index (see Box 1.2). While those persons whose expenditures are below the poverty line have a greater incidence of NBI than the population as a whole, as a targeting mechanism the use of NBI alone involves major errors of inclusion and exclusion (Table 2.4). Table 2.4: Basic Needs and Poverty Line by Geographic Area, 1993 Poverty Line estimates Rural Urban Basic Needs Estimates Extr. Poor Non- Extr. Poor T Non- Poor | | Poor Poor j Poor % with Crowding 77.9 56.2 34.6 85.2 68.2 33.9 % with Inadequate services 67.7 55.0 31.9 37.7 28.2 14.2 % with Children 7-14 not in School 57.2 37.5 17.0 47.3 18.1 8.5 % with High Dependency Ratio 71.2 59.5 38.8 77.9 61.0 34.1 BNI % Extremely Poor 87.0 67.6 33.4 81.6 59.0 23.8 BNI % Poor 10.8 24.7 38.0 17.5 30.5 36.1 BNI % Non-Poor 2.2 7.7 28.6 0.9 10.5 40.1 Note: The table is in percentage of individuls  - 6 - 14. By targeting specific unsatisfied basic needs, a large proportion of the poor are not recognized. For example, the indicator "child 7-14 years not in school" identifies only 57.2 and 47.3 percent of those living in extremely poor households in rural and urban areas, respectively. Overall, the NBI method identifies as such 87 percent of the extremely poor in rural areas and 82 percent of the extremely poor in urban areas. In addition, using the NBI index or its components, a large number of non-poor would be included in poverty alleviation programs, which would not only increase programs' cost but divert resources from those most in need. For example, over a third of those classified as poor according to the NBI index both in rural and urban areas have expenditures above the extreme poverty line. Box 1.3: Nicaragua - Inequality' The estimated Gini coefficient for Nicaragua is 50.3, which indicates that overall levels of inequality are close to those of other Central American countries. In 1989, the Gini coefficient for Costa Rica (0.46) was lower than Nicaragua's while those of Honduras and Guatemala were above Nicaragua's at 0.59 [Psacharopoulos et al,. 1992]. See Volume II, Poverty Profile for detailed data on distribution of expenditures. (4) Consumption Patterns of the Poor 15. This section identifies the important goods in the consumption basket of the poor. Such information is essential to ascertain the impact of price and trade liberalization measures on the poor. Food is the major expenditure category for all poverty groups, although its share falls as poverty declines (Table 2.5). For the extreme poor, food expenditures represent 61.7 percent of total expenditures, for the poor 58 percent, and for the non-poor 42.6 percent. Table 2.5: Distribution of Total Per Capita Expenditures by Poverty Group and Geographic Area, 1993 (Percent of total expenditures) I Extreme Poor | Poor [ Total Poor | Non-Poor |___________ |____ Urban Rural ] Urban Rural | Urban [ Rural U Ubn Rural | Total Food 50.6 64.8 50.7 64.3 50.7 64.6 40.1 52.7 51.1 Housing 32.7 21.5 25.9 14.4 27.5 17.8 22.3 12.9 20.9 Education 3.4 1.6 5.1 2.7 4.7 2.1 6.7 3.9 4.6 Health 2.2 2.1 2.6 3.1 2.5 2.6 3.9 3.9 3.2 Personal Goods 4.5 4.2 6.9 7.1 6.3 5.7 10.5 11.7 8.3 Household Goods 4.1 4.3 4.1 3.9 4.1 4.1 4.6 3.5 4.3 Transportation 0.8 0.2 1.5 1.4 1.3 0.9 4.1 4.8 2.6 Services 1.4 1.1 2.0 2.1 1.8 1.6 4.2 3.6 2.9 Transfers 0.4 0.1 1.3 1.1 1.1 0.6 3.6 2.9 2.1 Total Expenditures 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: See Volume 1, Poverty Profile, Appendix 6 products/services included in each category, and Appendix 8 "totals" for each poverty group. 16. Among the poor, the largest share of food expenditures is on basic grains (rice, corn, beans, tortillas and bread). They absorb over half of the food expenditures of the extreme poor and half of the food expenditures of the poor. Therefore, price increases in basic grains will have a significant impact  on household budgets, particularly on the poor. For example, if the price of all basic grains rose by 10 percent, the food expenditures of the extremely poor would rise nearly by 6 percent while those of the non-poor would increase by 3 percent (assuming the same quantities were purchased). However, to the extent that the poor are involved in the production of basic grains, price increases will benefit them. Table 2.6: Distribution of Monthly Per Capita Food Expenditures by Poverty Group and Geographic Area, 1993 (Percent of total food expenditures) Extreme Poor Poor Total Poor Non-Poor l Urban I Rural Urban I Rural Urban Rural Urban Rural Total Basic Grains 54.7 56.8 44.2 51.4 46.6 54.0 29.1 43.2 41.7 Sugar, Oil, Salt 18.8 19.5 17.2 17.4 17.6 18.4 12.4 13.9 15.4 Dairy 8.4 8.9 15.1 11.6 13.6 10.3 18.7 13.4 14.6 Meat 2.7 2.0 7.5 6.1 6.4 4.1 17.9 11.2 10.7 Vegetables 5.6 3.2 5.9 4.6 5.8 3.9 7.7 6.5 6.0 Fruit 2.2 2.9 3.7 3.7 3.4 3.3 5.5 4.3 4.3 Beverages 4.6 5.4 5.6 4.8 5.4 5.1 7.1 6.3 6.1 Total Food 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: Based on 30-day expenditures. Basic Grains = rice, corn, tortillas, bread and beans. Meat = red meat, poultry, fish and other meats. Beverages = non-alcoholic and alcoholic beverages. 17. For all groups, the second most important expenditure is housing."0 For the poor, food and housing expenditures take up approximately 80 percent of total expenditures compared to 65 percent for the non-poor. In short, for the poor, very little money is left over for all other expenditures. (5) Sources of Income of the Poor 18. Table 2.7 identifies the main sources of income of the poor and indicates that: (a) working in the informal sector is not associated with poverty; (b) there is a strong correspondence between poverty level and the share of agricultural income in total household income. In rural areas (after adding own production to wage or self-employed agricultural income), the extreme poor and the poor derive over 75 and 80 percent of total household income from the agricultural sector; (c) in rural areas, the importance of off-farm earning sources in total income increases as poverty levels decline. Twenty percent of the income of the extremely poor comes from off-farm employment, compared to 33 percent for the poor, and 36 percent for the non-poor. In urban areas, wages are the main source of income for the poor. The non-poor tend to rely more on income from self-employment; and (d) transfers from gifts and charities, which include transfers from government and donor programs are mostly received by non-poor households, suggesting that prevailing targeting of assistance programs is quite ineffective. Only 12.5 percent of the value of total transfers in gifts and charities go to the poor. 10. See Volume II, Poverty Profile, Appendix 3, for details on measurement of housing expenditures.  - 8 - Table 2.7: Household Income Sources by Poverty Group, and Geographic Area, 1993 (Percent of household income) Sources of income Extreme Poor Poor Non-poor Urban Rural Urban Rural Urban Rural Labor, of which: 67.7 47.6 69.9 53.1 66.2 50.9 Wage-agriculture 6.2 21.0 4.8 13.0 1.9 5.3 Wage-non-agriculture 38.9 12.9 37.7 19.8 33.8 18.4 Self-agriculture 1.1 6.1 0.9 7.2 1.8 9.3 Self-non-agriculture 21.5 7.6 26.4 13.0 28.7 17.9 informal sector 37.6 24.2 36.1 32.5 34.8 31.4 Own production 7.6 27.7 3.2 26.7 1.7 15.1 Imputed Rent 17.4 16.7 14.5 11.6 12.1 7.6 Non-labor income of which: 7.2 7.6 12.4 8.6 19.1 26.4 remittances 3.9 2.2 5.5 2.3 4.6 3.1 gifts, charity 1.4 2.5 1.9 2.9 7.2 15.9 pensions, returns to capital 2.0 2.9 5.1 3.2 7.3 7.4 Total Income 100.0 100.0 100.0 100.0 100.0 100.0 aote: Totals may not add up to lUU due to rounding. Labor income = all income obtained from primary, secondary an third jobs. Non-labor income = all income obtained from domestic and foreign remittances, gifts and charities, pensions, alimonies, insurance, interests on savings, stock dividends, rents, indemnities, lottery receipts, and loans. Informal sector = all income from wage or self-employed workers, employers in private sector businesses of 5 or fewer employees. See volume 11, Household Sources of Income, for detailed description of methodology used to calculate total income and specific income categories. B. The Social Conditions of the Poor (1) Characteristics of Poor Households 19. Larger households are poorer (Table 2.8). Poor households have 2 to 3 more members than non-poor households, mostly children under the age of 12. As a result, dependency ratios among the poor are between 50 and 100 percent higher than among the non-poor. Heads of poor households have half the education of their non-poor counterparts (3.3 and 1.4 years of schooling in urban and rural areas, respectively). Less education results in a lower earnings capacity for poor households and their larger family size means less income per capita. Table 2.8: Characteristics of Households by Poverty Group and Geographic Area, 1993 (Means and Percents) Extreme Poor Poor Total Poor Non-Poor l Urban [ Rural Urban Rural Urban Rural Urban Rural Total Mean size 8.0 7.2 6.5 5.8 6.8 6.4 4.8 4.5 5.5 Dependency ratio 1.4 1.3 1.1 1.1 1.1 1.2 0.7 0.8 0.9 Mean age: Head 46.2 45.0 44.5 42.7 44.8 43.7 44.3 45.3 44.3 Mean yrs. school: Head 2.3 0.9 3.6 1.7 3.3 1.4 6.3 3.3 4.1 No. of workers 1.9 2.0 1.9 1.8 1.9 1.9 1.8 1.7 1.8 % HH w/Fem. Head 37.7 19.5 36.2 14.9 36.5 16.8 34.3 23.3 28.5 Note: DeDendency ratio: number of 0-12 year olds + over 60 year olds divided by all other household members. No. of workers: mean number of household members who are economically active (employed and unemployed).  - 9 - 20. There is no evidence that female-headed households are at a greater risk of poverty (Table 2.9). Although female heads have less schooling than male heads and are less likely to be in the labor force, female headed households have, on average, the same number of workers as male headed households. Table 2.9: Characteristics of Female Headed Households by Poverty Group and Geographic Area, 1993 (Means and Percents) Extreme Poor Poor Total Poor Non-Poor l Urban | Rural Urban Rural Urban Rural Urban Rural 1 Total Mean size 8.2 6.8 6.7 5.8 7.0 6.3 4.7 4.4 5.4 Dependency ratio 1.4 1.4 1.1 1.3 1.2 1.4 0.8 0.8 1.0 Mean age: Head 50.0 47.5 47.3 49.4 47.9 48.4 48.0 53.2 48.6 Mean yrs. school: Head 2.0 0.7 2.9 1.7 2.7 1.2 5.0 2.6 3.7 No. of workers 2.1 2.0 1.8 1.8 1.9 1.9 1.7 1.6 1.8 Participation rate 49.0 33.6 56.2 39.5 54.8 36.6 62.6 49.5 55.6 % of all fem. Head HH 3.8 7.9 15.4 8.2 19.2 16.1 54.7 9.9 100.0 (2) Housing and Basic Social Infrastructure (water, sanitation, power) 21. Every indicator in Table 2.10 reflects the huge deficit in basic infrastructure, and depicts the difficult living conditions of Nicaraguans in general. The poor, however, particularly those in rural areas, are worse-off than all other groups. 22. Reflecting the country's housing deficit, housing expenditures are the second most important expenditure across poverty groups (para 17). They represent almost one third of total expenditures of the extremely poor in urban areas (Table 2.5). Across poverty groups, a large proportion of households does not have a legal title to their dwelling. Lack of legal ownership has major implications for investment as it limits access to credit (mortgages are the most common collateral required to obtain credit). The poor, and especially the rural poor, will have the most limited access as they have the lowest levels of ownership with legal titles. 23. Overall coverage of water is low, especially for the poor and rural areas. Less than 60 percent of the population has running water inside or outside their homes. The disparity in coverage between rural and urban areas is quite large. Almost 80 percent of the rural poor compared to one-quarter of the urban poor do not have access to running water. Lack of access to clean water is a major factor explaining diarrhea as the main cause of infant mortality in the country (para 46). 24. Coverage of sanitation services is generally low, but lowest for the poor and rural areas: 32.3 percent of the population has access to sewers, septic tanks or cesspools. Rural areas, in general, have no access. The urban poor have much more restricted access to sanitation services than the non-poor: almost 84 percent of the extremely poor have no access to such services compared to only 35.3 percent of the non-poor. Adding to the above health hazards, over half of the extremely poor in rural areas, one- third of the other poor, and, surprisingly, one-fifth of the non-poor have neither toilets nor latrines.  - 10- Table 2.10: Characteristics of Housing by Poverty Group and Geographic Area, 1993 (Percent of Individuals) l__ IUrTotal Poor Non-Poor l l ~~~~~~~~~Urban Rural Urban |Rural Total l Independent housing unit 92.4 76.6 95.9 88.1 88.4 Have legal title to housing 58.2 49.3 63.5 54.7 57.2 No piped water (inside or outside the house) 23.7 79.1 11.0 61.3 39.6 No sewer connection 71.5 98.4 35.3 92.2 67.7 Neither toilet/latrine 5.8 43.6 1.2 20.5 17.4 Inadequate trash disposal 12.8 34.3 5.5 16.5 17.1 Light: electricity 83.8 29.9 97.3 61.5 69.8 Light: kerosene 13.8 64.1 1.5 36.6 27.1 Use wood for cooking 84.2 99.0 47.9 89.8 75.0 Note: No sewer connection means the dwelling is neither connected to the public sewer system, nor has a septic tank or cesspool. 25. At 70 percent coverage, access to electricity is 20 percent higher than that estimated by the State Power Company, INE". Urban areas have the highest coverage. The lower uses of electricity by the extremely poor and other poor in urban areas is probably a function of the cost of this service as well as the concentration of the poor in new, marginal neighborhoods which are not yet connected to the public utilities. Interestingly, wood is the preferred fuel for cooking across poverty groups, which has implications both for time use of women and environmental degradation. (3) Access to Basic Social Services 26. This section examines the poor's access to education and health services, and the extent of child malnutrition. It is widely recognized that education plays a major role in an individual's ability to avoid poverty or to move out of it. Poor health and malnutrition are major consequences of poverty and limit productivity. The coverage of health and education services in Nicaragua compares well with most Central American countries [World Bank, 1993b]. To a large extent, this resulted from the GON's commitment to provide education and health services to the population in the 1980s. However, as shown below, major shortcomings exist regarding access and quality of primary health care (PHC) for the poor, and internal efficiency and quality of primary education. (a) Education Services 27. Poverty, illiteracy, and low educational attaimnent are strongly correlated. Over half of the extremely poor in rural areas and over a third in urban areas are illiterate. (Table 2. 11). The expansion of schooling in the 1980s increased the education levels of the younger cohorts across poverty groups and gender. However, overall school attainment remains low (Table 2.12). On average, a Nicaraguan has received 4.5 years of schooling. The extremely poor have received much less schooling: 3 years in urban areas and only 1.6 years in rural areas. 11. The reason for the disparity in estimates may be that the number of illegal connections is quite high. INE estimates that 23 percent of electric production is lost due to technical deterioration, tariffs not reflecting real costs as well as defective meters and illegal connections [World Bank, 1993aJ. LSMS data indicate that illegal connections may well represent a significant proportion of this loss.  - 11 - Table 2.11: Illiteracy by Gender, Poverty Group and Geographic Area, 1993 (Percent) Extreme Poor | Poor I Total Poor Non-Poor |o l Urban Rural Urban Rural Urban Rural U Total Illiteracy Rate 31.0 50.7 16.3 41.4 19.5 45.8 8.5 25.6 23.0 Male Illiteracy 27.6 49.9 13.9 42.9 16.8 46.2 7.1 24.3 22.6 Female Illiteracy 34.2I.6 T 18.6 40.01 22.0 4.4 9.6 1 27.0 23.5 Note: Individuals age 10 and up. Illiteracy is defined as being unable to read a book, letter, newspaper or magazine and unable to write a letter or message. Data are self-reported. 28. As shown in Table 2.13, access to primary school is alnost universal. A net enrollment rate of over 90 percent is very high compared to other Central American countries [World Bank, 1993b]. It is only among the extreme poor in rural areas that a sizable percentage of primary school age children (6-12 years) do not attend school (12 percent). However, students' educational progress through the school system is highly deficient, and affects the poor most. Fully one-quarter of all secondary age students are still in primary school. As a direct consequence of the poor's failure to complete primary education, their enrollment in secondary school is half that of the non-poor. For the same reason, enrollment at the university level is circumscribed to the non-poor. Moreover, attendance in pre-school programs, which is particularly valuable for developing learning skills among disadvantaged children, is low and largely caters to the urban non-poor. Table 2.12: Mean Years of Schooling by Gender, Age Poverty Group and Geographic Area, 1993 (Mean years) Extreme Poor Poor Total Poor | Non-Poor Urban Rural Urban | Rural Urban Rural Urban Rural Total National 3.0 1.6 4.4 2.4 4.1 2.0 6.5 3.8 4.5 Male 3.0 1.5 4.4 2.3 4.1 1.9 6.7 3.7 4.5 Female 2.9 1.6 4.4 2.5 4.1 2.1 6.4 3.9 4.6 Ages 10-19 2.9 1.9 4.4 2.8 4.1 2.3 5.7 3.8 4.1 Ages 20-29 4.2 2.3 6.1 3.3 5.7 2.9 8.4 5.3 6.1 Ages 30-39 3.5 1.3 4.9 2.1 4.5 1.7 8.3 4.4 5.6 Ages 40-49 2.1 0.7 3.2 1.5 3.0 1.1 6.3 3.1 4.0 Ages 50 and up 1.1 0.4 1.9 0.7 1.7 0.6 3.4 1.4 2.2 Note: Includes all people age 10 and up. 29. These findings are a matter of concern as the poor rely exclusively on public education (Table 2.14) and are spending significant amount of resources in sending their children to school. Among poor households, average expenditures on education represent 7.8 percent of the poverty line. At the same time, the few available Government scholarships (1.5 percent of students at the national level) are largely received by the non-poor.'2 12. See Volume II, Poverty Profile, detailed data on education expenditures (Tables 12, 14, 15).  - 12 - Table 2.13: School Enrollment by Age Group, Education Level, Poverty Group and Geographic Area, 1993 (Percent of age cohort) Extreme Poor Poor [TotalPoor] Non-Poor Age Groups Urban Rural Urban Rural Urban Rural Urban] Rural [Total Ages 0-5 In preschool 6.7 2.2 14.3 5.7 12.3 4.0 23.8 11.4 12.6 Ages 6-12 Not in school 7.7 12.2 3.8 6.0 4.6 8.8 2.0 1.1 4.4 Primary 92.3 87.5 92.6 92.4 92.5 90.2 91.2 96.2 91.7 Secondary . 0.3 3.7 1.0 2.9 0.6 6.7 2.7 3.8 Mid-level tech. . . . 0.7 . 0.4 0.1 . 0.1 Ages 13-18 Not in school 47.3 62.5 26.8 51.7 30.9 56.7 19.4 46.1 34.1 Primary 29.9 30.9 31.1 31.6 30.8 31.3 18.1 29.1 25.2 Secondary 22.8 6.6 41.9 16.3 38.0 11.8 60.9 24.9 39.8 Mid-level tech. . . 