ReportNo. 8643-BO PovertyReport

Public Disclosure Authorized October3, 1990 LatinAmerica and the CaribbeanRegion CountryOperations Division I CountryDepartment III

FOR OFFICIALUSE ONLY Public Disclosure Authorized Public Disclosure Authorized

Documentof theWorld Bank

Thisdocument has a restricteddistribution and maybe usedby recipients Public Disclosure Authorized only in theperformance of their officialduties. Its contents may nototherwise be disclosedwithout World Bankauthorization. Fiscal Y-ear January 1 to December 31

Curroncy Eauivalents Current Unit: Boliviano (Be) Exchange Rate Effective July 31, 1990 US$1.00 Bs 3.17 B51.00 = US$0.32

ABBREVIATIONS BAB Bolivian Agricultural Bank CEM Country Economic Memorandum CIAT Center for Tropical Agriculture Research COMIBOL Bolivian Mining Corporation CONEPLAN National Economic Planning Council CONES National Council for School Construction CORACA Peasant Agricultural Development Corporation DHS Demographic and Health Surveys EIH Integrated Household Survey ENSO Economic Management Strengthening Operation ENPV National Survey on Population and Housing EPH Permanent Household uurveys ESF Emergency Social Fund FDC Peasant Development Fund FDR Regional Development Fund GDP Gross Domestic Product HD Pealth District IBTA Bolivian Institute for Agricultural Technology INAN National Institute of Nutrition INC National Colonization Institute INE National Institute of Statistics MACA Ministry of Campesino and Agricultural Affairs NEC Ministry of Education and Culture MHCSP National Maternal Health and Child Survival Program HINPLAN Ministry of Planning and Coordination MPSSP Ministry of Health MTC Ministry of Transport and Communications NGO Non-Governmental Organization PDIs Private Development Institutions PSM Public Sector Management Program PVOs Private Voluntary Organizations RDCs Regional Development Corporations SENALEP National Service for Alphabetization and Popular Education SENAC National Road Service SENET National Technical Education Service SIF Social Investment Fund SNDC National Service for Community Development SU Sanitary Unit SVEN National Nutritional Surveillance System TGN National Treasury UDAPSO Social Policy Analysis Unit UN United Nations UNICEF United Nations' International Children's Emergency Fund VAT Value Added Tax FOR OMCIALUSE ONLY

This report is based on the findingsof variousmissions to Bolivia between June 1589 and March 1990. The report was discussedwith the Bolivian Governmentin June 1990. Bank participantsin the missionswere: Steen Lau Jorgensen (CountryOfficer, Task k,ager)Izumi Ohno (Economist),Linda McGinnis (Consultant,Nconomist), Julie VanDomelen (Consultant,Economist), Juan Carlos Aguilar (Economist),and William Shaw (CountryEconomist).

The following background papers were written for the report: "Poverty Report: Agriculture" James Cock and Douglas Forno; "Poor Rural Women in Bolivia" Ann. Webb; "The Importance of Rural Infrastructure for Poverty Alleviation"(in Spanish) RodrigoCisnerosl "Non-GovernmentalInstitutions in Core PovertyAreas of Bolivia" (in Spanish)Christina Mejia, et. al.; "Poverty in the Lowlands of Bolivia" Cowi Consult; "InstitutionalAspects of Poverty Alleviation"Linda McGinnis; "The Government'sStrategy and InvestmentProgram and Poverty Alleviation"Izumi Ohno; "Donor Activitiesin Poverty Alleviation" Izumi Ohno; "Education Reform and Poverty"Dan Newlon; "Annotated Bibliography on Poverty in Bolivia" Malene Hedlund.

The report was written by the task manager with substantial inputs from: Izumi Ohno (strategies and investment), William Shaw (macroeconomics), Linda McGinnis (institutionalissues), and Julie VanDomelen(efficiency issues). The preparation team is especially indebted to Emergency Social Fund staff, CEDRAGROand PROANDESin , and IPTK in Ocuri for arranging field trips. Several background papers were financed by trust funds including the consultant trust funds of Denmark, the Netherlands and Norway. Diana Cortijo did the typing and formatting and provided research assistance.

This documenthas a restricteddistribution and may be used by recipientsonly in the performance of their officialduties. Its contents may not otherwisebe disclosedwithout World Bank authorization. TABI or CONTENTS

SUNHARYAND CONCLUSIONS...... l-vii

CHAPTER1: INTRODUCTION AND AACKGROUND ...... 1 INTRODUCTION ...... * . .. . 1 Ratlonale for the study ...... 1 . OUTLINE OF THE REPORT ...... 1 BACKGROUND ...... * . . . . . * 3 Geography ...... 3 Hlstory ...... 4

CHAPTER 2s DI4ENSIONS OF POVE RTY ...... 6 INTRODUCTION ...... 6 The Data ...... * * * * . . .*..... *.... 6 DESCRIPTION OF POVERTY ...... 7 Poverty in General ...... 7 Poverty Over Time - A Paradox.. .. . 8 Poverty in Rural and Urban Areas ...... 12 Poverty by Region ...... 13 Poverty by Sector and Occupatlon ...... 18 Poverty by Gender...... 18 Ethnic Groups and Poverty ...... 20 POVERTY PROFILES ...... 21 CAUSES OF POVERTY ...... 22

CHAPTER 3: MACROECONOMIC POLICIES AND THE POOR...... 25 INTRODUCTION ...... 25 IPACT OF THE 1980-85 ECONOMIC CRISIS ...... 25 Government Policies and the Poor...... 26 Impact on Agriculture ...... 27 IMPACT OF THE STABILIZATION IN AUGUST 1985 ...... 28 IMPACT OF ADJUSTMENT POLICIES ...... 29 Public Sector Employment .. nt...... 30 External Sector Policies ...... 31 Impact on Agriculture ...... 32 Labor Markets ...... 34 Tax Reform ...... 34 CONCLUSIONS AND RECOMMENDATIONS...... 36

CHAPTER 4. INSTITUTIONS AND THE POOR ...... 39 INTRODUCTION ...... 39 GLOBAL INSTITUTIONAL FRANEWORK ...... 40 COMMON INSTITUTIONAL CONSTRAINTS ...... 43 Centralization and Urban Blas ...... 44 Lack of Demand-Driven Orientation ...... 46 Poor Planning and Policy Making ...... 47 Patronage, Poor Salaries and Poor Training ...... 48 Donor Dependence ...... 50 2

CONCLUSIONS AND RECOHNXIENTION ...... 51 The Demand-Driven Approach ...... 52 Decentralization ...... 55 Improved Policy Making, Planning and Coordination ...... * . . 56 Donor Coordination ...... so

CHAPTER S: HEALTH ...... 59 INTRODUCTION ...... 59 INSTITUTIONAL FRAMEWORK ...... 60 GOVERNMENTPOLICIES AND INVESTMENTS ...... 61 Sector Strategy and Expenditures ...... 61 Nutrition Pclicies .c.i ...... 63 Water Supply ...... 65 Investment ...... 66 SECTOR CONSTRAINTS ...... 67 Centralization and Inadequate Coordination ...... 67 Low Salaries and Poor Personnel Management ...... 68 Poor Financial Management ...... 69 InappropriateTraining ...... 69 Lack of Beneficiary Participation ...... 69 Insufficient Use of Non-Governmental Institutions...... 70 Women'sIssues ...... 72 RECONMENDATIONS...... 73 ImplementExisting Policies ...... 73 Shift Expenditures ...... *. . . 73 InstitutionalRecommendations ...... 73 Coordinate NGOs ...... 74 EncourageBeneficiary Participation ...... 74 Upgrade and Reorient Training ...... 75 IntegrateTraditional and Xodern Medicine ...... 75 Nutrition ...... 76 Women'sIssues ...... 76

CHAPTER6: EDUCATION ...... 77 INTRODUCTION ...... 77 INSTITUTIONAL FRAMEWORK ...... 78 EXPENDITURES AND GOVERNMENTPOLICIES ...... 79 Expenditures ...... 79 Reform Plans ...... I ...... s0 Public Investment ...... 81 SECTOR CONSTRAINTS ...... * . 82 Inflated School Staffing ...... 82 Low Salaries ...... 83 Weak Planning and Policy-Making Capacity ...... 83 Inadequate School Administration ...... 84 Legal Impediments to EducationSector Efficiency . . . . 85 Inadequate Non-Formal Education ...... 8S Under-Utilization of the NGO Option ...... 8S 3

RZCCOOENDATIONS ...... * . . #.* . * * * . * 87 Expenditures...... 87 Improving Teachere'Salaries ...... 87 Limit Rural/Urban Differences ...... 89 EfficientAllocation of Resources .. ... 90 Decentralization . . 91 The Role of NGOs . * ...... 92 Community Participation ...... 92 Women' Issues .. . .. 93 Phasing ...... 94

CNRITER 7a AGRICULTURE...... * . . . . .* 95 IWTRODUCTION ...... 95 INSTITUTIONAL RMWEORK ...... 97 GOVERNMENT POLICIES AND INVESTMENTS 98 Sector Strategy .. 98 Sector Investment . 99 SECTOR OONSTRAINTS AND ISSUES. . . 101 Land Use ...... 101 Land Titling ...... 102 Food Donations .. 105 Lack of Research and Extension . .108 Women' Issues .*.*.108 MACA Institutional Issues . 109 Politicization ...... 119 Week Human Res ources ...... 110 Poor Central/Regional Coordination . .110 Weak Information ...... lll Lack of Beneficiary Participation . .111 ECOWNENDATIONS...... 112 Subsidize Poor Communities through Infrastructure . . 112 Land Reform and Community Laws . .113 Appropriate Technology Development and Transfer . . 114 ImproveAgricultural Credit .*. .. .*.. . * 115 ImproveSmall-scale Irrigation . . . . 116 Linking the Poor to the Market Economy . .116 Migration ...... 117 Women'QIssues ...... 117 Strengthen National Policy and Planning . . .118

CMAT8R 8t !RAXSPORTATION...... * * . * 119 NTRODUCTION ...... 119 INSTITUTIONALFRAMEWORK ...... 119 GOVERNMENTPOLICIES AND INVESTMENTS ...... 120 Sector Str-ategy .. .. 120 SectorInvestment ...... 120 SECTORONSTRAINTS 8...... 122 Rural Roads ...... 122 Weak Planning and Coordination . ... 122 Scarcity of Operations & Maintenance Resources. .123 Low Salaries ...... 124 4

3ECOOM)NDATIONS...... 124 Seotor Priorities and Expenditures ...... 124 InstitutionalRecommendations ...... 125

CHUPTER 9s EFFICIENCYISSUES ...... 126 INTRODUCTION* * * .* .* ...... 126 CoSTS OF INVETENTS IN POOR AREAS ...... 126 EFICIENCY AND IMPACT OF INVESTMENTS IN POOR AltES ... . 129 Roadt .S.up ...... 129 errlgatCon...... 130 Water SupplyIVETNT ...... 131 Health Care ...... 132 Education ...... 133 IUMMINO INVE8 XNTS8 . . . o-*-*...... 134 Financial Resources . . 9...... 134 Community Contributions...... 135 CONCLUSIONS AND RECOMMENDATIOZ8 ...... 136

CHAPTER 10s RECOMMENDATIONS AND ACTION PLAN...... 138 INTRODUCTION...... *. . 138 THE CAUSES OF POVERTYREVISITED ...... 138 CRITERIA FOR SUCCESSFUL POVERTYALLEVIATION INTERVBNTONS ...... *. 139 ACTIONPLAN ...... 141

x Ideas for Poverty-Focused Projects and Studies ...... 151 II Various Poverty Naps of Bolivia...... 9..156 III Bibliography...... 161

TABLES:

2.1: Social Indicators and Expenditures . 11 2.2: o0cialIndicators for Rural and Urban Areas ...... 13 2.3: Poverty by Area from ENPV ...... 15 3.1: Agricultural Production ...... -.. . 33 5.1 Distribution of Health Resource. by Area ...... 62 5.2t Consultations per 8taff-hour...... 71 6.1: Distribution of EducationResources by Area ...... 79 6.2 NBC'sExpenditure by Department ...... 80 7.1: Food Donation Programs ...... 105 7.2: Food Donation Volumes ...... 106 7.3: Wh*at Donations and the Domestic Market ...... 107 9.1: Cost/Unit Variations by Poverty Areas for ESP Projects . . . . 127 9.2: Cost/BeneflciaryVariations by Poverty Area ...... 128 9.3: Counterpart Financial Contributions ...... 135 9.4: Counterpart Contributions to ESF...... 136 5

FIGURBS:

2.1s Poverty in 1976 ...... 8 2.2s Lorenz Curves for Urban Areas ...... 9 2.3s Value Added in Agriculture ...... 10 2.4: Distribution of the Poorest Provinces (Map) ...... 14 2.5s Relative Worsening of COre Poverty Areas ...... 16 2.6: Gross MigrationFlows Among Departments ...... 17 2.7: Native Languages and Poverty ...... 20 3.1s Development of Real Exchange Rate ...... 31 4.1: ESF Projects by Poverty Area ...... 39 4.2: Organization of the Public Sector ...... 41 4.3: Requests to ESF from Poorest Areas ...... 43 5.1: Infant MortalityRates ...... 59 5.2s MPSSP Expenditures by Department ...... 62 5.3: Chronic Malnutrition in Urban Areas ...... 64 5.4s NGO Requests to ESF ...... 70 6.1: Children's Health by Mother's Education ...... 77 9.1: ESP Standard Costs in Major Cities ...... 126

30138:

1.1: Social Cost of Adjustment ...... 1 1.2: Outline of the Report ...... 3 1.3: Bolivia'sNine Departments ...... 4 2.1: ODoftaMarta ...... 22 2.2s Reasons for Poverty ...... 23 3.1: The Shoe-shineBoy and Structural Adjustment ...... 29 3.2: Gabriela, Rafael and the Land Tax ...... 37 4.1: Working with NGOs ...... 47 4.2: Framework for Institutional Reform ...... 52 5.1: Maria and the Iodated Salt...... 65 6.1s Cutting the UmbilicalCord ...... 82 6.2: A Rural School ...... 86 7.1: To Migrate or Not, That is the Question ...... 96 7.2: The Guaranies Working for the Company Store ...... 104 7.3: Titling:A BureaucraticNightmare ...... 113 8UMMARY AND CONCLU8OXNS

Outline of the report

1. The focus of the report is on sectoralnolicies and povertv,with a special emphasison institutionalissues. The poor have poor health and educationindicators. They are largely involvedin agriculture,and have limited access to services and markets. Therefore,the report will focus on poverty and a series of sector related policies.

Dimensionsof Poverty

2. The voorestof the poor are rural smallholdersand agriculturalworkers of Indian origin. This is brought out by analysisof poverty indicators'by region, by urban/ruralareas, by gender and race, and by occupation,over time. Data from the last census (1976)show that 95% of the poorest 30% '.are rural peasants and day-laborers. Povertywas concentratedin the central Andean area of Northern Potoui and Chuquisaca, and Southern Cochabamba. Since then, agricultural productior. and prices of the peasants' products have fallen. Recent surveys show that the poor geographicalareas are worse off today relative to the national averagethan they were in 1976. Women and people of Indian origin are the poorest among the poor.

3. The core RoveAtvoroups are: (1) smallnolderneasants of Indian origin in the centralAndean region, (2) recent miarants to the Lowlands, (3) agriculturallaborers (herders,sugar cane workers, and brazil nut harvesters),and (4) Amerindiantribes in the Lowlands. Thus, the report focuses primarilyon rural poverty,as the poorest of the poor are in rural areas.

4. Poverty is fundamentallya Droblem of low ooductivitv and resultinalow income. The causes of low productivitycan be divided into four major categories.2 First, ceneral developmentDroblems: as Bolivia is a poor country,there are many poor people. Bolivia'ssocial indicatorsare worse than GNP per capita would seem to suggest. For example,Bolivia's infant mortality is higher than in most of Sub-SaharanAfrica. Second,the poor have insufficienthuman calpital.The problems of illiteracyand malnutritionare particularlyserious. Third, the poor have few Dhvsical assets. Even though a land reform in 1953 distributedland, many have no title to their land. Low productivityand lack of access to productivity-improvinginvestments or new technologiesare associatedwith the land to which peasants do have title. Fourth, it appearsthat there is discriminationplace on the basis of both gender and ethnic background.

1 Malnutrition,illiteracy, infant mortality,incidence of water supply and sanitation,and housing conditions.

2 Even though this linear causal relationahipis a simplification,it helps to focus the analysis and the four causes of poverty mentionedare equivalentto the problems expressedby the population. ii

Macroeconomics and the Poor

5. The stabilization oroaram probably heloed the relative osoition of the poor compared to what would have happened in the absence of otabilization. The Bolivian economy disintegrated rapidly in the first half of the eighties, resulting in inflation of 24,000% and a fall in GDP per capita of 20s over 4 years. In 1985, an orthodox adjustment and stabilization program was put in place. Inflation was controlled and growth resumed. Price stability and the reduction of state intervention in the economy were achieved. The rents available to groups with economic or political power were reduced and the burden on groups without such influence increased. Changes in the trade regime and the real appreciation of the boliviano probably impaired the competitive position of smallholders producing for the internai market. It was not possible to avoid the creation of a new class of poor when the international tin market collapsed and drastic reduction in employment in the state mining company was needed. The tax reform did not greatly affect the rural poor, who remain largely outside the tax system.

6. Bad macroeconomic oolicies hurt the poor. Interventionist policies increase the return to political or economic influence and are thus beneficial to the non-poor. But macroeconomic lolicies should not be used to improve the situation for any poor qrouM directly. Such instruments are much too blunt. For example, although the rural poor would benefit from an appropriate exchange rate for agricultural production, the present exchange rate is market-determined. Similarly, a more progressive tax structure may be desirable, particularly given the extreme inequality of income levels in Bolivia. However, the Government's ability to monitor highly-differentiated rates or to raise taxes on the rich is limited by administrative considerations and by the need to encourage greater private sector activity. Finally, increases in expenditures to assist the poo_ must be carried out within the overall budget constraints necessary to maintain economic stability. Even so, there is enormous potential within the available resources to reallocate expenditures to assist the poor and to improve the efficiency of programs which serve them. Increased tax collections would help fund more poverty programs without compromising other goals of fiscal policy.

Institutional Issues and the Poor

7. Typically, many Bolivian institutions are inept or their efforts to reach the poor are insufficient. Institutions are overly centralized, are plagued by poor salaries and political patronaae, and have little history of demand-driven policy making. Poverty-oriented institutions should draw lessons from the Emergency Social Fund (ESF), especially in terms of its demand orientation, incorporation of small businesses and NGO participation.

8. Decentralization of specific responsibilities could be a partial solu- tion to these institutional problems. The process should begin in a limited number of sectors and expand gradually. The substantial regional differences must be addressed. From the central government (or donors and NOOs) training and technical assistance should be made available to the regional and local levels. It would be appropriate to decentralize school management, management of primary health care and road maintenance. Decentralization is not IIi appropriate for all sectors. There will still be a need for centrally controlledhospitals and universities,serving several regions, for example.

9. The primary obiective of central am ernment should be to not overall p^oicy auidelines, while allowing decentralized institutions to assume more implementation responsibilities. The creation of the new Sub-seoretariat for 8ocial Policy in the Ministry of Planning is a positive first step. The Sub- secretariatshould be provided with sufficient financial and human resources to cre&te a small cadre of highly skilled experts in action-oriented poverty policy analysis. Similarly,it is essentialfor sectoralministries to strengthentheir policy and planningdepartments.

10. It is essentialto gaise the galariesof key civil servants to competi- tive levels through a reductionin the total number of public employeesa without this action it is unlikelythat public agencieswill be effective in any sector. This will imply greateremphasis on implementationof a public sector managementprogram. For such a reform to take hold, the present system of appointmentsby oolitigalostronace must be abolished. The shift towards a professionalcivil serviceworking for democratically-electedofficials is an essentialstep in the maturing process of Boliviandemocracy.

11. Greater effortsmust be made to coordinate2overtv strateaiesamono donors. governmentand NGOs. significantduplication of projects continuesto exist in some areas while in others there is virtuallyno interventionat all. Inter-donorcoordination should also be strengthened. The successfulexpe- rience of the regular coordinationmeetings in should be sustainedand expanded. A leadingdonor agency should be identifiedfor each sector and should remain in continuouscontact with the Governmentand other donors to discuss policies and interventionsin its sector,thus assistingthe Ministry of Planningwith coordination. gmalth

12. Health o2ligies hav a strong 22yertyjalleviation focus. Expenditures and investments, however, do not follow these policies. The first recommen- dation is to implementexisting policies and shift resourcesaccordingly. A number of NGOs work in health, but their activitiesare uncoordinatedand do not follow national standards(when standards exist, which they frequently do not). Greater beneficiaryparticipation should be encouraged. The poor today mainly use the traditional medical system. Thus, it is importantto integrate the modern and traditional systems through training of practitioners. Mal- nutrition is a seriousproblem in Boliviabut there is no central policy to deal with it. The first task would be to monitor ongoing programs and assess their effectiveness. A pre-school nutrition program could prove to be a good partial solution. In terms of women's issues, the Government should implement the very good Maternal Hbalth and Child Survzval Program that was presented in November 1989. Another serious health issue in Bolivia is the alarminglyhigh rate of life-threatening,illegal abortions. Bolivia needs a culturallyand religiouslysensitive reproductive education program. iv

Education

13. Education in Bolivia is in a crisis. Illiteracy is high, and surveys show that only minimal learning is taking place. Expenditures are skewed towards urban areas, and investment is scant except for ESF. Universities take up a large share of the budget, while rural schools go without blackboards or textbooks. Teachers are poorly paid, but salaries eat up more than 95% of total education expenditures. Efficiency is very low: most teachers work only half-time, and student/teacher ratioq are very low. Lme NGOQ have shown a capability to manage schoolo efficiently, but this option is not widely used. Administration is very poor and complicated by the existence of two separate systems for rural and urban education. Recommendations include shifting exDenditures toward primary education and increasino teaph,rs' salaries through higher efficiencv. This could be done by paying trained teachers more for working a full week instead of the 20 hours per week worked now. The differences between rural and urban education should be minimized through improvements in rural areas followed by a unification of the two systems. Improvements could include multi-grade teaching and pilot testing of boarding schools for dispersed populations. Multi-grade teaching would also increase efficiency and again allow for well-trained, full-time teachers to be paid more. Fees should be introduced for higher education.

Agriculture

14. Investment and exDenditures follow no discernible strategy and tend not to benefit the noor. The sector is full of semi-independent agencies, each following its own path. While the land reform distributed some land to the poor, tenure and titling remain serious issues. As communal titles are not recognized, some communities are in constant danger of being forced off their land. It is very costly and time-consuming to obtain a title, so many poor people do not have title to their land. There is no centrally organized extension service, and very little extension in the public sector at all. Food donations have been criticized for limiting local food production, although the data do not provide sufficient evidence to confirm this view.

15. Poor rural communities should be subsidized throuah the ofrovisionof infrastructure without reauiring full cost recovery. The infrastructure should be built with community labor and be designed in line with cultural norms. In general, small projects have been the most effective. Therefore, a number of small projects with beneficiary involvement shculd be promoted, instead of the massive integrated schemes currently in vogue in Bolivia. The role of women needs to be taken into account, especially as temporary migration removes the men from the day-to-day management of farms. Small- scale irrigation shows great promise and should be pursued. The land titling system must be revised to make it simpler and more transparent. The extension and research system should be reorganized. Joint private/public activities should be undertaken to improve market access for the poor with some marketable surplus. v

TransoDort

16. Government strategy should be revised to emphasize cost-effective construction of rural and feeder roads. The current strategy for transportation investments emphasizes connecting Bolivia with other countries and linking major cities in Bolivia. By emphasizing rural and feeder roads: the development of local markets would be facilitated. This can be accomplished, in part, through some of the community participation and labor- intensive means pursued by ESF. The transfer of responsibilities from ESF back to the sector agencies must ensure that the same types of poor commu- nities continue to be targeted with the same kind of labor-intensive projects.

Efficiency

17. Investments in poor regions, particularly rural areas, can be competi- tive in terms of efficiency. Economic viability of a project depends on the scale and type of technology implemented and the attention paid to issues of market structure and commercialization. For projects providing social services such as health care and education, dispersed populations are decided- ly at a disadvantage. Large efficiency gains are possible in both health and education by rethinking approaches to service delivery. In health, there is a need to build low-cost basic facilities and especially to improve the use of existing facilities. Outreach programs, education and community based service delivery are important in raising the efficiency of existing health infras- tructure. In education, focus should be on the better use of existing facilities, repair of structures and construction of boarding schools. In irrigation, projects should be followed up with training in new methods. In transport, new roads should be built only where there is a competitive transportation sector and where communities are organized, to assure that the rents accrue to the poor.

Action Plan

18. The poverty alleviation action plan is focused on imrrovina sectoral golicies and institutions. Actions are based on addressing the fundamental causes of poverty. To help improve general problems related to the overall low development of Bolivia, macroeconomic policies should remain conducive to private sector development. To address the poor state of human resource development, nutrition, health and education policies must be changed to improve efficiency and remove inequities in spending. The low asset base of the poor should be improved through improved titling, access to appropriate technology, and rural credit schemes (private-sector based). Improvements in public services should be demand driven. Improvements in infrastructure should focus on rural roads, small-scale irrigation, and community based water supply. Discrimination is harder tc tackle, but at a bare minimum the laws should be revised to remove institutionalized discrimination. Bolivia is too poor to limit che productivity of large segments of its population through discrimination. vi

A Summary of Recommendations

19. Key recommendations, by sector, are summarized below:

i) in fiscal policy, efforts to collect more taxes hould be inten- sified. This would enable more poverty programs to be financed, as specified below, without compromising other targets of fiscal policy.

ii) In health, the Government should implement the Maternal Health and Child Survival Prooram. This well-designed program has political support from the highest levels in Government and sufficient external financing. Hoslital and other tertiarv care investments should be reviewed and most likely scaled back.

iii) In education, an extensive reform should begn. A phased approach beginning with better pay for full-time teachers (primarily in rural areas), financed by savings resulting from the reduced need for untrained teachers. In the longer run, a thorough education reform is needed, but most short-term solutions could become ineffective without sucF _-reform.

iv) In agriculture, titlino Drocedures should be streamlined by limiting the number of offices involved in approval of documents and eliminating the need for the President to sign individual titles. Small sc5le irriaation schemes should be sum,orted. Such schemes have proven to be both effective and efficient. Government subsidies should be shifted towards providing poor communities with productive infrastructure such as flood control, storage facilities, irrigation, etc. Communities should provide labor and local materials for such works.

V) In transport, a vroaram of economically iustified rural roads should be develoDed and carried out. There are several large road projects in the existing public investment plan with low rates of return that should be scrapped in favor of rural roads with higher rates of return.

20. Recommendations that apply to all sectors are summarized below:

i) A civil service reform should be initiated. Key staff in planning and policy roles in areas important to poverty alleviation must be assured stable employment free from political interference at sufficiently high salaries. This key cadre of highly trained professionals could be financed through efficiency improvements in public administration.

ii) A leading donor aaencv should be identified for each sector. The agency should remain in cortinuous contact with the Government and other donors to discuss policies and interventions in "its" sector, backing up the work of the Ministry of Planning. Agreement should be reached by the 1990 Consultative Group. vii iii) The Government and external donors should encourage a multitude of activities, programs, solutions, rather than centralized, national, global solutions. National schemes have generally not taken the concerns of the poor into account. An exception to this rule is the natior.alvaccination campaigns and some of the other national health campaigns. iv) Activities should be small-scale. Small projects have a much better track record in terms of poverty alleviation and efficient implemen-tation than large multi-million dollar projects. Such small projects include irrigation, local road projects, and simple water supply schemes. v) 8chemes must take into account the cultural heritage of the affected communities. Communities and their superstructures (e.g., peasant unions) can become very efficient partners in development if projects are developed in cooperation with the beneficiaries. Many projects have failed when cultural norms were not considered. The indigonous cultures have practiced sus- tainable agriculture for centuries. These traditions should be built upon, not discarded as inferior. vi) Proiects should be demand-driven. One way to involve communities is to react to demand for interventions, and not rely on a centralized, top-down planning process. A demand-driven system must be modified to include substantial outreach activities. Communities without the necessary capacities to develop project proposals must be helped. The modified demand-driven system presently in place in the Social Investment Fund could serve as a model. vii) Execute oroiects with local orivate oroanizations. Locally based NGOs, communities, neighborhood councils, mothers' clubs, etc., could all play an important role. At times community organizers such as NGOs might be needed. This could be an NGO interested in expanding its geographical or sectoral scope. viii) The fundamental principle in a strategy based on the outlined principles is to help the poor help themselves. Only if this route is followed will interventions be effective, efficient and sustainable. CRHr-R 1t !NTRODUCTIONAND BACKGROUND

INTRODUCTION

1.1 This chapter provides a brief outline and introduction to the report, followed by a section with background information. First, the rationale for the report is laid out. Second, the outline of the report is presented. Third, background information on geography, economic history, and administra- tive matters is presented. Readers familiar with Bolivia can safely skip the section or background.

Rationale for the stugy

1.2 Since the resumption of Bank activities in Bolivia in 1985, there has been a need for a thorough assessment of poverty. In previous economic reports, various aspects of poverty have been discussed including the avail- able data on poverty (Updating Economic Memorandum, 1988), an assessment of the Government's effort to alleviate the social cost of the crisis and sub- sequent adjustment through the Emergency Social Fund (Country Economic Memora- ndum, 1989), and analysis of expenditures in the social sectors (Public Sector Expenditure Review With a Special Emphasis on the Social Sectors, 1989). When the Bank resumed lending, its strategy was focused on assisting the stabilization program through improving the balance of payments. Once stabilization had been achieved, the Bank shifted its strategy toward the more long-run goals of social development and poverty alleviation.

1.3 Many international financing agencies have asked the Bank to play a leading role in the analysis of poverty issues. Especially following the successful collaboration in the Emergency Social Fund (ESF) program, many external aid agencies felt the need for further collaboration and coordination. The Bank was asked to take the lead in this exercise. The Government of Lic. Jaime Paz-Zamora, which took office in August 1989, stressed its continued commitment to sound fiscal and monetary policy but wanted to move forcefully to alleviate poverty at the same time.

OUTLINE OF THE REPORT

1.4 The first order of business is to assess the extent of poverty in ,Bolivia. Poverty will be analyzed from various angles. As data are notoriously poor in Bolivia, the goal is to identify core poverty groups by assembling brief poverty profiles. These profiles are based on an analysis of poverty indicators in the past, their development over time, and their distribution by ethnic background, gender, region, sector and occupation. It is k Sxlt " i oa of dlutrntin also important to analyze the : tfio h tn sjin thet.a pe*ncan tt causes of poverty to identify Ifti a vination gakmt tvpholud,ithounntin possible solutions. This the v1ntionth pen*hai might bec analysis is presented in wy ckorewn die.- Chapter 2. Box 1.1: Social Cost of Adjustment 2

1.5 Following the UNICEF reports on the social costs of adjustment, such themes have been at the forefront of the discussion of structural adjustment. It has been argued that the poor were hurt by the policies necessary to restore order and growth to the economy. Although the social costs of adjustment debate is rather general and not directly concerned with the poor, it is still very important to analyze how macroeconomic policies affect the poor. This report looks at the effects of macroeconomic Policy both before and during Bolivia's adjustment program to assess the effects on the poor of various macroeconomic policies. This analysis is included in Chapter 3.

1.6 Often poverty alleviation efforts are not implemented due to a lack of institutional capacity. Increasing capital resources and redirecting public policy toward the sectors directly affecting the poor is an important, but not sufficient, condition for alleviating poverty. Without appropriate institutions to implement the policies and direct the investment, these efforts will rarely accomplish their objectives. "Appropriate" institutions are important for two main reasons: first, in their absence, the benefits from major inflows of physical capital tend to accrue to the elite and rarely reach the poor; second, appropriate institut.ons not only help to distribute benefits more equitably, but also contribute to the sustainability of projects and policies, and thus to sustainable growth. Very little effort has been made in the past to identify, strengthen or create institutions to reach the poor effectively. This report will focus on institutional problems such as excessive centralization, urban bias, low salaries, political patronage, and lack of planning and coordination. The general institutional problems will be analyzed in Chapter 4 while the sector specific institutional issues will be discussed with other sector issues in Cbapters 5 through 8.

1.7 The report looks at poverty alleviation from a Policy perspective. In Bolivia, discussions of poverty alleviation are often carried out in a spatial context, emphasizing the geographical distribution of the poor, based on any one of several poverty maps. There are several problems with such an approach. First, the provinces identified as poor are very heterogenous internally. For example, the capital of the province is often much better off than the rest of the province. This makes it hard to justify spending more in the province as a whole. Most inistitutionsare concentrated in urban areas and thus are more likely to present projects and programs that would benefit these areas. Second, there are important pockets of poverty that poverty maps do not capture. These include the indigenous groups in the Lowlands and rural areas with little agricultural potential within non-poor provinces. The poor are mainly rural peasants of Indian origin with poor access to markets and social services. Thus, a better match can be made between the needs of the poor and policies and interventions, if a sectoral perspective is taken. Analyses include policies in health (Chapter 5), education (Chapter 6), agriculture (Chapter 7) and transportation (Chapter 8).

1.8 In a situation of very scarce resources, as is the case in Bolivia, efficiency must be addressed along side of equity. When policies and interventions that would help the poor have been identified, and institutions have been created that could implement the policies, the question of efficiency of interventions arises. It is often argued that interventions in 3 poor (rural and remote) areas are more costly'and have lower benefit/cost ratios. This assumption is analyzel in Chapter 9.

1.9 The final outcome of the report is a framework for an action plan for , n poverty alleviation. Because of the l d toawty 2 focus of the report, the action plan Nc0>wpit* c ptr focuses on policies and institutional c-aptr4 issues instead of a specific list of projects, each chapter contains a C;apter?aglcPiir. section on recommendations. These lsuttatcn Chaptr recommendations are summarized and -ChaPt 9 presented in a matrix form in Chapter 10. &tm Plan a 1

1.10 In sum, the outline of the report Box 1.2s Outliae of the Report is presented in Box 1.2. Each chapter is fairly self-contained, with an introduction, discussion, and conclusions and recommendations. For readers with special sector interests, and with a basic knowledge of Bolivia, it is thus possible to read only the particular chapter concerned maybe combined with Chapter 2, which presents the various dimensions of poverty.

BACKGROUND

*5eoaralphv

1.11 Bolivia has a population estimated at approximately 7 million (the last census was in 1976). The country covers an area of 1,098,581 square km. The three major ecological regions are the Highlands (the Altiplano) in the south and west, the Valleys (Valles and Yungas) of the mountain chain in central Bolivia and the Lowlands in the east and north. The average altitude of the Highlands is 3,700m and the areas has an average temperature of 10 C. This region includes the major city, La Paz and the major mining areas of the country. Although this semi-arid region occupies only 23% of the total land area, 38% of the population lives there. The Valleys region consists of small areas of flat valleys and mountains. The major flat areas are the three semi- arid valleys of the Cochabamba region with a total area of about 100.000 ha. The Valleys occupy 27% of the area with 42% of the population. The vast Lowlands (below 500m in altitude), the frontier lands of Bolivia, cover 50% of the total land area but are sparsely inhabited with only 20% of the population. The Lowlands are below an altitude of 500 meters.

1.12 Administratively Bolivia is divided into nine Departments and 99 provinces. Three of the Departments (La Paz, Oruro and Potosi) are mainly in

Annex 1 contains a preliminary list of projects which would complement the action plan, but this is not the primary goal of the report. 4

the Highlands; 2 three (Cochabamba, Chuquisaca DEPT. CAPITAL POPULATION and Tarija) belong mainly to the Valleys (thwANXaIS region; and three (Beni, Pando and Santa Cruz) La Paz La Paz 196.6 are in the Lowlands. The three most populated Oruro Oruro 388.3 Departments are La Paz, Cochabamba and Santa Cochabaa Cosaiba 982.0 Cruz. Six of the nine Departments have the ChuquisacaSucre 442.6 same name as their Capital, this is Tarija ToriJa 246.6 occasionally a source of confusion. Santa Cruz Sants Cna 1110.1 Beni Trinidad 215.4 Pando Cobija 41.0 History

- n - 1.13 In pre-Colombian times under the Box 1.3: Bolivia's Niue Tiwanaku and Inca dynasties, the area around Deparments Lake Titicaca flourished. The Tiwanakus are today's Aymara Indians, and the descendants of the Inca are the Quechua. The languages of these two groups continue to dominate among the indigenous population. Agriculture was highly developed and the production most likely supported a much larger population than today. Under the Inca rule most of present day Bolivia was colonized and social structures that survive to this day were established, especially as concerns communal land use (Chapter 7). The Spanish conquerors established many of the cities that exist today. La Paz was founded because of the discovery of gold, Potosi because of the rich silver ore in the famous mountain Cerro Rico. In the seventeenth century Potosi was the largest city in the Americas. The Spanish rulers introduced a centralized urban-based bureaucracy. Large groups of Indians were brought from the area north of Lake Titicaca to the high valleys around Potosi to provide a labor pool and produce food for the mines. The vast Lowlands were not settled, except during a period of Jesuit influence when several important missions were established. The Jesuits quickly became too powerful for the King's liking and he forced them out.

1.14 Following independence, the centralized, urban- and Highlands-biased development continued. By the end of the last century, Bolivia was basically a feudal society with a small group of all-powerful landlords controlling large numbers of Indians, who were little more than serfs. When tin mining became very profitable, the industry was concentrated in a few large companies, controlled partly by foreign interests. Until the 1952 revolution, most Bolivians could not vote and lived on large estates or around the mines. Governments came and went as power shifted among the mining and land "barons."

1.15 In 1952, an armed uprising of farmers and miners defeated an attempt by the armed forces to annul the results of an election. The winner of the election, Dr. Paz-Estenssoro, came to power. The party he headed, the National Revolutionary Movement, was held together mainly by opposition to previous military governments. The Government nationalized the large mining companies, introduced universal suffrage and implemented land reforms that broke up the large estates in the Highlands.

2 The northern part of La Paz is in the Lowlands and forms part of the Amazon Basin. 5

1.16 From 1960 to 1981 various military governments controlled Bolivia. Each Government (regardless of political conviction) increased the role of the state in the economy. The 1952 revolution's promise of a more equal income distribution was slowly eroded as new powerful groups formed. These new groups were the (reorganized) armed forces, the miners' unions and various private interest groups. All parties agreed that the state was the vehicle for improving the lot of themselves and their supporters. Thus, instead of focusing on social reform, social service delivery, and infrastructure, the state became involved in productive activities. Most of these activities were heavily subsidized. The Bolivian Mining Corporation (COMIBOL) in particular became a heavy burden on public finances. Employees of COIBOL gained substantially relative to the rest of the population. They enjoyed substan- tial benefits in addition to an income in the upper 30% of the income distribution.

1.17 During the seventies substantial hydrocarbon discoveries, relative political stability and cheap money on the international markets following the first oil shock allowed Bolivia to borrow heavily to finance questionable inves..ents. Most investments benefitted the friends of the people in power. Around 1980, serious questions concerning macroeconomic performance and political stability arose. At the same time money became tight on inter- national markets, the inflow of foreign exchange stopped and Bolivia entered a very deep crisis during which the fiscal situation went out of control. Tha results were a hyperinflation and a fall of more than 20% in GDP per capita. In 1984 inflation reached 24,000% p.a. and the public deficit reached 25% of GDP.

1.18 Coming to power again in 1985, Dr. Paz-Estenssoro introduced an orthodox stabilization and adjustment program. Prices were freed; restrictions in the labor and capital markets removed; the currency was devaluated and liberalized, COIBOL was closed for restructuring, and the role of the state was redefined. The state should be a provider of infrastructure and social services, rather than a producer. Following a change in Government in 1989, bringing a coalition of the socialist Left Revolutionary Movement and the right wing National Democratic Action to power, the general thrust of the adjustment policies have been continued.3 The new Government has pledged its intent to improve social conditions in Bolivia.

For an analysis of the Government's economic policies see Updating Economic Memorandum (8623-BO). CHAPTER 2s DIMENONg OF POVER5Y

INTRODUCTIO

2.1 Definitions of poverty and resulting poverty indicators are many, and there is substantial disagreement on which are most relevant. In this report a household is considered poor if it lacks sufficient income to cover its basic needs. Lack of income earning potential can have many causes: lack of good education, poor health, inadequate housing, insufficient nutrition, poor initial endowments (e.g. land), discrimination based on gender or ethnic background, etc. An improvement in one of these conditions can, by itself, be regarded as an improvement in welfare. For instance, good education brings status, knowledge and social skills, and better housing or health conditions are beneficial in themselves. often poverty is described as a vicious cycle instead of the linear causality expressed here. Low income leads to lack of education as children must work and to poor health. Poor human capital development leads to low income, etc. (paras. 2.42-2.46).

2.2 This Chapter describes the various aspects of poverty in Bolivia. Poverty indicators are analyzed in global terms using international comparisons, over time, by rural and urban areas (including a look at migra- tion), by geographic region, by sector of activity, by ethnic group and by gender. The indicators used include: income, production, health, education, nutrition, public services and housing. Finally, poverty profiles are drawn up based on both qualitative and quantitative evidence. The conclusion is that the core Dovertv arOuRs are rural and-involved in small-holder agriculturM. This includes the peasants in the Highlands and Lowlands and the Amerindian tribes ir.the Lowlands.

The Data

2.3 Unfortunately the quality of data is extremely pour in Bolivia, especially for social indicators. The last census was done in 1976, and discussion continues about the reliability of the results. Apparently many questionnaires were lost and results for some areas were generated based on a sample that could not be checked for representativity. Lately, three major surveys have been carried out by the National Institute of Statistics (INE): the 1988 National Survey of Population and Housing (ENPV) supported by the UN, the 1989 Integrated Household Survey (EIH) covering the major urban areas supported by IDA,I and the 1989 National Survey of Demography and Health supported by the USA and done jointly with Demographic and Health Surveys (DHS). Several priva.e and international organizations have also sponsored surveys of specific regions, e.g., UNICEF's 1989 survey of poor areas of the Central Highlands.

2.4 The two national surveys (DHS and ENPV) have been heavily criticized for not reflecting the Bolivian reality. The results show very rapid improvement in social indicators even during the crisis. In the preface to the report on

This survey is comparable with the previous Permanent Household Surveys (SPH) done yearly through most of the eighties in the nine Departmental Capitals. 7

FNPV the investigators mention that "remote and isolated" areas of the country were not visited. This would suggest that the indicators for the rural areas are probably too optimistic, making the national averages too optimistic. On the other hand, the urban numbers from ENPV appear consistent with the numbers from EIH, while the DHS survey comes out more positive than any other source. In this report, the DHS numbers are used mainly in a relative sense, e.g., how children's illness relates to the mother's education. The absolute levels of indicators are not used. The non-official sources will be used inter- rmittently, mainly to illustrate points where the official sources are inadequate.

DESCRIPTION OF POVERTY

Poverty in General

2.5 Most Bolivians are poor by international standards. GDP per capita is estimated at US$580 placing Bolivia in the very low end of the lower middle- income countries world wide and at the bottom of the South and Central American countries. Social indicators would place Bolivia even lower. ENPV estimated infant mortality at 102 per thousand during the first year of life comparable to countries such as Zaire and much worse than Ghana, Kenya or India. Projections based on the 1976 cenaus show an infant mortality rate of 117 or comparable to countries in the low income group such as Bangladesh and worse than many countries in Sub-Saharan Africa. Maternal mortality is 48 per 10,000 live births, more than 10 times higher than other Latin American countries. The average population growth rate is estimated at between 2.2 and 2.7%, and the median age of the population is estimated at between 15 and 18 years.

2.6 Literacy defined as the ability to sign one's name or as any attendance in school is 82%, whereas functional literacy has been estimated as low as 50%. 23% of households have access to sanitation, while 60% receive (largely non-potable) piped water.2 Malnutrition is widespread. In 1980, 47% of children under the age of five were malnourished according to Gomez's classification.3 Recent indicators show a daily calorie intake of 65% oG requirements in (the poor city next to La Paz) and 59% in rural areas. Chronic malnutrition4 is also widespread, over 90% in some schools.

2.7 According to the last census, 80% of all Bolivian household were poor in 1976. A household was defined as poor if income could cover 70% or less of a

2 "Sanitation" includes any kind of waste-water or human excreta removal including public latrines. "Piped water" includes standpipes and other water supply outside the home.

Based on weight/age

4 Chronic malnutrition is measured as low height for age (stunting). |Dis3tri but ion of Ho use holdisI

20% 2 69

Total Population Agriculture 0 Non poor (non salarled) Poor R9 Very Poor 58% 1791 57_ ga13 Extremely Poor

Construcf Ion Services (non salarled) (non salcaried) Source: Morales. R.(83)

Figure 2.1t Poverty in 1976 basic needs basket.5 Two subcategories of poor were identified: the very poor whose income could cover 80% or less of the basic needs food basket, and the extremely poor who could only cover 30% or less of the basic food basket. Dividing the population by these criteria gives a distribution as in Figure 2.1 where "poor" refers to poor but not very poor. More than 95% of the extremely poor are non-salaried persons working in agriculture. This does point to the problem of the valuation of auto-consumption of production.6 However, the fact that rural, agricultural communities are among the poorest is supportad by other indicators as well, as discussed below.

Poverty Over Time - A Paradox

2.8 Data on real income7 in urban areas show a decline in the eighties. Between 1980 and 1986 real urban income fell by 37%. From 1986 to 1988, real income increased about 40%, resulting in a total fall from 1980 to 1988 of 12%. The fall is worse in commerce, services and manufacturing. This mainly reflects an increase in the share of employment offered by the informal

5 The basic needs basket used is developed by the Latin American branch of the International Labor Office (PREALC).

6 Estimates for consumption of own production are based on numbers for agricultural production and sales.

7 This section is based on Susan Horton's chapter on Bolivia for the forthcoming volume: Mazumdar and Horton (Zda.): Labor Markets in an Era of Adjustment. 9

Percentilesof Income 100

70 ...... --.------

90 ...... -...... --...

I0 aD------.. ------......

2030 ------. ------. . . - . -- ...--.------t- - -...... ,...-.; ...... ------

o .- ......

20 . .. . --. . _ .__ .

10 _ _...... ==== =....

00 0 10 20 30 40 50 60 70 80 90 100 Percentilesof Urban Vorking Population £982 1985 1988

Figure 2*2t Lorenz Curves for Urban Areas sector. 8 From 1980 to 1988, informal employment grew from 54% to 64% of total employment in the major urban areas. There have been important shifts in the sectoral shares of the informal sector. The share of informal employment in mining has grown from 16 to 20% between 1982 and 1988, while the share of informal employment in manufacturing grew from 59 to 68% in the same period.

2.9 Lorenz curves for distribution of income in urban areas in 1982, 1985 and 1988 are shown in Figure 2.2. There is very little difference between the income distribution for employed people between 1982 and 1988. From 1982 to 1985, the poorest 50% increased their share of total income from 18 to 22%, while the richest 5% doubled their share. However, this data was collected in September, immediately following the stabilization of prices and thus might reflect unsustainable relative prices. Between 1985 and 1988, the poorest 30- 50% have lost ground together with the top 5. Between 1982 and 1988, the poorest 3C, are slightly better off but still only earn around 12% of income.

Informal sector is defined as workers on own account (excluding professionals) and domestic employees, and employees, employers and unpaid family workers in establishments with less than five workers. 10

The richest 5% (who gained from the hyperinflation) were slightly worse off in 1988 compared to 1982. The income distribution has changed very little, but it does appear that the poor and the middle classes are slightly better off.

I Value Added in Agriculture 1980 - 1987 Agricultural Products Total and Selected Crops Million 1980 Pesos 25,000

20,000 __ / ......

20,000 --

5,000

0 1980 1981 1982 1983 1984 1985 1986 1987

2 Potatoes E Maize g Vegetables g Fruits i Leaf L Other Source:INE

Figure 2.3: Value Added in Agriculture

2.10 Data on rural earnings are virtually non-existent, but it appears that production has fallen over time, especially for the crops grown by the small Highland farmers.9 INE has estimated value added in agricultural production for a series of products. These numbers (presented in Figure 2.3) show that while agricultural incomes have increased overall, the increase is exclusively due to an increase in the value of coca production. Even though these numbers are of a doubtful quality in general and for coca in particular, they do confirm anecdotal and micro study evidence.

2.13 Value added from the principal commodities produced by the poor show a continued decline, with an extraordinary low in 1983 due to drought. If the 1980-82 average is taken as the base, real value added from potatoes (main income source of High-land farmers) dropped by more than 5% per year up to

9 See Chapter 3. 1.

1987. Vegetable and maize production (produced in the Valleys region by small farmers) fell by It per year. In sum, this would suggest that earnings of rural smallholders fell in real terms, except for those who have taken part in the coca boom. These farmers, while better of in monetary terme, are faced with all the negative social consequences of growing an illegal crop. For the purely subsistence farmer only production volume is important not value-added. However, very few Bolivians are in a purely subsistence economy. Even the poorest depend on the market for goods such as salt and cooking oil. Anyhow, both volume and value added of the goods produced by the poor have decreased, indicating a fall in welfare.

2.12 All social indicators show improvements over time, while social spending has declined. Table 2.1 summarizes these results for 1980 to 1988. The 1980 numbers are projections based on the 1976 census. The expenditure data is 1980 1998 Cing based on the Bank's recent Educaton: Publ.c Sector Expenditure JiMit Rate 36.8% 18.3% -50% Revic- (PSER).A° All UAM ~ 15.2% 7.6% -50% indic.ators improve and all Runl 53.1% 31.1% -50% expenditures fall, a ExpcndkW (tmWin1980 p@s) situation common to Tota 5020 3011 40% several Latin American Bsic 3212 1807 -44% countriesil and to many of -pr studen 3339 1416 -58% the countries facing (rml 1980 pmo) economic crisis and/or adjustment. This would infantMoua ty 123 102 -17% suggest huge productivity Empeadivr.s: (1987 US$ per capiu) improvements in social TON to MPSSP 6.6 2.5 462% service delivery. Som: PSE WNE 2.13 Several explanations can be put forward for ! !v_5o}**'.*-.+!.9-..-E=t- St4.zw-W-this apparent paradox. Table 2.1: Social indicators and Expenditures First, there are serious data Droblems, especially in Bolivia. The expenditure data have to be deflated through a hyper- inflation and the social indicators are questionable, as discussed above. Some expenditures might not be reflected in the official expenditures as Non- GovernmentalOrganizations (NGOs) (or Private Development Institutions (PDIs), as they prefer to be called) have played an increasing role in the social sectors financed mainly from abroad. The substantial increase in food donations (see Chapter 7) is also only recorded to a limited extent. In some countries it has been argued that the distribution of expenditures is important. The fall seen in expenditures is mainly a fall in real wages of

10 Bolivia: Public Sector Expenditure Review With a Special Emphasis on the Social Sectors, September 1989, Report No. 7746-BO.

Social Spending in Latin America: The Story of the 1980s, (Report No. 8450-LAC), The World Bank, forthcoming 12 the employees in the sector. This will have longer run negative incentive effects, but not necessarily immediate, negative effects on the indicators. However, in Bolivia non-salary items have fallen more than salary items, so that is not a reasonable explanation. A related argument is that inputs, not expenditures, are important, i.e., though inputs furnished by the public sector fell, most i-.putsin schools are provided by the parents and are thus not reflected in the statistics. However, as incomes fell over the period, it is unlikely that private expenditures increased substantially.

2.14 Another hypothesis, best explained in a production function framework, is that an improvement in knowledge of diseases and the importance of educa- tion is more like a ca2ital stock. Once a mother knows about oral rehydration or the importance of vaccinations that knowledge stays with her. Once a person is literate that will not change over his/her life. As long as expen- ditures are high enough to allow for "reinvestment," i.e., high enough to allow the capital stock not to depreciate, social indicators will improve. At times, this has been called the "inertia" effect. Social indicators show much auto-correlation, and changes over time are relatively non-dependent on short run fluctuations in expenditures. A related explanation in based on time- laos. The fall of expenditures seen in the eighties will not have an effect until the nineties. By the same token, the improvement in social indicators seen lately is due to higher expenditures in past decades.

2.15 In conclusion, available data show that social indicators have improved over the last decade while social sector expenditures and urban real incomes have fallen. Rural incomes also appear to have fallen especially in the Highlands. Thus, the findings of the 1976 census that 80% of the population earn incomes below the poverty line and 60% are very poor are likely not to have improved. The relative situation of the poorest of the poor has probably gotten worse in rural areas and stayed relatively constant in urban areas. If these trends continue, it is likely that the gains observed in the social indicators will be reversed. Chapter 3 will discuss issues related to the social costs of the crisis and adjustment.

Poverty in Rural and Urban Areas

2.16 The poorest areas in Bolivia are rural. Table 2.2 compares a series of statistics between rural and urban areas. The data for water and sanitation are indicators of urbanization, and thus should be interpreted with care. Rural populations are younger, more children are born in rural areas, and infant mortality is higher. Illiteracy is a more serious problem in rural than in urban areas. Household sizes, both in absolute terms, and relative to the number of rooms are alike in rural and urban areas. A larger share of the rural population speaks a native language. Households headed by women are more prevalent in urban areas. This is probably because men keep their relationship with the rural villages even when they move away, and thus (formally) remain as heads of households.

2.17 These average figures mask huge differences within urban and rural areas. The indicators for some smaller towns (between 2,000 and 10,000 inhabitants) are a lot worse than the average for urban areas, but still nowhere near the rural levels. As an example of the differences within rural 13 areas, a recent survey of four communities found s.let i2xnArr below-five mortality rates ranging from 127 to 415 Indicat,r Total Urb' wtuf$ per 1000 live births. Jfnt ortatity (r thouand) 102 a3 120 Some well-established BIQsl fertility Rate(per inmw) 5.0 4.2 6S1 communities in the oA gdi 18.4 19.0 18.2 Lowlands and Valleys have Percent of ChiWren OUner10 2 27X 302 good indicators and well t iteItry Rate (X 14 years) 19% SX 31X functioning social SchootAtteakne (ZX6-19 yas) 74X1 84 64& services while some communities in the Houslhots USth Piped Water 601 89 31X Highlands or among the HehotdW With Sanitation 232 42% 31 indigenous groups in the Person per Hoehold Roe 1.76 1.74 1.80 Lowlands have no access to Sixe of Hosehois (person) 4.56 4.44 4.49 public services and face a miserable existence based Houeotd Speakitg Native anguage 51X 361 68x on very low productivity Noehotdsfeoded by Wasn 14X 18X 14% agriculture. Source:INE, 198 2.18 The average annual growth rate of the urban Table 2.2: Social Indicators for Rural and Urban population between 1976 Areas and 1988 was 4.4% while the rural population only grew 1.2% per year. In 1988 it was estimated that 51% (3.3 million) of the population lived in urban areas, up from 44% in 1980. Most of the rural-urban migration goes to the Capital of the Department from the interior of each Department. Only the city of EL Alto12 attracts people from other Departments in greater numbers.

Poverty by Reaion

2.19 Several poverty maps have been generated based on the 1976 census. The poverty maps are based on a variety of indicators at the Provincial level, including infant mortality, illiteracy, water supply, migration, and agricultural production." Figure 2.4 shows such a map developed by the National Council for Population Policy. A group of provinces in Southern Cochabamba, Northern Potosi, and Northern Chuquisaqa, the Central Andean Region, is depicted as poor in most of the maps.

2.20 Most of the poorest provinces are inhospitable, remote areas. Despite their poverty, few institutions work there. The public sector is totally

12 1EAlto was separated from La Paz in 1987. It is a recently settled city consisting of largely unplanned urban areas surrounding the international airport of La Paz. is Annex 2 contains a description of five different poverty maps and presents another map dividing the country into S poverty areas. These five areas will be used later in the report. 14 IBRD21869

PoorestProvinces

.- |ProvinceBoundaries -RLIntemational Boundaries

2~~~~~,,

ARGENTINA......

,. _ ^ .L f A~~~~~~~~~AeaofMa IPAAIJ1TARAUA ERAZIL

0imes100 200 300 BO LMAl _EC~~~~~~~~~~~~~~Ae0 100 200 ofMap >

I ARGENTINAKh

Jigure2.4 DISTRIBUTIONOF THE POOREST PROVINCES SPINRw 15 absent, with the possible exception of works financed by the ESF, and PDIs are only active if the areas are not too remote. Even then they generally do not reach the poorest areas, but remain close to the provincial capital. However, there are a couple of notable exceptions where PDIO have managed to obtain almost total coverage of a Province with very important results (see Chapter 4).

2.21 Province-level data obscure many intra-provincial differences. There are generally large differences within a province, especially between the provincial capital and isolated rural areas. Recently, efforts have begun to break the 1976 information down by Canton, the next administrative level down from a Province. Such efforts are unlikely to yield much new information as the numbers are 14 years old and of doubtful quality (para. 2.3).

2.22 In 1976, the poorest Departments were Potosi, Area Large Children/Iltite- No Poor* Pando, Chuquisaqa and Oruro. household household racy schoolin water Aggregating at the level of Chuqufsaca-Rural 13 5 I 1 3 Department covers even more Potosi-Rural i5 13 2 2 7 heoterogenousareas. As Cochabamba-Rural17 9 3 3 5 heterogenousareas. AS ~Tartja-Rurat 3 4 4 4 1 shown in the poverty maps, LaPz-Rural 16 17 5 5 4 many departments cover both Oruro-Rural 6 3 6 8 9 poor and non-poor areas, and Beni-Rural 2 1 7 6 8 Pando-Rural S 6 8 7 2 provinces from 1 to 5 on the Santa Cruz-Rural 4 7 9 9 6 poverty scale (see Annex 2). PotosSiUrban 10 8 10 10 15 Unfortunately, much data is Chuquisaca-Urban 9 16 11 12 16 only available on a Depart- Tarija-Urban 8 12 12 11 13 La Paz-Urban 18 18 13 Is 18 mental level, and thus Oruro-Urban 12 IS 14 17 17 generalizations such as Bent-Urban 1 2 15 13 12 "Potosi is poor" or "Santa Santa Cruz-Urban 11 10 16 14 14 Cruz is not" need to be Cochabn"-Urban 14 14 17 18 10 made.* Panda-Urban 7 11 is 16 11

- poorwater" refers to bad sources 2.23 The ENPV data can only be broken down by rural and SrceNPY CNE) urban areas within each - Department. Table 2.3 shows Table 2.3: Poverty by Area From ENPV the ranking of these areas according to several indicators. A ranking of 1 is the poorest area; a ranking of 18, the richest. Rural areas in Chuquisaqa and Potosi are still the worst off in many aspects. Only in housing are there substantial problems in Beni, both in rural and urban areas. Most indicators are worse for rural areas than for urban areas.

2.24 Regional differences have increased over time. Numbers for the Central Andean Region for 1988 show substantial improvements in the absolute level of infant mortality and illiteracy but deterioration relative to the national average. I.e., these poor areas have improved less than the national average. Figure 2.5 shows infant mortality and illiteracy rates in 1976 and 1988 for seven provinces in the Central Andean Region. The numbers presented are indices using the national average as a base. All provinces except Zudanez were relatively worse off in 1988 than they were in 1976. 16

Relative Worsening of Core Poverty Areas Illiteracy and Infant MortalIty Rates Comparedto NationalAverage

Index (National Average - 1) 4

2-....~~~~~~~...... I 3-* .------1------IS--.-...... -N--......

Arce Mlzque Campero Oropeza Zudanez Bllboo Charcas Ibanez Provinces Infant Mortollity~ Infant MortalityfS Illiteracy\ Illiteracy t 1976 V2J 1988 131976 |1988 Sources±IE.SVSST

Figure2.5: RelativeWorsening of Core PovertyAreas

2.25 Reaional migration takes place mainly from the Highlands to the Valleys and Lowlands. Figure 2.6 shows gross migration patterns among Departments. $he positive numbers indicate the number of people in the Depa.rtmentin 1988 who said they lived in another Department in 1983. The negative numbers show the number of people living outside the Department in 1988 who lived in the Department in 1983. The main recipients of migration are the Departments of Santa Cruz and Cochabamba. In Cochabamba this could be due both to better soils and to the attraction of the main coca growing region (Chapare). Host of the migrants to Cochabamba come from the Highlands (La Paz, Oruro and Potosi). Migration is also important between neighboring departments (e.g., oruro and La Paz) mainly reflecting rural-urban migration to the larger cities.1' There is also a tendency to move gradually down through the al-

14 The migration from Oruro Department to La Paz Department also figures in the data for migration into the city of El Alto (para. 2.18). Otherwise, as was stated above, the largest share of migration to the cities is from within the Department and, thus, is not reflected in the inter-Departmental migration numbers. 17

GrossMigrafton Plow AmongDpfef

70 60 it 50 To OpWl 40 30 20 1.~~~~~~~~~~~~~~~~~~~~~~.

0O 10

'U

(0)

( 60) CHQ LPZ CBB ORU POT TAR SCZ BEN+PAN . ESa~~~Cituqulsoqo eD La Paz E9 Cochabantba 1Oruro

Figure 2 . 6: Gross Migration Flows Among Departments titudes, e.g., the migrants to the Valleys come mainly from the Highlands, while the Lowlands attract people from both the Valleys and the Highlands (e.g., the largest single group of migrants to Santa Cruz is from Cochabamuba).

2.26 2igration of the population from the Highlands to the Lowlands is not new. Traditionally the Highlands were closely linked with the Valleys by temporary migration.os Migration was stimu'ated on a more permanent basis by the various "boom"eperiods of crops with high labor requirements in the Lowlands. The first of these was the collection of quinine in the 19th century, which was curtailed by the rising production in Africa. This was followed by the rubber boom in which the native plants in the jungle were tapped. This boom ended with the large rubber plantations in Southeast Asia. More recently temporary and permanent migration has been stimulated by the coca boom. In these cases the migrants had ready access to the production base, either in the form of wild native trees or land to cultivate. These examples suggest that when a high value crop, which has high labor

15 See paras. 7.17-7.28 on land use and land reform. 18 requirements and an expansive market is identified, and migrants have access to the production base, migration to the Lowlands will occur spontaneously.

2.27 In conclusion, poverty is geographically concentrated in the Altiplano and high valleys in the Central Andean Region. Outside this region there are poor provinces in Pando, in the Chaco region of Santa Cruz and in southern Potosi. Regional 'ifferences in social indicators have increased over time though most areas have seen improvements in absolute terms. The regional differences have induced substantial inter-Departmental migration mainly from the Highlands to the Lowlands and Valleys.

Poverty by Sector and Occuipation

2.28 The poorest of the poor are the households of rural smallholders. Since the last census their real incomes have fallen, their social indicators have worsened relative to the national average, and they have continued to migrate to the cities. Typically the urban poor are found in small family businesses involved in commerce or simple manufacturing.16 Among urban occupations, the domestic workers earn the least, followed by people in family businesses and workers on own account. Among sectors, commerce pays the least followed by manufacturing and personal services. Average earnings in the Departmental Capitals for employees in household enterprises were Bs.125 per month in 1989.

Poverty by Gender 1 7

2.29 Women have lower educational levels and higher illiteracy rates than men. ZNPV put female illiteracy at 25% (44% in the 1976 census) versus 14% for men (21% in 1976). Of an estimated illiterate population of one million, 65% are women. Rural women exhibit the lowest literacy rates of all groups. Women earn less and have fewer training and employment opportunities. Ninety- five percent of women (and 60% of men) do not have access to credit. Some 81% of women in marginal urban areas are working in the informal sector.

2.30 Women's poor health and nutrition status and low standards of education are leading causes of poverty. Poor health diminishes women's productive capacity and endangers their personal and family well-being. Illiteracy constrains women's social and economic participation. Long-standing social discrimination based on gender denies women equal access to social, political, and economic power. Women are constrained by cultural norms and behaviors (notably "machismo") that accord men rights and privileges not enjoyed by women. This touches upon all aspects of family and social life. Women also confront discriminatory practices in the areas of employment (pre-employment pregnancy tests and lower wages), housing (national Housing Fund self-help housing program directed at male heads of household), and law (family code stipulates that husbands must give wives permission to work). The agrarian reform discriminate against rural women in that only men and widows have the

16 Data based on EIH and other urban household surveys.

17 This section is based on the background paper: "Women and Rural Poverty in Bolivia." 19 right to obtain land through the reform process, and rural syndicates are established for male heads of household.

2.31 Rural women bear more children than their urban counterparts. Rural women have poorer health and nutrition status than urban women. Seventy-five percent of the rural illiterate population are women. Rural women in the Highlands perform an estimated 80% of agricultural and animal husbandry labor (with children), working on average 100-120 hours per week. In the Lowlands, women's time spent in production, domestic labor, and child care averages 14- 16 hours per day. The usual pattern is for women and men to work together during the short periods of planting and harvesting (3-4 months). Then the men migrate temporarily to exchange products in other areas or to secure casual labor. In men's absence, women and children perform all activities related to agricultural and animal husbandry besides domestic chores and child care. Rural women from the Highlands have the advantages over their Lowland counterparts of the relative stability of Andean society and culture and what is characterized as a "benign" environment. Migrant women living in settlement schemes may enjoy a slightly higher income level attributable to a larger proportion of production directed at the market and greater off-farm opportunities for men. These do not, however, obviate the daunting task of adapting to a new environment.

2.32 Although women's agricultural role is equal to or greater than men's, they do not participate in agricultural development activities because of language barriers; low educational attainment; dominance of men in leadership positions; the "double day" of production work (heightened by male migration), domestic labor, and child care; and the lack of attractive incentives and support structures to overcome male resistance to women's participation, especially with regard to decision-making.

2.33 Mining women constitute a rural poverty group because some mining families have yet to be successfully reintegrated into the economy following the 1985 lay-offs, and because miner's wives have been abandoned as their husbands relocate to seek work. Good data on this population is not available, but one example suggests the gravity of the situation. The Red Cross in Potosi has identified 300 women who have been abandoned by their husbands and are living in mines with their children. The women collect mineral waste to sell to buyers who pay in either cash (at 1/10th of subsequent resale price) or goods (bread, rice). The women and children do not have access to medical care or education. Once a month, the Red Cross delivers 40 grams of powdered milk, oats and margarine to each woman. No other agency provides assistance. 20

Ethnic Grouns and Poverty

- Language Spoken and Social Indicators 110% 1 0 0% ...... 969s ...... 90%-S 80%i~6^7---... . 70% _ 70. 60%- 50% .....- ...... - ... ..5 ..

30% . .. ' ...... 2. ...1 ...... 20% _...._...... T..

LI Spanish Only S Spanish/NativeLang. Native Languages Source: INE (ENPV)

Figure 2.7: Native Languages and Poverty

2.34 The relationship between culture/language and poverty is striking. One- half of the population is of Indian origin (largely Quechua or Aymara). Figure 2.7 summarizes the relationship between language of head of the house- hold and various social indicator.. While 69% of Spanish speaking households have access to piped water, that is the case for only 21% of households speak- ing native languages. Ninety-six percent of native language heads of house- holds have no education, and none have technical or university education. Caution should be used in positing a causal relationship. The argument reflects an unfounded, deterministic prejudice against the indigeneous culture, i.e., a Quechua/Aymara speaker is somehow "destined" to be poor. One recent studyl concludes that the inequality between the Indians and the Spaniards is due to class and not ethnicity. Different factors influence occupational success for the Indians and Spaniards. For the Indiana the

19 See Jonathan Kelley: Class Conflicts or Ethnic Onr2esion? The Cost of Beina Indian in Rural Bolivia, Rural Sociological Society, 1988. 21 father's occupation is crucial whereas for Spaniards education is crucial. According to the author, these two economic worlds are equally rewarding. The (few) Indian sons born into families with an education and occupation typical of a Spanish, did just as well as Spanish sons. So, the argument goes, inequality in Bolivia is a matter of class and not ethnicity.