0.1 . 0.1 . 0.2 . 0.1 University . . . . . . 1.1 . 0.5 Adult ed. . 0.1 0.1 0.3 0.1 0.2 0.3 0.9 0.2 Ages 19-22 Not in school 90.7 93.3 75.9 92.6 78.9 92.9 63.3 84.3 74.9 Primary 1.2 3.7 1.3 1.4 1.3 2.3 2.1 3.0 2.1 Secondary 6.8 3.1 22.2 5.9 19.0 4.8 25.7 11.0 18.5 Mid-level tech. 1.2 . . . 0.3 . 0.5 . 0.3 University . . 0.6 . 0.5 . 8.4 1.8 4.3 Note: Includes private, public and subsidized schools. Preschools include MOE and FONIF (previously INSSBI) programs. indicates no students in category. Table 2.14: Public School Enrollment by Level, Poverty Group and Geographic Area, 1993 (Percentage of Students) Extreme Poor Poor [ Total Poor Non-Poor Urban] Rural Urban I Rural Urban Rural Urban Rural Total Preschool 86.8 100.0 85.4 89.4 85.6 92.3 52.0 68.3 62.5 Primary 98.6 99.7 93.6 98.4 94.6 99.0 78.9 94.9 89.7 Secondary 100.0 97.7 86.0 88.9 87.5 91.1 69.5 81.1 75.1 Mid-level tech. . . 34.2 75.0 25.5 75.0 43.6 44.4 47.2 University . . 100.0 . 100.0 . 71.4 100.0 73.8 Note: "." indicates no students in category. (b) Health Services"3 30. This section assesses the access of the poor to health care services by examining LSMS data on use of health care services and health expenditures in the 30 days prior to the survey. Although official figures estimate that the health care system potentially reaches 70 percent of the population, actual household level data suggest that the poor, particularly in rural areas have significantly less access to health care services than the rest of the population. 13. Source [J. Fiedler 1994], and Volume II, Poverty Profile.  - 13 - 31. In 1989 both diarrhea and respiratory infections were the leading causes of mortality among young children"4. Today, infections and communicable diseases along with diarrhea are still the most common ailments. Over 18 percent of all children under the age of 5 reported an incidence of diarrhea in the 30 days prior to the survey and the vast majority of the 34 percent of children reporting other illness suffered from respiratory ailments. Of those who reported feeling ill, including children under 6 with diarrhea, the majority did not seek care. Overall 23 percent of those not seeking care when ill gave distance to the medical center and the high cost of care, direct proxies for access, as the reasons for not seeking care, as did 18 percent of those with diarrhea. However, among the rural poor distance and cost were a deterrent for 34 percent of all those not seeking care and 28 percent of children with diarrheal disease. 32. Table 2.15 shows consultation, coverage, and concentration rates, which are conmmonly used as indicators of the adequacy and performance of a health care system. In each of these indicators, the extreme poor and the poor show greater barriers to access than the non-poor. Table 2.15: Consultation, Coverage and Concentration Rates by Poverty Group, 1993 Poverty group Consultation rates Coverage rates Concentration rates (% of ill) (% of ill) l Extreme Poor 35.7 26.2 1.36 Poor 55.5 38.1 1.42 Non-Poor 91.5 51.5 1.78 Note: consultation rates = total number of consultations/persons reporting illness. Coverage rates = persons with at least one consultation/persons reporting illness. Concentration rates = total number of consultations/number of persons with at least one consultation. 33. Consultation rates show that a non-poor person who perceives him/herself to be ill is three times as likely to obtain care compared to an extremely poor person who perceives him/herself to be ill. Coverage rates show that the proportion of non-poor receiving at least one consultation is twice as high as that of the extremely poor. Finally, the mean number of consultations per person is also lower for the extreme poor than for the non-poor. 34. Table 2.16 presents the same indicators by region, reinforcing the inverse relationship between poverty and access to health care services. The poorest regions in the country, particularly the Northern region, suffer disproportionately from differential access to health care services. Table 2.16: Consultation, Coverage and Concentration Rates by Region, 1993 Region Consultation rates Coverage rates Concentration rates (% of ill) (% of ill) I Segovias 60.8 36.7 1.66 2 Western Region 70.6 43.8 1.61 3 Managua 80.1 47.8 1.68 4 Southern 101.0 53.7 1.88 5 Central 63.1 39.9 1.58 6 Northern 44.2 31.9 1.39 7 Atlantic Coast 56.3 38.8 1.45 14. According to MOH estimates, acute diarrheal disease was responsible for 40 percent of infant mortality and, for one to four year olds, the rate of mortality from acute respiratory infections was 30 per 100,000 [World Bank, 1993b].  - 14 - (c) Child Malnutrition'5 35. Almost a third of Nicaraguan children are malnourished, and most malnourished children are poor. Three malnutrition indicators were estimated for children aged 0-5 years old. The first is height for age (HFA), a proxy for chronic malnutrition. The second is weight for age (WFA), a proxy for underweight and the third is weight for height (WFH), a proxy for acute malnutrition. Of all 0 to 5 years old, 28 percent suffer from some form of malnutrition. Children in extreme poverty are at the greatest risk, particularly in rural areas. Four out of ten extremely poor children living in rural areas are malnourished. Child malnutrition is not a recent phenomenon as indicated by the predominance of chronic malnutrition. Table 2.17: Malnutrition Among Children under Five by Poverty Group and Geographic Area, 1993 (Percent) Extremely Poor Poor Non-Poor Malnutrition Indicator Urban I Rural Urban I Rural Urban I Rural Chronic (HFA) 28.0 41.1 18.1 31.0 12.7 19.4 Underweight (WFA) 18.6 18.6 10.4 15.4 5.7 8.8 Acute (WFH) 4.6 1.8 1.7 2.3 1.1 1.0 Total (any form) 34.5 44.4 22.0 35.1 14.3 22.1 Note: Chronic undernutrition is defined as height-for-age < -2 SD from the NCHS gender-specific reference median. Underweight is defined as weight-for-age < -2 SD from the NCHS gender- specific reference median. Acute under-nutrition is defined as weight-for-height < -2 SD from the NCHS gender-specific reference median. Table includes all children ages 0 to 59 months. 36. There are significant regional differences in the extent of malnutrition. The rural areas of the Northern and Segovias regions, the two areas with the worst poverty (Table 2.3) have the highest prevalence of malnutrition (Figure 1.1). The concentration of malnourishment in the poorest regions indicates that location should be an important consideration for targeting nutrition programs. Figure 1.1: Percentage of Malnourished Children by Region and Geographic Area en- x 40 -~~~~~~~~~~~4. a 0 Sag. West. Man. South. Cent. North. AtUan. Region |Urban Rural | Soufl: Phwz, 1993. LSLS, a, 1993, Maie: Periw mahoubhed hndude ia chidran wlany kwnn of nainAritkn 15. Source [E. G. Piwoz 19941.  - 15 - C. Employment and Povertyt6 37. As the primary source of household income, earnings are a critical determinant of poverty. This section focuses on the employment status of the poor. Specific attention is paid to: (i) labor force participation and the relationship between unemployment, underemployment and poverty, (ii) where the poor work, and (iii) the extent to which low earnings, and, hence poverty arise from labor market segmentation as opposed to low human capital. (1) Labor Force Participation 38. Unlike other countries, in Nicaragua, labor force participation rates for both males and females have not increased during the adjustment period"'. Participation rates are lower for the poor than for the non-poor, but most of the difference is due to the lower participation rates of poor women. And poor women participate less in the labor market because they face high demand for household work stemming from a large number of children coupled with general lack of daycare, lack of access to basic services and the ensuing need to collect wood for cooking and fetch water for drinking (Table 2.10). Table 2.18: Labor Force Participation and Unemployment Rates by Sex, 1993 (%) Extreme Poor Poor Total Poor Non-Poor | Urban Rural Urban Rural Urban Rural Urban R| _ Rua{l Participation Rates Men 61.6 77.6 61.2 78.7 61.3 78.2 64.8 76.1 69.4 Women 23.9 15.9 32.7 21.7 30.9 19.0 41.6 33.0 32.8 Both 42.3 47.7 46.5 50.3 45.6 49.1 52.3 54.9 50.3 Unemployment Rates Men 14 11 23 6 21 8 17 11 14 Women 21 12 17 10 18 11 12 11 12 Both 16 11 21 7 20 9 15 II 13 Note: L.F.P. = percentage of the working age population (12 years and older) employed or unemployed. Unemployed = all those who did not work in the 7 day reference period but actively looked for work during the period. (2) Education and Training 39. Education. Table 2.19 shows the educational attainment of the Nicaraguan labor force. A number of poverty-relevant features emerge. First, women participating in the labor market are more educated than their male counterparts. This shows that for women to participate in the labor force they must have a comparative advantage relative to household work. Second, and contrary to what is commonly believed in Nicaragua, the unemployed have more schooling than the employed: 5.3 years of schooling. It is the underemployed who have the lowest educational attainment of all: just 2.7 years of schooling. 16. For details see Volume II, Poverty Profile and [Carola Pessino, 1994a]. 17. In 1989 labor force participation was 67.9 percent for males and 32.5 percent for females. For experience in other countries see [A. Cox-Edwards and J. Roberts, 19941.  - 16 - Table 2.19: Average Years of Schooling of the Labor force, 1993 Mean years schooling Total Males Females Total 4.1 4.1 4.1 Employed 4.8 4.4 5.6 Underemployed 2.7 2.5 3.9 Unemployed 5.3 5.3 5.4 Out of L.F 3.1 3.1 3.2 Note: Employed = all those who wored for a minimum of one hour in the 7 days prior to the interview whether paid or unpaid. Underemployed = person working less than full time and who looked for another job in the 7 day reference period. 40. Training. Only 16.5 percent of the employed report having taken any kind of occupational training, which seems low compared to other Latin American countries'8. Given the changes underway in the Nicaraguan economy, it was expected that more people would take training to align their skills to the needs of a market economy. In Nicaragua, however, only 5.3 percent of the employed took some training since April 1991, suggesting that adjustment in terms of productivity growth has yet to happen or, less likely, that Nicaragua has enough skills to meet the change. 41. There is a strong positive correlation between schooling and occupational training, indicating that formal education and training are complementary and not substitutes. That is, the higher the level of schooling, the higher the probability of training: only 1.2 percent of those without schooling took training, 5.6 percent of those with incomplete primary took training against 40 percent of those with a secondary education or more. The likelihood of training increases for each successive educational level, peaking for those with complete secondary education. Even after controlling for educational attainment, the non-poor and those living in less poor regions are those more likely to take training. As the majority of the poor have no schooling or less than primary education (Tables 2.11 and 2.12), and occupational training requires a certain amount of formal schooling, these findings imply that occupational training is not an effective strategy for helping people to move out of poverty. (3) Unemployment and Underemployment 42. At 13.4 percent, the overall unemployment rate is not too high compared to other Latin American countries during adjustment [World Bank, EDI, 1994]. In Nicaragua, unlike most countries, female unemployment is not higher than male unemployment, probably because only women with high advantages in the labor market participate (para 61). 43. Unemployment is not strongly correlated with poverty, but underemployment is. National unemployment figures mask huge variations at the regional level, where the less poor the region, the higher its unemployment rate. As shown in Table 2.20, while labor force participation rates are similar throughout the country, unemployment rates go from 18.5 percent in Managua and the Western regions, to 3.2 percent in the Atlantic and 9.5 percent in the Segovias regions. The three regions that comprise over 80 percent of the country's unemployed (Western, Managua and Southern) have the lowest incidence of extreme poverty. Multivariate analysis of the determinants of unemployment indicate that even after controlling for individual characteristics, urban residents in these three regions are more likely to be unemployed. 44. National underemployment estimates also mask huge variations at the regional level: the poorer the region the higher its underemployment rate. This might indicate that the poorer the region, the lower 18. [Heckman and Hotz, 19861 [Arriagada and Stelcner, 19871  - 17 - workers' reservation wages and hence the higher the probability they will accept any job rather than remain unemployed. That is, they might be too poor to afford to be unemployed. Table 2.20: Labor Force Participation, Unemployment, Underemployment, and Poverty by Region, 1993 (percent) Region Labor force Unemployment Under Region's extreme participation rate Region Country employment rate poverty headcount I Segovias 49.3 9.5 5.7 26.9 38.2 11 Western 49.6 18.5 20.3 10.8 19.0 III Managua 54.3 18.5 47.5 10.0 5.1 IV Southern 51.1 16.7 18.5 14.1 14.9 V Central 49.2 4.8 3.3 14.4 24.6 VI Northern 48.2 3.6 3.5 24.9 39.0 Vn Atlantic 48.3 3.2 1.2 27.3 19.6 Total 50.3 13.4 100.0 15.4 19.4 Note: Definitions under Tables 2.18 and 2.19. 45. Underemployment exhibits three additional features relevant from a poverty perspective. First, it is almost 2.5 times higher for men than for women (18.7 percent versus 6.6 percent), which signals the lower costs for men of moving to areas providing seasonal work, and is consistent with higher unemployment among men. Second, by region and geographic area, underemployment is inversely correlated with unemployment. The wealthier regions have at the same time the highest proportion of unemployed and the lowest proportion of underemployed. Having more wealth allows people to search for better jobs. This increases the duration of unemployment, but raises the probability of finding a better job, and therefore results in lower underemployment. Third, the highest rates of underemployment are found in rural areas where employment in the agricultural sector predominates. Almost three quarters of all underemployed are found in rural areas. Table 2.21: Underemployment Rates by Region and Geographic Area, 1993 (%) | Region | Underemployment Rate I Urban Rural Total I Segovias 14.4 35.6 26.9 H Western 5.6 17.3 10.8 m Managua 5.6 28.3 10.0 IV Southern 8.7 23.7 14.1 V Central 6.7 17.9 14.4 VI Northern 7.6 31.1 24.9 Vll Atlantic 10.5 36.7 27.3 Total 6.9 26.5 15.4 (4) Sector of Employment 46. Agriculture is the main sector of employment; one third of all workers are in this sector. And there is a strong correlation between agricultural sector employment and poverty (Table 2.22). In rural areas, almost three-quarters of all poor workers are involved in agriculture. Even with a greater diversity of sectors of employment and a higher proportion of people in wage employment in urban areas, a high  - 18 - proportion of urban workers still work in agriculture and, as in rural areas, the greater the proportion of workers in agriculture, the greater the level of poverty. Table 2.22: Distribution of Employed by Economic Sector, Poverty Group and Geographic Area, 1993 (Percent) Total Poor Non-Poor Sector Urban [Rural Urban Rural Total Agriculture 16.2 73.4 5.7 42.8 31.3 Mining 0.5 0.2 0.3 0.0 0.2 Manufacturing 16.5 4.6 18.1 7.7 12.7 Elect., gas, water 1.2 0.3 1.7 0.5 1.1 Construction 5.9 1.8 3.5 2.0 3.2 Commerce 21.7 7.8 31.3 23.4 22.1 Transport 3.4 1.1 5.4 2.7 3.5 F.I.R.E. 1.2 0.2 3.2 0.3 1.7 Services 33.0 10.5 30.3 20.1 23.8 Note: Table refers to primary job. Includes all those ages 12 and older who are employed. "F.I.R.E." = finance, insurance and real estate. (a) Agricultural Sector Employment'9 47. Within agriculture, there are two groups of rural poor in Nicaragua: the self-employed or small farmners, and the landless agricultural wage laborers. Self-employed households are those that farm agricultural land and have one or more members working in agriculture. Landless households have at least one member in agricultural wage employment, but do not own or cultivate any agricultural land. As shown in Table 2.23, landless households are significantly less poor than small farming households. Important factors explaining such a difference are the level of education of the household head and dependency ratios. Agricultural policies have only recently started to address the issue of low productivity of small farmers through the expansion of research and extension services (para 107). In the design of these programs, the low levels of education among small farmers must be considered. Finally, farming female headed households are not found to be at higher risk of poverty. 48. Farming households. Poorer farming households have lower productivity than non-poor farming households. Productivity in agriculture is affected by the amount of land under cultivation, the crops being planted, and the mix of inputs. Poorer farming households use less land than their non-poor counterparts. Interestingly, the difference in holdings between the poor and non-poor does not seem very large (on average < 4 has. versus > 6.5 has.). Unfortunately, we did not control for other variables such as land quality or location. The majority of the land used by the poor is devoted to producing food crops (primarily basic grains). Typically, corn and "millon" sorghum (for human consumption) are produced as subsistence crops and beans as a cash crop by the poor. Corn alone, represents over 40 percent of the area planted by the poor. The non-poor devote a greater share to export crops (tobacco, sugar and coffee) and to raising livestock (Table 2.24). 19. Note that a specific agricultural production module was not included in the LSMS-93 because: (i) an agricultural census was planned for 1993; and (ii) ownership of agricultural land questions were deemed highly sensitive given the property rights situation at the time. Thus, the LSMS only collected data on the use of agricultural land by households, regardless of ownership.  - 19 - Table 2.23: Comparison of Agricultural Households with and Without Agricultural Land, 1993 (means) Self-employed WI Characteristics Small Farmers Landless, Agr. Wage Years of Education of Head 1.9 2.4* Age of head (years) 43.8 42.7 Percent heads who are female 12.5 24.5** Percent households living in urban areas 16.9 33.9** Percent household workers in agriculture 76.7 86.7** Monthly per capita expenditures (cordobas) 225.4 274.8* Percent households with running water 14.6 27.6** Dependency ratio 1.2 .9* Notes: * -significant at .05 level of contidence, ** -significant at .01 level of confidence. Table 2.24: Distribution of Crops by Poverty Group and Geographic Area, 1993 (Percent of Total Hectares) |Extreme Poor | Poor R Total Poor Non-Poor UUrban [ura Ub Rura Ub Rural Total Food Crops 70.2 79.3 69.8 73.1 69.9 75.8 52.3 64.1 70.8 Export Crops 2.7 3.4 6.5 3.6 5.2 3.5 8.9 4.1 4.2 Livestock 15.4 10.4 9.6 11.3 11.5 10.9 25.1 14.1 12.9 Other Crops 11.8 6.9 14.2 12.0 13.4 9.8 13.8 17.7 12.1 Mean Hectares Used 2.4 3.3 5.1 4.1 4.2 3.8 24.0 6.5 6.3 Note: Mean hectares refers to total per household. Fiigures tor eiac crop represent the average percent of total hect.ares use for that crop. Note that all figures are for agricultural land used by the household, regardless of ownership. 49. All indications are that the poor are the most likely to be rationed out of the credit market. As shown in Table 2.25, poor households receive far fewer loans than non-poor households and the loan amounts are also strikingly lower. Given interest rate and subsidized credit policies of the past decade, the poor's lack of access to credit is probably not a new phenomenon.0 If this is the case, low productivity among poor agricultural households is probably not new. Table 2.25: Loans Received by Agricultural Households in Previous 12 Months, 1993 (Percent and means) Exlreme Poor Poor J Total Poor Non-Poor _ _ _ _ _ _ _ Urban Rural Urban Rural Urban Rural Urban | Rural Total Percent of Households Receiving 0.0 2.4 4.1 3.4 3.0 3.0 13.6 |6.9 4.7 Loan(s) _ __ _ _ _ _ _ Annual Amount of Loan . 897 3,533 4,827 3,533 | 3,460 29,367 | 14,336 13,368 Ndote: Table refers to all househo ds using agricuitural iana. Loans reter to all loans received by the househoid in the previous I months, be they for agricultural purposes or not. Mean loan amounts in current cordobas. '. ' indicates no loan in category. 20. While no evidence exists for Nicaragua, the experience in other countries is that subsidized interest rates and interest rate ceilings lead to credit rationing and result in the small borrower being squeezed out of the formal credit market. See [Vandell, 19841, [Wolken and Navratil, 1981] and [Anderson and Khambata, 1985].  - 20 - (b) Fonnal and Informal Sector Employment 50. Informal sector employment21 tends to swell during adjustment, and Nicaragua has not been an exception. It has been increasing at an accelerated pace, at least in Managua, since 1991. Prior estimates by INEC and FIDEG show that the informal sector increased its share of employment from 43.6 to 47.6 percent between 1970 and 1989 its share jumped to 52.1 percent by 1991 and to 64.5 by 1992. According to LSMS data, in 1993, 70 percent of the employed worked in the informal sector22. 51. The increase of informal sector employment in the 1970-89 period appears to be associated with the war and instability in the context of a centrally planned economy with formal worker protection. The increase experienced since 1989 is more associated with hyperinflation, which makes informality more profitable (given the absence of explicit contracts). The increase experienced since 1991 could be associated to several factors. First, the stabilization and adjustment program initiated in March 1991 included the reduction in government jobs (Programa de Conversi6n Ocupacional), the privatization of state-owned firms, and the reduction of the army, which likely created an excess supply of labor that was faced with the choice of informal sector employment, un- or underemployment, or leaving the labor force. Additional factors that may explain the rise in informality are taxes and uncertainty. A quick review of tax rates suggests that costs of doing "formal" business are high in Nicaragua by Central American standards [SIECA, 1993]. Finally, the political risk and uncertainty involved in doing business in Nicaragua possibly also provides additional incentives for keeping it "small". And small operations (5 workers or less) are by definition classified as being in the informal sector. 52. Contrary to conventional beliefs, informal sector employment is not a good predictor of low earnings and hence of poverty. Table 2.26 compares average monthly earnings in the formal and informal sectors. First, although monthly earnings are higher in the formal sector, earnings in both the formal and informal sectors are well above the poverty line23. This is a clear indication that it is the high dependency ratios that makes these earnings inadequate on a per capita basis. Second, there is a strong relationship between educational attainment and earnings for males and females, and the pattern of earnings growth by educational level is significant both in the formal and informal sectors. Estimated wage equations show that the earnings difference between the formal and informal sectors disappears when controlling for individual characteristics, especially education and job-training, even when controlling for selection into formal or informal sector employment24. 21. The definition of informal sector used here corresponds to that most used in Nicaragua, and includes: employees in firms with less than 5 workers + self-employed workers + domestic workers. 22. See main characteristics of workers in the formal and informal sectors in Volume II, The Informal Sector in Nicaragua, Table 1. 23. The comparison of log hourly wages of workers in the formal and informal sectors shows the same pattem to hold. See Volume II, The Informal Sector in Nicaragua, Table 2. 24. See Volume 11, Labor Markets in Nicaragua, Tables I and 2 for estimated wage equations for the whole country, Managua, Other Urban Areas, and Rural Areas.  - 21 - Table 2.26: Mean Monthly Earnings in the Formal and Informal Sectors, 1993 (current cordobas) Characteristic Formal Informal Total Total 1,382 1,047 1,172 Males 1,444 1,117 1,252 Females 1,252 955 1,047 Education: No schooling 652 571 593 Primary incomplete 1,108 1,023 1,045 Primary complete 1,277 1,074 1,136 Secondary incomplete 1,408 1,367 1,384 Secondary complete 1,292 1,673 1,418 Higher 2,000 3,450 2,850 Industry: Agriculture 790 622 679 Manufacturing 1,669 1,079 1,299 Elect/gas/water 1,631 1,812 1,642 Construction 1,786 1,074 1,424 Commerce 1,404 1,358 1,367 Transportation 1,715 1,469 1,580 Financial Services 2,395 1,971 2,294 Personal Services 1,245 854 1,030 ote: See volume 1, The Informal Sector tn Nicaragua, Table 3. 53. To further test whether informality is the result of a high degree of segmentation of the labor market, separate wage functions for the formal and informal sector were estimated by gender, controlling for sectoral choice. No evidence of segmentation arising from fornal or informal sector employment was found. Nicaraguan workers sort themselves out into the sector where they receive higher wages according to their individuals skills. 54. The above findings indicate first, that the key issue in the labor market is not one of unemployment but one of underemployment, which results from low skills, and probably low capitalization and use of complementary productive inputs. They also show that low earnings and hence poverty are mainly determined by the human capital endowment of individual workers, rather than by labor market segmentation. This implies that improvement in the skills of Nicaraguan workers is key to bring about higher earnings and thus contribute to poverty reduction. (5) The Structure of Wages 55. While human capital accumulation is one of the major components of economic growth, it is also one of the main factors explaining earnings inequality and poverty. This section examines the returns to the key components of human capital: education, on-the-job experience, and training. To this end, wage equations were estimated for prime-age males and females (25-64 years old) that work and report positive wages from employment, including self-employment income. 56. As shown in Table 2.27 there are considerable regional wage differences, particularly among males, which are largely explained by regional disparities in workers' educational attainment. (a) Male Wages 57. Schooling pays off in the Nicaraguan labor market. The average private rate of return to an additional year of schooling among males is 14 percent, which is very much in line with similar estimates for other developing countries [Psacharopoulos, 1994]. These returns are 18 percent for primary schooling, 13 percent for secondary schooling, and 16 percent for higher education.  - 22 - Table 2.27: Mean Years of Schooling and Log Hourly Wages by Region and Sex, 1993 Region Males Females Years of school Log hourly wage Years of school Log hourly wage I Segovias 3.44 4.68 5.97 4.35 II Westem 4.96 5.46 5.78 4.49 m Managua 6.81 8.13 6.40 6.29 IV Southern 5.74 7.13 5.75 5.26 V Central 3.57 5.14 4.85 3.85 VI Northern 2.54 3.15 4.76 4.07 VII Atlantic Coast 3.17 5.10 4.59 4.06 Total 5.11 6.26 5.85 5.26 58. Flat age-eamings profiles indicate that Nicaragua has low amounts of both general and specific training. First, having taken a training course raises wages by 19 percent, which is relatively low compared to other developing countries5. Second, the effect of potential labor market experience (defined as [age - 6 - years of schooling]) on wages is also low. On average, one more year of potential work experience raises wages by about 3 percent for those with no experience, and then the effect decreases to zero growth at age 48 and 37 years of potential work. This feature, similar to other Latin American countries during the 1980s, reflects the low level of capitalization both in physical and human capital resulting from stagnation or even negative growth. This could also be the result of heavily regulated labor markets in the 1980s where wages were set with no correlation to productivity growth. Finally, years in the current job, which capture returns to investments in specific skills, show that 10 years in the job raise the wage of the average worker by only 11 percent. Compared to other developed or developing countries this is very low. For example, in the US a typical male earns 25 percent more after 10 years of job specific experience [Topel 1991]. [Pessino 1993], and [Arriagada and Stelcner 1987] found similar results for Buenos Aires and Peru (24 percent and 22 percent, respectively). 59. National level wage equations mask pronounced regional variations, where the poorer the region, the higher the return to human capital investments. The poorest regions (Segovias and Northern) have the highest returns to education and to job-training courses'. The effect of urban location is also lower for the less poor regions, signaling the wider differences between rural and urban areas in poorer regions. Part of the explanation of these huge variations lies in sizable disparities in workers' human capital. 60. Wage equations by age cohort show rising rates of return to education, indicating higher demand for education on the part of firms and a low elasticity of substitution between skilled and unskilled labor. On average, an additional year of schooling increases the wage rate of a 25-34 years old worker by 17 percent. This declines to 15 percent for those aged 35 to 44 years, to 14.4 percent for those aged 45 to 54 years, and to less than 10 percent for older workers. Note that this increase in returns to education took place in the face of expanding educational attainment during the 1980s (Table 2.12), underscoring the potential of education investments to assist people grow out of poverty. (b) Female Wages 61. Since about 70 percent of the women 25 to 64 were not in the paid labor force, wage equations were corrected for self-selection. As expected, the probability of a woman being in paid employment 25. Panama [Heckman and Hotz 19861; Peru [Arriagada and Stelcner 1987]; Argentina [Pessino 1993]. 26. See, Volume 11, Labor Markets in Nicaragua, Table 3.  - 23 - significantly increases with education27. Also, the higher the earnings of the husband and the larger the number of children under the age of 6 in the household, the lower the probability that a woman will participate in the paid labor force. 62. Returns to education among females are lower than among males: the average private rate of return to an additional year of schooling is 8.6 percent. These returns are 7 percent for primary and for secondary schooling, and 13 percent for higher education. Returns to training courses for females (27 percent) are higher than for males. Returns to work experience both potential and in the current job are not statistically significant. This might be due to the fact that women largely work in occupations (services) where wages show less growth. The question of why women work in these occupations in part has to do with the sexual division of the labor market, where women concentrate in jobs that are compatible with their family roles (jobs with flexible hours, jobs where interruptions are not penalized in terms of wages, etc.). Note that the pattern of regional variations found among males also holds for females: the poorer the region, the higher the returns to education and training. These results do not imply that there are less payoffs to investing in the human capital of females. The effects estimated here refer only to labor market outcomes, while there is an ample literature showing the strong and positive effect of mother's education on children. As will be shown in Chapter V, the education of the mother is positively related to her children school attaimnent and nutritional status (paras. 133 and 149). D. The Probability of Being Poor 63. This section examines the probability of a household being poor. This analysis is useful, first, to verify the relative role of the different factors shown in the poverty profile in determining poverty status, and second, to assess the impact that a policy induced change in a factor is likely to have on the probability of being poor, holding everything else constant. 64. Table 2.28 presents a probit model of the determinants of household poverty. The "marginal effects" column shows the percentage change on poverty associated with a unit of change in the explanatory variable. The estimates clearly show that the key correlates of poverty in Nicaragua are: spatial location, access to running water, household size (particularly number of children under 12), educational attainment, and work in agriculture. Contrary to what is commonly believed in Nicaragua the probability of a household being poor does not increase if the head of household is a female', if the head is unemployed, or if household members work in the informal sector. 65. To compare the probability of being poor for different values of a factor, we estimated predicted probabilities29. That is, the probability of being poor was estimated by assigning different values to a selected characteristic (such as number of children) and maintaining all the other variables constant at their national mean values. 27. Only 20 percent of women with no education participate compared to 50 percent of those with secondary schooling and 60 percent of those with higher education. See Volume 11, Labor Markets in Nicaragua, Table 4. 28. These results are similar to those from South Asia [Lipton, 1983]; [Kakwani, 1980]. 29. Predicted probabilities were estimated on the basis of the means and coefficients of the variables used in the poverty probit.  - 24 - Table 2.28: The Probability of Being Poor, 1993 Variable Marginal Effects CONSTANT -14.7* RURAL 17.0* REGIONI': SEGOVIAS 24.4* WESTERN -0.3 SOUTHERN 2.0 CENTRAL 8.3* NORTHERN 14.9* ATLANTIC 3.7 HEAD YEARS OF SCHOOLING 4.1 HEAD UNEMPLOYED 3.7 HEAD INACTIVE 2.9 HEAD FEMALE -3.2 HEAD AGE -0.2* NUMBER OF CHILDREN 0-12 10.5* NUMBER OF ELDERLY (60+) 3.1 HH HAS RUNNING WATER -15.4* % WORK INFORMAL SECTOR -2.4 AGRICULTURAL HH 11.2* Log-likelihood -1944.1 Sample size 4155 Mean dependent variable .461 Note: * significant at the 0.05 level or better. 1/ Dummy variables. Omitted region is Managua. 66. Spatial Location. Large differences in the incidence of poverty by area of residence and region remain after controlling for key characteristics of the household. Living in a rural area increases the probability of being poor by 17 percentage points (Table 2.28). The predicted probability of a rural household being poor is 54.1 percent, compared to 37. 1 percent for a similar household in an urban area. Households located in the Segovias, Northern and Central regions are also significantly more likely to be poor than those located in Managua. For example, the difference in the probability of poverty between a household in Segovias and one in Managua, holding all else constant, is 24 percentage points. 67. Access to Running Water. Households with running water are less likely to be poor even when controlling for area of residence and region. This indicator probably reflects the level of infrastructure investment existing in a community and is a proxy for the quality of life in that community. 68. Household size. As shown above, households with more dependents, particularly children are more likely to be poor than smaller households. Each child increases the household probability of being poor by about 10 percentage points. Table 2.29 shows the increase in the predicted probability of poverty with each additional child in the household, while all other household characteristics are held constant. Table 2.29: Probability of poverty by number of children < 13 Number of children Zero 23.4 One 32.3 Two 42.3 Three 52.8 Four 63.2 Five 72.6 Mean number of children 2.19  - 25 - 69. Education. As Table 2.28 shows, the probability of poverty is substantially affected by educational attainment and signals its potential impact for poverty reduction objectives. As shown earlier (para 28), the poor do not face constraints to participate in school as access to schooling services is adequate. What is needed is to improve their school attainment. On average, the probability of being poor decreases by 4 percentage points with each additional year of schooling of the head of household (Table 2.28). Table 2.30 shows the predicted probabilities of poverty by educational attainment of the head. The likelihood of a household being poor when the head has no schooling is 60 percent, which declines to half for a household whose head has completed primary schooling. Table 2.30: Probability of poverty by years of schooling of the head Years of Schooling, Head of HH No formal schooling 60.1 One year 56.0 Two years 51.9 Three years 47.8 Primary completed (six yrs) 35.8 Basic Cycle (nine years) 25.1 Mean years of schooling of head 3.87 70. Agricultural households30. The poverty profile shows that