2.35 it is, however, very hard to believe that no discrimination is taking place as the output indicators show an almost perfect correlation with the incidence of indigenous languages. If access were equal, such differences would disappear over time. Whether discrimination is based on class, race or ethnic background is almost purely of academic interest. The fact is that a large group of poor Indians of a native ethnic background has remained at the bottom of society.

2.36 The indiaenous population of the Lowlands is another core poverty group. Twenty-eight different languages are spoken among the indigenous people of the Lowlands. However, many groups are bilingual as Spanish is used in the education system. Although some groups have abandoned their tradi- tional culture, there is an increasingly strong sense of consciousness or "reevaluated nativeness" among the indigenous groups.

2.37 Lack of property rights to large enough territories is the single most serious problem facing the Lowland indigenous peoples. It implies that many are not able to maintain their traditional way of life combining hunting, fishing and gathering with small scale slash-burn farming. Previously, many indigenous groups were nomads but today almost all grotupshave settled in small communities and live from mainly subsistence agriculture often combined with hunting and fishing and/or off-farm employment for commercial farmers, cattle ranchers or logging companies. Most indigenous groups have limited or no access to public services within health, education, agricultural extension, construction and maintenance of roads, and provision of drinking water and electricity. The assistance that is taking place is almost exclusively financed by NGOs operating in the area.

POVERTY PROFILES

2.38 The living standards of the groups identified ka poor in the 1976 census have deteriorated relative to the population as a whole, or (especially for income Indicators) have deteriorated in absolute terms. Thus, the characte- ristics of "the poor" identified in 1976 remain largely valid. Based on that information updated with the more recent data presented above, the following characteristics increase the likelihood of being poor:

a) living in a rural area,

b) owning little land,

c) being female,

d) being of Indian origin, 22

e) living in the Central Andean Region, and

f) working in agriculture or household industries.

2.39 As the 30% poorest in 1976 were rural smallholders, and as their production has fallen since, it appears reasonable to assume that rural poverty remains the most severe. Core povertv arougs thus include:

a) Rural smallholder families in the iahilandsespecially in ioolated areas where Aymara and Quechua dominate;

b) Rural women, especially heads of households;

c) Indicenous croucs in the Lowlands,

d) Recent miarants to the Lowlands.

2.40 Whether the recent migrants to the Lowlands should be included is a contested issue. One argument states that people migrate voluntarily to the Lowlands because life is better there, i.e., recent migrants to the Lowlands are better off. This, however, assumes no transaction costs (that people could move back freely without costs) or perfect foresight. Unfortutunately, very little information is available on incomes to compare levels of welfare. A sample of farm budgets from the Lowlands show good potential for non- mechanized agriculture in terms of food security, but only limited possibilities for cash income. As long as cash income remains an unrealized potential, the migrants remain poor.

CAUSES OF POVERTY

2.41 The "causes* of poverty are many. They can be described as in There is a peasant out the Wo.i8 Box 2.2, i.e., as a linear causality *inq. There is8 culture out there that from the four basic problems isbe tosurive. ItisestruegtswiO specified to low income. A more bast survat evey day.te mod totee,hov common way is to present a cycle of w O1ILnpart of thepftithir sutatio 1 poverty where each issue is tleir needs and their survival strategies as interrelated. For instance, poor jfven a bel bein to hetpthem hetp health leads to low income, which leads to no education, which leads to scS tv f te wee tlmr poor nutrition, which leads to poor 4fscWs1#* cause Vd effect, theortical health. No matter which causality Xaeectsof povertyandieological links are postulated, the causes ,5 Thor 1* n ime fo debate "Winte, T Isn8 t1Dei fordXete; hat4s' specified in Box. 2.1 are the reasons r*W ais for their poverty given by many poor and their representatives. This Gartaarcia,CEEAGRO report will take a pragmatic view and discuss how each of these reasons for Box 2.1t wDoia Narta" poverty could be affected by a given set of policies. 23

1) Poorly2e2welorae4d HIM Fasital.

A. Low tevelof education Lackof educationlowers productivity and inhibitssocial organization B. Badhealth conditions Goodheatth is an improvementin livingstandards by itsetf,and is directlyrelated to productivity. C. Poor NutritionalStatus A matnourishedchild cannot tearn and is more tikelyto fallitl. Adultmalnutrition lowersproductivity.

2) Lackof PhysicatAsset..

A. Land The problem is secure access to good land. S. Technotogy Espectallya problem in agriculture where there is a serious lack of appltied research and extension C. Credit Creditto the poor is ontyavailable from the informalsector at prohibitivelyhigh interestrates.

3) No Provisionof Infrastructure

A. Roads Rurat roads are non-existent or in very poor condition. B. Water Besidesthe direct welfare improvement of cleanwater, poor water suppty is one of the waincauses of poorhealth C. Irrigation Nainlya problemin the Highlands.

4) Discrimination(ender or race).

Results in a loss of output as resources are not altloedto move to theiroptimal use.

Box 2.2: Reasons for Poverty

2.42 Most of the reasons described in Box 2.1 are not specific to Bolivia, some of them take different forms or are especially serious in Bolivia. The lack of sufficient development of human capital is more serious in Bolivia than in other countries at comparable levels of economic development. Human capital is so poorly developed, that it will become the major obstacle to growth in the medium-term. As long as students learn virtually nothing if they attend school at all, large parts of the population are ill and malnourished, Bolivia will be mired in a low productivity and thus poverty (see Chapters 5 and 6).

2.43 In terms of asset distribution, Bolivia has a better land distribution than many other countries, at least formally. The problem is lack of secure property rights to good land. Without such rights, there is no incentive to attempt sustainable farming methods, and productivity suffers (see Chapter 7). Technology transfer is a serious problem especially in agriculture. Again productivity suffers and thus incomes remain low. Credit is a problem for the 24 urban informal sector and for the farmers with marketable surpluses, not for the subsistence farmer. This report will deal only with rural credit. 1 9

2.44 Insufficient provision of infrastructure is very serious in Bolivia. Through the first half of the eighties, public investment almost disappeared as the public sector lost revenues. Since 1986 the ESF has helped alleviate the need for the provision of public services and infrastructure to the poor. It is often argued that it is prohibitively expensive to provide infrastruc- ture to the poor. The conclusion of Chapter 9 does not corroborate this hypothesis. If the technological level of infrastructure is adapted to local needs and capabilities, it is equally cost effective to provide services in poor and in non-poor areas.

2.45 Discrimination is a serious issue in Bolivia. Unfortunately, it is hard to evaluate exactly the extent of discrimination. Discrimination would usually take place on the input side where it is hard to detect, e.g., lack of access to education, lack of health care, discriminatory practices in the labor market, etc. On the output side, the results are very clear. As discussed above, people of indigenous ethnic background are more illiterate, have higher infant mortality rates, etc. The high level of maternal mortality (the highest in the Hemisphere) shows the lack of concern and action to improve the health situation for poor women.

For urban credit, the basic keed is to subsidize the setting up of private credit institutions through technical assistance for the non- banking (development oriented) activities. Several PDI experiences have been positive in this field. 25

CIAPTER 3t MACROECONOMICPOLICIES AND THE POOR

INTRODUCTION

3.1 This Chapter describes the macroeconomic situation and macroeconomic policies and their effects on the poor. The focus is on the eighties. The presentation is divided in two main sections: (i) the interventionist policies and serious economic crisis of 1980-85; and (ii) the liberal policies and economic recovery after September 1985. The conclusion compares the effects on the poor of the two periods.

3.2 The basic thrust of this analysis is that the inappropriate macroecono- mic policies of the early 1980s had a deleterious impact on the poor, both in absolute terms and relative to other income groups. Although the Government should avoid distorting key prices in the economy (for example, interest rates, exchange rates and public sector prices) to help the poor, it should take into account the impact of its macroeconomic policies on the poor. Increases in taxes can assist the poor by financing programs to alleviate poverty. Devaluation can have a considerable role in increasing the incomes of poor agricultural producers, although the magnitude of this effect is difficult to measure.

IMPACT OF THE 1980-85 ECONOMIC CRISIS

3.3 Bolivia experienced a disastrous fall in output and income in the first half of the eighties. GDP fell by 10% between 1980 and 1985, and by 24% in per capita terms. Per capita consumption is estimated to have dropped by 16% over the same period. There is some reason to believe that these estimates overstate the fall in output, however, extreme rates of inflation coupled with extensive price controls provided a growing incentive for the channelling of production through the informal, and mostly unrecorded, sector of the economy. As the illegal trade in cocaine rose considerably during this period, the fall in the recorded economy was (at least to some extent) compensated for by the rise in coca profits. The difficulty in interpretation is compounded by the lack of reliable time series data on welfare or employment. Still, it is very unlikely that the errors of measurement were so great as to reverse the generally-accepted judgment that the Bolivian economy suffered a drastic decline in output and welfare in the early eighties.

3.4 Although we lack time series data on either the distribution of income or welfare, the poor probably suffered relatively more than other groups during the pre-1985 crisis. Economic chaos and hyperinflation provided a premium to access to information and political power. Since the poor lacked both, they tended to lose out in the sharp competition to maintain living standards in the face of the economy's decline. The process by which the poor were penalized during the crisis can be seen in the distribution and financing of Government subsidies (see below). 26

Government Policies and the Poor

3.5 The Government had a large role in the al,ocation of goods and services in the economy, both through production by state enterprises and through sub- sidies. Direct transfers from the Treasury to support certain productive activities, directed credit at subsidized rates, and access to foreign exchange at the controlled (and heavily overvalued) rate were allocated based on compromises among the important political actorss the Government, the military, the labor unions, the banks, and some richer Bolivian families.

3.6 While data on the distributional impact of Government subsidies is not available, anecdotal evidence suggests that the poor rarely benefited. For example, the allocation of agricultural credit was heavily skewed to the large estates in the Lowlands. Little of the credit provided by the Govern- ment found its way to the campesinos. The subsidy provided by agricultural credit alone was estimated to equal about 1% of GDP from 1982-85. Similarly, banks (or their owners) were in the most advantageous position to profit from investments in real estate and other real assets (the prices of which rose sharply during the hyperinflation), and to make speculative profits through access to foreign exchange at the overvalued official rate and through access to credit at subsidized rates. Thus, a recent review of the Bolivian banking sectorl concluded that some banks survived the hyperinflation in relatively good financial condition. The unions representing civil servants and workers at state enterprises could at least soften the fall in the purchasing power of their salaries, particularly through the proliferation of bonuses and in-kind benefits. By contrast, public sector spending on programs to benefit the rural poor, including agricultural extension services, health services, and education (particularly spending on basic education) fell sharply in real terms over this period (see below).

3.7 While the poor had little success in participating in the allocation of resources by the state, they could not avoid helping to finance the ever- growing budgetary deficit that resulted. As the economy entered hyperinfla- tion, it became impossible to maintain the real value of tax receipts, while cuts in expenditures did not keep pace with the precipitous fall in revenues. As a result, the budget deficit shot up from 9% of GDP in 1980 to 31% by 1984. As the Bolivian Government lacked access to either domestic or external loans, this deficit was financed largely through inflation, a regressive tax that penalizes those most unable to protect the real value of their assets or incomes. The size of the implicit inflation tax is generally rendered as the rate of inflation (which is the rate of thrjimplicit tax) times the money supply (the tax base). From 1982-85, the inflation tax averaged about 14% of GDP a year.2

3.8 The incidence of tax on the poor varied greatly, depending on geogra- phical location and relationship to the market economy. Campesinoa far from

Bankino Sector Study (Report No. 6765-BO, dated November 1988).

2 The inflation tax is calculated from monthly data on the consumer price index and the money supply. 27 urban areas who depended on the market for consumption goods were most severelyaffected, as they lacked the informationnecessary to anticipate price movements. On the other hand, campesin2osnearer to the cities and on the frontiersmay have benefittedfrom the opportunitiesfor speculation. Price controls on many agriculturalproducts, including coffee, milk, butter, rice, sugar, oil, wheat derivatives,meat, chicken, and eggs, maintained officialprices well below the market. These controlsmay have reduced campeslnoincome, but were to some extent evaded as inflationincreased. However, campesinoproducts are bought by intermediaries(most often) at the farm. Thus, camnpesinoscould only benefit from non-controlledprices, as far as the intermediarypassed on the rent gained by evading the controls. In suam,it is thereforedifficult to measure the impact of price controls on agriculturalproducers.

rmnpacton Agriculture

3.9 Some aspects of the crisis benefitedagriculture, which may have cushionedthe decline in campesinoincome. The hyperinflationresulted in a sharp real depreciationof the peso in the parallel market, which provided some protectionto domesticproduction of agriculturalgoods. While the officialrate was highly overvaluedduring most of the hyperinflationary period, the parallelexchange rate was a more accurate index of the market price of foreigjnexchange. 3 The parallel rate with the us dollar shows a considerablerise in real terms from 1982 to August 1985, when stabilization took place. This real depreciation,despite the dramaticrise in inflation, occurredbecause hyperinflationengendered a strong demand for foreign exchangeto protectthe real value of assets. Respondingto this shift in demand for assets, the parallel exchangerate rose more rapidlythan would have been necessaryto maintain equivalencebetween the real price of trade- ables in Bolivia and abroad.

3.10 In addition,agricultural production benefitted from other policies and effects of the hyperinflation. For example, it is possible that the hyper- inflationincreased the demand, at least temporarily,for some agricultural goods as hedges against inflation. In 1983 agriculturalprices were liberalized,freeing output prices while controllinginput prices. Specifically,agriculture was subsidizedheavily by the Government'sdecision to maintain the price of gasolinewell below internationallevels, because the marketingof agriculturalproducts uses transportationintensively. This subsidyprobably affordedfurther protection to campesinoproduction, although it is not clear whether the campesinoaor the transportmonopolies reaped the bulk of the benefit.

3.11 In conclusion,the economic policies followed prior to August 1985 had a severe impact on the rural poor, althoughsome aspects of the macroeconomic environmenthelped to cushionthe loss in welfare. The identityof winners and losers from the hyperinflationand extensivestate controls on production

Althougjhthe parallel market was illegalduring much of this period, the rate was widely used. 28 depended greatly on geographical location (including rural/urban), market access, and particular market conditions.

IMPACT OF THE STABILIZATION IN AUGUST 1985

3.12 The abrupt stabilizationof the economy and comprehensive liberalization of markets, which took place in late 1985, imposed a sudden and severe change in the structure of economic incentives facing Bolivian producers. This section analyzes the impact of these policy changes on the poor. It is useful to differentiatebetween policies aimed at stabilization and those designed to liberalize markets, although in practice the two policies are strongly inter- related, and individual policies affect both. The specific policies under- taken and their general macroeconomic effects are analyzed in the recent Country Economic Memorandum for Bolivia (7645-BO). Here we will focus on their consequences for the poor.

3.13 The Government of Mr. Paz-Eotenssorostopped the hyperinflation by limiting public sector expenditures to cash on hand (inducing a fall in central administrationreal salaries) and by establishing a realistic exchange rate through a 93% depreciation of the peso. As a result, the consumer price index, which had risen by 57% in August 1985, fell by 1.9% in October 1985. With a few exceptions, price increases have remained below 20% at an annual rate since then. An end to the hyperinflationwas essential to avoid an even greater collapse of the economy than actually occurred, and was a prerequisite to long-term growth. The control of inflation is generally recognized as a major contribution of the Paz Estenssoro Government to Bolivian development.

3.14 It is less widely understood that stopping inflation probably had a positive impact on output and welfare.4 Stabilizationmade it unnecessary to engage in numerous time-wastingactivities that had become common during the hyperinflation. For example, the rapid fall in the purchasing power of currency led people to spend money as soon as it was received. By contrast, low inflation enabled the consumer to space purchases more convenientlyand reduced the accumulation of real assets as a hedge against rising prices. Stabilization allowed people to spend more time in productive activities and less in speculative ones, facilitatinga rise in output.

3.15 Unlike the experience of many countries, but similar to earlier hyper- inflations, there was no tradeoff between stopping inflation and growth. Economies with high rates of inflation frequently suffer a fall in output with stabilization. Economic agents under inflation typically make decisions and

Output did fall by 2.9% in 1986, but this reflected to a large extent the collapse of Bolivian tin mining exports owing to the sharp drop (50% over the year) in internationaltin prices in late 1985. The prices of Bolivia's other major export-good, natural gas, fell by 12% in the same year. Export earnings from these two goods fell by US$130 million in 1986, equivalent to 3% of GDP. Some demand effects must have had negative consequences as well as real income fell during 1986 before picking up in 1987. 29 enter into contracts assuming continued inflation. An abrupt Poncho is 10 ot 11 years otd. He is a shoe-shine boy change in the monetary on one of the sqares in the otd part of La Paz from 7 in the regime that stabilizes mornfngto S at night, interrWted by a few hoursof school prices can result in in the afternoon. Panchodoes not wakeauch but his income rshis family's most ifrt,nt sore of incoW. His family unexpected changes in consits of his motherad foursiblings: "Wedon't havea relative prices; for ex- father, I'm the man of the house." ample, increases in real Pancho said that "thirgs are muchbetter nowthan interest rates or real wa- before." Whenasked about the hyperlnflation, he gave a ges. Such unexpected biank stare, not understanding the words. #But," he said changes can severely re- eagerly. *the people in power before ruined mymother's duce the profitability of money." Full of Indignation he explained that his mother had savedup her moneyto buy a pit. AS she did not trust the existing firms. However, banks, she hid her moneyin a tin can under the bed. When this effect of stabiliza- she cams*own to the market to bu her pig, the farmer tion was not relevant at laughedat her and said that her moneybas worth nothing now. the extremely high rates "So wedon't like the people whodecided then. They cheated of inflation experienced us out of a pig." in Bolivia. By August of Panchogoes to a center for shoe-shine boys every day. 1985, the population had The boys seta hot moal at noon foltowed by "academic largely abandoned the nutritiontf in the form of help with their homework,but only If they go to school in the afternoon. "tThat's a0," says Bolivian peso as a store Pancho, "people are at work anywayduring those hours, so of value or unit of there is nothing eise to do." The boys leave their sear at account. Even contracts the oenter every night. Previously they had to pay up to lasting a short time were half a day's wages to store theie things. The center has a bakery, a tailtors' shop and a wood-workingshop where the denominated in dollars, boys can learn these crafts. The incomefrom the sale of the not pesos. Since prices bread wndprtvate donations pay for the recurrent costs of were not fixed for even the center. £SFpaid for repair of the center and furniture. short periods of time in terms of pesos, stabi- ox 3.1s The Shoe-shine Boy and Stabilization lization had little impact on relative prices (except for activities dependent on Government subsidies or access to subsidized credit or foreign exchange), and did not reduce output. While recorded real interest rates did rise, this reflected a shift from quantity to price rationing of capital, not a rise in the cost of capital except for formerly subsidized activities. It is likely, that the position of the poor itproved greatest, relative to what would have occurred by continued hyperinflation, as price stability reduced the premium for access to information.

IMPACT OF ADJUSTMENT POLICIES

3.16 The impact on the poor of the comprehensive liberalization of markets begun in 1985 is more difficult to judge. The Paz-Estenssoro Government undertook several drastic policy changes to increase reliance on the market, rather than the state, for the allocation of resources. Briefly, the initial stabilization package removed all price controls on goods produced by the private sector (except wheat); eliminated controls on interest rates; removed restrictions on the buying and selling of foreign exchange; established a market-based auction mechanism to determine the exchange rate; drastically reduced tariff rates and moved toward a uniform tariff structure; and eased 30 restrictions on wage setting by the private sector.5 The Government also took steps to reduce the role of the public sector in direct production, increase the efficiency of public sector operations, and reform the tax system to reduc. Bolivia's dependence on hydrocarbons earnings. In addition, the collapse of the international tin market in late 1985 made necessary a drastic reduction in employment at COMIBOL, the state mining company, and greatly reduced Bolivia's export earnings in 1986.

3.17 The reforms vary greatly in their impact on the poor. The elimination of price controls improved efficiency by reducing opportunities for corruption and increasing the incentive to channel goods through the formal sector of the economy. However, most small scale producers, particularly in rural areas, already evaded the controls. Removal of interest rate controls probably had little impact on the poor, whose access to loans from the formal banking system was, and remains very limited.

Public Sector Employment

3.18 The reduction in public sector employment created a new class of poor. Approximately 23,000 COMIBOL emnloyees, who had benefitted from public subsidies prior to 1985, lost their jobs. Opportunities for other employment for the miners were few, owing to the lack of transferable skills for most miners,6 and the depressed state of mining and domestic production. As a result, mining communities have suffered a drastic rise in unemployment. Further, the fall in mining output has depressed incomes of communities which provided goods and services to the miners. The most dramatic example of the depression in mining areas can be seen in Potosi Department. In theory, the easing of labor market restrictions could have increased the demand for labor and eased the absorption of newly-unemployed workers. However, slow growth in domestic demand and high interest rates have limited the investment necessary to take advantage of more flexible labor policies. Further, the easing of restrictions on firing may have led to the dismissal of redundant workers. Therefore, the new labor regulations may have had little impact on demand for labor and made little contribution to easing the burden of the fall in mining employment.

3.19 It is important to realize, however, that the collapse of the tin market made the cut in public sector mining unavoidable. Even before the collapse of tin prices in late 1985, COMIBOL was losing about US$2 million per month. It would have been impossible to maintain employment in COMIBOL without massive state subsidies that would have destroyed the economic program and could not, in any event, have been sustained for any significant period of time. Thus, while the creation of a new class of unemployed poor was a human tragedy, it is difficult to see how it could have been avoided.

See the CEM (7645-BO) for a review of the liberalization measures and their implementation.

6 COMIBOL workers such as carpenters, mechanics and other specialists had less problems finding new employment. 31

External Sector Policies

3.20 The changes in the trade regime may have reduced the competitive position of poor producers. Effective rates of protection on consumer goods were extremely high prior to the 1985 reforms. The average rate of effective protection in mid-1985 was 44%, with relatively low rates on capital goods imports. Many consumer goods enjoyed effective protection rates in excess of 100%. The Government reduced tariff rates on coneumer goods to only 20%, which contributed to the substantial rise in consumer goods imports experienced in 1986-87. In particular, strong competition from agricultural products of neighboring countries, some heavily subsidized, probably reduced the demand for camp.aino production. The elimination of the subsidy to gasoline products further exacerbated the problems affecting campesino agriculture, which is heavily intensive in transportation to reach markets.

Real Exchange Rate (Bolivian Pesos per US$)

Index (January 1982 = 1)

2.8

2.4 __ ......

2.2 ......

1.4 ...... --......

1.2 ...... 1 F~~~~~~~~~~~~~~~...... 0.8- 1982 1983 1984 1985 1986 1987 1988 Source: Staff calculations based on Tnfornotlon from INE

Figure 3.1s Developmeat of Real Exchange Rate

3.21 The stabilisation and unification of the exchange rate also may ihave impaired the competitive position of domestic production. Prior to stabiliza- tion, the parallel exchange rate (the one used in most private sector transac- 32 tions) was relatively high.7 The Government increased the official exchange rate by 93% in August 1985, and for most of the time after that little difference existed between the parallel and official rates. Comparing the parallel rate prior to August 1985 with the official rate afterwards (both represent the rate actually used in private sector transactions, particularly in small scale production), one can see a significant appreciation of the domestic currency, due to a rise in demand for domestic currency owing to stabilization. Figure 3.1 shows the development of the real exchange rate with the US dollar from 1982 through 1988 in the unofficial (parallel) market. Although the real exchange rate shows considerable variability during the hyperinflation, the level of the index averaged about 2 (January 1982 = 1) from 1983 through August 1985. The real exchange rate index fell by almost 60% with stabilization in September 1985 (implying an appreciation of the domestic currency), and thereafter averaged about 1.1, well below the levels experienced during the hyperinflation.

Impact on Acriculture

3.22 These factors have resulted in a decline in the real price and quantity of agricultural products, particularly affecting production by campesinos. A price index of agricultural products declined by 29% from the stabilization to the end of 1988 relative to overall consumer prices. Some of this relative deterioation in prices are probably due to the substantial appreciation of the currency in the parallel market. Further, campesino production, particularly for the internal market, has fallen since 1985. Table 3.1 shows the evolution of agricultural production since 1985, expressed as a percentage of the average for 1980-85 (excluding the drought year 1983).8 Total agricultural production fell 17% from 1985 to 1988 in volume terms, and in 1988 remained 15% below the 1980-85 average. The decline in 1988 was in part due to a drought that affected some areas of the country. Nevertheless, the same declining trend is apparent through 1987.

3.23 Production of many goods produced by campesinos for the internal market (corn, wheat, potatoes, onions, sugar cane, and alfalfa) have fallen sig- nificantly since 1985, and in 1988 production remained well below the 1980-85 average. By contrast, production of export crops has increased relative to the early eighties, particularly quinoa, fruits, soya, and coffee. This increase, despite the relatively low level of the exchange rate, is due to increased yields per acre and a rise i- international prices of some key export crops. This trend largely refljcts increased output from the larger estates in the Lowlands, although campesino production of quinoa and coffee,

7 The Latin American Region defines the exchange rate as units of domestic currency per unit of foreign currency. Therefore, an increase in the exchange rate implies a depreciation of the boliviano (prior to January 1987 the Bolivian peso).

Table 3.1 is based on data on tons produced, rather than constant price data, owing to the lack of consistent series for production and prices. Thus, the weights used in calculating the total are tons produced, rather than the value of production in the base year. 33 which is partly for export, has also Hwdred oetric Tong Inrx, 1980-85=100 benefited. Overall, (merW 1980-85) 198 1906 19 190 1989 the appreciation of Comas 710 104 99 92 84 74 the exchange rate Rice 125 138 109 108 129 94 following stabiliza- torn 478 116 96 96 83 66 tion heand reult- Quisina is 138 138 158 151 128 tion (and the result- 7uirea1 104 114 107 90 ST ing fall in the relative prices of Tubers 1121 99 98 97 94 A agricultural Potato 809 95 86 93 87 67 products), and the Yucca 239 110 130 101 110 MA Increase in gasoline Frits andvog. 890 98 102 83 92 NA prices clearly had a Onion 34 102 105 93 96 #A substantial negative Tohto 30 96 156 1U 129 NA impact on campesino orangsh 63 91 t10 99 90 NA production. Fodde 5' 79 83 82 83 VA 3.24 The impact of Alfalfa - 72 73 56 57 u - these changes on the Onions '2 94 101 124 126 NA welfare of the poor is Other 3160 104 99 90Q 94 XA more difficult to Sgar cane 3042 104 94 78 70 63 determine, as the Swa 69 120 195 169 236 353 appreciation of the Coffee 22 107 111 116 116 NA exchange rate also Totsl 12165 103 98 91 as MA would have reduced the r domestic price of Note: Subcate"ories iiithin each major groW arenot exhaustive. imported inputs and Sourcet Ministry.ofAgriculture consumer goods. The ______poor farmers who Table 3.ls Agricultural Production produce food for market and whose major purchase is domestic energy experienced a significant fall in welfare, as production and prices of their crops fell while the cost of energy rose, relative to pre-stabilization levels. The urban poor would have seen a rise in their purchasing power, reflected in the reported, although poorly documented, sharp increase in unrecorded food imports.

3.25 The conclusion that the exchange rate since 1985 has been too low to permit the level of agricultural production prior to 1985 does not imply that the Government has maintained an artificially low exchange rate. The present exchange rate is determined by the market, with only minimal attempts to influence the rate directly through the auction system. The parallel market still exists, and with very few exceptions the parallel rate has been within 1-2 percentage points of the official rate since September 1985. An important reason for the low level of the exchange rate is the inflow of coca money, which in the short term is largely outside the control of the Government. Still, Government policy has contributed to the low exchange rate by running a considerable fiscal deficit financed by inflows of foreign assistance, and by setting relatively high interest rates on Certificates of Deposit. Together these factors have reduced the real exchange rate to the point that domestic production of some agricultural goods formerly produced in Bolivia is no longer competitive. Increasing the efficiency of customs to reduce the amount 34 of goods which evade import duties and other taxes would improve the profitability of domestic production. Fundamentally, however, increases in taxes and some reduction in interest rates may be necessary to increase the ability of poor producers to compete.

Labor Markets

3.26 In an economy with a high degree of mobility of labor and capital, those hurt by the abrupt changes in the relative profitability of different economic activities described above would suffer only temporary losses. Indeed, Bolivian labor has reacted to the change. instituted since 1985 by switching to other activities. A degree of labor mobility traditionally has existed even in poor communities. Short-lived booms in commercial agriculture in the Lowlands (quinine in the 19th century, rubber, cotton, and coca) have attracted migration from the Highlands. Further, the common practice of sending family members from the Highlands to work in the Valleys has provided mome flexibility in the sources of family income.

3.27 Even so, the significant constraints on labor mobility persist in the Highlands. Alternative employment for miners who have lost their jobs and farmers who can no longer earn an acceptable living is extremely limited in the Highlands. Therefore, changing jobs generally means moving to the cities or Lowlands, either permanently or for extended periods of time. In those areas where land is still held communally, migration of a family may result in forfeiting property rights. Leaving one's traditional community may also reduce access to systems of support that had afforded protection from the vicissitudes of the market.9 The danger of disease to Highlands people unaccustomed to the Lowlands is significant, as are the social problems related to growing coca in the Chapare province of Cochabamba. Further, the lack of growth in the economy has generally limited new jobs to extremely unproductive activities. For example, the hormiga ("ant") trade, under which contraband imports are carried by hand across the border, has enabled many workers to survive, but the returns to such activities are very low. In sum, labor has moved in reaction to the changes in the economy since 1985, but the costs involved have limited opportunities for the poor.

Tax Reform

3.28 The Paz-Estenssoro Government instituted a drastic tax reform in 1986. The hundreds of domestic taxes that existed in the old system, including a complicated personal income tax with numerous deductions, were replaced by a simplified group of taxes. Emphasis was shifted from direct to indirect taxes, with a 10% value added tax (VAT) the major source of revenue.10 A program of administrative reforms was undertaken to improve the administration of the tax system. The new system has been successful in greatly increasing domestic tax revenues compared to the period of the hyperinflation. Domestic revenues equalled 6.1% of GDP in l;S8, compared to less than 1% in 1985.

9 More in Chapter 7 on Agriculture.

10 See Chapter 3 of the CEM for an analysis of the new tax regime. 35

3.29 The tax reform has been criticized as regressive (in comparison to the old system), because it relies heavily on indirect taxes rather than a progressive income tax. In fact, the system as a whole is somewhat progres- sive, probably more so than the system in force prior to 1986. The VAT is imposed in tandem with the complementary tax, a 10% tax on income. Receipts for payments under the VAT are deductible from the taxpayer's liability under the complementary tax. The complementary tax is designed to encourage complia;ce by the taxpayer and the retailer with the VAT, rather than to collect additional income. Together, the two taxes constitute a 10% tax on all income, whether consumed or saved, and are therefore neutral with respect to income. Further, the tax system includes excise taxes on luxury items which are bought only by the middle and upper classes, and persons with incomes below a certain minimum are exempt from the complementary tax.

3.30 That is not to say that tax collections are the same share of income for all income groups. Data are not available on the distribution of tax payments by income levels. It is likely that the poor pay little in the way of taxes, as they are exempt from the complementary tax, do not purchase many of the items subject to excise taxes, tend to work in the informal sector in which taxes are evaded, and tend to patronize stores in the informal sector that do not pay the VAT. The tax burden falls relatively heavily on workers in the formal sector of the economy, since the complementary tax is deducted directly from their salaries. Considerable anecdotal evidence exists that the rich evade taxes on their dividends and interest income. A major effort is underway to collect taxes from the richer taxpayers, through separate offices in the three largest cities. While these offices have had considerable success in capturing greater tax revenues, clearly more work needs to be done. The proposed land tax has begun to be implemented but only slowly. The land tax has generated much fear among the campesinos even though they would not have to pay the tax, as smaliholders are explicitly exempted.

3.31 It would be possible to modify the present tax structure to make it more progressive. For example, multiple VAT rates could be introduced to increase the tax burden on luxury goods, or the complementary tax rate could be raised for higher-income wage earners. Such changes would complicate the administra- tion of the tax system, but would be worth investigating if a greater degree of progressivity is desired. In summary, the tax system is somewhat progres- sive, and is superior in terms of progressivity, efficiency and taxes collected, to the reliance on the inflation tax prior to August 1985.

3.32 Public investment had fallen dramatically between 1980 and 1985. Furthermore, many of the projects implemented were of questionable quality. This fall was reversed with the stabilization program. By 1988 public investment had doubled compared to its 1985 level. At the same time continuous effort was put into "cleaning up" the investment program and improving control and planning mechanisms. 11 The improved quality and quantity of public investment have been an important factor in turning around the economy and setting the stages for future sustained growth. The reforms

More about 1980-88 public investment in Public sector Expenditure Review with a Special Emphasis on the Social Sectors (Report No. 7746-BO). 36

in the public investment program and its management need to be continued and deepened. It is a very slow process while the old projects are being phased out. The following chapters analyze the public investment program in four sectors, and make recommendations for changes. On the macro level each new project ought to be screened for its social (employment) effect as an integral part of the economic evaluation. This is one key area where government policy can influence the development path of the country.

CONCLUSIONS AND RECOMMENDATIONS

3.33 The stabilization Probably helped the relative position of the Poor compared to what would have happened in the absence of stabilization. While it is impossible to measure precisely the impact on the poor of the policy reforms initiated in August 1985, the impact of different policies can be judged in qualitative terms. Price stability and the reduction of state intervention in the economy reduced the rents available to groups with economic or political power, and reduced the burden on groups without such influence of financing the deficit. Changes in the trade regime and the real appreciation of the boliviano probably impaired the competitive position of campesinos producing for the internal market, compared to the degree of protection they had enjoyed prior to the 1985 reforms. The collapse of the international tin market and subsequent drastic reduction in COMIBOL employ- ment created a class of "new poor," but this was unavoidable. The tax reform did not greatly affect the rural poor, who remain largely outside the tdx system.

3.34 What emerges most clearly from this analysis is that macroeconomic policies have considerable impact on income distribution. It is important for the Government to consider the impact of its overall economic program on the poor. However, this does not imply that interventions which distort key prices are justified in the name of alleviating poverty. Indeed, a major lesson of the early 1980e is that such interventions tend to penalize the poor by making political influence the determinant of economic success, rather than the market. Thus, our central recommendation is to strenqthen the liberalization of markets and pursue fiscal balance. Fiscal stability is essential to avoid the regressive inflation tax. As the poor are p^werless, it is important not to introduce constraints in the economy that i.- subject to manipulation by the powerful.

3.35 The rural poor would benefit from an exchange rate which is more appropriate for agricultural production. The present exchange rate is relatively low due to the large fisc&l deficit, high level of interest rates, and the inflow of money from the illegal coca trade. To the extent possible, the Government should maintain an exchange rate which takes into account the long run competitive position of agriculture, and certainly should avoid supporting an unrealistic exchange rate through an outflow of reserves. Increases in taxes to reduce the fiscal deficit would help to depreciate the exchange rate and improve the position of the rural poor. However, exchange rate policy must be determined on the basis of macroeconomic considerations. Attempts to adjust the exchange rate through administrative actions to assist 37

particular sectors or groups are likely to have heavy costs in terms of efficiency and engendering increased infiacion.

It wasSunday, but thatdid not make any difference to Rafaelend his wife, Gabriela. They had left thevillage early in themorning, with Libardo the youngest of theireight children, wrapped in his mother'sshawl and slungacross her beck, to walkto theirterrace in themountains overlooking the greattake Titicaca. Rafset and his brothers had atreeady cultivated the land using hand hoes and he hadplanted rows of coca a monthbefore. Todayhe was goingto planthis rowsof potatoes, with the samllestpotatoes that he hadsaved from last year's harvest. Re trudgedup thehill poking a hole in theground with a stickas Gabrielafollowed behind picking the littletubers out of a foldin herskirt and placingthem in theholes letft by herhusband. She hoped that it wouldrain soon so thatthey coutd get a goodharvest. They did not have enough money to buyfood end depended on the harvestto eat,and if theywere very lucky, to sella fewpotatoes or fava beans in thelocal market.That was not very likely. With so littletewd and all the children to feed,they would be luckyjust to produceenough to filltheir stomachs for another year. Gabrieta also worried about thenew landtaxes they had to pay. Rafaelhad spent 530 to go to townto findout about the taxes andalthough he wasnot sure he thoughthe andhis brothers would have to payover $300 in taxes. Shewould not be ableto pay thisamount with thet little extra money she made from knitting and if shedid she would not be ableto buy more wool.

lBox 3.2s Gabriela, Rafael and the Land Tax

3.36 Similarly, a more progressive tax structure may be desirable, par- ticularly given the extreme inequality of income levels in Bolivia. However, the ability of the Government to monitor highly differentiated rates or to raise taxes on the rich is limited by opportunities for evasion, administra- tive considerations and by the need to encourage greater private sector activity. It would be beneficial to review the collection of taxes to ensure that the rich do pay taxes. Estimates by the Economic Policy Analysis Unit of the Ministry of Planning show that 45% more revenue could be raised without changing the tax code. This would mean a continued emphasis on collecting taxes from the large taxpayers. The recent establishment of offices for large taxpayers in the major cities is an important first step. The progressive land tax should be implemented with a sufficiently large exemption for small land-holders to allow the majority to pay nothing (see Box 3.2). At the same time, tax rates should be increased to limit speculation and improve the distribution of land, especially in the Lowlands. The customs service should be improved to ensure an adequate level of Government revenues while providing some protection for campesino production against subsidized goods from neighboring countries. The public investment program should continue to be monitored and improved to reflect the Government's strategy of improving social conditions.

3.37 Finally, it is also important to point out that increases in expendi- tures to assist the poor must be carried out within the overall budget constraint necessary to maintain economic stability. Therefore, increases in taxes will be necessary to implement some of the recommendations of this report. More importantly, enormous potential exists within available resources to reallocate expenditures and public investment to assist the poor, and to improve the efficiency of programs that serve the poor, issues which will be dealt with in Chapters 5 through 8. Therefore, although the scope for manipulation of macroeconomic policies to assist the poor directly is limited, the Government has other policy instruments to do so. Shifting expenditures 38 is not an easy task. Powerful interest groups will strongly resist any reduction of their subsidies for the benefit of the poor. 39

CBAPTER 4s -X8TXTUTIONS AND jORSTHE

INTRODUCTIO

4.1 The ourvooe of thie chapter is to focus on the institutional aspects of poverty alleviation. By focusing on institutions, the analysis highlights the implementation aspect of poverty alleviation. Increasing capital resources and redirecting public policy toward the sectors directly affecting the poor are necessary, but not sufficient, conditions for alleviating poverty. Without appropriate institutions to implement the policies and direct the investment, these efforts will rarely accomplish their objective. In this context, the chaoter is divided into three narts: (1) an outline of the existing institutional framework; (2) an analysis of the common institutional constraints found across DISTRIBUTIONOF ESFPROJECTS BY POVERTYAREA sectors; and (3) a I = RTchest,5 = Porad series of general recommendations to improve the capacity of institutions to reach the poor and to broaden 11.4 the scope of institutions involved in implementing poverty- oriented programs. 5.6% Chapters 5-8 contain a201.2 more detailed assessment i.L*6% 21.M of the specific institutional Distributlonacross DtstrlbutTonof constraints and P Areas Populon recommendations relevant I O AREAI IN AREA2 3 AREA3 0 AREA4 U AREA5 to each sector. AreasRefer to the Pov"ey Map Source: ESF 4.2 Until the creation Figure 4.1t ES? Projects by Poverty area of the ESP,I many remote areas of Bolivia had never seen central or regional Government intervention. Yet not even the ESF has extended a proportional share of its investment to some of the Poorest

The positive aspects of the ESF experience will be highlighted in this report as these should be adapted by other public institutions. This report will thus not present a balanced review of the ESF experience, but rather only highlight lessons which are applicable to other institutions. 40 orovinces (see Figure 4.1).2 Local government agencies are of very low quality (where they exist), and are often controlled by local elites. In many cases, these elite groups show even less regard for the poor than central government agencies. Although NGOs have filled the gap in certain regions, many of the very poorest areas remain almost untouched by outside service- delivery institutions, private or public. In sum, a viable poverty-oriented strategy must include efforts to reform the current institutional framework for reaching the poor or benefits will continue to bypass those who need them the most.

4.3 Careful thought mi 3t be given, however, to the political and economic constraints inherent in modifying any system in which there are considerable vented interests. Recommendations which focus on ways to reach more effec- tively the poor (decentralization with local institution building, increased use of NGOs, and demand-driven investment) are all actions that will ultimately reduce the political and economic power of some groups. Besides, benefits from investing in institutions often require a longer term prespective and a firm commitment to the overall objective of reaching the poor.

GLOBAL INSTITUTIONAL FRAMEWORK

4.4 Within the public sector, institutions with the potential to affect directly the poor can be subdivided into four broad categories: (a) central administration; (b) regional governments; (c) local governments; and (d) the ESF. While further detail is provided in subsequent chapters, the general description is as follows (see Figure 4.2 for organizational chart):

Central Administration consists of central government, universities, and the socia] security system; it accounts for about 80% of general govern- ment expenditures. Presently, there are 18 Ministries, but the number changes over time. Line ministries' investment is funded by the National Treasury (TGN), cost recovery, and external resources provided by donors. Some miriat-riesoversee "deconcentrated"3 agencies, which are still considered a part of central government, yet often receive theiz own donor financing.

2 It should be noted that any given province is very heterogenous (see Chapter 2). Thus, a project in a non-poor province might benefit a poor population group. ESF targets its projects mainly by assessing the poverty levels of the project beneficiaries and not through geographical targeting.

Deconcentration is the handing over of some amount of administrative authority or responsibility to lower levels within central government ministries and agencies. It is a shifting of the workload from centrally located officials to staff or offices outside the national capital, but is not equated with the more far-reaching measures of decentralization. 41

INon-financial Public Sector l

F_~~~~~~

Public General Enterprises Government

Central Local j Regional Administration Governments Governments

Municipalities I _____,___ I

Central Social Regional Govt. lUniversity Security Prefectures Dev. Corp.

|3ESF | |Ministries

Figure 4.2: Organization of the Public Sector

Reaional oovernments include regional development corporations (RDCs) and prefectures. RDCs' projects are financed by the central government, their own resources (revenues generated from taxes and royalties, etc.), and external resources. While autonomous in their management, they fall under the jurisdiction of the Ministry of Planning (MINPLAN) and their 42

Directors are political appointees. Their performance and impact vary widely by department, depending largely on their capacity to generate their own resources. Prefectures are the regional arm of the executive branch of government and play a limited, almost purely ceremonial role. Current proposals for decentralization, however, have suggested that the prefectures be targeted as the key institutions for increased regional responsibilities.

Local aovernments are municipalities in urban areas and similar local government entities in more rural areas (all called "Municipios"). Leadership positions are occupied by officials elected by residents.

Emergency social Fund. Beginning operations in 1987, the ESF was establiehed as a short-term emergency funding agency independent of sectoral ministries and acting in response to requests from communities to finance projects. It is scheduled to terminate its disbursements by March 1991. Since April 1990, a permanent agency forming an integral part of central government--the Social Investment Fund (SIF)--has taken over the ESF's functions in the social sectors.

4.5 Within the private sector, the institutions with the greatest capacity for reaching the poor have been the non-governmental oroanizations (NGOs). NGOs, have had an increasingly visible profile in Bolivia and their impact on the poor has been impressive in certain areas. NGOs began to emerge in force in Bolivia in the seventies with the general mandate of targeting those populations not reached by existing institutions. Most of these organizations held the conviction that the Government had neither the intention, the means, nor the capacity to meet the basic needs of the poorest segments of the popu- lation. Many external funding sources viewed NGOs as viable alternatives to reaching the poor and provided support for their activities. Consequently, in nearly 20 years of existence, NGOs have strengthened and legitimized their presence in civil society. They have contributed significantly to delivery of services to rural, migrant and marginal urban populations.

4.6 There are three general types of NGOs operating in Bolivia: religious organizations, international private voluntary organizations (PVOs), and national NGOs, or Private Development Institutions (PDls). The church has been active for the longest period and has a large infrastructure in health and education. The international PVOs focus their activities in a number of clearly delimited geographical areas, and generally seek to provide integrated basic services including health care, education, water supply and sanitation, and agricultural development. PDrs are extremely heterogeneous. Most operate small projects with an emphasis on community education and participation. Many have a political or ideological edge honed during the years of dictatorship in Bolivia. Recently, PDIs have begun to form associations and networks along sectoral, geographical and thematic lines.

4.7 To illustrate the poverty focus of NGOs as compared to other types of institutions in Bolivia, it is useful to examine the vroject requests received by the ESF by type of institution and across poverty regions. Although not even NGOs have reached sufficiently the most critically poor areas, they far exceed most other organizations in the share of project requests in the two 43 poorest regions (Poverty Areas 4 and 5 of Map IBRD 22289 in Requests to ESFfor Projects in Poorest Areas Annex II). As illustrated in IP Typo of IflUttlon Figure 4.3, the share of KoO Shweof Totdl Requsts project solicitations in these 5AB regions is much higher (nearly 40%) than any other type of institution except the National 3a _ Road Service (SENAC), the principal institution respon- 205 l 11-9ass sible for roads.

4.8 Other importantinstitu- ox tions to consider are the com- q^' 5 0 < o f munity and grassroots organiza- *Ox-° 1 0-r tions, or "organisaciones de "Pews As" e amas 4an wV mQP base." organizations such as Source:ESF Mothers' Clubs, Neighborhood Councils, Farmers' Unions, Figure 4.3t Rquests to ESP fron Poorest Popular Health Committees, Areas Parent-Teacher Associations, and so on abound in Bolivia and can be instrumental in mobilizing local resources and generating community parcipa- pation in the identification, design and delivery of services to the poor. While various NOos, the ESF (to some extent) and a few other public sector institutions work with these groups, much of their potential remains to be tapped.

COMMON INSTITUTIONAL CONSTRAINTS

4.9 Most of the sectors involved in Povertv alleviation share a common set of institutional constraints to meeting their poverty alleviation objectives:

a) centralization of decision-making and urban bias;

b) lack of demand-driven orientation and beneficiary participation;

c) weak plannina and volicv-making;

d) ineffi2ienvy characterized by overstaffing in administrative jobs, low salaries, poor training, and patronage; and

e) donor dependence created by a lack of sufficient resources.

It is argued that poverty alleviation efforts will never adequately reach their objectives without effective institutions at all levels of intervention, from central government to community organizations. An analysis of institu- tional constraints should, therefore, include both those that affect the poor directly, and those that, by improving the general efficiency of the public sector, will affect the Government's ability to focus on issues of poverty alleviation. 44

Centralization and Urban -BEs

4.10 The Bolivian public sector has a long history of highly centralized planning, decision-making, and investment. This has resulted in: (1) a substantial bias toward urban and central government demands and institutions at the expense of regional and rural development priorities; and (2) a highly inefficient execution of public service delivery incapable of reaching the poorest areas. Those who have suffered most under this centralized scheme are, of course, the rural poor.

4.11 Urban bias can be found across all sectors. In health, urban hospitals, social security, and curative care receive much greater emphasis than rural health posts and primary health care. In education, universities are guaran- teed a significant portion of annual tax revenues, while rural schools go without books and their teachers' salaries remain abysmally low. in trans- port, public investment centers on a few large projects, while the RDCs have limited resources to meet even a small percentage of demand for rural access and feeder roads. In agriculture, the larger (urban based) producers from politically powerful regions have received the greatest assistance.

4.12 Centralization has created a public sector virtually paralyzed by excessive concentration of responsibilities. This prevents the central sectoral ministries from assuming their normative and coordinating roles, e.g., the Ministry of Education is charged with both setting overall policy for the education sector, and with approval of the purchase of pencils for remote primary schools. At the central level, this has resulted in overstaff- ing, no clear lines of responsibility, inefficient program execution, lack of clear priorities, proliferation of self contained project units, and insuffi- cient sectoral policy guidelines. Decisions are generally made on an ad-hoc basis, depending on whose interests are best represented at the center, and thus disregarding the interests of the poor. Central Government staff cannot know the complex variety of factors that affect the success of projects in local communities throughout the country. This often leads to inaction on essential decisions.

4.13 At the local level, centralization has contributed to the weakening of institutional capacity to carry out even the most minor implementation responsibilities. The inefficiency associated with central execution is exacerbated by severe communication constraints leading to inadequate informs- tion flowe and monitoring capability. An absence of guidelines for program implementation is then combined with lack of accountability to perpetuate the weak technical and programming capacity at the local level.

4.14 The result is a stratification of political, economic, social, and cultural life along central/peripheral and urban/rural lines. Given that most state institutions are found in the large urban centers, they have a signifi- cant influence over the delivery of social services and provision of infras- tructure. By contrast, because rural institutions are not a legitimate part of the state system, they often lack the legal, financial, and political means to initiate or sustain local development efforts. This dichotomous system has remained remarkably stable since colonial times. Colonists imposed a system of government based on "urban" centers, designed to control the indigenous 45 population. The indigenouspopulation was--and is--foundmostly in rural areas, and continuesto constitutethe poorest group.4 Even the reforms in land tenure and educationfollowing the revolutionof 1952 were based on the idea of an all-powerfulcentral governmentcreating national and vertically integratedinstitutions to developthe rural sector. Entirely dependenton external (urbanor foreign) initiativesto take action,the rural sector remains at the bottom of the hierarchy.

4.15 The limitationsplaced by history on the rural sectors are severe. Without being able to establishformal units of governmentat the local level, agriculturaland rural sectorsface insurmountableobstacles to taking such actions as building and maintainingfarm-to-market roads, or constructing irrigationor drainage infrastructure. The rural populationcannot legally tax itself to provide this kind of infrastructureor rural service. It cannot infringeon private land for access or drainageways. The local community cannot open a bank account,hire staff, or enter into any kind of contract.

4.16 In 1985 the Governmentbegan to transfer several importanteconomic resDonsibilitiesto the RDCe. These measuresmet with limited success, however. Today, the RDCs are formallyresponsible for a variety of public services and economic activitiesformerly under the purview of the central government (e.g.,rural electrification,secondary roads, forestryservices, communitydevelopment, and some agriculturalextension services). The problem is that responsibilitieswere transferredwithout taking into accountthe technical,administrative and financialcapacities of RDCs. SeveralRDCs have virtuallyno financialresources of their own. These RDCs are generally located in the departmentswith the greatest incidenceof poverty. Conse- quently, they continueto be too dependenton scarce and unreliableTGN support and central governmentmicro-management. This has further exacerbated already significantinterregional disparities by favoringthose who receive royaltiesfrom the exploitationof their natural resources(e.g., Santa Cruz). The public investmentprogram continuesto be the primary channel for project formulation. This leaves little room for each departmentto address its distinctivepoverty problems.

4.17 The Government's1990 pronosalsfor decentralizationmust be very carefullyassessed in this context. The transfer of responsibilitymay well help to redress the ingrainedcentralization described above. If that is the case it will help address regionalpoverty issues more effectively. It must be ensured, however,that any delegationis accompaniedby sufficienttechni- cal and financialassistance to the weakest (and poorest) sub-nationallevels of government. Furthermore,poverty reductionpriorities should be given a high profile. Decentralizationgives rise to a host of institutionaland ad- ministrativeissues that affect the design and implementationof sectoral programs and projects: (a) the difficultiesinvolved in implementingreforms in many local jurisdictionswith widely differingtechnical and administrative capacitiest(b) the need to integrateplanning and budgetingbetween the central Governmentand the sub-nationallevels; (c) uncertaintiesabout the future role of some centralagencies; (d) uncertaintiesabout the role of

See Chapter 2. 46 departmental authorities; (e) the absence of policies and mechanisms for extending credit to local governments; and (f) the lack of a coherent financing strategy between central agencies and local governments. Each of these issues must be carefully addressed before wholesale decentralization takes place.

Lack of Demand-Driven Orientation

4.18 The constraint most directly related to efforts at alleviating poverty in Bolivia is a lack of a demand-driven orientation. Successful poverty- oriented programs are carried out by institutions that provide services in response to soundly assessed demand. This contradicts the supply-driven policies of many service organizations that try to induce clients to consume what is judged best for them. Because virtually all ministries and public institutions in Bolivia are structured along the same hierarchical and centralized lines, there exist few mechanisms by which to incorporate the demands of the poor into any level of policy-making or even project design. Except for the efforts of the SIF and ESF, the demands of the poor are inadequately understood and improperly linked to supply in most sectors in Bolivia. The ESF has confronted numerous difficulties associated with the lack of technical expertise among the poor to present their demands as project proposals leading to improved outreach activities in the SIF.

4.19 Local oraanizations of the Poor, such as community associations, cooperatives, peasant unions, and women's groups abound in Bolivia, yet remain largely untapped as intermediaries between government agencies and commu- nities. On one hand, representatives from these groups are not included in regional planning efforts. On the other hand, they often do not have suffi- cient technical capacity or contacts to initiate and promote their proposals. Even if they had such capacity, local organizations generally do not have sufficient financial resources to implement projects. No formal local govern- ment institution is accountable or responsive to the needs of the rural inhabitants. Thus, the rural poor have no institution with the capacity or responsibility to address the provision of public goods.

4.20 Community labor input, in particular, is an important mechanism for beneficiary participation, a cost-effective means of executing projects,5 and an opportunity for the community to "own" projects. Different organizational arrangements can also lead to responsiveness from different types of clientele. The broader and larger the organization, the more likely it is to have heterogeneous clients and to be dominated by local elites. The smaller and more vulnerable the organization, the more likely it is to need intermediary organizations that are particularly dedicated to its existence. This is where NGOs can play an important role.

4.21 Little has been done to examine the most effective ways to intearate NOgs into national poverty alleviation strategies. Although many donors use NGOs as intermediaries, there is little coordination of strategies or even sharing of information among them. NGOs can bring the following strengths:

5 See paras. 9.32-9.37. 47

(1) detailed knowledge of local conditions, (2) experience in "Inthe baginning it was hard for ESf community development, (3) a to work with MWOs,they did notwant state willingness to experiment with non- interference, they didn't knowhow it vas traditional forms of service going to work. Later they saw that we onty provision, and (4) staff commitment evatuatedthe tychnficioprt.nt to work with to poverty alleviation. Much greater NGos.Especialty in heatth, NGOsare very effort should be made to ensure that well organized, they can give a continuity to these strengths be effectively the projects that the public sector often combined with those of government and cant. community groups. This would make OBynow, we have gotten the NGOs to basic social services more accessible work as fast as we do.... to prepare to the poor and let NGOO serve in an reports... they never used to prepare reports. We say: "if you went us to give you money, you intermediary function by assessing or have to give us regular reports, we wilt visit generating the demand for services the project, you have to work fast, you have from the poor. The challenge of to have records of howmany children get involving NGOs in large-scale poverty breakfast every morning, you have to showus how the fofey is spentu. The NGOswere not programs will be twofold: first, to usedto that." encourage their cooperation with a Government that has historically uThe best supervisor of a project is generated resentment among NGOs for the colnudity, the nmunityhas to get involved,that's the onLy way we can have a excessive intervention; and second, goodproject" to assess the comparative advantages- -and disadvantages--of the different Renate Ctaros,Supervisor for ESF types of NGOs, and utilize them .______accordingly. Box 4.1 Working with NOOs

Poor Plannina and Policy Making

4.22 If a national poverty alleviation strategy is to be effective, the institutional capacity to formulate Policy and coordinate its implementation is of primary importance. Within the public sector in Bolivia, this capacity is extremely weak and past attempts at reaching poverty alleviation objectives have often failed as a result. This weakness can be attributed to several factors:

a) excessive centralization (see above);

b) Door communication with ad little planning input from poverty- stricken regions and institutions (e.g., NGOs) working in these areas;

c) no capacity or nower to coordinate the plethora of donor and NGO intervention;

d) a oroliferation of self-contained nroiect units with virtual in- dependence from the 'coordinating" ministry, yet carrying out line functions; and

e) a general lack of will to maintain cood nolicv units at the center. 48

The result has been duplication of poverty projects in certain areas and complete absence of intervention in others. Many ministries have become totally ineffective in providing the normative role they were intended to play. Sub-national levels of government have little authority to fill the gap.

4.23 From a cross-sectoral perspective, the President of the Republic and the National Economic Planning Council (CONEPLAN) are jointly responsible for defining overall development objectives and national development policy. CONEPLAN relies heavily on MINPLAN to coordinate technical planning at the sectoral and regional levels. There is very little reliance on the RDCs in policy formulation, and the coordination of planning among the RDCs, the sectoral ministries, and the MINPLAN is very weak. Until 1989, no specific institutional emphasis had been placed on poverty-oriented strategies. In 1989, however, a new Subsecretariat for Social Policy in MINPLAN was created with this objective in mind. Still in its formative stage, it is difficult to assess just how influential this Subsecretariat will be. It does not yet have the capacity to incorporate a demand-driven mechanism into policy formulation, its primary focus is on the theoretical aspects of poverty, and its staff is small and were originally hired for other tasks.

4.24 At the departmental level, RDC coordination of the planning of sectoral institutions is very difficult. In agriculture, for example, the Ministry has regional offices in each department, w.hichreport directly to headquarters in La Paz. In addition, seven decentralized public agricultural sector institu- tions have independent planning units over which the RDCs have no formal control. Thus, while the RDCs are responsible for coordinating departmental planning, they have little authority to do so. In practice, coordination is a function of the goodwill of the parties involved. In some departments, various sectors take pains to ensure that their collective proposals of programs and projects form an integrated whole. In other departments (mainly the poorer ones), annual operating plans are proposed in isolation from each other. In summary, RDCs have become decentralized planning entities without authority or normative guidance.

Patronaae, Poor Salaries and Poor Training

4.25 No poverty strategy can be successful if the institutions responsible for delivering services are characterized by inefficiency and weak human resources. One of the most fundamental constraints confronting the Bolivian public sector as a whole is that of surplus employment combined with low wages. Poor salaries have driven away many of the best employees and many institutions are unable to attract qualified new personnel. Patronaae has inflated personnel rosters throughout the government with people whose skills are poorly matched to the demands of the respective positions. To make matters worse, there is little emohasis Placed on in-service training, thereby maintaining the poor skill level already in place. Real wages in the public sector plummeted during the first half of the eighties and have only recovered in parastatals. Today they are well below wages received for comparable, competing jobs in the private sector. Central government salaries range between US$31/month for a messenger to US$210/month for a director-general position, and US$960 for a Minister. These salaries are roughly half that 49 paid by state enterprises (which also add other benefits) and even less than that paid by the private sector.

4.26 The effect on the poor, while indirect, has been a public sector constrained by a burdensome wage bill, with insufficient resources for supplies and services to reach the poorest populations, and with under- qualified staff to deliver what few services exist. By 1985, the Government wage bill accounted for approximately 80% of total public expenditures. This was a particularly severe problem in the health and education sectors. The Education Ministry, for example, devotes 99% of its budget to salaries, with no public resources for textbooks, writing materials or teaching aids. Although reforms aimed at resolving some of these pay and employment problems were undertaken in 1985, available data Fuggest the reforms have not achieved the souaht after efficiencies. Efforts to contain the wage bill have been stalled in recent years. Personnel costs jumped from 48% of total expenditures in 1986 to 58% in 1988, varying inversely with expenditures on materials and supplies, which declined from 15% to 11% over the same period.

4.27 Compounding these difficulties are the distortions created by both Government and donor provision of salary supplements and non-salary bonuses (see next section). Because enforcement of a rational, equitable system for awarding bonuses is weak, international donors and the Government find themselves competing for the scarce human resources available in Bolivia. Depending on the sector, the result can be close to anarchy in planning, evaluation and implementation of projects (e.g., agricultural sector). Top public administration talent is perhaps scarcer overall than sector-specific technical personnel. This is because of the virtual absence of training for public administration either at the university level or within the public sector itself.

4.28 A related problem is the lack of continuity in leadership positions and the concomitant change in institutional objectives. This problem is par- ticularly acute in Bolivia because of the high number of political appointees at all levels. Under the 1985-89 Government, there was some tendency to hire technical staff based on merit instead of political affiliation. This positive tendency has been reversed under the present administration. The continual changeover of those in power in the planning, and even executing, positions creates a serious underlying sense of discontinuity. This results in a reluctance of lower level civil service staff to invest any significant level of long term commitment in specific initiatives from above, and a tendency for higher level officials to fire all previous appointees and to start over with their own. In agriculture alone, this has resulted in the development of seven restructuring plans for the Ministry, none of which has even begun to be implemented. This has a particularly deleterious impact on politically sensitive poverty programs. Although it appears that very few of the present Government's overall policies have changed substantively from the previous Government, the underlying sense of discontinuity and uncertainty prevails. Under the current admin.istration, for example, most o. the RDC Directors, their sub-directors and, in some cases, virtually all staff have so been changed6 under the new administration, and the adjustment period has created substantial delays in project design, implementation, and evaluation.

Donor Dep-endence

4.29 aolivia is becoming increasingly dependent on donor funds and interven- tions. On one hand, because the resources from the TGN are generally insuffi- cient and difficult to depend on, most public institutions are largely dependent on external funding to carry out their programs. In health, for example, foreign aid financed 20% of recurrent costs and 100% of investment of the Ministry in 1988. On the other hand, because of the considerable institu- tional weaknesses in the public sector in general--and in targeting the poor in particular--donors have tended either to create their own implementing agencies, or have provided salary supplemente to existing public sector officials.

4.30 The result is twofold. First, it creates a plethora of independent agencies with uncoordinated Poverty agendas. This contributes to the weakness in policy formation and cohesion in sectoral programs. The dilemma is particularly evident in the agricultural sector. While the urgency of aid agencies to do something to alleviate poverty is understandable and commen- dable, these same actions are creating major institutional distortions. Not only are many

4.31 Second, donor dependence further weakens the public sector by attracting the most aualified officials away from existina Dositions. It also encourages staff to follow the directives of the aid agencies rather than the official ministry policies (if such exist). Until 1989, aid agencies provided substantial salary supplements to critical project personnel receiving inadequate remuneration under the government's pay policies. These supplements often reached -Žisurdlevels where the topping off was several times the official salazy. Although there was an official prohibition of such supplements in early 1989, many donors continue to pay them on an informal or in-kind basis. It is even more common to employ local consultants in line positions in key government agencies. In these situations, the national government official with good skills (and no private income) either leaves the public sector altogether to become a private consultant or seeks aid projects for which to work within the public sector. The fundamental issue is that

6 In CORDECRUZ (the Santa Cruz RDC), a list was drawn up of all staff (including secretaries and drivers) including party affiliation and any one who did not belong to one of the ruling parties was fired. 51 donor financed employment is unsustainable and antithetical to the notion of building solid and effective public sector institutions.7

4.32 NGOs are not immune to this distortion. Many NGOs have noted that because of donor dependence, staff are unable to meet internal priorities. One example is in the area of food donations. While certain NOOs continue to provide and administer some form of food donations, many have concluded, that these donations negatively distort production, consumption and general incentive patterns among the poor.8 As a consequence, many (who can afford to do so) are refusing to work with these donations. For others, however, food donations represent a major source of income, and refusing to work with them represents too big a financial risk. Consequently, their priorities continue to be driven by external sources, despite their internal reservations. Food donations can also create distortions in community OarticiDation. Specifi- cally, once a community has benefited from projects in which food is exchanged for labor participation, it expects similar compensation for all projects. NGOs have reported that communities are refusing to participate in projects for which no food is provided in return. This, despite the fact that these projects provide skills or assets important to the community. Over time, a succession of food-related projects can leave communities locked in a cycle of donor dependence.

CONCLUSIONS AND RECOMMENDATIONS

4.33 In conclusion, many Bolivian institutions are inept or insufficient in their efforts at reaching the poor. This can be attributed to several factors. First, many institutions nominally charged with assisting the poor are overcentralized in their decision-making and apt to impose models that do not sufficiently consider the needs or demands of the poor. This has lead to a concentration of development efforts benefiting the privileged urban areas at the expense of the rural and peri-urban poor. Second, many of these same institutions are overly denendent on external funds and are thereby driven by the agendas of donors and overburdened with an excess of autonomous or semi- autonomous project management units substituting for line agencies instead of strengthening them. Third, nearly all public sector institutions are plagued by the dual-edged dilemma of Door salaries and vatronage creating an under- qualified, yet overstaffed, civil service incapable of reaching the poor. Finally, and most important, there is little history of demand-driven nolicv- making in that which affects the poor. Most policies are made in a top-down fashion and take little advantage of the wealth of existing non-formal and non-governmental institutions existing at the local levels.

Even though only donor financed positions provided continuity in staff in several agencies following the advent of the new administation in 1989.

S See Chapter 7 for an analysis of the effect of food dorations on agricultural production. 52

FRMANEWORKF2 INSTITUTIONAL REFORN

1) Ircorsorate a demand-drivenap_roach into Poverty alteviation 2olicy; e strengthen local organization capacity to propose projeets * increase use of non-publicsector institutions * establishspecific linkages between institutions (e.g., Government/NGO) * increase beneficiaryparticipation in planning * draw lessonsfrom the ESF 2) Decentralizespecific manasement and invlementationresponsibilities * support a phased approarh * establish mechanismto mitigate regional disparities provide training and technical assistance to subnational levels of goverrnent * consider mechanismsto utilize labor union networks 3) Strengthen Policy-makina. PLanningand Coordjnation * support §INPLANSubsecretariat for Social Policy Cand UDAPOS) * incorporate demandmechanism into formutation of policy * redirect sectoral ministry resources towards policy and planning and away from program and project execution * raise the sataries of civil servants to competitive tevets * strengthen implementation phase of Public Sector ManagementScheme * reform system of appointments by political patronage * increase emphasison training in public sector administration and economics 41 Increasedonor coordination * coordinate poverty strategies * pursue greater identification, monitoring and utilization of NGOactivity * strengthen donor monitoring unit under NINPLA1Subsecretariat for Public Investment e identify leading donor agency for each sector

Box 4.2: Framework for Institutional Reform

4.34 Institutions which have an effect on the poor in Bolivia range from the large national planning and sectoral ministries to small local grassroots organizations ("organisaciones de base"s). Many are not only inefficient in reaching the poor, but also inefficient in all their activities. Consequent- ly, in certain cases, internal efficiency improvements are the first step in developing the capacity to tackle the broader issues of reaching the poor more effectively. In this respect, certain recommendations delineated below will concentrate on improving the efficiency of institutions internally, while others will focus on the more global issues of how to reach the poor more effectively. Box 4.2 illustrates a general framework for institutional reform from which specific sectoral recommendations are drawn in subsequent chapters.

The Demand-Driven Anproach

4.35 A demand-driven approach is the most effective means for reachina the poor. Specifically, this approach includest (a) strengthening local organizations; (b) supporting institutional pluralism; (c) building institutional linkages between all levels of intervention; and (d) fostering beneficiary participatic*nand local resource mobilization. In general, the goal of poverty intervention should be to help the poor help themselves, which will require not only resource commitment, but willingness to support local capacity to use the resources effectively. 53

4.36 The importance of local organizations is twofold. Not only do they provide a mechanism for participation to the poor, but they also improve the access of the poor to resources and enhance their bargaining power. Commonly, they will take the form of membership organizations (e.g., user groups in irrigation or credit, or women's groups), local government, service organiza- tion (e.g., NGOs), or even private business. In health cars, for example, Mother's Clubs, Popular Health Committees and Neighborhood Councils are existing community organizations with a broad knowledge of local health needs combined with experience in community mobilization. In education, parent- teacher associations and community councils are already active in certain areas mobilizing resources for supplies, monitoring quality of teaching, and building additional schools. Given this experience and knowledge of local needs, these groups should also be placed on Departmental planning boards. In transport, the regional engineers affiliated with SENAC and RDCs have exten- sive knowledge of rural road demanas, yet have insufficient funds to implement projects. Moreover, they were rarely consulted in the past. In each of these cases, institutional support can be provided by Government and donors alike through specific project intervention. Where community organizations do not exist, it may be necessary to employ community organizers (e.g., NGOs with experience in other areas) to help them be formed, to design incentives for participation in a given project, or to train the poor in basic organizational skills.