the incidence of poverty is highest among households that derive a large part of their income from agricultural activities. Even after controlling for household characteristics and spatial location, the estimates presented in Table 2.27 confirm that work in the agricultural sector increases the probability of a household being poor by 11 percentage points. An agricultural household has a 51 percent probability of being poor compared to 40 percent chance of being poor for a similar non-agricultural household. III. INCOME-EARNING OPPORTUNITIES FOR THE POOR: THE ECONOMIC FRAMEWORK 71. Without accelerated economic growth, there will be no poverty reduction in Nicaragua. This chapter examines, first, the 1990-93 economic reform program of the Government, second, its likely impact on the poor, and third, the challenges ahead for enhancing poverty-reducing growth. A. The 1990-93 Economic Reform Program 72. When President Chamorro assumed office in 1990, Nicaragua was a country devastated by a decade of civil war and natural disasters, and with an economy which mirrored the collapse of the Soviet block economies. The GON initiated ambitious reforms to support Nicaragua's transitions from war to peace, from dictatorship to democracy, and from central planning to a market economy. In 1991-92, the Government implemented a strong stabilization program supported by an IMF Stand-by arrangement, and began to implement a structural adjustment program aimed at transforming Nicaragua into a market economy capable of achieving sustained growth. IDA supported these efforts through an Economic Recovery Credit (ERC I) approved in September 1991. 30. A household was classified as agricultural is any of the following are true: (i) the households farms land regardless whether the land is owned, rented, or just used by the household, (ii) the household raises cattle, and (iii) one or more household members work in the agricultural sector.  - 26 - 73. The stabilization measures introduced in March 1991 were remarkably successful. Inflation was reduced from hyperinflationary levels in 1989-90 to single digit annual rates from April 1991 through December 1992, and the economy recorded a very modest output growth (0.4 percent) in 1992, thus arresting the continuous decline since 1983. The key factors for this success were a major reduction of the fiscal deficit, maintenance of credit discipline-particularly with regard to state-owned enterprises-and a fixed exchange rate system backed by foreign aid flows. The Government also succeeded in implementing the first phase of its structural adjustment program, complementing and helping consolidate the stabilization program. Main achievements of the program are presented in Box 3.1. Box 3.1: Achievements of the Structural Reform Program Downsizing the public sector: public expenditures were reduced from 46 percent of GDP in 1990 to 25 percent in 1991-93. Military personnel were reduced by over 80 percent and central government employment by 12 percent through an occupational conversion program (POC). During 1991-2, the GON divested 233 out of an initial 351 state-owned enterprises identified for reforms; * Liberalization of foreign trade through the unification of the exchange rate system, reduction in tariffs to a 10-40 percent range, elimination of most non-tariff barriers, and abolition of state trading; * Liberalization of domestic trade through the elimination of price controls and privatization of state- managed trading activities by end 1992; * Liberalization of the financial sector through the elimination of interest controls, the creation of an autonomous superintendency of banks, restructuring public banks, and opening the financial system to private banks. By 1994, nine private banks were operating, one state-owned bank was liquidated, and the staff and offices of the other state-owned banks reduced by 50 percent; * Creation of safety nets such as the Emergency Social Fund (FISE) and programs for war-affected populations (Chapter VI). Social sector budgets have been protected against expenditure reductions. The share of total ordinary central Government expenditures allocated to the social sectors increased from 36 percent in 1992 to 46 percent in 1993, and ambitious reform programs have been initiated in the health and education sectors (Chapter V). 74. Although the measures implemented in 1991-92 succeeded in stabilizing the economy and arresting the fall in output, they were not sufficient to trigger the resumption of growth. The social impact of the program was largely cushioned by a large influx of foreign aid that allowed a 12 percent expansion of private consumption per capita in the same period. However, the political situation deteriorated in 1993 and the US suspended aid disbursements. Political and social instability, coupled with a drought and adverse terms of trade resulted in a 0.4 percent decline in GDP in 1993. Overall fiscal performance was below target during 1993 and total aid inflows declined significantly compared to 1992 levels. In 1993, gross disbursements of grants and loans were 35 percent below the 1990-92 level, a drop equivalent to 12 percent of GDP. B. The Effects of the Economic Reforn Program on the Poor 75. Lasting poverty reduction requires sustainable growth, without which, attempts to reduce poverty will have little impact. Nicaragua's economic reform program aims to create the conditions for sustainable growth. The effects of economic reforms on the poor are difficult to trace with certainty. They may affect the incomes of the poor through their employment conditions, as owners of assets and factors of production, and their household expenditures, as consumers of goods and services. The mechanisms by which these measures affect household living standards vary, and their short term and long term effects may have differing distributional impacts.  - 27 - 76. Trying to determine how the poor have fared during the reform process is particularly difficult in Nicaragua. This is not only because of the massive transformation involved in moving from central planning to a market economy, but also because this process has been undertaken as the country emerged from over a decade of civil strife and was beset by hyperinflation. Under central planning, prices bore little relation to scarcity values, and access to goods and services was not a function of purchasing power. Scarcity, rationing and black markets were the norm, increasing search time and encouraging non- productive rent seeking. Price trend data is of little use from a welfare perspective because prices did not affect consumption decisions and because of the distortions caused by hyperinflation. Evaluation of price trends alone cannot capture a shift from a situation of extremely limited product access to one of abundant variety and immediate availability. The analysis in this section is unavoidably speculative. (1) The Incentive Framework 77. The success of the stabilization program in arresting inflation unequivocally benefitted the poor. Since inflation erodes the real incomes of wage earners and the value of unprotected assets, it is highly regressive in its distributional impact. It also grants windfall gains to holders of unindexed debt-bank credits or delayed payments. Since the poor are the least able to hedge their earnings and assets, or to benefit from the erosion of debts, they are the most vulnerable to the losses and disruption of consumption associated with inflation. 78. Relative price realignments. Devaluations may not have fully eliminated Nicaragua's exchange rate overvaluation3', but the current relative price regime is far less of a burden on exporters than the previous tangled structure of levies and controls. Domestic trade reforms allowed relative prices to reflect costs and scarcity values. Reduced indirect taxation has eased the incidence of taxation on the poor. And while high input subsidies are no longer available, producers are not required to accept the often low producer prices offered by state monopolies during the 1980s32. Overall, the reforms have already substantially improved the incentive environment for labor absorbing growth in all sectors, including export agriculture. 79. Remaining trade barriers aim to protect domestic industries on a declining basis until 1999 to ease the transition to competitive production. While protecting domestic employment and production, the barriers create domestic monopolies and higher prices for consumers. An additional concern in the structure of trade taxation is a possible relative protection of industry versus agriculture. Temporary protection is applied primarily to industrial products. Although certainly an improvement over the pre- reform situation, the evolution of effective protection of agriculture versus industry and its effects on consumer prices, warrants close monitoring. 31. The amount and adequacy of the real devaluation since 1991 is a subject of debate. Calculations of the real devaluation since March 1991 range from a 20 percent appreciation [FIDEG] to a small 0-7 percent devaluation (IMF) and a 10 percent devaluation [Vos 1994]. However, there is consensus that the real exchange rate remains overvalued in trade competitiveness terms [Edwards 1992, Corbo 1993, FIDEG, Vos 1994]. While real devaluation would be desirable, it is hampered by the significant inflow of donor aid and considerable dollarization of the economy, reflected particularly in the indexation of salaries to the 'canasta basica' under collective agreements, and of the financial markets to the dollar for virtually all transactions. The costs of overvaluation are partially offset by the benefits of cheaper imports to smooth consumption during the transition. Overvaluation, however, is an implicit tax on agriculture and other exportables. 