4.37 Strengthening these local organizations is an important part of a larger strategy that suggests the need for complementary and competitive organiza- tions in the public, private, and NGO sectors as well. In other words, the poor benefit from institutional pluralism and the Government should increase the scooe of the tvyes of institutions it currently utilizes to deliver services to the noorest seaments of societv. Different organizations have varying capacities to work with people living in different conditions of poverty. Organizational comparative advantage calls for cooperation between agencies with different skills and resources. In this context, there is much room for increased Government-NGO cooperation, with the Government taking advantage of NGO flexibility and grassroots orientation and NGos benefiting from central resources, increased information on donor priorities and funds, access to scarce information on regional and sectoral priorities, and national norms. Similarly, there should be increasing consideration of possible uses of the for-profit private sector in the delivery of specific services.

4.38 Institutional linkages and access to local information have important implications for implementation of poverty programs. They should be struc- tured so that communications can flow back and forth between central, inter- mediate and local levels. This requires emphasizing linkages that encourage representation of interests, and include a combination of assistance and control. Control linkages enable one institution to determine some aspect of another's performance, while assistance linkages ease program implementation by compensating for gaps in the performance of implementing organizations. An effective institutional development strategy should stress the latter over the former. Examples of linkages the Government should pursue include: (a) a formal mechanism by which lGOs and local representatives can participate in regional planning (e.g., the regional planning groups included under the EMSO decentralization proposals); (b) technical and personnel assistance to local 54 organizations in areas such as designing project proposals, undertaking needs assessment, or forming credit groups; (c) technical assistance among RDCs, such that those with effective technical capacity can set up programs and train personnel in weaker PDCs; and (d) regulation and monitoring of all levels of intervention to avoid duplication and improve efficiency (e.g., audits or registration of local organizations).

4.39 Finally, increased beneficiary participation and local resource mobili- zation are important because they increase the influence of the poor in decision-making in poverty reduction programs, while cont-ibuting to the sustainability of intervention. Poor communities have provided large counter- parts to ESF, for example, as labor inputs.9 Participation will have several benefits. First, it will enhance the project effectiveness as poor people's input helps to identify needs and local constraints. Second, participation will improve project targeting by reducing leakages of benefits to elites. Third, resource mobilization will play a very important role in generating local commitment to program goals. Fourth, peoples' involvement will reduce the financial and administrative pressures on overextended governments.

4.40 Taking all these factors into consideration, Poverty-oriented institu- tions should draw on the positive lessons from ESF. The major lasting developmental impact of the ESF may well have been to strengthen Bolivia's institutional response to poverty reduction. ESF was for the most part well- administered, fast, and responsive to initiatives from local institutions (both public and private). It channeled more money to more poor people with wider coverage throughout the nation, reaching areas until now untouched by any government program. The following factors could be particularly relevant for other public institutions:

a) its demand orientation: by responding to proposals written for and with communities by local governmental agencies or NGOs, ESF ensured that projects would be adapted to local preferences and needs;

b) the incorporation of small businesses (often in construction-related activities): ESF activated this segment of the private sector in an otherwise stagnant economy, giving them experience in social projects;

c) NGO narticioation: NGOs came to work with Government as a result of ESF experience as they never had before.

4.41 While ESF was not a substitute for line ministries, it has actse as a catalyst to their strengthening in a number of ways. First, it has shown that the poor can be reached with a minimum of red tape (ESF staff totals 87, and administrative costs are 3%). Examples have already been reported where communities have referred to the positive experience with ESF in conversations with line ministry personnel to upgrade the quality of service they receive. Second, it should also serve as an example for civil service reform to demonstrate what levels of salaries actually do attract competent staff and

9 See Chapter 9. 55 motivate it to perform (ESF salaries are about three times normal public sector salaries).10 Third, it has provided a model of how the center may be supportive of local institutions. Finally, the ESF has demonstrated that government programs that reach the poor--in a soundly reformed economy-- attract substantial foreign assistance.

Decentralization

4.42 One means by which the preceding objectives can be achieved is by bringing the poor into direct contact with institutions that have the power and responsibility to intervene in poverty-stricken areas. Given the substan- tial ineffectiveness of Bolivia's public sector in reaching the poor within its highly-centralized system, decentralization of specific resMonsibilities to subnational levels of aovernment apnears to be at least a partial solution to the problem. However, it is important to remember that centralization and decentralization are not mutually exclusive or dichotomous arrangements for governance. The challenge is to find the proper balance between them, to promote poverty &lleviation most effectively. Carefully managed decentraliza- tion can achieve improvements in resource distribution, local participation, extension of public services to rural areas, project identification and implementation, and employment generat.on. Of equal importance, decentraliza- tion can open the door to significant rationalization of the public sector at the central level by transferring responsibilities (and resources) of program management and implementation to the local level.

4.43 While this recommendation appears appropriate for Bolivia in theory, and in fact, seems to have substantial support from many key decision-makers, it is important to proceed with caution. Certain key issues must be resolved before any sweeping plan (such as that proposed to Congress) can be implemented effectivelys

a) The Phasing of decentralization must be addressed in the early stages. Based on Bolivia's own experience and that of other Latin countries who have undergone wide decentralization, it is recom- mended that the Process begin in a limited number of sectors, usina a pilot scheme avoroach, and exDand carefully and incrementally, realizing at the outset that progressive change will take a long time to achieve.

b) Mechanisms to avoid increasina already sianificant regional and subreaional disparities during the process of decentralization must be explored. 7lthough further assessment is required, it may be possible to utilize the existing Regional Development Fund (Fondo de Desarrollo Regional) as such a mechanism.

c) Trainina and technical assistance should be made available from the central government (or donors and NGOs) to the regional and local levels. Particular emphasis should be placed on: planning (and t0 Salaries are, however, in line with those of other donor financed positions. 56

mechanisms to incorporate community-level demand into planning efforts): project design, implementation and evaluation; cost- recovery mechanisms; and options for contracting the private sector for specific service delivery. RDCs and Municipalities with sub- stantial experience could be contracted to train those with less technical capacity.

d) Greater continuity in key iositions must be achieved in sub-national levels of government. This could take the form of longer terms for elected positions and establishment of political civil service positions for key technical staff (see below for more details). This would be especially important for the proposed regional sectoral committees of the prefectures.

e) Because of the negative response of many of the more powerful labor unions to proposals for decentralization, it may be useful to pursue greater dialogue with them in decisions regarding the implementation of decentralization measures and in utilizing their local branches as one option for delivery of certain services.

4.44 Donor assistance in supporting decentralization efforts should be concerned principally with how best to help local government units build-up technical and administrative capacitv to assume increased responsibilities and to target the poor mioreeffectively. This is likely to take the form of: (a) technical assistance and training components of particular sectoral projects; and (b) working with the Government to devise flexible and pragmatic arrange- ments that take account of the enormous disparities in administrative and fiscal capacity among regions and municipalities. Donors should help the Government ensure that this institution building objective not lead to the duplication of the number and skills of staff at the center and the sub- national levels. In particular, decentralization should not result in a larger government sector or in decentralizing institutional weaknesses and insufficient resources to the local level.

Improved Policy Makina. Plannina and Coordination

4.45 The primary objective of central government should be to set overall policy guidelines, while allowing decentralized institutions to assume more of the program and project implementation responsibilities. From a poverty perspective, the creation of the new Subsecretariat for Social Policy in MINPLAN is a Positive first step. but should be further surnnortedand promoted as a guiding force in poverty alleviation throughout each of the sectoral ministries. In doing so, this Subsecretariat should develop the capacity to incorporate a demand-driven mechanism in the formulation of policy. Now, the focus of the Subsecretariat appears to be on theoretical aspects especially the health and education sectors. While these sectors are essential in targeting the poor, others such as agriculture, transport (i.e., roads), and housing are equally important and should receive assistance in developing practical poverty alleviation strategies.

4.46 It is, therefore, recommended that the Subsecretariat should be provided with the sufficient financial and human resources to create a small cadre of 57 highly skilled experts in poverty policy analysis (the proposed Social Policy Analysis Unit)"1 to prove the cross-sectoral guidance necessary in a global effort to reach the poor. In this way, the Subsecretariat should not only assist in proposing a concrete practical poverty alleviation policy, but it could act as a technical assistance resource to those ministries whose planning resources are too stretched (or non-existent). Similarly, it is essential for sectoral ministries to redirect their resources toward strengthening the technical capacitv of their Policy and nlannino departments, while de-emphasizing the current focus on proiect and Proaram execution. This combination of reforms would allow for the elevation of poverty alleviation concerns to the sectoral policy level, while allowing the sub-national units of government to focus more on delivery of services.

4.47 Any efforts to improve the efficiency of public institutions in Bolivia are inextricably linked to overcoming a much pervasive constraint in the public sector--a weak civil service. The Government will find it much harder to attract and/or retain qualified staff for a more efficient management of the various sectors unless it pays competitive saiaries and provides its staff with an adequate work environment. It is, therefore, essential to raise the salaries of civil servants to competitive levels; without this action it is unlikely that public agencies will be effective in any sector. This should be done even if it means reducing the scope of activities to be undertaken by the public sector. The public sector continues to lose qualified people to higher-paying private and donor-funded organizations and projects. This problem is perhaps most acute in those areas where the greatest poverty exists. Only when incentives (preferably financial) are provided to work in these areas will the constraint of under-qualified staff be overcome.

4.48 In this context, greater emphasis should be Dlaced on the imnlementation phase of the Public Sector Management Program (PSM'fto meet the lproaram's obiectives of: (a) rationalizing the existing uneven wage structure; and (b) creating conditions of stability in the public sector which would attract qualified personnel to the public sector. Although it is recognized that such a program will take at least ten to fifteen years to reach these objectives in all public agencies, it is essential that they continue to be pursued to establish a stable cadre of technical and professional public sector servants. The reform can take place in several phases. The first step, recently agreed upon, is for donors to implement unilaterally a salary scale for all donor- financed positions. Second, a similar salary scale must be adopter for all positions financed by counterpart funds. Third, emphasis must be placed on the creation of a Civil Service Fund (CSF), as the first step to increase salaries for key staff (including those with responsibility for poverty alleviation), and provide a mechanism for providing continuity for civil service staff. Finally, once this system is in place and sufficient rationalization has taken place, the CSF norms should he adopted as the country's civil service code.

4.49 For such a reform to take hold, however, the prese c system of appoint- ments by nolitical patronage must be abolished. A first step would be an

11 "UDAPSO" (Unidad de Analisis de Politica Social) 58

agreement among the major political parties to protect the central cadre of professionals to be formed under PSM from political removals. The second step should be an agreement limiting the number of political appointees to a certain level in each agency (e.g., only levels of Director and above will be politically appointed). The shift toward a professional civil service working for democratically-elected officials is an essential step in the maturing process of Bolivian democracy.

4.50 Finally, improved planning and policy-making can only take place if there is increased emphasis on strengthening human resources throuch trainino and nrofessionalization of existina Eersonnel. Particular emphasis should be placed on the technical skills associated with planning, monitoring and evaluation of specific projects and programs within each sector, and coor- dinating activities of implementing agencies. There should be greater efforts at the university level to training graduates in areas directly related to public service, particularly in public sector administration, public finance, applied economics, and project evaluation.

Donor Coordination

4.51 Greater efforts must be made to coordinate 20verty stratecies amona donors. aovernment and NGOs, and identify weaknesses in the coverage of the Poor at all levels of intervention. Significant duplication of projects continues to exist in some areas while in others there is almost no intervention at all. In general, it is the poorest regions that receive the least attention from any institution, external or domestic, private or public. Greater coordination can be pursued in a number of ways--the Consultative Group meetings, mobilizing increased cofinancing, greater utilization by donors of the donor monitoring unit under the Subsecretariat for Public Investment--but all efforts should concentrate on maintaining a cohesive and coordinated framework for a poverty alleviation program. These efforts should include greater identification and monitoring of NGO activity, as there is a certain amount of duplication at the local level that goes unnoticed. The ESP experience showed that international agencies can be convinced to finance a government-sponsored and coordinated program instead of the usual insistence on detailed involvement in selection and design of individual projects.

4.52 Inter-donor coordination should be strengthened as well. The successful experience of the regular coordination meetings in La Paz should be sustained and expanded. A leading donor agency should be identified for each sector and this agency should remain Jn continuous contact with the Government and other donors to discuss policies and interventions in "its" sector as an assistance to the efforts of MTNPLAN. For coordination purposes it might be useful to formalize sub-groups with regular meetings under the chair of the lead agency. Whenever a donor wants to enter a sector, has a programming mission in the field, etc., a meeting with the lead agency should be high on the agenda for the new agency. Finally, if any additional donor activities result from this report, utmost effort should be made to coordinate efforts so that policies or institutional arrangements proposed in one project are not inconsistent with those recommended under others. 59

CEAPTRR5:t NEU=

INTRODUCTION

5.1 Bolivia's health conditions are worse than those of countries with similar levels of income per capita. Figure 5.1 shows two regressions of infant mortality rates on GDP per capita for all low and lower-middle-income countries who reported indicators to the 1989 World Development Report. Bolivia is above the level of comparable countries (as indicated by the

International Comparison of Infant Mortality

Infant Mortality (per thousand live births) 170-

150_i6O --- +

140 + + Bolivia (Census) i30 ++ + +

*1 {1- + vOlivia (ENPY) ice + +\ + +

so - + |

60- so - E1 Salvador tea a ia + +Iexico 40 - + +Paraguy +

20 + Cil 20 - + + Chi +

£00 300 500 700 900 1100 £300 1500 1700 £9c0 GDP per Capita Regressiom Lines + Actual Figure 5.1: Infant Mortality Rates regression lines) even when the newest--and most optimistic--figures are used. The result holds both for the linear and the exponential models. The linear model gives best fit. It is, however, plagued by heteroscedasticity making the significance levels unreliable. This problem is reduced in the 60 exponential model.' It is noteworthy that the two South American countries with strongly above average levels of infant mortality, Bolivia and Peru, have large indigenous groups who are largely outside the power structure, and among whom poverty is concentrated. The health problems are worse among children and childbearing women. Poor health conditions are related to lack of clean water, poor nutrition and (for women and infants) frequent childbirths.

5.2 This chapter analyzes policies, expenditures and institutions in health with a view to improve the delivery of health services to the poor. The gy- line of the chapter is as follows. First, the institutional framework cover- ing both public and private sector entities is described briefly. Then sector policies and expenditures are analyzed. Policies have a strong poverty focus (except those regarding nutrition), but expenditures do not foliow the policies set. Third is a section on sector constraints. These include excessive centralization, inadequate coordination, low salaries, weak manage- ment, lack of beneficiary participation, poor use of and coordination with NGOs, and women's issues. Finally, recommendations focus on implementing existing poverty alleviation policies; shifting expenditures toward rural areas and primary health care; increasing the involvement of NGOs and com- munity groups; improving training of personnel; integrating the traditional system of medicine into the modern one; and assessing the ongoing nutrition programs to improve targeting.

INSTITUTIONAL FRAMEWORR

5.3 Institutions of the health sector can be divided into four major groupos

a) The State apnaratus: Ministry of Health (MPSSP) and the Social Security system. MPSSP serves about 30% of the population, mostly in cities. The Ministry includes a sub-secretariat of Public Health and one of Social Security. The Public Health Sub-secre- tariat includes eleven Sanitary Units (SUs) with regional hospi- tals responsible for ambulatory and inpatient curative care; 93 health districts (HDs) for specialized health care and referral centers; and 451 health posts in charge of outreach and basic health care. The social security system provides curative care benefits for the households of insured wage-earners employed in industry, commerce, mining, and government services. Benefitting only about 20% of the population, this system rarely reaches the poor.

b) State iurisdiction. but subiect to autonomous manaaement: RDCs, Municipalities, local water and sanitation companies, and univer- sities. Only a few municipalities and RDCs are active in the

The fundamental result holds for slightly more sophisticated models explaining infant mortality rates using more explanatory variables (e.g., fertility rates); further transformations of the variables to reduce heteroscedasticity; and instrument variable approaches (to take the endogeneity of regressors into account). 61

health sector, with less than 1% of their respective investment portfolios targeted toward health. In the larger cities, municipalities have formal titles to almost as many health facilities as MPSSP, but do not operate them. Regionally, the wealthier RDCs (e.g., Santa Cruz) have played a limited role in planning, executing, and maintaining health facilities, principal- ly in rural areas.

c) N00s. orivate institutions, and traditional medicine: The non- profit private sector serves about 20% of the population, mostly in peri-urban and rural areas, and is comprised of about 300 NGOs working in health and nutrition. Accounting for approximately 28% of total health expenditures in 1988, these NGOs contributed nearly $20 million to the health sector (including both investment and current expenditures). Recent initiatives to promote coor- dination among them have yielded modest results. Funded primarily from North America and Europe, only a few NGOs have national coverage. Most are limited to the financing and management of a single or a few service centers. The private for-profit sector serves a small portion (no more than 5%) of the population that has decreased in recent years with the general impoverishment of the country. Traditional medicine is the only recourse of large segments of the population, particularly in rural areas. Traditional health practioners number some 12,000 and include herbalists, healers, and midwives.

d) The Emercencv Social Fund: ESF has played an important role through supplementing the extremely low level of MPSSP investment directed at vulnerable groups. To reach the poor, ESF-funded health projects have focused on primary health care programs targeting mothers and infants, vaccination programs, nutrition intervention programs, and rehabilitation and construction of health posts. The Social Investment Fund (SIP) will soon replace some functions of the ESF, as the ESF winds down its activities.

GOVERNMENTPOLICIES AN1DINVESTMENTS

Sector Strategy and Exoenditures

5.4 Until recently, Bolivia lacked well-defined official health policies in attacking the country's long-standing health problems. In the past, the NPSSP (supported by some external donors) considered hospitals the center of health care and directed health resources toward hospital construction in urban areas and provision of curative care programs. While primary health care and rural health services were supplemented by other external agencies, weak interagency coordination resulted in sporadic and un-targeted program design.

5.5 Recent policy statements show a renewed effort to develop more clearly defined health policies in order to address structural constraints facing the health sector. An increased emphasis on decentralization as well as on 62 primary health care - has become the center of the Distr(bjttnofb Hel Iuses ANS (IWS) health sector development 0toflmmnt A t strategy in the medium- and long- Urba Rurat Urbn trel Ubn Q*tI term. The highest priorities have Chuatfaca 132 (71.?) 52 (28.9) 834 (73.2) US (26.8) 19.3 $1? been placed on: (a) Cads 156 (778 ) 45 (22.2) 475 (69.) 20 30.5) 3S.5 64*5 a multi-sectoral a" Cear 210 (70.?) 87 (29.3) n.ae na. 48.0 52.0 approach aimed at improved planning Soure: Mo. Juw 1989. and coordination Percnta ereinintG:prernthems. among various health service providers; and (b) Table 5$.1 Distribution of Health Resources by Area improving efficiency and equity of public health service deliveries through attentior to primary health care for the most vulnerable groups. Most notably, recent statements have emphasized a need to allocate Increased health resources to the extremely poor areas--the northern part of Potosi, the southern and western parts of Cochabamba, Chuquisaca, and Pando (for rural areas) and E1 Alto (for peri-urban areas). Special efforts will be made to lower infant mortality, malnutrition, and the incidence of prevalent diseases. Furthermore, the intent is to provide access to simple curative care and basic sanitation services.

5.6 Expenditures have yet to be MPSSPExpenditures by Department in 1988 shifted toward the US_per copffa most vulnerable c groups. Health per- t .9 sonnel and hospital 6 I0 5 beds are still con- 5 centrateddispropor- 4'$0 tionatelyin urban 3 ~~~~~~~~~~~~areas.During 1984- 87, on average, 51% 2 of the HPSSP's health I ~~~~~~~~~~~~~expenditurewas directedtoward 0 J o 0 4 health care services

4,e ? e o in rural areas. For 6¢ 9 g iN example, while 80% of the population of Ndot: Chuquisaca is rural, Excludes Not. Projects and Central Admln. Exp. only 29% of doctors Source: ESO are in rural areas L______(Table S.1) and only Figure 5 * 2t*PS5P Expmeditures by Department 20% of health expenditures go to 63

rural areas. Furthermore,geographical distribution of the MPSSP's expenditures has been skewed, without sufficient regard to the population size or the poverty level of each department. While the Potosi Department suffers from the poorest health conditions, its per capita allocation of health expenditures was only US$1.6 in 1988 (national average US$3.2), the smallest of the nine departments (see Figure 5.2).

5.7 The distribution among types of expenditures does not reflect the new strategy either. Between 1984 and 1988 the share of curative care increased in the three largest sanitary units (there is no time series information available for the rest). Cost recovery for public health services and supplies has increased significantly over time (600% in real terms between 1984 and 1988). These funds have been used mainly to pay bonuses to health workers. Simultaneously, the wage bill has decreased (even including these bonuses), while the number of persons employed in the sector has increased by 7%. Thus, the health sector provides a classic example of the pervasive pro- blems in the Bolivian public sector--a large number of poorly paid staff. On a positive note, spending on medicine has increased over time, both as a share of total expenditures and in absolute (real) terms.

Nutrition Policies2

5.8 Even though nutrition indicators are very low in Bolivia and poor nutrition is a major cause of the poor health status, the Government does not have a clearly formulated nutrition policy. Data on nutritional status are of very poor quality. Estimates of malnutrition using height for age vary widely. The DHS survey found that 38% of children between 3 and 36 months were malnourished,3 the number was 31% in urban and 45% in rural areas. A 1981 survey by the National Institute of Nutrition (INAN) found similar results (33% in urban areas and 46 in rural areas). Nutrition indicators follow the poverty map closely. Malnutrition is higher in the Highlands than in the Lowlands. A recent urban survey by the National System for Nutrition Surveillance (SVEN) found high rates of chronic malnutrition resulting in stunting (low height for age) among first graders in seven urban areas. Figure 5.3 eummarizes the results of this survey. The Figure shows the average prevalence of stunting among first grade students, as well as the rate in the worst and the best (public) school in the seven Departmental Capitals covered by the survey. As is clear from the graph, incidence of malnutrition varies substantially within a city. For example, in one school in Potosi 93% of students were malnourished, while only 10% were malnourished in the best school. These numbers show the difficulty of attempting to target nutrition interventions by geographical areas.

5.9 Assessments of existing local nutrition programs vary. Limited experiences from El Alto with school feeding suggest that there is a strong substitution effect within the family when a child gets a meal in school. The

2 Food donations are discussed in paragraphs 7.29-7.35.

3 Defined as those with a height for age at or below two standard deviations below the norm. 64

Malnutrition Among First Grade Children (% of children 2 standard dev. below standard) (height for age)

_~~~~~~~

Potosi '4 93

Trinidod

La Paz 77 Torija _77 57

Sucre so

Oruro 48

Cochobamba 34:

0 20 40 60 80 100 120 (%) 0 Average D MaxImumg Minimum Source: SVEN,MPSSP

Figure 5.3x Chronic Malnutrition in Urban Areae child may not receive any food at all at home and may actually end up worse off than before. However, numbers from NGO-managed health posts with nutri- tion programs4 show substantial improvements over a short run as far as weight/age indicators are concerned.

5.10 The Government is planning to introduce a national school breakfast program. School breakfast programs are traditionally not a good type of nutrition intervention. First, as the damage done by malnutrition has already happened by the time children reach school age, a much larger effect would occur if the nutrition was given earlier in life. Second, the poorest of the poor don't go to schocl. On the other hand, the schools provide a very convenient distributicn channel, as the physical and human infrastructure is already in place.

4 One experience, also in El Alto, shows a substantial improvement in child nutrition during the strictest part of the stabilization measures. This indicates that it is indeed possible to protect the urban poor during adjustment, if the correct mechanisms and institutions exist. 65

5.11 The school breakfast program should be carefully analyzed, taking into account the possible effects of spending the same amounts on other schemes with potentially higher benefits. INAN is working with the RDCs to introduce national system of information gathering and processing for nutrition The results of this exercise, combined with the information already availaAle in MPSSP, could provide a basis for targeting nutrition interventions.

5.12 The composition of the breakfast suggested depends heavily on the use of milk. This is a potentially serious issue as many poor Bolivians are physically unable to absorb lactose. Furthermore, milk is a very expeneive source of calories. The extra protein provided by the milk will not M=Q'3X== be absorbed as long as the children Maria lives inthe high parts of the Cochabamba receive insufficient calories. Valley. She is the mother of four chitdren from Instead, serious consideration ought a small baby girl to a seven year old bay. High to be given to INAN's newly developed up in her "pantry" (one shelf in a corner of the basic food basket, which relies on Salt. Proudly she shose)ith e ba of saltad traditional cereals and tubers. This explafns that she knowshow irportant it is to diet is better suited to the local use the salt to avoid "a serious disesse." then population. While there are no asked why the bag istunopened, she answers: pouation.lo Whgionlethverenaeno '"out Sir, none of my children are sick now. national or regional Government This salt is for ule henIthey arE sirl; As sponsored nutrition programs at long as her children are not iltwith fever she present, many NGOs are involved in continues to use salt from the block she bought community nutrition programs. There at the local market last month. is a dire need to assess theae experiences and coordinate the Box 5.1 Maria and the Iodized Salt activities.

5.13 As far as micronutrients are concerned, there is a generally well- regarded national iodine supplement program in place. The first stages of the program have had good effects. The main remaining issue is to iodate the blocks of salt from the salt flats. These blocks of salt are the traditional form in which salt is sold, and though many mothers are aware of the need for using iodated salt, they are skeptical about the "new" packaging in little sacks. There are some technological problems as the iodine tends to evaporate, before a block of salt is used up. A solution does appear to have been found to this problem. Another micronutrient problem is iron deficiency (especially for women). There are no programs in place to alleviate this serious problem.

Water Suonlv

5.14 Lack of clean water is a major cause of poor health conditions, especially In rural areas. At any time about 30% of children under the age of five suffer from diarrhea, mainly related to the poor quality of water. While relatively good water companies take care of water supply in the major cities, rural water supply is often without an institutional "home." Experiments using hand-pumps and other appropriate technology have shown that with technical assistance in the formation of water coopitratives,substantial cost recovery, i.e., good sus tinability, is possible even in remote rural areas. Rural water supply is the responsibility of the MPSSP's Directorate of Environmental Sanitation, which is a very weak institution. Some RDCs are 66 involved, but only to a limited extent. There is a need to build on the positive experiences of some projects funded by the World Bank, UNICEF, EEC and others. The financing of rural water supply will be the responsibility of the Social Investment Fund, which could restore the importance of the sub- sector. It will be important for the SIF to assess the various experienceo, especially in the Highlands, and help develop a standard package for water supply.

Investment

5.15 Public investment in health totaled US$3.2 million (US$4.5 million was programmed) or 0.9% of public investment in 1988. Except for three non-poor Departments, RDCs were inactive and most investments was undertaken by MPSSP on a national scale. Primary health care projects accounted for the largest share (76% of the total), followed by hospital-related projects (16%). External agencies provided 92% of the sector's investment.

5.16 In the investment program for 1990-92, investment should increase to US$15.1 million per year (or about 3.8% of total public investment). UNICEF and USAID plan to allocate increased resources to ongoing projects and a new IDA-assisted project will start during this period. External funds will continue to cover most investments (88% of the total). Therefore, donor coordination with regards to Bolivia's health sector development will be critical in determining the composition of projects.

5.17 Over the next few years, there will be an important shift in the tvpe of oroiects as several new integrated health care projects will be added. These projects are supportive of the Government's general health strategy and will serve to improve health and nutrition status of mothers and children, espe- cially in poor rural areas. The 1990-92 public investment in health will be distributed as follows: primary health care (38%), integrated health care (30%), hospital-related health care (19%), and others (13%). Most integrated health care projects are to be implemented by the MPSSP with the assistance of external donors. The IDA-assisted Integrated Health Development Project will help enhance efficiency and effectiveness of the health care system through improving management of the sector.

5.18 Although the general direction of future public health investment is commendable, care should be taken in treatment of several hosnital-related investments. The major hospital-related projects include construction of Bracamonte Hospital in Potosi and modernization of hospital equipment and facilities in Pando and Beni, financed mainly from external sources. In particular, economic justification for the Bracamonte Hospital is highly doubtful. The wisdom of devoting the entire health-related budget of the Potosi department in 1989-92 to this single project, forgoing all other projects, is questionable. Especially as an externally financed study showed that a hospital at one fifth of the cost would meet the expected demand for hospital services. Although not included in the 1990-92 public investment program, the Government envisages constructing two additional hospitals in Cochabamba and Pando. Even if externally financed, these large projects would require sizable counterpart funds and operating expenses that would compete 67 with more useful projects. MPSSP's budget is already incapable of staffing four hospitals built through bilateral donations over the last several years. 5.19 Since 1987, AU has supplemented )PSSP investment. As of mid-1989, ESF had financed US$2.3 million for 95 nutrition projects, US$1.3 million for 34 health care projects, and US$3.4 million for 121 health infrastructure projects. ESF-funded health projecis have focused on primary health care programs targeting mothers and infants, vaccination programs, nutrition intervention programs, and rehabilitation and construction of health posts. In 1988, ESP disbursed US$2.6 million5 by mobilizing various organizations belonging to the public and NGO sectors. This amount was equivalent to about 80% of investment provided by MPSSP and RDCs.

5.20 SI1 started its operation in April 1990, and will extend the work of ESF. The health sector is expected to be granted US$26 million a year for assistance and infrastructure projects during the first two years.6 If the SIF contribution is added to the annual public investment planned by the MPSSP and RDCs, the health sector will receive US$38.4 million in 1990 and US$43.5 million in 1990. Each of these amounts is more than six times what the health sector received in 1988 (including ESF-funded investment). One area that SIF is considering is how to enhance its effectiveness in reaching the poorest of the poor. The demand-driven mechanism will be adjusted with outreach mechanisms to enable the least fortunate to prepare and present projects.

SECTOR CONSTRAINTS

Centralization and Inadeouate Coordination

5.21 Excessive centralization has prevented MPSSP from assuming its normative and coordinating responsibilities. Executive functions in MPSSP are con- centrated in the National Directorates in La Paz, and the SUs can only decide the allocation of about 20% of the budget, consisting of the revenues generated locally from cost recovery. The inefficiencies associated with central execution are exacerbated by severe communication constraints among levels of government. This leads to inadequate information flows and monitoring capability. Lack of guidelines for program implementation and lack of accountability perpetuate the weak technical and programming capacity at the local level.

5.22 The proliferation of self-contained project units in MPSSP also disrupts the fulfillment of its coordinating responsibility. Besides the work of the MPSSP, much of the public health care available comes from Social Security, which in turn is composed of independent agencies. The National Directorates have gained virtual independence through the execution of projects financed by

This figure does not include disbursements for nutrition programs, which amounted to about US$1 million.

6 The official figure in the 1990 investment program is only $1.5 million, a more realistic figure as SIF is only getting of the ground slowly. 68 donors. The General Secretariat, which hag not been the object of specific cooperation programs, is weaker than the directorates under its supervision.

5.23 The weak enforcement of MPSSP's plan of activities prevents the Ministry from integrating and coordinating donor support and the work of the private sector. An obvious problem resulting from this complex structure is the difficulty in coordinating the work of these groups into an efficient and equitable health care system. This has led to the development uf 16 major national programs, with overlapping objectives and beneficiaries, and inade- quate monitoring.

5.24 Coordination between MPSSP, the Social Security System and NGOs is starting to be addressed in the SUs. At the central level, coordination is difficult because of incomplete inter-institutional information, the many national programs, and the separate legal and operational system of the Social Security System. The relationships between NGOs and the MPSSP have often been difficult, and occasionally marked by mistrust. Impediments to Government/NGO coordination includes (a) lack of an effective identification and selection system for NGO presence and activity combined with cumbersome and contradictory registration procedures; (b) lack of SU credibility among NGOs as relevant partners in delivery of services; and (c) inadequate information about MPSSP and SU services. In 1988, a survey in four major cities showed that only 45% of institutions providing health services were aware of MPSSP's programs. General awareness is much lower in rural areas. More recently, coordination and cooperation with NGOs and local governments have been achieved, helped by the renewed confidence in the public sector from funding of over 300 NGO projects through the ESF.

Low Salarieb and Poor Personnel Manaaement

5.25 Health workers' salaries fell by 30% in real terms over the last decade. The salaries in the Public Health sub-secretariat, ranging from US$31 to US$240 per month, are the second lowest in the public sector. Accordingly, staff seek better salaries and working conditions (especially in NGOs), and tend to move away from those regions which are the most difficult--and poor. Personnel turnover is 30% annually. Doctors in HDs, for examplc, have lower status and pay than doctors with comparable training in hospitals or with administrative functions. This discourages able staff to move to the areas of greater needs. Weak personnel policy &nd management c'-.ound tne ealary issues in the public health sector. SUs and many hospitals have no adminis- trative committee, no organizational plan, no job description for personnel, and erratic hiring and personnel procedures. Not only are the demands of the poor difficult to consider in such an inefficient system, but patronage and co-ruption abound at the expense of service dalivary.

5.26 Medical personnel working for NGOs typically receive higher salaries or benefits than those in the public system. Often, doctors and nurses receive iase pay from the MPSSP and a supplement from the NGO. Base salaries for doctors in NGOs are typically 20% to 50% higher than public sector wages. According to NGOs, this salary policy is justified since personnel working in NGOs are expected to work longer hours and provide additional services, such as community education and extension services. To compensate for the hard- 69 ships of Bolivia's rural areas, medical personnel also receive support from the community, often in the form of free housing and food donations organized through the NGO. Similar incentives for personnel working in poorer areas should be considered in the public sector if implementation of poverty objectives is to be sustained.

Poor Financial Manacement

5.27 Financial management of the health sector is weak because of the inadequate distribution of financial responsibilities, the outdated structure of information used for budgeting, 'd the inefficient procedures for budget preparation and execution. The inadeqjuatedistribution of financial respon- sibilities stems from a combination of excessive centralization of functions within MPSSP along with an inadequate management information system. Combined with inadequate accounting practices, these factors inhibit any meaningful control of budget implementation and make formulation difficult and inac- curate. Budgeting is also separated from planning, and budget categories reflect inadequately the sources and uses of funds. At the local level, HDs follow only rudimentary financial procedures and the health posts have no financial planning capacity at all.

5.28 MPSSP probably depends too much on cost recovery to ensure adequate coverage of, particularly, basic health care. Programming and execution have been complicated by the dependence on project-specific donations from the international community. Until such projects and programs are coordinated and set within a national health policy framework, financial management will remain problematic. Uses of funds favor secondary and curative care and the share of basic health care does not reflect the government priorities. In sum, too much cost recovery and too little primary care place a high price on health for the poor.

Inappropriate Training

5.29 Although the capacity to train the necessary health staff exists in Bolivia, the design of curricula does not respond to the staffing requirements for basic health care, and the education process is not conducive to the extension of serZ'icesin low-income areas: the autonomous medical schools in universities offer a curriculum for medical doctors and nurses that is not directed to the epidemiological and social reality of the country. In particular: (a) maternal and child care, community medicine and nutrition do not receive the necessary attention; (b) no training is given on the adminis- trative and team management responsibilities of doctors and nurses; (c) the health perzornnl are often ill-prepared to deal with illiterate and Non- Spanish-speaking patients; and (d) the educational process is not supportive of the objectives of basic health care in impoverished communities. In addition, the professional ties between teaching personnel and the hospitals are much stronger than those between teachers, personnel, and MPSSP authori- ties responsible for basic health. Too few auxiliary nurses are trained, few staff are prepared to serve in remote areas, and no manual of standard procedures for the provision of health care is distributed to the graduating students, or updated for their uso. 70

Lack of Beneficiary Particloation

5.30 In Bolivia, organized groups of beneficiaries could overcome the lack of institutional capacity in the formal public sector. Their potential, however, remains largely untapped. Despite the heavy emphasis placed by MPSSP on community participation in its policy statements, effective participation has been mostly limited to vaccination campaigns. The support and participation of the community in the health delivery and referral system is essential for the system to work. Such participation is effective and sustainable only when the community is present in the design, execution, evaluation and control of activities. It is useful to examine the type of beneficiary participation enjoyed by some NGOs in the sector. This support has three forms: direct payment of fees for serviceJ, community provision of health infrastructure, and community provision of benefits to health care workers (e.g., housing, food, etc.). Often, NGOs operate community-owned infrastructure and thus avoid the capital cost of constructing their own facilities.

Insufficient Use of Non-Governmental Institutions

5.31 NGOs have taken on an increasingly important DTstrTbutlon of Requests from Health NGOs to ESF role in providing primary health care services in Bolivia. In many rural areas, they are the sole providers of services. However, their services 22.2 \ are uneven and lack coor- dination. It is estimated 67% that of the 300 NGOs in the health sector, fewer 30.t than one-third have for- malized relationships with Dlstrlbutlon across the Government. In some Poverty Areas regions (e.g., central Co- o chabamba), NGO presence is C0 AREAI (rchest) AREA2 E AREA3 large, while in others 2 ARE 4 S AREA5 (poore3i) (e.g., northern Potosi), Source ESV they are almost non- existent. Unfortunately, Figure 5.4: GO Requests to ESP although NGOs are more likely to be established in poorer areas than most public sector institutions, they also have bypassed many of the poorest regions (see Figure 5.4). Because of lack of coordination between the various types of NGOs, there is often substantial duplication of services in some areas and non-existent coverage in others. In one region of the Altiplano with 80,000 inhabitants, for example, an estimated 30,000 people are receiving duplicate services from more than one NGO, while in Northern Potosi, there are communities that have never seen any intervention, public or private.

5.32 As with most other institutions in the health sector, NGOs have faced significant obstacles to efficient service delivery. These obstacles include 71 cultural beliefs and customs among indigenous groups, a strong presence of traditional medical practitioners, economic constraints on paying for items like medicines among largely subsistence populations, and a lack of infra- structure for access to facilities. They have been more successful at providing primary services outside health establishments, such as mobile vaccination campaigns.

5.33 There is generally a low rate of return on NGO investment in infrastruc- ture, particularly in rural areas. On the positive side, NGos appear to be the most efficient of the institutions in the health sector if measured in terms of consultations per staff per hour. The figures for the three main institutions are presented in Table 5.2. Furthermore, there is some evidence that NGOs can provide health services at a lower cost than other institutions. The primary factors that contribute to this are the concentration of NGOs within lower cost, primary health care. There are also Ministry personnel working with NGOs whose salaries are not included as part of NGO expenditures. Finally, NGOs have a general ability to stretch resources further and secure donations for their needs.

5.34 Not all NGOs working Uti±erI of Consutatiom per Staff per ftw in health have proven ef- by Tvpe of 0rgAlzathisn ficient, however. There Center Center Ave- is a wide variance in Post w/oSaed, /Beds Wospitat rage efficiency among NGOe and among projects. In a MPSSP 0.29 0.45 O.ZO 0.06 0.11 study, which reviewed scialSec. 0.76 0.29 0.38 0.17 0.3 thirteen health projects 9GOs WO0.37 n.s.0.17 0.38 located in six different Source: PSER departments, it was shown ______that NGOs varied Tablo 5.2: Consultations per Staff-hour dramatically in the number of patient visits per post per month and in their per capita expenditures. 7 The average number of patient visits per day to NGO health posts in the rural area ranged from one to three. Urban-based NGOs showed a much higher concentration of patient visits, even where NGOs operated under a -1-icy of full cost reccvery.

5.35 Finally, NGOs have historically had unproductive working relations with MPSSP, although their relations with the SY8 are better and more concrete, Onle reason for these poor links is the failure by MPSSP to make sensible use of the agreements signed to coordinate and rationalize the activities of NGOs. Moreover, a history of mistrust and skepticism exists on both sides. Among NGOe, there is a consensus that MPSSP is inefficient and ineffective due to its l_rge n-aSber of pzograr,, poor'.ytral-.a.6, u..da.paidand poorly distributed staff, vertical management, and lack of community participation. On the other hand, MPSSP faults some NGOs for not following national health policy or for not being fully integrated with the SUe. This, in part (according to MPSSP and some critics), is due to an operational style of NGOe, which pays more

I From PSER, September 1989. 72 attention to its financing agency, admiiistrative budget and community counterpart than to the national health system and policy.

Women's Issues

5.36 In comparison to other ministries, MPSSP has the best-articulated and most far-reaching program reflecting women's concerns. The priority policy is maternal-child health care, and 83% of ministry programs are directed at women and children. Seventy percent of health service beneficiaries are women. Many health committee members are women. In 1989, the Ministry established "Women, Health and Development" program to "increase women's participation in defense of the right to health, contribute to the improvement of quality of life, and give renewed value to the role of women in the development process."

5.37 MPSSP's program goal is the active participation of women in the design of health policies and strategies. The program's four action lines are: (1) social support for child care and home-making through integrated centers offering supplemental feeding and child care; (2) non-formal education, including training for community health workers, traditional midwives, commu- nity child care workers and community organizations; (3) health activities focusing on women such as pre- and post-natal care, detection and treatment of uterine/cervical cancer, and mental and occupational health care; and (4) the training of women's health care providers.

5.38 The National Maternal Health and Child Survival Proaram (MHCSP) was issued in November 1989. The plan's objectives include universal immuniza- tions and the eradication of polio by 1990, increased coverage of growth monitoring and development for children under five years, increased prenatal coverage, reproductive health services in all health facilities by 1992, universal human sexuality and family life education by 1993, development of local integrated health systems throughout the country by 1993, and national application of integrated health attention norms for women and children (ages 0-5). The goals to be reached by 1993 are a 50% reduction in maternal mortality, 30% reduction in infant deaths within the first month of life, 50% reduction in infant mortality (0-12 months), and 30% reduction in mortality in children 1-4 years old. In the months that have passed since the highly publicized introduction of the MHCSP, nothing has happened. Unfortunately this is becoming an all too familiar problem. Very good policies or strategies have been presented but implementation is lacking.

5.39 MPSSP estimates that 30% of deaths related to pregnancy and births are due to induced abortions. Though abortions are illegal, the Ministry esti- maw.ea that up to 20,000 were performed in 1989. Of this number, 30% will be admitted to a hospital subsequently. An estimated 80% of female hospitalizations are abortion related. The Departments of Oruro, Cochabamba, Sucre, and Trinidad show the highest abortion rate in the country: 20 per thousand women of child-bearing age (15-49 years) per year. Poor women may induce abortions because family income will not support another child. Poor women in marginal areas in Santa Cruz report that they perform abortions on their daughters because the latter are victims of rape or incest, are unmarried or cannot afford to have children. CIES, an NGO that offers 73 maternal-child health services, found that 40.4% of 2,776 low-income women who visited its La Paz clinics during 1987-88 had had at least one abortion. Reasons for abortion ranged from lack of knowledge of human physiology and knowledge/access to birth control methods to the economic crisis and the related perceived necessity of limiting family size.

5.40 The high rate of abortion seems to show a large, unmet dewand for reproductive health services. Better access to reproductive health services would have many other positive effects. For example, the DHS survey showed that infant mortality was three times higher for children born within two years of a previous birth than for children born after four years or more since the last birth. The seventh child born to a family is 35% more likely to die within the first year than the first born child. 44% of children below the age of three born within two years of the last birth were malnourished, while 29% were malnourished when four or more years had passed between births.

RECOMMENDATIONS

Implement Existing Policies

5.41 Several well-designed programs and policies are already in place, but are not being implemented. A good example is the Maternal Health and Child Survival Program (MHCSP). This extremely important program, which is ready to be supported by several external agencies is still just a program. Nothing is happening in the field. Many other national campaigns are working well, like the vaccination campaigns. Anecdotal evidence shows that supplies for these campaigns are the only supplies that ever reach remote areas. The Ministry should build on these positive experiences and begin implementing other promising policies.

Shift Expenditures

5.42 The first priority in the health sector should be to ahift expenditures toward Primary health care, rural and peri-urban areas, with a special focus on mothers and children. This has been the Government's stated intention, but actuai expenditures ano investment programs do not reflect these concerns. Intermediate steps would entail implementing the MHCSP (see preceding para- graph), nna lirmitingnecondary health care expenditures to a certain share of total health spendiig, in th-e shott-Let--n,an inventory of health facilities and staff should be madeB to assess needs and priorities; and donors should be required to follow the Ministry's plans to avoid more empty hospitals with faincy aquApuIfiat thit# hbe eih- q,alnkaXeed nSo 4t ete*t

The Ministry should coordinate with SIr's ongoing activities in this area. 74

Institutional Recommendations

5.43 Institutional improvements must be centered on better financial manage- ment and imgroved human resources. A first important step will be to imple- ment the financial control systems envisaged in the integrated health develop- ment model, supported by several external agencies. This system would improve budgeting and control at all levels within the health system. Salary issues are as hard to resolve in health as in any area, but without substantial improvements in salaries the effective implementation of policies cannot be ensured.

Coordinate NGOs

5.44 The Ministry should assume its normative and coordinatina role. while leaving the implementation and service delivery responsibilities to the Sani- tary Units (SUs). selected NGOs. and community organizations. Also, it should develop a coherent program to govern, assist, and coordinate NGOs in the health sector. Improved working relations with and among NGOs are essential to extend coverage to rural and peri-urban areas, to avoid duplication of efforts and wasted resources, and to achieve a truly national health policy.

5.45 NGOs are showing an increased willingness to cooperate with MPSSP, principally at the SU level, and incentives should be pursued to further encourage this participation. Practical recommendations include: (a) NGO participation in the Departmental "Consejo Tecnico"; (b) NGO participation in the development of national and regional health strategies, possibly through an NGO network; (c) increased contracting out to NGos and certain private businesses for service delivery through effective "convenios" (agreements) with the Ministry; (d) assignment of budgeted staff slots to NGOs that have demonstrated a capacity to work efficiently with poor populations; and (d) systematic use and follow-up of information provided by NGOs on the needs of the rural poor.

Encourage Beneficiary Particination

5.46 The support and participation of the community in the health care delivery and referral system, iz cc^:ntialfor intcrvcntionz to reach the poor. The community should be present in the design, execution, evaluation and controi of activities directed at improving its well-being. Freauentlv the as6essments of a communitv' health needs are not presented to the community. This should be the first step toward organizing the community into action. There are three stable community organizations that could be increasingly Lnvolved in health care:

(a) Xotlar Clubsh a 20-year I;i6Lciy n ace presently involved primarily in channeling food donations.

(b) Popular Health Committees (CPSs) have been promoted by the MPSSP since 1982 as a means to mobilize community participation in the organization and delivery of primary health care. They are elected by Neighbors, Councils for matters related to health. The CPSs are organized at the district, departmental and national 75

level and assist in the definition of priorities, planning, execution, and evaluation of health activities.

(c) Neighbor's Councils are traditional civic groups organized around community interests and are the basis for participation of the community in municipal and labor activities, particularly for the execution of self-help projects.

Upgrade and Reorient Training

5.47 Community health workers are the first point of contact and, thus, the critical link between the community and the system. There are two kinds of community workers: the auxiliary nurse and the health promoter. The auxiliary nurse provides medical support services supervised by health center/post doctors and nurses. The health promoter is usually selected by the community, trained in the promotion of primary health care and is generally not paid. MPSSP and SUs should u2crade the training prooram for auxiliary nurses. as well as doctors and nurses. including courses on community iarticipation. The role of the health promoter should be clearly distinguished from the auxiliary nurse to avoid duplication of efforts. One possibility is to train the promoter in traditional medicine to serve as a bridge between the traditional and modern medical systems (see next section).

5.48 Training of health staff and education of medical students should be reoriented to more effectively meet the country's requirements for basic health care. This would include: (a) increased emphasis on maternal and child care, community medicine and nutrition; (b) training to deal with illiterate and non-Spanish speaking patients; (c) increasing links between teaching professionals and authorities responsible for basic health; (d) internships for students, and increased incentives for staff to work in remote rural areas.

Intearate Traditional and Modern Medicine

5.49 Given that many rural poor continue to rely on the traditional medical system--by choice and by necessity--an integration of traditional and modern mdicit-e systema zould prove wore eaffctive in reaching these populations. Such integration could include: (a) training community leaders and traditional health practitioners as health promoters; (b! training traditional midwives in modern medical practices regarding prenatal care, childbirth, etc.; (c) coordinating training between MPSSP, the Medical College and the Bolivian Society for Traditional Medicine; (d) requiring doctors and nurses to travel periodically to the communities to understand the social context of health and Illness; (e) incorporating the option of using the system of labor exchange-- instead ot casn, doctors can request work obiigatiuns, whieh would iower Lhu public sector coats of providing, e.g., an improved sewerage system;9 and (f)

9 This could be implemented through each community having an "account" in SIF with labor owed for services rendered. Thls "account" would then be activated as community contributions to lower the financial costs of SIP projects in the community. 76 increasing use by doctors of medicinal plants as much as possible, recognizing their efficacy and the widespread unavailability and high cost of pharmaceutical products.

Nutrition

5.50 Nutrition interventions should be closely monitored and coordinated. MPSSP has the formal responsibility to coordinate and monitor nutrition interventions but has not fulfilled its role. The first step would be to study the effects of present nutrition interventions to understand the effects of current programs and to better target future programs. Such efforts should include strengthening of the National Nutritional Surveillance System and a reassessment and possible strengthening of the National Institute of Nutrition. The ongoing iodine program should be continued and expanded to include the traditional block salt. Other nutrition interventions much as school breakfasts and other education related programs should be evaluated to obtain better targeting. Over time nutrition programs' reliance on food donations should be limited, allowing for purchases in local markets, and thus stimulating local production.

5.51 Specifically, individual community nutrition Droarams should be improved by assessina the nutritional value of foodstuffs and usina local products. Women should be trained in how to care for severely malnourished children and community-based intervention programs for these children should be established. For communities dependent on food donations, a short-term scheme to rationalize the use of foodstuffs should be implemented in which certain donated items are used to complement, not replace, local products.

Women's Issues

5.52 Immediate actions to improve the situation of women would be the imple- mentation of the MHCSP. Another immediate need is to follow up on the recent- ly held conference on abortions, with an action plan agreed upon by all parties involved. At a minimum such action Dlan should include a simple but effective oroaram of revroductive education. rn the longer run, it is hard to see how the alarming maternal mortality and infant mortality rates can be brought down without a substantial increase in child spacing. Family planning and sex education are very sensitive issues, and extreme care should be exer- cised in introducing programs that are sufficiently sensitive to cultural and reli2jous beliefs. There is some hope to be found in the increasing female literacy rates (see next chapter). Cross-section analyses and results from other countries show that increased female literacy will lead to improvement in both maternal and child health. 77

CHAPTER 6 i EDUCATION

INTRODUCTION

6.1 Educational achievement is very low in Bolivia. Official illiteracy ratee are among the highest in Latin America. The last nationwide testing (in 1979) showed no statistically significant learning' for rural education in both reading and mathematics. Urban schools did show statistically significant learning in reading but not in mathematics. Since then the education system has deteriorated according to anecdotal evidence from parents, students and universities. For instance, the university in La Paz has to have its students take basic mathematics and Spanish before allowing them to start a university degree. The number of students failing such courses has been increasing over time. Rural poor areas are especially hard hit by the crisis in education. Poor rural areas have lower participation rates and higher repetition rates, and higher illiteracy rates.

6.2 Besides the direct posi- tive welfare effects of educa- Children's Health by Mother's Z'ducation tion, good education is highly x correlated with other social 50 indicators. As shown in Figure 6.1, women with better educa- 40l tion have fewer sick children, and when the children become 30 lI sickthey are treated more of- ten. There is also a strong 20- positive correlationbetween the incidenceof vaccinations 1o and the level of educationof the mother. 0 Cough Dkrrhea Voccinctod -ofwhich saw doct.r -of whichoral rehyd 6.*3 The fundamental problem Nne 0 Basic a Intermediate 0 Higher 0 Total in the Bolivian education sys- SIa,=ENsP tem is inefficiency. Student/ FI teacher ratios are low. Figure 6.1s Children's Realth by Mother's Teacher salaries are low, but Education the total wage bill covers 98% of all non-university spending. The average rural student takes five years to complete grade 1 through 3. School administration is doubled all the way through because of a dual rural/urban school system. In an inefficient system such as this, equity improvements are hard to come by.

6.4 This Chapter focuses mainly on how to improve efficiency. First, the principal institutions in the sector are described. Second, educational a2endino is analyzed, focusing first on the inequity of spending (the bias in favor of urban areas and universities). Third, the Government's draft reform

1 The answers given on the tests were not significantly different from random answers. 78 p.= is analyzed. Fourth the public investment program in education is reviewed, contrasting the investments by the Ministry with those of the ESP. Fifth, a series of sector constraints and issues is discussed, mostly relating to the low efficiency of the system. The constraints include weak planning and policy-making capabilities, inflated school staffing, low teachers' salaries, inadequate school administration, legal impediments and under- utilization of NGOs. Sixth, a set of recommendationa is presented. These include improving equity of expenditures by spending more on non-university education, raising teachers' salaries through efficiency improvements; improving the allocation of human and physical resources, improving planning and financial management, increasing the role of the communities and of NGOs, improving women's access to education and increasing emphasis on bilingual education. Finally, the recommendations contain a brief outline of a possible phasing of the daunting reforms needed to improve education for the poor.

INSTITUTIONAL FRAMEWORK

6.5 Within the education sector, the principal institutions affecting the poor can be categorized into four broad groups:

a) The Ministry of Education and Culture (MEC) and its corresoondina agencies. Education in Bolivia is the responsibility of the government in general and the MEC in particular. MEC formally controls almost every aspect of public and private education except for the autonomous university system. In the public system, MEC defines the content of the curriculum, pays the teachers, finances the construction and repair of schools and administers the operation of the schools. MEC is also responsible for the private school system. It negotiates and monitors agreements on education with the Catholic church, other religious organizations, private schools, educational cooperatives and secular NGOO. There are three independent agencies within MEC: the National service for Alphabetization and Popular Education (SENALEP), established in 1983 to help achieve the goal of eliminating illiteracy by the year 2000, an agency that did have some good programs in the past but is presently without funding; the National Technical Education Service (SENET) and its fourteen technical education institutes, created in 1978 to improve the quality of industrial, agricultural and commercial education; and the National Council for School construction (CONES), responsible for the infrastructure of the educational sector. Until 1987, earmarked taxes financed CONES. Since then, it has depended on grants from the ESP for most of its investments.

b) Denartmental Education Administration. Each of the nine depart- ments has its own administrative structure for education. Each departmental educational system is headed by an Executive Council, and includes two directorates responsible for the separate rural and urban school systems, respectively. 79

C) Church organizations and other NGOs. By supervising and con- structing schools in areas in which they are active, organizations such as Fe y Alegria, Accion un Maestro Mas, and Escuelas de Cristo provide educational access to poorer children. Although the private for-profit educational system in Bolivia is also very active and diverse, it reaches virtually none of the poorer segments of the population.

d) The Emergencv Social Fund. ESF has met specific construction, materials and training needs of local communities easing the inequities of the uneven public investment program in the education sector.

EXPENDITURES AND GOVERNMENT POLICIES

Expenditures Distribution of vd=atfon Resoureot byArea (1988) 6.6 Expenditures Perseml Expenditures Nuber of Teachers Population are skewed toward (in afUions of as.) (X} urban areas and Dertment Urban Rural Urban Rural Urban Rural non-poor Depart- -______ments. It is SantaCruz 16.6(75) 54 (25)7,02 (61) 4,251(39) (48) (52) estimated that 80- Cochaba*9 23.1 (65) 12.7 (35) 6,490 (64) 3,719 (36) (36) (64) 90% of rural school cubquf ac 6.4 (68) 3.0 (32) 2,703 (58) 1,946 (42) (19) c81) children, as compared to 40-50% ourcett'S0 Project,J e 1,§ of urban school Mote: Percentage share in parenthese. children, lack textbooks. In the

3a-E;vXsF-ByE=ne< W tv;

2 Since the tax reform in 1985, univeraities have received an annual transfer of 5S of gross internal tax revenues. 80 groups, the present allocation policy heavily favors the rich at the expense of the poor, accentuating educational inequity in Bolivia.

6.7 MEC expenditures are used almost exclusively for salaries. EC's Expmdittreby Deprtmt (198) From 1980 to 1987, the percentage of the MECbudget used to finance oepartment NECEExpditure Percapita salaries de reased only from 99% to (thousands of U) (US) 98%. Meanwhile the share of the MEC's expenditures on materials has .huquisca 79037 15.9 gradually decreased to only 0.1% of La Paz 28799.2 13.5 the total education expenditures in Cochabama 17226.1 16.0 1987.3 Orgro 7085.1 Potosi 10640.2 12.' Tarija 5038.2 17.2 Reform Plans SantaCruz 17788.8 13.5 Bni 5190.0 18.3 6.8 In November 1989 NEC was Pa6o 07.9 11.1 asked by President Paz Zamora to Centrat 41Z1.2 prepare an educational reform plan within 30 days. The plan has not yet materialized, but a prelimi- TOTAL 104400.4 Aer. 14.4 nary di.scussionpiece was issued. The components under discussion Source: E4SOProject, Jun 1989. most relevant to poverty issues are higher priority for rural and peri-urban areas; higher priority Table 6.2: NEC's Expenditure by for the education and the rights of Dpartment women; national mobilization of all groups in Bolivia (Armed Forces, Church, NGOs, private enterprises, unions, etc.) to combat illiteracy; higher priority for basic education; Integration of rural and urban education with higher priority given to raising the quality of rural schools; availability of a school breakfasts program in all zones of critical poverty; construction of annexes to rural schools where students could work when not attending class; more volunteer activities to aid the disadvantaged, sick, children and elderly; visits and assistance by private school students to public schools in a "Colegios Hermanous program (n-o emphasis given to helping public schools in poorer areas); and more centers, cooperatives and private establishments for training the handicapped and reintegrating them into society. The specific plans for basic education include coverage of all urban and 60% of rural areas,4 extension of the basic curriculum down to age one or two in rural sectors, one day off for extracurricular activitiee or work p(e week, reform. of the grade achocl curriculum to emphasize core needs, and larger absolute and relative budgets for grade school education. The reform plans for basic education are

3 See PSER for a review of social expenditures.

4 If INE statistics on coverage could be trusted, then this target has already been reached. What appears to be happening is that children in the communities without schools eventually go to relatives in raighbor- ing communities with schools. 81 potentially very dangerous. As long as numeracy and literacy is not achieved, children should not be given a day off for extracurricular activities.

Public Investment

6.9 Owing to the government's tight budget situation, public investment in education amounted to US$0.9 million (US$0.4 million was initially projected) in 1988, down more than 40% from the level of 1987. Public investment in education was stagnant in 1989 (US$0.9 million or 0.2% of the total), and will not increase much in the 1990-92 period. Investment during the next two years will be mainly financed by external donors increasing the dependence on external funds from 59% in 1988 to about 80% in 1990-91. Many previous projects did not address the national priority to enhance the coverage and quality of primary education, particularly in rural areas. Out of 12 projects undertaken in 1988, only two appeared to have had major positive impact on the poor. The other ten projects were weak in poverty focus, aiming at improvement in urban education system, university education, technical training, and cultural preservation.5

6.10 This insufficLent poverty focus in the education sector is likely to continue over the next years. Except for two UNICEF-supported projects, it appears that all projects planned for 1989-92 pay insufficient attention to providing primary education to the neediest groups. More than half the investment program for 1989 was planned for strengthening the televised education system. Separately, the Government intends to initiate a school mapping project in 1991 that would be a very important project providing an informational basis by matching under-utilization and shortages of educational facilities, and preparing for providing adequate and equitable education Infrastructure across the country. However, neither the funding source nor the basic concept has yet been determined for this project. Several important projects, which could have had an impact on poverty alleviation, have been excluded from the core program, downgrading them to lower priority. These projects would aim at development of bilingual textbooks for primary education (Spanish, and Aymara or Quechua) and promotion of intercultural education to address disparities in educational opportunities among different ethnic groups.

6.11 ES has been providing resources for the expansion of primary education. In 1988, EST financed US$6.0 million for the education sector, which was about seven times as much as the MEC and RDCs. As of mid-1989, ESP had disbursed US$13.3 million including USS3.2 million for 100 projects for educational assistance (i.e., school desks, textbooks, teachers' training, and other education programs) and US$10.1 million for 415 act:oolrepair and construction projects. In 1989, ES? financed US$9.2 million worth of education projects (about ten times the rest of the public investment in education).

6.12 3SF-financed education projects were well distributed among departments (except for Pando). Especially, the Cochabamba and Potosi departments, which were largely neglected by the MEC and RDCs, received a fair portion of funds.

5 Restoration of museums and churches in La Paz and Potosi. 82

3SF-projects have focused on primary education and have mobilized partici- pation of various organizations in the public, private, and NGO sectors. oiUvtet' adjustnt, en thogh Itwas written In Oconyfc ters, due dep into the imnsemosbebviwral patterns of cwiety. we 6.13 Upon the terminaticn of ESF, 1 d* 4. C 1, * r SIF will take over and expand ESF ac- A th wbit{ct cordthat has bn devltoped tivtties in the social sectors. In throug .tnwies between the state and the edacation, US$14 million is expected tta..R goingtI$ to help you with whatever to be available for education m""", OMean,adhoveI,btfirst ecOndty I"m only I golrgtv to heaphe you " infrastructure, education materials, In thoo roas were you ask for heLp. I'm and program assistance annually not goinm to tryto figure out your needs during the first two years. If the a8yfcO. Forget abou patenalIs; the SIP contribution is added to the co'etsoateotlonia-xetrorys rid already planned public investment, forer". the total investment in public education will amount to US$16.3 Ferncw ottSfu million in 1990 and US$19.0 million in 1991, which is two to three times as large as that in 1988 (including . -- ,: NSF). Box 6.1: Cuttingthe UzblilcalCord

SECTOR CONSTRAINTS inflated School Staffina

6.14 The average student/teacher ratio in Bolivia is about 17, much lower than in comparable countries. School staffing is inflated in large part by three main factors* teacher specialization, allocation of "Items," and "ghost" teachers. First, there are five special fields in Bolivia's primary schools and eleven special fields for middle and superior schoole. A teacher trained for one specialty, such as music, will not teach any other specialty if it can be helped. This means that a fully staffed primary school should have five teachers for each grade, while middle and superior schools should have eleven teachers per grade. As a consequence, Bolivia's primary school teachers are supposed to work loes than twenty hours a week, though paid for forty. Similarly, teachers for higher grades are supposed to spend only two hours and ten minutes per day in the classroom. The remaining time is supposedly spent grading homework and tests (although there is considerable evidence that little homework or tests are given) and improving teaching skills (although there are no teacher training programs). In reality, about 20% of the urban grade school teachers use this time to teach in private schools. In most rural schools, teachers are forced to teach several specialties despite training and a curriculum that assumes specialization. These schools usually have small student populations served by one or two teachers. These teachers have to cover the entire range of courses and several grades. They also usually do not have access to the same employment possibilities as urban teachers and are often sepazated from their families for long stretches of time. Yet they have to work twice as long for the same basic pay as the urban teachers. 83

6.15 Over-staffing is also due to the way the NEC allocates Items (teacher equivalents) to schools. once an item is given to a school, it is almost never withdrawn even if the community served by the school changes and most of the families migrate elsewhere. Politically, MEC is unwilling to face the consequences of closing a school and technically it lu unequipped to re- evaluate the justification for items. Each year several thousand requests cannot be filled because the budget is not large enough to finance the requests. These are usually small, rural, communities that are willing to build a school by themaelves if the Ministry would finance the item. Finally, the budget is in essence overstuffed by "ghost" teachers receiving checks although they are not on MEC records. In 1988, the number of teachers paid by the Ministry of Finance was 14,500 or 23% larger than the number said to be teaching according to MEC. The Government is implementing a computerized tracking system to make sure that everyone who receives a check is in fact teaching.

Low Salaries

6.16 Bolivia can no longer afford large numbers of low paid teachers. Almost every problem in the education sector can be linked to the poor performance aitdthe low incomes of teachers. As salaries already absorb the vast majority of expenditures, substantial increases in salaries can only come about through increased efficiency of the system. The previous Government developed a plan which places more emphasis on raising teacher incomes than the current draft plan. The previous plan sets an explicit goal of recapturing for teachers their real salaries of 1980 by improving the efficiency of the educational system. It proposes an explicit target percentage of GDP for educational expenditures. The new draft plan attacks the "myth" that higher teacher salaries are essential to effective reform. This plan does promise higher salaries and more prestige for basic education, but the promises are vague and, given the recent conflicts between the teacher unions and the Paz Zamora Government, probably unpersuasive. The demands for higher teacher salaries crowd out other budget items, so there are inadequate funds for supervision, construction and repair of school buildings, texts, classroom supplies, training and special projects. The low income of teachers creates a shortage of qualified teachers and makes it difficult to set reasonable teaching standards. These problems are worse in the rural areas and the poorer regions of the country, so low productivity increases the inequity of the educational system.

Weak Plannina and Policy-Making Caoacitv

6.17 NEC is faced with serious problems in administration, planning and coordination between the central and regional levels. Important contributions to this problem are: (a) the organizational structure of the Ministry that concentrates decision-making in high-level central administrators, (b) the low level of professional training of many Ministry employees, and (c) the limited non-salary budget that makes travel and communications between the Ministry and the regions extremely difficult. The administrative structure of MEC is highly centralized and hierarchical. Educational policies are made by the Minister and in the four Sub-secretariats for Coordination, Financial Ad- ministration, Rural Education, and Urban Education. The National Office of 84

Educational Planning monitors the effects of MEC decisions and assembles statistics on the educational sector. While these offices are physically located in La Paz, there are two parallel administrative systems, one for the rural sector and the other for the urban sector. Both systems stretch down from a national director in La Paz to the departments, zones, regions, local districts, and individual schools. As a consequence of this structure, MEC's administration often substitutes administrative routine for responsive decision-making. This is due in part to the low quality of most of MEC's administrative staff; it is easier to do the same thing year after year than it is to try to improve or adapt to changing conditions. This inertia also reflects the centralization of authority in La Paz, combined with weak planning capacity and the overlapping and often redundant layers of admini- stration within NEC. Finally, the separate rural and urban systems, the autonomy of the Universities, and the teacher unions create additional barriers to change.

6.18 At the planning level, the Plarning Division within the Ministry lacks qualified staff and effective ties with other internal offices, particularly the Budget ofZice. It lacks control, for example, over how many teaching positions to request for each fiscal year. The Ministry of Finance simply allots a given number and increases it by a small percentage each year. Each region is then allocated a given number of teaching spots rather randomly. Another problem is the lack of a sufficient data base from which to plan. Lack of updated census information, as well as an inadequate information system in the Ministry of Education, make it very difficult to determine accurately where teachers, materials, and school buildings are needed. At the budget level, the Ministry's Budget Office has responsibility for drafting the budget but has little knowledge of actual needs because it lacks the data base and skills for conducting the necessary analyses. Consequently, inadequate justifications for budgetary items are offered to the Ministry of Planning, which in turn reduces Ministry of Education proposals to little more than salary requests for the existing teachers. Another problem is that at both the central and departmental levels many units are over-staffed, including: curriculum development, evaluation, vocational counseling, and student services departments. To carry out the principle of providing different curricula to urban and rural students, there is a tremendous amount of duplication in all line positions.

Inadeauate School Administration

6.19 Individual school administration is theoretically the responsibility of each school principal. These officials, however, spend most of their days teaching and have no real decision making power in matters of personnel, budget, or obtaining materials. There is no special training available to enable them to provide instructional leadership In many locations, schools are grouped together into a "nucleus* under one director. Howeiier,nuclei directors are provided no means to visit their schools on a regular basis.

6.20 All positions and procurement for local schools are filled or carried out centrally, thereby removing any responsibility or power from local school directors to run schools in a manner they deem most efficient. Each year, all teaching poeitions are filled during a three month period called the time of 85 permuta or barter. Vacancies, candidates for those vacancies and requests for transfers to new positions are passed on from the local school director to the zonal supervisor, to the regional supervisor, to the departmental director, and then to the general director of the urban or rural educational systems. If no qualified candidates can be found within the centralized system, then the school director is given permission to hire an interino, someone who is not a graduate from a teachers' college. In theory, the Interi-iocan be replaced by a more qualified teacher during the next barter period, but in practice this happens infrequently. Teachers' unions claim that the interino system is used for political patronage and that many Interibos never teach.