32. According to the Uni6n Nicaraguense de Cafetaleros, the average producer price for coffee paid by the state monopoly, ENCAFE. during the 1980s was USS40/quintal, while the average world price during the same period was US$140/quintal.  - 28 - 80. Financial Sector reforms have little direct impact on small producers or businesses in the short- term, as they are generally outside the formal system33. For borrowers, the elimination of credit subsidies has increased real interest rates, while the GON has reduced net credit expansion to sustainable rates. Since repayment rates are low34, these measures reduced operating liquidity, possibly forcing salary cuts or layoffs and harming labor incomes in the short term. In the medium term, reforms should price capital at its scarcity value-reducing credit rationing, and encouraging credit allocation to sound investments rather than politically attractive schemes. These effects improve sustainable growth prospects, and would benefit the poor. (2) Public Expenditures 81. Tax reforms raised government revenue through improved compliance and extension of the base. Income tax reforms have a limited direct impact on the incomes of the poor as they do not pay corporate and/or personal income tax. The General Value Tax (IGV, a consumption based value-added tax) is applied to sales in the formal sector, with some coverage of the Managua informal sector. Although general consumption taxes are regressive in their incidence, the IGV replaces a cascading tax, of even more regressive incidence. Because collection in rural areas is limited, this tax is likely to affect the urban poor more than the rural poor. Exemptions from the IGV include basic food products, medicines, school books and many intermediate inputs. 82. Public expenditure reductions have been substantial. In the face of declining real per capita budgets in these sectors, GON has focussed on improving efficiency and reallocating available resources toward basic services most used by the poor. Nevertheless, the poor have been hurt due to the introduction of user fees, not always well targeted to protect the poor, who according to the poverty profile are already spending significant proportion of their incomes in health and education. See Chapter V for more detailed coverage of this topic. 83. During 1990-93 there was a drastic reduction in public sector employment and demobilization of military personnel. These workers received an average of 20 months' salary as severance payment, which probably cushioned the initial social impact but their current welfare levels are not known.35 They could have found salaried employment, joined the informal sector, remained unemployed, or dropped out of the labor force. All indications are they have gone into informal sector activities or remain unemployed. Private employment has grown rapidly, though outside of official statistics and especially in the infornal sector (para 50). Although at the national level unemployment is about 13 percent, it is much higher in Managua, Western and Southern regions, where a large part of the civil servants worked (Table 2.20). At the same time, labor force participation does not appear to have changed during the adjustmerit period (para 38). 33. In 1993, only 3% of poor households report having received a loan in the prior 12 months. International experience shows that informal credit systems serve the poor at market rates. Starting in the 1980s, however, they have increasingly incorporated a subsidy element-often donor financed-to become as distortionary and unsustainable as the formal systems. 34. Repayment rates in the 9 private banks and in one of the public banks (Banco Popular) are acceptable. But while the private banks now account for more than half of all deposits in the system, they still account for only a small share of overall lending of the financial sector. The two big public banks, BANIC and BANADES, account for most of the loan stock currently outstanding, and are where repayment rates are low. They bring down the average for the whole financial sector. Since the private sector is not lending much to small producers or businesses it does not serve the credit demands of the poor. The Banco Popular lends to cooperatives, via the MAS, and with guarantees from donors, so their own money is not at risk. Otherwise, it is BANADES that lends for rural activities. 35. Severance packages were granted to state employees released through the Programa de Conversi6n Ocupacional (PCO) or CORNAP privatizations, including up to a maximum of three years of salary as a lump sum. Indemnizations under the PCO were C$208 million, of which 87 percent was disbursed in 1991. This reflects an average beneficiaTy compensation of CS8300. or US$1660. Indemnizations were scaled to salaries eamed. In some CORNAP enterprises, workers received an average of 33 percent ownership in the privatized enterprises.  - 29 - (3) The Regulatory Environment 84. Elimination of state controls over international and domestic trade, and the reduced role of the state in the banking sector has had mixed results on the poor. The private sector has substantially replaced state marketing structures and financial services in urban areas, and in the rural activities of largest profit and quickest turnover. These areas already reflect considerable efficiency gains for consumers and producers with radically increased variety in goods available. For isolated rural areas and products with a slower turnover or smaller margins (such as basic grains), commercial links and financial services have been slower to reappear. Since Nicaragua's poorest are concentrated in the latter group, they have been hurt by this process. 85. Labor Code reforms which substantially restrict labor market flexibility were passed by the Assembly in late 1994, after two years of debate. Consequently, the executive branch has vetoed many articles. The bill is now back in the Assembly, which is deciding how to proceed with this legislation. This nearly 3 year delay leaves uncertain the rules under which employers must operate, and may be a disincentive to hiring labor, as the huge growth experienced by the informal sector since 1991 would suggest (para 50). (4) Trends in the Incomes of the Poor 86. Real wage and producer price series offer indications of winners and losers during the transition. In particular, real producer prices of basic grains, and real wages in agriculture, construction and services are indicators of trends in labor incomes in the sectors where Nicaragua's poor are concentrated. As shown in the poverty profile (para 52), no significant earnings differences between the formal and informal sectors exist, with workers sorting themselves out in the sector they receive the highest wage according to their skills. Wage trends therefore can be said to reflect earning trends in both sectors. Figures 1-3 present real salary trends for formal employment in 13 sectors. Virtually all sectors reflect real wage gains in relation to April 1991. 87. Real producer and nominal price trends for basic grains (Figures 4-5) reveal significant variability, especially for beans. Sorghum and corn show the most regular gains, but do not consistently outperform the national average wage. Price variability reflects thin marketing structures, and vulnerability of poor producers, as well as external market conditions. 88. Since the majority of the poor rely on agriculture for their primary source of income (Table 2.7), the weak relative performance of agricultural wages and basic grains producer prices suggests that the poor and more seriously, the extreme poor, have been relative losers during the economic transition. Wage trends by crop are not uniform, however, confirming that there are winners and losers within the sector. Wage and producer price trends also indicate a strong urban bias in the economic transition process. Salary gains in urban areas far outstrip rural trends. Moreover, since agricultural wages are the lowest in absolute terms, their continued decline relative to other sectors displays a widening of a preexisting gap between rural and urban incomes. 89. Several factors help explain the divergence of income trends. During the years of greatest economic distortions, those able to profit from rationing, negative real interest rates, unindexed debts to state enterprises and utilities, and wholesale debt forgiveness, were primarily urban or non-poor-favored clients of the state, large agricultural producers, and public sector employees. Elimination of these distortions was undoubtedly pro-poor, but it left a legacy of high wealth disparity between urban and rural areas.  - 30 - Figure 1 Index of Average Salaries plus Benefits in Constant April 1991 Cordobas (CPI and RCPI deflators) 500.00 450.00 Sugar 400.00 - , -- Sugar 350.00 ' -z Rice 300.00 _ - Coffee 250.00 200.00 --,,, - ' National 150.00 * g. * Agricultural 50.0 _. Fisheries 0.00 . _ N CN N N NM M m m m 0 e It OO a,o, 0Q n O rn d ) ) 0 a d a, < <