Leocal Impediments to Education Sector Efficiency

6.21 The charter for the Bolivian educational system is the Bolivian Educa- tional Code of 195S. The Code has been modified and elaborated by Presiden- tial decrees, laws and practice. The most important are the Educational Reform of 1969 and the Law of Bolivian Education of 1973. Some loftier goals of the Code were not realized, but it remains the key legal document for Bolivian education. The centralized administrative structure, the dual rural and urban educational systems, and the hierarchical curriculum ere all mandated by it. The rules make employment depend on the supply of teachers and not demand for teachers. Graduates fzom teachers' colleges are placed on the payroll on receipt by MEC of a diploma from a Bolivian.teacher's college and cannot be removed from the payroll unless tried and convicted of serious moral or disciplinary offenses. Promotions and pay increases follow seniority rules (the Eacalafon) without any regard to job description or merit. Teachers can choose to take a test every four years that will award them one extra year of seniority if they pass. The tests are entirely voluntary and no penalty is imposed for failing. Administrators are paid according to their seniority as teachers as well, i.e., an administrator does not get paid extra for extra responsibility.

Inadeauate Non-Formal Education

6.22 The NEC is supposed to provide a number of educational opportunities for those ineligible for the regular school cycle. Among other things, it is supposed to provide formal education for adults 15 years or older, specialized centers and vocational institutes for the handicapped, and special educational programs and extension services outside the school system for the general public. These laudable goals are still unrealized. The existing institutes are very expensive and reach on]-,a small portion of the adult population. For example, there is no adult education in the rural sector, and although the Government is now proposing to expand educational programs for the handicapped and provide non-formal education, it was supposed to provide this service over 35 years ago. Currently, the only significant programs in non-formal education are conducted by NGOe.

Under-Utilization of the NGO Oction

6.23 The Bolivian Government is not utilizing fully the option of allowing NGOs, religious associ&tions and private schools to compete with the public school system. NGOo, in particular, can often provide a type of education 86 that is much cheaper and more effective for the poor than the usual public monopoly,6 but inadequate Government support and financing limit their potential and leave significant demand for their services unmet. For example, some observers estimate that the number of public schools managed by the Catholic Church would double if the government approved all community requests for their services. In 1988, over seventy communities petitioned for Fe y Alegria schools (one was approved). There is also a similar queue of com- munities that would like to have schools managed by Eacuelas de Cristo. On the other hand, Accion Un Maestro Mas (AUMM) had a tradition of starting schools that eventually were incorporated into the public education system through an agreement with the Government. In 1987, however, they stopped turning over their schools to the public school system because of complaints from communities about a decline in the quality of the education after their AUMM schools became public. Furthermore, none of the public schools that were once AUMM schools have retained the integrated approach to education and agricultuzal production that was and is AUMM's tradition and strength.

TheEscueta de Cristo, located in the agricuLtural communityof Totorapampa,20 km. from Potosi, is sdministeredby the Franciscanreligious caummuity.There are 185 students (74 girls) and ten Governoentteachers. The schoot offers classes from kindergarten through eighth grade and serves as the central school for surrounding,smaller communitieswith "feeder" schools which each have one teacher for grades1-4. Thetotal student populatfon In this systemis 550. The systemis welt oranized and follows the state rural educationcurricutum in the context of Franciscanpoputar education methodology. The curriculum includesagriculture/animal husbandryand religious Instruction. Of special interest is the active participation of parentsand the comnwity in general in school affairs. Theparent/coommity school auxiliary is led by a Boardcomposed of a president, vice-president, treasurer and secretary. The Boardsupervises three school committees chargedwith, respectively, CiOthe coltection of monthtyquotas from families (Bs.50 or one arro of agricultural products) used to purchasematerials, pay for building upkeep,and remwweratethe womenuho prepare the school breakfasts; (ii) the organization of manuallabor for school maintenanceand work in the school gardenwhose potatoes and beansare sold with profits returrnedto school; and(iii) the administration of the school breakfast program which provides a glass of tkand bread to eachstudent, using dried milk, flour andofl donatedby Caritas. TheBoard is sisoresponsible for student enrollment and school attendance, a particularly difficult task at the start of eachschoot year becauseFrancisean polity requires a mininum of 20 studentsor elsethe schoolis closed.This happens frequentlty in the aalIlercomamuities. The Board meets weekLy and submitsa workplan to school authorities oncea month.

Box 6.2t A Rural School

6.24 Currently, about 12% of the rural students are enrolled in schools managed by NGOs (including the church) and a much larger number benefit from a variety of special education programs run by NGOs. Of these providers, only the Catholic church competes with public schools on an even ground. Catholic churches and organizations have agreements in which the government agrees to pay for teachers in schools that are not managed by the Government. Under

6 One measure of efficiency is the teacher/student ratio, which is between 17 and 18 for the public sector. In three NGOs teacher student ratios ranged from 26.5 in (Accion Un Maestro Mas) to 33.5 (Fe y Alegria). 87 pay for teachers in schools that are not managed by the Government. Under these conditions, the subsidy from the Government is the eame whether the school is managed by the MEC, a Catholic association or Diocese. At the same time, however, the Government limits the supply of these schools by withhold- ing approval. For other NGOs, religions and private schools, the Government does not pay the salaries. These organizations have to finance the entire cost of the school from student tuition, donations, or other sources of funds. In the poorest communities, those in the rural and peri-urban areas, there are many NGOs (over 120 according to a recent survey) that provide a wide variety of educational services. However, lack of public financing limits their ability to compete with the public educational system. Although the recent education reform plan proposes a government/NGO partnership in education, many of the potential options within such a strategy need to be further explored.

RECOMMENDATIONS

Expenditures

6.25 Expenditures in education should be shifted toward basic education In essence, the stated strategies should be reflected in the actions. This would require some serious rethinking of education management, where highly vocal groups would feel threatened. The earmarking of five percent Df tax revenues to universities is a positive net transfer to the middle classes from the rest of society. Projections show this system would in the year 2000 result in more than a third of spending in education going to the richest 1% of the student population (the university students). Any issues of fairness or income redistribution aside, that is a monumental waste of money.7 Within basic education more money needs to be set aside for materials and maintenance of schools. The Social Investment Fund (SIF) should be allowed to play a major role in this. In the short run, SIP should be charged with all financing of school construction and repair, to ensure that local demands are taken into account. A first step toward greater attention to the poor in education expenditures would be to freeze university expenditures at the present share of total education expenditures.

Imorovina Teachers' Salaries

6.26 Teacher income must be explicitly linked to the success or failure of reforms that improve the efficiency of the system. If the system becomes more efficient, teacher incomes increase, and vice versa. A couple of figures illustrate this effect. All the examples assume a fixed MEC budget. If the

7 University educations should be reformed as well. Test should be introduced to avoid students staying in school for ever, today about 10% of students graduate. Curricula needs to streamlined to focus on Bolivia's real xleudssuch as agricultural and health degrees, public sewtor management, and business degrees. Private sector participation should be encouraged to improve the fit between teaching and private sector needs. 88 number of teachers paid were brought into line with the number of teachers on HEC's records that would increase salaries by more than 20%. If the student/teacher ratio increased from the current 17.4 to 35, a level still below that of many other countries, and at or below the level of private schools in Bolivia, then the average salary would rise from Bs.220/month to 8s.584/month. If, furthermore, transfers to Universities were eliminated completely, teachers' average incomes would rise to Bs.72'/month. While no government could or would implement such drastic reforms as the latter two, it is still possible to provide many teachers with incomes two to three times last year's incomes, approve thousands of requests for new schools in the poorer areas that are now turned down for lack of funds, and adopt effective programs of teacher training and evaluation. Specifically, there are three possibilities: through a fixed budget share, efficiency bonuses, and selection on merit.

6.27 The fixed budget share idea is that Government could dedic&te a fixed percentage of treasury revenues to salaries of its teachers. If the total salaries paid to teachers were less than this target percentage, the teachers would receive a bonus. Such a policy would give teachers an incentive to be more efficient because teacher income would be a function of the student/ teacher ratio.

6.28 Teachers could be paid bonuses financed by some or all savings due to educational reform, thereby linking pay to efficiency. The bonus could be part of an improved reward system in which pay increases the longer the teacher's workweek, the more students the teacher teaches, the less desirable the conditions under which the teacher works (e.g., many poorer rural areas), and the more efficiently the entire educational system functioned. In this context, the Government should aleo consider treating teacher specialization as luxuries allowed only in the largest schools and reward those teachers already effectively teaching multiple subjects (e.g., rural teachers). Simul- ations (see paragraph 6.26) have shown that 20,000 positions could be dropped and there would still be enough items left over to start several thousand new schools in poorer regions of Bolivia. MEC should check student enrollments against the number of items for larger schools in regions of changing popu- lations periodically. Cutting back on the number of teachers is less likely to provoke community protest than closing a school down completely.

6.29 The third way that efficiency and teacher salaries could be linked would be to select on merit a core of proven. capable teachers to be expected to work full-time (40 hours a week or more). They would have to be flexible, willing to cover different specialties and even handle several grades at once. They would receive training and much higher pay as well. Their higher pay would be financed by their greater productivity. The entire transformation of the educational system could take place gradually and voluntarily. There is a shortage of (part-time) teachers at current low salaries. About a third of the teacher positions are filled with interinos, usually members of the comunity who have few qualifications to teach but are willing to fill the position in the absence of trained teachers. Instead of hiring inter1nos in the future, NEC could use vacancies to create new positions for full-time teachers. Interinos and incompetent teachers could be phased out of the system and gradually replaced by more productive highly paid teachers. The 89 pressure for more full-time positions by existing teachers would offset the traditional reluctance of teacher groups to see teachers let go for any reason. The shrinking number of teaching positions also will help the poorer communities get qualified teachers. Interinos are disproportionally con- centrated in the poorer regions and the more isolated communities. Replacing them with trained teachers would produce a more equitable system.

Limit Rural/Urban Differences

6.30 As the dual education system has clearly favored the urban system, it is hard not to propose a unification of the systems. However, a unification now would totally destroy whatever (little) is working in rural areas. Qualified teachers would move to urban schools were the work hours are less and opportunities for private employment are better. The rural schools would loose their "identity," as whatever little statistics are collected presently are collected separately for the two systems. Instead, short run measures should focus on strengthenino the rural system. Such improvements could include:

a) Introduce multi-grade teaching, where several grades are taught simultaneously. Teaching in such a system would mean higher pay and better benefits in return for teaching full-time and attending relevant training courses.

b) Make the school year more flexible to fit the needs for children's labor around harvest or planting seasons.

c) Introduce bilingual education8 through a national campaign to raise awareness of the country's cultural and linguistic diversity; rural teacher training in intercultural and bilingual education; primary school curriculum development with input from universities.

d) Imorove rural educational infrastructure using target mechanisms established under the Social Investment Fund. Many schools are in very bad disrepair, have no furniture, no materials, often not even a blackboard. Infrastructure such as offices and maybe even housing for the teachers should not be forgotten. Here there is great scope for community involvement in terms of inputs of labor.

e) Test system of rural boarding schools. For the rural dispersed populations boarding schools are often the only solution. Such boarding schools should be established with plots of land to teach students about agricultural techniques and to supply the students with food. Such schools also could be among the recipients of food aid.

6.31 After such improvements, especially in teachers Income and living standards, the dual career paths of the two systems should be abolished,

UNICEF has recently developed a proposal for introduction of bilingual education which could serve as a basis for future work. 90 sllowing the better qualified teachers to move into the higher paid, higher prestige rural jobs. The next step could be to unify curricula, unifying the requirements at each level, i.e., after first grade a child should now how to add and subtract, after second grade how to multiply with 1 through 5, etc. The actual textbook systems should not be unified, instead a variety of text- books should be developed. There should be a choice of the actual text-book system to allow each school district to adopt teaching to local needs within a national standard. Another step would be to unify the administration of the two systems, which could save substantial amounts of money. Each school would still be operating under different rules, but would all report to the same unit in the administrative structure.

Efficient Allocation of Resources

6.32 Internal efficiency in basic education can be improved through:

a) providing educational leadership through trained and effective school administrators and supervisors; and

b) introducing, in phases, a national testing system.

6.33 Effective allocation of teachina resourceg can be achieved through:

a) reducing redundant teachers and urban administrative personnel;

b) reducing the number of teachers' colleges, converting some into retraining facilities for teachers; and

c) making in-service training on subject matter and methodology available to teachers (e.g., through distance education or mobile units) and provide incentives (e.g., in the pay scale) that will motivate teachers to undergo in-service training.

6.34 To improve the Ministry's 2lannincLand budoetina capacity, three areas must be emphasized:

a) creating a stronger management and information systems for budget- ing, programming, and financial controls in the Ministry and in the regions, including a comprehensive school mapping system to conduct an accurate and updated needs assessment;

b) developing the management capability of sector personnel at both the central and regional levels; and

c) decentralizing decision-making and budget control to the regional and local levels.

6.35 To establish a more cost-effective use of exibting financial resources and to mobilize additional resources, three general objectives should be highlighted: 91

a) developing a comprehens4 ve education information system, including data based on school ma! Lng, to determine teacher and facilities needs and redundancies;

b) gradually shifting financial resources to the lower levels of the system, consistent with procedures to improve efficiency and quality;

c) instituting cost-recovery mechanisms at the tertiary level through fees for students who can pay and grants, or loans, for those who cannot; and

d) providing selected subsidies for private schools willing to serve low-income populations.

6.36 one of the greatest obstacles facing schools in poorer areas is the lack of a minimum amount of educational materials. To make these materials more widely available, a strategy of competitive bidding for textbook printing would reduce unit prices. This, combined with the development of a cost- recovery system to sell and/or lend reasonably priced textbooks to those who can afford it, would allow for special provisions to be made for the neediest families. Such a progrr-mcould build on the ongoing "One Student - One Book" program.

DecentrAlizati2n

6.37 The general recommendations presented in Chapter 4 (paras. 4.42-4.44) for implemntation of decentralization hold true for education. Specifically the following conditions should be met prior to decentralization or regionalization:

a) determination of which decision-making functions are to be de- centralized,

b) adequate staffing and training of regional offices,

c) development of workable relationships between regions and central administrations that will help avoid duplication;

d) development of information, reporting and auditing systems.

The Departmental Education Departments should become autonomous from the central level in financial and educational administration. In this context, they would be delegated clear responsibility for administration, control and supervision of school activities in both rural and urban areas. In matters of financial administration, they would need the authority to formulate and manage their budgets based on their respective regional demands and priori- ties. This would allow local authorities the flexibility to distribute expenditures according to regional necessities. This would avoid the current centralized and irrational assignment of teaching positions. 92

The Role of NGOs

6.38 Some NGOs have demonstrated that they can provide cheaper and more effective types of education for the poor than the usual public monopoly. There are also some large organizations, such as the umbrella organization representing Bolivia's farmers, that would be willing to manage parts of the public system. The following changes should be considered to increase competition in Bolivia' educational sector First, the Government could contract out the manaaement of schools in especially poor or troubled areas to NGOs with successful track records. Intsrnational financing could provide an incentive to private schools or NGOs to improve the management of public schools in poor communities. The Government could find willing partners in groups that had representatives in the communities and a long history of concern for the poor. Second, the Government could set up community councils for school districts. These councils would have the right to make their own contracts for the management of the schools within their districts with NGOs accredited by the Government. These councils already exist in many com- munities, but they have little power. Third, schools run by the Catholic Church and NGOs could be used as laboratories for government initiatives in public education. Many of these organizations have experience with teacher training, teacher evaluation, bilingual education, non-formal education, etc. The Government should inventory this experience and recommend ways to general- ize the successes.

6.39 Successful NGOo should be rewarded in several ways, such as with favorable publicity and media attention, prizes for their accomplishments, or through managament contracts with the Government. More important, all successful NGO schools should be accredited. If the performance of these schools is tracked by the Government and publicized to all, competition among schools will automatically increase. Those who were not accredited could continue to be monitored by the Government and NGOs could appeal for the right to contract directly with school community councils if they feel they have a successful track record.

Community Particination

6.40 Community representatives should be placed on Departmental planning boards. Other efforts to increase the power of local communities over their schools also should be encouraged. School community councils and the teachers in a given school should be encouraged to form their own NGOs and manage the school. This would be a form of school-based management, an innovative approach that works well in some parts of the U.S.A. The Government could also act a check on potential abuse by allowing MEC to retain its responsibi- lity for Bolivian education and track the performance of school-based NGOO, terminating contracts if performance were poor. Special members of the community, such as the alcalde escolar, could be given the responsibility of verifying the presence of teachers in the classroom. Absenteeism is apparent- ly a widespread problem both in urban and rural schools. Community members should be regularly consulted when teachers are evaluated.

6.41 Informal institutions as parent-teacher associations should be recog- nized and strengthened. These organizations have shown in many rural areas 93 that they can mobilize community resources for school operation and improve- ment. This has been confirmed through development experience in other countries in which the greater the degree of parental participation, the greater flow of local resources for teachers and schools. A greater focus on achieving primary education for all could be facilitated, therefore, through a trilateral partnership of government, the private sector, and parents.

Women's Issues

6.42 Increasing educational opportunities for girls and improving primary education in rural areas are the principal strategies to attack the causes of poverty for women. A secondary strategy is tu incre&se literacy among rural women. The benefits of this approach would ba enhanced women's health and family well-being, increased productivity of women and better human resource development. A girls' scholarship Program could be tried on a pilot basis to keep rural girls in school for a minimum of five years. The program would be based on paying a stipend to parents for the loss of their daughters' manual labor. The community would select scholarship recipients, using criteria to reach those at highest risk.9

6.43 A pilot project establishing girls' boarding schools in rural areas should be designed and implemented with the objective of providing rural girls with access to secondary education. Presenrly, the Franc.scans operate 20 boarding schools in the Department of Potosi. Some of their schools have mixed student populations, but most students are boys because cultural attitudes favor boys' education. A boarding school exclusively for girls would take parents' concerns into account while providing girls with an incentive to stay in school.

6.44 A technical school scholarship groaram including room, board and tuition, should be developed so that rural girls can attend technical schools such as CIPCA's training school in agriculture and animal husbandry in the Department of La Paz (where most students are young men) and Fe y Alegria's integrated production and education boarding school programs in Yungas, Sucre, and Chuquisaca. Such a program would provide girls with an incentive to finish secondary school, increase the number of technical personnel in rural agricultural development and redress the imbalan e between male and female educational opportunities. The program could include a component of return community service for participants.

6.45 As women make up 75% of the rural illiterate population, priority should be given to female literacy training and adult education. The Government could allocate resources to SENALEP to upgrade and expand its services in r.aral areas. SENALEP's programs have functioned well in the past. It has developed well received materials in Quechua, Aymara, and Guarani. Rural radio education, such as Fe y Alegria radio, should be expanded. All adult literacy programs should have technical/vocational education and health

9 A program of this type is working in Guatemala, and the reported cost per student per year is US$15.00. 94 education components. Literacy training for women should be practical (how to build solar greenhouses, improve nutrition, production strategies).

Phasing

6.46 The needs for reform in the education system are so many and diverse that serious consideration must be given to issues of phaoing. In the short term, institutional development efforts should concentrate on a management needs assessment of the sentral ministry functions and regional offices. (Emphasis could be placed on the three most populous geographic departments) Because decentralization appears to be a political priority for the Govern- ment, this assessment should be accompanied by management training of central and regional ministry employees. The school facilities mapping efforts already underway should be fully functional as a basis for needs assessment. Also in the short term, a program to strengthen the ruril system should be implemented, including multi-grade curriculum and teaching accompanied by higher pay, pilot testing of rural boarding schools among dispersed popula- tions and improvements in materials and facilities in rural areas financed through SIP.

6.47 In the medium-term, the focus should be on the increase of quality and equity of basic education. This could be achieved by the following:

a) Rationalizing human resources by linking teachers pay to efficiency improvements; reallocating teachers to areas of unmet needs; providing in-service training as appropriate; and strengthening the management and supervision of the teacher force.

b) Developing and implementing with the participation of local communities low-cost means of school construction, repairs, maintenance and furnishing. New construction should be minimal, undertaken only when clear alternatives do not exist.

c) Financing, publishing, and distributing textbooks and guides to students and teachers; distributing educational materials to teachers and classrooms; developing cost-recovery or low-cost means to finance the stock of books and materials.

d) Reducing the number of teachers' colleges; updating and unifying the curricula and admissions practices; providing instructional materials for the students.

e) Continuing support for educational management strengthening programs at both central and regional levels; improvement in levels of financial resources for che Ministry of Education and its decentralized agencies.

f) Combining the dual education systems. 95

CNAP&ER7s AGRICULTURE

INTRODUCTION

7.1 Agriculture plays an important role in the Bolivian economy. The incidence of extreme poverty is particularly high in rural areas where agri- culture is the predominant economic activity. In 1976, 95% of the poorest 20% in Bolivia were smallholder peasants (campesinos) and since then, their pro- duction has fallen.' The vast majority of the campesinos are of Indian origin (Quechua or Aymara). In 1988, the agriculture sector accounted for about 22% of GDP. Another 20% of GDP originated from related economic activities, including processing and distribution of agricultural products. About half the population is engaged in agricultural production. Of these, 60% live in the Highlands, and 20% each, in the Valleys and Lowlands.

7.2 Generally, the causes of low household income are low productivity of land (especially in the Highlands), small farm size, low ratio of commercial crops to total output, long distances to markets, low quantity and quality of inputs, and no access to technical assistance. A close relationship exists between rural poverty and regional inequality. In the Highlands, difficult climatic conditions, including droughts and frost, limit agricultural potential and productivity is low. Furthermore, as fertilizer use is minimal, two-thirds of the land has to be kept idle in any given year to allow soils to regenerate. Most of the small farmers produce staples and vegetables (pota- toes, maize, beans, quiroa, etc.). They have poor links to markets and credit institutions, and sell (on average) less than 30% of their output. The far- mers in the Valleys produce basic staples but are also engaged in cash crop production, including fruits, flowers and timber as well as dairy farming. Despite the low productivity levels, relatively favorable weather and soil conditions allow two-crop farming in the region. A major obstacle to improved agricultural productivity in the Valleys is poor market access, mainly as a result of lack of transport. In the Highlands and the Valleys, 61% of the farmers own less than 5 hectares and no farm is larger than 1,000 hectares.

7.3 The tropical Lowlands is the most dynamic region for agriculture. Major crops grown are sugar, rice, cotton, soybeans, and oilseeds which are sold in urban markets in the Highlands as well as locally. The average landholding of the Lowlands farmers is much larger than those of other two regions: 83.8% of the farmers own more than 5 hectares; and 37.0% own 50 hectares or more (only 3.0% of farmers belong to this category in the other two regions). However, the recent migrants in the newly settled areas (the "colonizers") can often only use 1-2 hectares and have no title to their land. They are normally able to produce sufficient food for their own sustenance but are faced with savere problems in marketing surplus. The major poverty problems in the Lowlands are found among these groups and among the indigenous tribes. The problems of these people are not adequately reflected in the statistics. Because of their

1 See Chapter 2 96

Pedro had difficulty throwing the last spadefut of stones and earth out before clfmbing from the well he had dug so deep that he could no longer see out of the top. He sceiewedup his eyes as fIe looked to the sky to see if there were anyclouds on the horixon that would bring the long awaited rain that should have moistened the soiltwo fult moonsbefore. Whenhe was youngthe rains were not so irfportatnt as novedays. Longapg he wouldtake his Ccesrallamas, strong beasts of burden with poor wool but goodmeat, to the South to cut salt and then lead them eastward over the hills and downto the warm, lush Valleys. There he wouldbarter the salt for waize, coffee and coca leaves before trekking back to his land and family. There his wifecared for the children, wovecloth, looked after the few llamas he had left behind and cultivated potatoes, coca and beans in the moist soil fed by the undergroundsprings and the surnarrain. In bad years wlhentheir crops were dvmagedby the dry coot air, withered by the intense cold of a suiirer frost or were destroyed by hail, he could always rely on the maizeand somedried Ccara meat and hope that the next year would bring a bounteousharvest. But now his frail body and failing eyesight dependedon the fewpotatoes and broad beans that he watered with the bucket he hauled out of the well. At least he only had to feed himself: his wifehad died several years ago and his children had all movedto the valleys in search of a better life. His eldest son had suggested thqt he go to tive with him in Cochabamba.He thought to himself that perhaps that was what he should do: he could not survive another year alone in the harsh,desolate land that he so lovedbut which offered so little.Then again he thought that this was his land and it was all he possessed.

Box 7.1: To Migrate or Not, That is the Question precarious legal title to their land, and their geographic isolation, they are simply not recognized in official statistics.

7.4 This chapter follows a similar outline as the two previous one. First, the institutional framework is described. Institutions in agriculture are very weak, and have a very bad track record in reaching the poor. Second, investment and policies in agriculture are analyzed. Sector policies are geared towards improving commercial agriculture and the investment portfolio also pays most attention to the needs of this subsector. Third, sectoral issues and constraints are laid out. These include problems in land use and land titling. The basic thrust of the argument is that the breakdown in communal structures, the lack of legal recognition of communal landholdings, and the slow and inefficient titling procedure are major constraints on improving the situation for the campesinos. Then the issue of food donations is discussed, concluding that it is now clear that the food (especially wheat) donations presently inhibit domestic production. There is no nationally coordinated agricultural extension taking place in Bolivia and what there is does not take into account the problems of women. A further issue for women is the lack of representation through campesino organizations. Institutional issues include the politicization of the Ministry, low salaries, lack of coordination and planning, and lack of information. Finally, the section on constraints assesses the need for gr-later beneficiary participation. The last section of the chapter contains the recommendations. These include: subsidizing of poor communities through infrastructure with less than full cost-recovery, reforming the titling system including recognition of community rights, appropkiate technology development and transfer, linking the poor to the market economy, migration, women's issues, improving national policy and planning, and encouraging small-scale operations. 97

INSTITUTIONAL FPAMEWORR

7.5 The agricultural sector is confronted with the most difficult institutional challenges of all the sectors, the foremost being its principal actor: the Ministry of Campesino and Agricultural Affairs (MACA). While MACA has undergone more restructurings than any other ministry, it remains extreme- ly inefficient and ineffective at least partially as a result of its diffuse mandate. This includes the politically loaded issues of the campesino which has served to further complicate matters, making MACA one of the most politicized ministries in Bolivia. Interest groups have taken turns enjoying the spoils of political patronage, while those intended to benefit from such a focus are forgotten. Despite these weaknesses, MACA has the responsibility for 13 decentralized agencies as well as for coordinating a significant amount of agricultural activities with the nine Regional Development Corporations (RDCs). In sum, although the greatest incidence of poverty exists in rural areas and among the campesinos, the corresponding institutions are the most ill-prepared and poorly coordinated to address the needs of the poor.

7.6 The principal institutions in the agricultural sector are constituted within four categories:

a) MACA and its "Deconcentrated" Affiliated Aaencies: MACA is the central agricultural authority in the country with the responsibility for planning and coordination. Over the years, MACA has set up a series of deconcentrated agencies for specific agricultural interventions, which have often evolved outside the direct control of MACA. Similarly, donor agencies have favored creating new public entities to execute their own aid projects. Combined with considerable institutional weaknesses internal to MACA, this has resulted in a proliferation of implementing bodies with little or no coordination among their respective programs.

Among the "deconcentrated" MACA agencies whose formal mandate it is to reach the small farmers, three stand out:

(i) the Bolivian Agricultural Bank (BAB), the only national credit institution for which the majority of loans are supposed to be directed towards small farmers. In reality it has provided subsidized credit to powerful friends of (especially) the military Governments, mainly in the Department of Santa Cruz;

(ii) the National Service for Community Development (SNDC), whose mandate is to target peasants with productive infrastructure projects. This institution is plagued by high administra- tive costs (about half of its budget) and poor project implementation; and

(iii) the Bolivian Inastitut) for Agricultural Technology (IBTA), whose research stations are at least nominally targeted to improving production for both small and large farmers. 98

b) RDCu: At the departmental level, the RDCs are beginning to play an increasingly important role in agricultural planning. Agricultural projects have become the principal component of most RDCs' investment activity.

c) NGOc and Private Enterprises% An estimated more than 80 NGOs (both local and international) have participated actively in the sector in recent years, although their impact is uneven and dispersed. Weaknesses in internal organization and lack of interagency coordination have generally precluded NGOs from having an impact on a national scale on alleviating rural poverty. However, at the local level they have been relatively successful and their potential for expansion should not be overlooked.

d) SFs The total amount of ESF investment in the sector is modest compared to MACA and RDCs. ESF did finance a number of small, but important projects at the community level by funding physical infrastructure and technical assistance projects necessary for increasing agricultural production.

e) Peasant Development Fund (FDCI: A recently created institution charged with providing credit to poor farmers through group credits with cross guarantees.

GOVERNMENTPOLICIES AND INVESTMENTS

Sector Strateav

7.7 The agriculture sector has traditionally received low priority in the national development plans. Agricultural policies lacked focus and their stated objectives were not well-defined in operational terms. This lack of strategy and direction combined with the institutional weaknesses of MACA, meant that public sector involvement in the sector was no more than the result of uncoordinated initiatives taken by foreign aid agencies and pressures from domestic interest groups. With such ad hoc policies, most important changes in Bolivian agriculture over the last thirty-five years have taken place without specific policies designed by the Government.

7.8 Most public projects have benefited middle- and large-sized farmers producing cash crops. For example, in 1982, the sector received less than 7% of total public investment and only 16% of that amount was specifically targeted to small farmers. Agricultural credit policy has historically favored relatively well-to-do commercial agricultural producers in the Department of Santa Cruz. These producers were the main beneficiaries of a substantial amount of external financing made available through agricultural 99 credit usually at subsidized terms.2 By contrast, improving agricultural productivity of the campesinos has received low priority.

7.9 At present, Government policies are not directly geared towards poverty alleviation. The strategy encourages a market-oriented economy, promoting dynamism in the private sector, and boosting the production of new agricul- tural exports as the engine of the country's future growth. The three principal objectives of the agriculture sector are to: (a) expand agricul- tural exports; (b) increase food production for domestic consumption, in order to decrease reliance on foreign food aid; and (c) increase production of high value-added products, including fruits, flowers, rubber, and Brazil nuts. In addition, the strategy proposes continued efforts to support programs to substitute other crops for coca leaf production through creating marketing channels, providing extension services and infrastructure improvements to increase the productivity of alternative crops.

Sector Investment

7.10 The 1988 public investment in agriculture was US$52.2 million (US$21.5 million was initially programmed3), accounting for 15% of total public investment. Investment in 1989 was much smaller (US$34.7 or 8.4% of the total).4 The sector's share of programmed expenditures fluctuates between 8% and 12% of total between 1990 and 1992. A modest increase in MACA's contribution to total investment is expected in the coming years, from less than 30% to about 37%. However, RDCs will remain the largest provider of agricultural investment in terms of both number and cost of projects. Meanwhile, dependence of public investment in agriculture on external funds will rise to 74% during 1990-92 compared to 58% in 1988.

7.11 A dozen or so large projects absorbed the lion's share of investment funds in 1988. The 13 largest projects, out of a total of 136, accounted for 6S% of total public investment in agriculture. While resource concentration on large projects is likely to continue in the future, the Government has stated that it will redirect the focus of these large projects toward poverty alleviation. Large-scale projects in the revised plan include ongoing integrated rural development projects and provision of irrigation facilities in poor areas, as well as agricultural extension, technical assistance to the colonizers and technology transfer in animal and vegetable sanitation. In fact, however, the Government seems to be reverting back to the large projects

2 During the expansion of cotton production in the seventies, cotton producers failed to repay a considerable portion of the loans they had contracted with commercial banks. The Government forced the Agriculture Bank to buy these non-performing assets from commercial banks, representing an estimated US$40 million of income transfer to cotton producers.

3 Most of the RDC investment was not included in the initial program.

4 However, part of the decline is due to under-reporting of the proposed projects by the Santa Cruz RDC (CORDECRUZ). 100 of the seventies which will only benefit the already powerful. One such "white elephant" is the Villamontes project in Tarija, a billion dollar program to clear forests, establish agro-industrial enterprises, and finance large irrigation schemes. This project is promoted by very powerful interest groups and, as a result, the Government has not been able to pursue its stated goals of helping the poor.

7.12 The large-scale integrated rural development projects are usually complex and difficult to manage. A typical example of a project with defi- cient design is the ongoing Chuquisaca Norte Project (started in 1983, with a total cost of US$21 million). Although this project is aimed at the poorest parts of Chuquisaca, it has had limited impact on the targeted population due to the complicated project administration (consisting of 10 components) and the lack of interagency coordination (involving 11 government agencies). Based on this experience, a recently started Chuquisaca Sur project tries to rectify the situation by reducing the project components to four and setting up a unit specifically charged to coordinate interagency activities as well as to encourage wider participation of farmere in the targeted area. The problem remains, that farmers are only involved in implementation and not in design.

7.13 Apart from these large projects, most projects are small in scale, typically costing less than US$1 million. More than two-thirds of them purport to promote either production of traditional crops (e.g., maize, potatoes, cereals, etc.) or high value-added production for domestic market (e.g., fruits, flowers, milk, fishery, timber, etc.). While the basic concept of these small projects appears supportive of poor farmers, the effectiveness is highly questionable. For example, out of 15 small rural development projects implemented by CORDECRUZ in 1988, only three have been judged successful (by CORDECRUZ) in improving productivity of poor farmers in the targeted areas.

7.14 The level of investment by departments is ad hoc year by year, and there has been little consideration in allocating resources according to the poverty level or the size of rural population of each department. For example, the received a disproportionate share of total public investment in agriculture in 1988 (34%), but the share of its planned investment shows a drastic decline to 19% in 1989 and less than 13% during 1990-92. The per capita allocation6 of the has also fluc- tuated greatly between the smallest (1988) and one of the largest (1989 and 1990-92). Furthermore, the amount of concern given to poverty alleviation in the planned projects varies across RDCs. The problem of misplaced emphasis is particularly serious in Pando.

The accuracy of information on RDCs' projects as reported by the MINPLAN is highly questionable. CORDECRUZ has consistently failed to report many of its planned investment to the MINPLAN (see above).

6 Per capita allocation here is calculated on the basis of rural population only. 101

7.15 IEE has funded physical infrastructure and technical assistance projects necessary for increasing agricultural production. As of mid-1989, ESF had disbursed US$9.6 million for agricultural projects,7 covering 61 irrigation projects (US$2.9 million), productive infrastructure, erosion control and forestation (US$4.4million) and rural credits (US$1.3 million). Geographically, ESF-funded agricultural projects have been concentrated in a few departments. The Cochabamba Department alone received more than a third of ESF's investment in 1988, while 1eni and Pando did not receive any investment. This illustrates the dilemma of ESP's demand-driven mechanism. After ESF closes its activities, the Regional Development Fund (FDR) was supposed to take up the task of funding small agricultural projects. The project bank of ESF will be transferred to FDR at the end of 1990. Unfortunately, FDR has not assured any financing for agricultural projects and has no poverty targeting, so the future of "ESF-type" projects is very uncertain.

SECTOR CONSTRAINTS AND ISSUES

Land Use

7.16 Many of the traditions of the Inca period still prevail aaxong rural populations, particularly in the assignment and use of communal land. In the pre-Colombian era, land was communally held and the communities (ayllus) provided part of their produce and also labor for the maintenanc, of the Inca empire. The laws of the Inca society were such that people were born into an ayllu and were forbidden to migrate to other areas, unless they were chosen by the Inca officials for some specific task or job. There was no cash economy as such. While part of their produce was turned over to the Inca regime, Inca officials also stored food and this was made available to the ayllu8 when crop failure occurred.

7.17 Today in many parts of the Highlands, the land continues to belong to the ayllu, as opposed to the individual. It is assigned to family units according to their ability to work the land. Cultivated land use is carefully controlled in these communities. Part of the area is left in fallow for grazing and recuperation of fertility and the rest is cultivated in blocks. Each block is planted to only one crop with each farmer having an area which he/she cultivates and harvests. In this manner the rotations are carefully c:ontrolled. Cooperation is often strong with the formation of work groups which work together on such tasks as planting and harvesting. Payment is in the form of returning services rather than cash payment for the work carried out. The organization of the community system is based on community service rather than paid officials. Within the social structure of the community, service is important in determining a persons status: to serve the community means more than to be financially successful. The community leadership is rotated between families of the community with the family of turn electing the most capable member of the family. If a family rejects the responsibility of

7 The average cost of ESF-funded agricultural projects was about US$50,000. 102 being the leader it forfeits the right to land and membership of the ayllu. The land that is forfeited cannot be sold, it is returned to the community. Hence a person who leaves his community loses his rights to what may be his only capital. This service requirement is a crucial factor in tying people to their original community.

7.18 Communally grazed areas were carefully controlled in the Inca period,8 but control has disappeared today. Areas in the Highlands more distant from Lake Titicaca and with less rainfall were used as extensive communal pastures for cameloids such as llamas, vicunas, and alpacas. Readily domesticated, llamas and alpacas continue to provide transport, meat and fibers today. Vicunas, on the other hand, are difficult to breed in captivity and run wild throughout the Highlands. In the times of the Inca, hunting of the vicuna was carefully controlled so as to maintain herd populations. This system has since broken down and indiscriminate hunting is reducing the vicuna popula- tion, particularly south of lake Titicaca. After the Spanish conquest sheep and cattle were introduced into the area. Today, control of grazing on communal land has all but broken down, leading to overgrazing and degradation of pastures.

7.19 With the arrival of the Spaniards, two major structural changes occurred in the Highlands. First, large tracts of communal land where given to the "conquistadores." Second, the opening up of silver mines, particularly in Potosi, led to a large increase in population in inhospitable areas little suited to agriculture. Miners had to be fed, however, and agriculture was developed in a hostile climate, severely limited by frosts, hail, and lack of water. Production in these areas was never sufficient to feed the population that worked in the mines, and food importation, particularly of wheat, became a feature of the food balance of the region which exists to this day. These forced migrations also created pockets of Quechua Indiana in otherwise Aymara areas, which eventually led to the breakdown of social structures.

7.20 Although the communal land system still exists in many areas, it is breaking down in others. The breakdown of the communal systems of providing families with as much land as they can work and control of crop rotations is leading to problems of fragmentation of land, shortening of fallow periods and poor maintenance of such communal structures as the terraced areas around Lak3 Titicaca. In many areas farmers are now cultivating their own lots on a permanent basis. The individual land plots tend to be smaller in the area around Lake Titicaca where production potential is greater, often being less than one hectare. In the drier and more inhospitable areas, land holdings are often in the range of two to five hectares, however much of this land cannot be cultivated on a regular basis and is left in pasture or scrub.

Land Titlina

7.21 The inappropriate nature of laws relating to land titles and lack of consistent government policy, have a damaging effect on the poorest

It has been suggested that grazing land without permission was an offence punishable by death during the rule of the Incas. 103 populations. Until the 1952 revolution, the majority of the peasants lived under feudal systems with no right to vote. In 1953, land reform laws proposed that "the land belongs to those who cultivate it". To date, it is estimated that the land reform has provided S50,000 campesino families with access to 4 million ha of land. However, in many cases, these families have no official titles to this land and are not able to legally buy, sell or use it as collateral for credit. Furthermore, the land reform law was specifically aimed at individual landholdings, and did not resolve the problems of communally-held land. This contrasts with Peru, which has a cultural heritage similar to Bolivia, yet has a specific general law for communities. This law recognizes the campealno communities as "democratic institutions, autonomous in their organization, with communal labor and use of the land, as well as in their economy and administration." In Bolivia, however, campesino communities and indigenous tribes in the Lowlands are significantly hampered by inappropriate laws that were designed for other purposes.

7.22 In the Hiqhlands, land is typically handed ,own within the family and title x supported by wills or testaments. These documents, however, have little legal meaning as they are not accompanied by land titles. Neverthe- less, land continues to be bought and sold on this tenuous basis. To further complicate the system of property rights, small landholdings are handed down from generation to generation and are often subdivided into lots. Each membeir of the family gets the rights to a number of rows, often in different patches of the field. Although there has been discussion on laws to regulate the use and title of communally-held lands, no laws have yet been promulgated. The communities that d. facto have access to land do not have persona juridicae or title to the lands. In many cases, this limits the potential for both on-farm and more general improvements in rural infrastructure.

7.23 In the Lowlands, very few individuals in recently settled areas have title to the land they are using. However, efforts are now being made to obtain titles; normally on a communal basis and sometimes with direct support from NGOs. When given a plot at a colonization scheme the farmer receives a certificate of settlement, which is issued by the National Colonization Institute (INC). This is not a document of legal ownership. It is not certified by the National Council of the Agrarian Reform and thus gives no security against claims to the land presented by others. In areas which increase in value because of agricultural potential or strategic location (e.g., where a new road is planned), several claimants often appear for the same plot of land.

7.24 Lack of well-defined territorial rights is also the most serious problem facing the lowland indicenoug oeoole. Although they have occupied land and forests for centuries, they normally hold no land titles. Their exclusive rights to use these resources have never been acknowledged and in many cases they have been forced to abandon the best agricultural lands. With the expansion of economic activities in the Lowlands, the indigenous groups are increasingly being pressured by logging companies, cattle raisers and large- and medium-size commercial farmers as well as new smallholder-settlers. The agrarian reform had little effect in protecting the land rights of these indigenous people because, for them, land is a generally a communal asset and 104 not individual property. Some groups have made claims for land on a communal basis with little success. One such request was ready for approval before the Government took office in August 1989, but no action has been taken since then. In the southern Santa Cruz Department area of Chaco the Indians of Guarani origin are being pushed away from their land by large cattle ranchers (see Box 7.2). This is a problem both in terms of equity and efficiency as the yield under crop production as practiced by the Guarani is much higher than the yield of cattle ranching on the same lands. If the indigenous groups are to maintain their traditional way of life, combining hunting, fishing and gathering with small scale agriculture, they need to control much larger areas than are normally allotted to smallholder settlers.

The GuaraniIndians in the lowlandarea of the provinceof Cordilters in the Departmentof Santa Cruzare facedwith severeproblems of drought(mean anrwal rainfalt is only450-600 mm), land disputesand racialdiscrimination. Only 9X of the landbelongs to the Guaranicommuities and of the cultivableland, they own onlyarourd 7T. Thereare about 5,500Izoceno-Guarani in the loe altftudepart of Cordillera.They are forcedto tivenear the riverParapet(, the only source of watersupply. Forcenturies they have dug canals fromthe riverto theirfarms but in recentyears serious problems of landdispute have arisen because the cattleranchers have obtained title to most of the land along the r-ver which provides water for their herds. Slowly the Guaranis have been forced off their land.

The Guaranisused to be self-sufficientfarmers, growing a widevariety of cropssuch as maize, beans,yuca, sugar, cotton, melons, fruits and grouridnuts. Todaythey are almost cospletely dependenton largescale agriculturaL enterprises for employment as theyare no Longerabte to surviveon theirsmall plots far away from the river. Instead,the Guaranis are forcedto work as sharecropperson the largefarms, as seasonallaborers during the sugar harvest or as day laborers on thecattle ranches. Working conditions are extremely poor: they are not coveredby any kindof laborlegislation, receive no medicalservices and livein insufficienthousing. They become wore and more indebtedto theenterprise each year as theyhave to purchasetheir food in the company store,use companytrucks for transportation,etc.

Box 7.2t The Guaranis: Working for the Company Store

7.25 Recent estimates indicate that less than 100,000 ha (0.3% of area distributed by the land reform) of the land owned by the agricultural compa- nies is cropped, whereas the campesinos cultivate over one million ha (about 25% of the land distributed by the land reform). The 1953 law does recognize the danger of land not being used and held for speculation. It stipulates that land that is not used for two consecutive years should revert to the state, however, the onus of determining whether the land is used or not is on the public who denounce to the state those owners who do not use the land. This system has not been effective. Hence the land reform has only been marginally effective in giving the land to those who cultivate it.

7.26 Apart frorathe inappropriate nature of the laws covering land tenure and rights the campesino has in general a mistrust bordering on feat of the official legal sector. This has arisen as a result of centuries of exploita- tion by the ruling classes and exists to this day in spite of the radical changes in the structure of society wrought by the 1952 revolution. This point is illustrated by the reaction of the campesinos to the land tax. Apart from being applied only to the larger land holdings and even then at very low 105

levels (the actual rate depends on the area), the campeainos are fearful as to how it will be applied to their communal lands and many fear that their land will be taken away from them.

Food Donations

7.27 Food donations are often blamed for depressing local production of food. Table 7.1 provides summary data on food donations programs in Bolivia. The USA is the largest donor of foods to Bolivia, regulated by Titles 1, II, and III of Public Law 480 (P.L. 480). These programs provide funds for Bolivia to purchase US agricultural commodities. Under Title I, the US provides credit at concessional terms for purchasing wheat, which is then resold to domestic millers at international prices. Title II provides in-kind food contribu- tions, which are used as payment for work on local development projects, commonly referred to as food for work programs. The Title III program also provides a grant for food purchases from the USA, conditioned on the use of the counterpart funds for agricultural development activities.

7.28 The European Community and the World Food Program are the other significant donors of food to Bolivia. These donors provide in-kind food contributions, which are used in a variety of project targeted at poorer communities, including food for work activities, emergency assistance for areas suffering food deficits, and programs designed to increase health, education, and nutrition levels among the poor. Bolivia received a significant amount of food aid from other bilateral donors during the 1983 drought. Since then the other bilateral programs have either disappeared or declined significantly. Presently, France, Canada and Spain provide small amounts of donations.

7.29 There is no question that donations are important in supporting consump- tion, particularly, although not exclusi- vely, among the poor. Approximately 21% of Bolivia's population US TftteI Aeat Ctdlt providedat receives food dona- tions, either free or .:UStitteil fIOUF,grar4,0tt t>d*< w in payment for work on rice,bw*s, powderditk community development projects. Donations USt~~~~~~~fltt were-t criticaliefit in f drfremdte - staving off widespread hunger during the 1983 XEuopean Cai. fetotr dud ; tktlS doti drought, and since silk,orl,"s -IS f then have allowed some

oridFood Program ftowt, it, beasw' ' nan d4 poor communities to enred flab, powderd ailk achieve much higher levels of consumption. 'RedCross powderedmsik in-kind dwition ______7.30 Donations also Table 7.1: Food Donation Programs provide financial sup- port to the Government 106 and the economy. About USS20 million per year goes directly to the Government to finance administrative expenses. In addition, the Government earns money from the sale of donated foods which is then used to finance development programs. A portion of these funds does not represent additional resources, as for example the funds used to cover US embassy expenses which would in any event be provided. A large proportion, however, is used for general develop- ment programs over which the Bolivian Government has some control.

7.31 An intense debate over food donations has raged in Bolivia since the sharp rise in donations during the 1983 drought. Despite the support to consumption and to government revenues, it is argued that donations have a number of undesirable effects on the Bolivian economy and society, largely through their impact on incentives. The greater supply of food lowers food prices and diaccurages domestic production, induces shifts in consumption patterns to imported rather than domestic foods and inc._aasesdependency. Further, the provision of free food reduces the 3icentive to work and breaks down the tradition of collective self-help efforts in rural communities. While analysis of price and quantity data can do little to help resolve the latter argument, the available data do shed some light on the former.

i2 1980 SK I92 (thousands of metric tons) U.S. Titil I 71.2 77.4 -- - -- Titte 11 4.4 18.7 25.4 25.2 30.4 42.9 42.6 139.0 Tftte Ifi -- 74.1 173.9 64.5 -- 163.0 180.9 124.5 World Food PrIorm 1.2 7.8 4.1 11.1 1.6 7.1 14.7 16.4 European Cnmmtyt* 7.4 8.6 10.3 .. 22.6

Others 10.5 39.3 1.5 7.0 25.0 10.0

TotaL 5.6 100.6 213.9 218.7 119.5 230.3 263.3 312.5

Source, "Apuntes Sobre Las DonacionesAlimentaries y Su impecto en Bolivia" Jutio PrudmncioDohrt, 1989

Tablo 7.2: Food Donations Volumes

7.32 Table 7.2 shows the large rise in food donations over the past two decades. From a mere 5600 metric tons in 1970, donations rose to over 100 thousand in 1980, and then doubled by 1983 to meet the demand growing out of the severe drought of that year. Despite the recovery in agricultural production since 1983-84, donations have risen further, and are now more than 33% higher than during the drought. Data on the relationship between dona- tions of wheat (the largest single category of donations) and domestic production and consumption are shown in Table 7.3. Clearly, donations are extremely important sources of wheat, making up 37% of total consumption and equalling 2.5 times domestic production in 1988. The large size of wheat donations relative to domestic production, combined with the decline in 107 domestic wheat production, are often cited as evidence for the negative impact on production incentives discussed above.

WheatDonations and the Doaestic Market (thousands of metric tons)

1975 1980 1983 1984 1985 1986 1987 1988

TotalStpply 147.2 403.2 371.6 352.3 44T.? 339.2 451.9 362.0

Production 61.8 60.1 40.3 68.5 67.7 69.8 56.6 54.9 Comwrcial Imports 65.3 255.3 92.0 194.6 256.4 92.? 191.8 171.8 Donatfons 0.1 24.8 239.3 78.9 123.6 176.? 203.5 135.3 Change in Stocks -20.0 -63.1 0.0 -10.4 0.0 0.0 0.0 0.0

Neuor.ndumItems tonations/Prod. 0.0 41.1 593.1 115.3 182.5 253.0 360.8 246.6 bonations/Cons. 0.0 6.1 64.4 22.3 27.6 52.1 45.0 37.4 Conm.Imp./Cons. 60.9 92.1 24.8 58.7 57.3 27.3 42.4 47.5

Source: Tabte7.2

Table 7.3: Wheat Donations and the Domastic Market

7.33 Given the great importance of wheat donations, it is likely that they have reduced domestic production incentives to some extent. In the case of wheat sold domestically, a large proportion of this is sold directly to millers, formerly at subsidized prices. This policy was changed in 1987 to sell wheat at its opportunity cost, and partly as a result acreage planted in wheat increased significantly in 1989. Although the program now is designed so that the millers pay a competitive price, the assured delivery of donations and centralized distribution system make it difficult for domestic producers to compete effectively. Further, a significant proportion of wheat is provided as in-kind payments for work on development projects (US Title II and the World Food Program). The availability of food as payment to under- employed workers may have reduced the demand for domestic food production, particularly as the provision of transportation in areas with poor access to markets probably biased consumption towards donations.

7.34 Still, the evidence that wheat donations had a significant impact on domestic production is not conclusive. Commercial imports, presumably at international prices, exceeded donations in most years of the eighties. Despite the availability of substantial donations, commercial imports were necessary to cover 42% of consumption in 1987 and 47% in 1988. Thus, while domestic prices (and thus production incentives) were no doubt lower than they would have been in the absence of donations, it is not clear whether Bolivia is a partizilarly competitive producer of wheat. In any event, the often-made claim that donations have been largely responsible for the poor performance of Bolivian agriculture is not credible, at least in the case of wheat. Wheat production in Bolivia from 1986-88 averaged 60 thousand metric tons, or about the same as in 1975, when donations were at trivial levels. 108

7.35 It is true, that there is substantial potential for increased wheat production in Bolivia, particularly in the Lowlands in conjunction with soybean production. However, it is not necessary to cut donations to realize this potential, at least in the short term. Improvements in infrastructure and an effective extension service would permit a sig.ificant expansion in domestic production, allowing the country to reduce its dependence on commer- cial imports. As production increases and imports decline, a conflict between producers incentives and donations may arise. It therefore makes eense for the Government to continue to monitor food donations and to reduce them as domastic production takes a larger share of consumption. However, it is not necessary to undertake a sharp curtailment in donations, which would be extremely painful to recipients and deny the Government access to significant financial resources.

Lack of Research and Extension

7.36 Although IBTA has the official responsibility for research and extension (except in the Department of Santa Cruz), in reality it does not do any extension at all. Instead extension is done by a multitude of NGOs (e.g., CEDEAGRO and CIPCA in Cochabamba, ACLO in Chuquisaca). IBTA has a Btaff of about 250, only a few of which are qualified researchers and extensionists. The average salary is US$50 a month and thus staff turnover has been between 20 and 30% in each of the past years. IBTA has had some success in research in the past, even though research efforts have been diluted by a lack of focus (i.e., research has been "focused' on more than 25 crops). A restructuring of IBTA is necessary and a program is presently being prepared by FAO and IDA.

7.37 In the Lowlands, official colonization policies have by and largo not been accompanied by any delivery of services to facilitate the agricultural production of the settlers. Of the RDCs only CORDEBENI has been involved in extension to the poor. A number of NGOo are involved in agricultural development in the Lowlands, mainly in the Santa Cruz area, where the oldest colonozation schemes are located. Although the NGO0 generally work with the poorer sections of the population, they form an extremely heterogenous group because of considerable differences in resources, origin and ideologies. Consequently, their programs differ widely in approach (in the same general area) and the problems they intend to solve. In spite of the several (semi) official and NG0 agricultural extension services to small farmers in the Lowlands, the general picture is one of sporadic and insufficient activities, uncoordinated and lacking continuity.

Women's Issues

7.38 Government policies have not addressed the issue of the integration of women in rural development. Development schemes tend to focus on men. The introduction to men of new technology (fertilizers, pesticides, artificial insemination of cattle, new varieties of crops and pasturage) has ignored the traditional and considerable role of women in agricultural production (especially in animal husbandry). Traditional women's activities such as the cultivation of certain crops and production of crafts (artesanla) have not benefitted from development because public and private agencies do not view these as "marketable." The substantial temporary and permanent migration of 109 men are creating special problems for women. The only grolapsrepresenting campesinos are the unions. Only widowed women can become members of the union. This means that even though women will be taking care of the family farm while the man is away her interests and needs as a producer are nct recognized.

MACA Institutional Issues

7.39 The magnitude of the problem of rural poverty is so great that it wili require massive human, financial and other resources to achieve a measurable impact. Somewhat paradoxically it is the small proiects with active partici- gation of the poor communities and a consensus approach to decision-making that a22ear to be most effective in alleviating Poverty. Successful develop- ment programs should thus imply support to a large number of small, effective projects. Such efforts must be closely coordinated to avoid duplication of effort, or even worse conflicting objectives of different projects. The obvious agency to set overall policies, plan and coordina-q activities is MACA. However, it is so weak that its own internal constr-aintsmust be overcome. The principal constraints to the effective func-ioning of MACA can be broken down into four broad categories: (a) politicizat:on; (b) weak human resources; (c) poor central/regional coordination; and (d) absence of information.

7.40 Foliticization. Over the years, MACA has increasingly become a political tool, focussed on political patronage nominally related to concerns for rural or campesino affairs. Because of its orientation, MACA has been at the center of a political struggle between campesino unions, polit3cal parties, and military governments. Due to the enormity and political sen- sitivity of these issues (e.g., land titling and distribution), technical agricultural questions have often lost priority, and there continues to be no resolution to those issues most affecting the poor. With every shift in the balance of power, most personnel also shift, along with policy priorities. While these appointees can be effective in carrying out the wishes of central authorities, they often have little credibility with the farm sector and rural communities in the regions. To address these constraints, MACA is attempting to establish regional development councils to include representatives from both the public and private sectors at all levels of agricultural production. Because only a few of the councils are actually functioning at present, it is still too early to draw conclusions as to their effectiveness.

7.41 Another result of politicization is that national leaders and donors wishing to carry out viable agricultural programs generally have done so through parallel institutions outside the real control of MACA. These affiliated, or odeconcentrated" institutions place special emphasis on specific kinds of services. Most have structures similar to those of MACA, with head offices in La Paz and centralized, vertical systems of management and administration. The directors of these institutions are appointed by the minister, yet they obtain their budgets directly from the Treasury or outside sources. Many operate autonomously and do not hold themselves administratively responsible to MACA. Donors often deal directly with them for funding projects in the agricultural sector and basically ignore central MACA. The affiliated institutions have achieved their independence and 110 operational capability as a response to the weakening of MACA. Some, such as IBTA and BAB have larger budgets and number of personnel than the central ministry. The result has been lack of coordination and conflicting sector priorities.

7.42 Weak Human Resources. Virtually all of the offices in MACA are overstaffed with under-qualified personnel, have limited operating funds, and are generally out of touch with the sector at large. Despite the efforts at restructuring, not only do there remain too many employees in the sector, but they are also the wrong kinds, in the wrong places and with the wrong types of background for their particular functions. MACA is virtually unable to conduct planning and policy analysis for the sector. In addition to low salaries, the approximate 50/50 split between administrative support personnel and the technical/managerial staff results in an inordinate amount of salary monies going to an inflated administrative staff. In total, the amount left for travel, office supplies, and other expenses is so limited that the central MACA becomes practically inoperative.

7.43 Although MACA is nominally responsible for national agricultural planning and policy-making, its scope and influence has been rather limited. As a practical matter, the major role of agricultural sector planning in Bolivia is one of "programming" the sector. Broader questions of strategy and the applicability of policy instruments, other than specific projects, tend to receive relatively little attention. For example, even when the state has intervened in pricing of agricultural commodities, the responsibility has rested with the Ministry of Industry, Commerce, and Tourism, not with MACA. At the working level within MACA, the Agricultural Sector Planning Office is charged with the responsibility for sector planning and policy analysis. The office's analytical capability and stature are generally low. It has contributed virtually nothing to sector development or strategy and only marginally to development, monitoring, and evaluation of sector programs and projects.

7.44 Poor Central/Regional Coordination. RDC coordination of the planning of the range of agricultural institutions in their departments is difficult within the existing structure. MACA has regional offices in each department which report directly to its headquarters in La Paz. In addition, seven MACA- affiliated agencies have independent planning units over which the RDC has no formal control. Thus, while the RDCs are responsible for coordinating departmental planning, they have little authority to do so. In practice, coordination is a function of the goodwill of parties involved. As a rule there is an inverse relation between the degree of RDC dependence on La Paz for technical approval or funding and the ease with which a project passes from design to execution.

7.45 In sum, although RDCs have become decentralized planning entities, assuming the former role of MACA in both rural and agricultural development, the result has been a plethora of independent regional activities with no clear sense of regional or national direction. Because of this lack of mandate and coordination, those with the weakest capacity to voice their concerns--the poorest--are the first to be overlooked. ill

7.46 Weak Information. Another constraint is the absence of information about the sector, particularly the poorer segments. Basic sectoral statistics on areas planted and harvested, production, and prices are not only weak, but generally compiled too late to inform policy decisions. The source of most sector information is at the regional level, in farmer associations and RDCs. While effective planners and analysts exist in many of the wealthier regions, the poorer regions do not have the same capacity. This only serves to exacerbate the already serious problem of regional disparities. Furthermore, MACA is underdeveloped in manual and mechanized data systems. Personnel records are often incomplete, inaccurate, and not computerized. Financial balances are done manually, the most recent available figures being at least two years old. There is no viable information gathering system that can assist in management decision-making, or even which can provide an accurate description of what types of projects are being undertaken in the sector as a whole. This will constrain the restructuring of MACA significantly by forestalling readjustment of human resources and perpetuating inefficient and uninformed sector service delivery.

Lack of Beneficiary Participation

7.47 Programming of the agricultural sector has traditionally been a top- down, La Paz-based exercise. Efforts to strengthen national policy and planning within MACA have not led to positive action because these have been largely divorced from the farm community. With the government's professed shift to heavier reliance on free-market signals and private sector initia- tive, however, this mindset must change towards greater client orientation. The clientele in question--Bolivian farmers--must be furnished opportunities to have their concerns heard, to have a say in setting funding priorities, and to contribute as partners in the overall process of agricultural policy formulation. Policy analysis and planning require experienced staff who cannot only evaluate the potential impact of different policy alternatives on the macro economy, but who can also help leaders in the farm and rural sectors identify common goals and define the means to take action.

7.48 Rural people pay few direct taxes, and have little say as to the magnitude and nature of the services provided. Opportunities to support agricultural production and marketing systems are continually identified in rural areas, but are not acted upon. Participation from the rural and agricultural sector has two weaknesses at the national level. First, the farm community is not organized in a uniform, formal manner across the country. There is no official form of representation or delegation of authority with defined limits. Second, participation implies considerable travel costs and investment of time. This has been a severe constraint to organizing producer associations at the national level.

7.49 In each region, however, the representation of the farm community has been accommodated in a few committees and boards where the public and private sectors cooperate. The types of organizations that represent farmers vary significantly among regions, ranging from cooperatives to chambers to peasants' unions. As farmers become more organized, these varied forms of representation are becoming more proactive. This is, however, a gradual process, which is still in its early stages of development. The creation of 112 the Peasant Agricultural Development Corporation (CORACA) is an indication of changing attitudes among campesinoa. Yet, despite the apparent effort to create a depoliticized, technical campOeiio organization, the increasingly independent political postures of peasant unions and parent organizations makes the future of CORACA and other such organizations uncertain. Without mobilization and generation of their own resources, any institution is unlikely to succeed as an independent and autonomous entity. Some producer associations are run by small farmers with the assistance of NGOo. These often coexist with unions and other communal institutions, but are maintained strictly as technical and economic (as opposed to political) institutions. Farmer cooperatives have been promoted over many years by NGOs and interna- tional donors, although severely limited scope and coverage have generally been the result.

7.50 Lately there have been some examples of involvement of campesino unions and other grass-roots organizations in the design of rural programs. One is the development plans for the province of in Cochabamba Department, the other a development scheme for the central part of Chuquisaca. The Chuquisaca plan is interesting because an international agency (the Swiss Development corporation), local and international NGOs and grass-roots organizations were all involved in the design of the program. In the case of Mizque the initial needs assessment was done by the unions, and then turned into a development plan for the Province by an NGO (PROANDES). In Mizque, coordination at the local level has generated large amounts of financing, or at least promises of financing. This indicates that financing agencies are willing and able to support locally developed plans. These experiences and other similar ones should be evaluated and guidelines established to allow for a true bottom-up planning process.

RECOMMENDATIONS

Subsidize Poor Communities Throuah Infrastructure

7.51 Government subsidies should be concentrated in the Poorest rural communities. Subsidies should not be in the form of handouts, or credits, but rather in the orovision of infrastructure, extension services, and technoloav transfers without reauirements of full cost recovery. Experience has showed that the poorest communities often provide the highest degree of counterpart financing in the form of labor. Thus provision of infrastructure does not imply a free gift. Providing subsidies to these groups should not mean disregarding cost-benefit and cost-efficiency concerns. On the contrary, such tools are necessary in choosing among the many possible interventions in poor rural areas.

7.52 Such a oglicv of subsidized orovision of infrastructure and oroduction related services should be based on the exoressed needs of the communities. Recent experiments with micro-regional planning based on proposals from campesino unions are showing great promise. It is important that such subsidies be transparent. In the post, richer farmers have been able to receive large subsidies, mainly because they have been hidden, or even labeled as poverty alleviation. Assuming that 30% of investment costs would be 113 covered by the communities in the form of labor, a subsidy as large as that given to the few cotton producers in Santa Cruz in the seventies (US$40 million) could provide poor communities with almost 800 irrigation systems increasing the total area under cultivation by almost 400,000 hectares, an expansion of more than 30% in total cultivated area.9 All other agricultural investment should only be undertaken under full cost recovery or in the private sector.

7.53 This would imply a fundamental redefinition of the role of the state in aariculture. The state should refocus on its efforts in areas where economic analysis shows the existence of public goods or externalities. Such areas include agricultural research and extension for the poor,10 efficient and secure titling systems, and land use mapping and planning in the Lowlands.

Land Reform and Community Laws

The tottomingdocunents sre aSmeof the onesrequired to obtaina titlefor a recenttysettled fanrerIn the lowlands:

- A copyof the idontitycard. (manypeople do not even have a birth certificate becauseof the costsinvolved)

- 'SettemenntCertificate* issued by NationalInstitute of Cotonization(INC) (inthe caseof unorganizedsettlements, the process abortshere)

- A statisticalfoam with personatdata anddata an the parcetof lard

* Certificatefrom theNationat Council of the Agrarian Reform (NCAR),certifying that nobody else baa a titleto the tand. (TheNCAR is severalyears behind in its processingof suchforms)

- A reportfro INC fottowinga directinspection on the spot. (The INChas no vehiclesand a veryamell travet budget)

A gewnralmap of theccoammnty made by en authorizedsurveyor. Oftenthe process aborts because of sirple bureaucratic errors such asa misspelling of names of pe*Pleor places or becnuse atl membersof the coammity have to sign and cooplete someforts. if rw person is teaoyrarily away, thatstalls the wholeprocess. The tftleprocess goesthrough many publicoffices nd all tftleshave to be signedby thePresident of the Republic.

Box 7.3t Titling: A Bureaucratic Nightmare

7.54 The problems of land titling and the inappropriate nature of laws related to communal land holdings and the structure of traditional communities makes it difficult to consolidate and build upon existing social structure. In the past the traditional structures were obviously successful in maintain- ing an extremely organized society. Even today many of the organizational

9 If ESF's unit costs were applied.

to For other products joint efforts with growers' associations and universities should be pursued. 114 features still exist on an informal basis, but do not provide a sound legal basis for rational development efforts. Rationalization of the laws related to these areas and more effective land titling are prerequisites for rural development. A methodologv for land titlino based on a comorehensive study of the Dresent situation needs to be developed and used as the basis for a national effort to rationalize the laws and the land title situation in Bolivia. The Government of Bolivia is not at present in a position to carry out such a project without external support, which appears well-justified.

7.55 Only few of the small peasants in the Lowlands have a legal claim to their land. Land disputes are common, and there is a general sentiment of insecurity, which is not conducive to long term investments. A solution to this problem has to be found, as the insecurity makes it difficult to en- courage more farsighted behavior by the peasantry. The existing, extremely slow procedure is not exclusively due to excessive bureaucratic requirements, but is also related to the complexity of the problem and the weak administra- tive system. The right to issue land titles. both preliminarv and final. should be unified in a single entity. Although this action is probably not very expensive in money terms, it may, nevertheless, be very difficult to implement due to political constraints.

7.56 The most important claim made by indigenous organizations is the right to communal land for agriculture, hunting, fishing, and other economic acti- vities such as timber extraction and exploitation of minerals. If these groups are marginalized on too small territories, they will have to abandon their traditional form of life. In spite of the problems with Bolivia's first experience with 'debt-for-nature' deals other debt-for-nature swaps could be implemented both for relieving the heavy pressure of foreign debt, for conserving the environment, and for securing the indigenous peoples' terri- tories. Swift stels should be taken to consider-the indigenous groups' claims for land. These cases are probably best treated on a case by case basis. The first step should be to ratify the agreement that was ready to be implemented in August 1989. Since the change in Government no progress nas been made.

Aporooriate Technoloav Development and Transfer

7.57 IBTA must be reorganized"1 to be an effective research institution. Such a reorganization should include:

a) A refocusing of research activities to fewer crops, taking into account the specific needs of the campesino;

b) improvements in salaries and other incentives, financed in part by reassignment of staff, and in part by donor financing, in accordance with the rules to be set down by the Public Sector Management unit under MINPLAN;

IDA is presently studying the possibility of financing such a restructuring based on recommendations from FAO. 115

c) IBTA should not do extension by itself, but rather contract out extension to NGOs, RDCs, producers' associations or whoever io interested and has the capacity.

7.58 For poor farmers some promising integrated schemes have been implemented where, for example, poor communities higher up in the mountains (where there is a comparative advantage in seed production) produce seed for the poor communities in the Valleys, who produce the potatoes.12 These experiences, however, remain isolated and are not followed up by applied research. For the farmer with some marketable surplus appropriate technology could help overcome his/her very low productivity and make him/her more competitive vi8 a via imported foodstuffs.

Improve Agricultural Credit

7.59 The need for credit exists among the campesinos with some marketable output. For the subsistence farmers credit is not feasible. Either it is turned into a grant, creating dependency or it forces the family to sell even more of its insufficient production. Public sector involvement in rural credit has been a disaster. Thus, the first order of business should be to liquidate BAB. It is a drain on public resources and its clients should be served by the private sector. Of other public sector institutions the FDC should be carefully studied"3 and oossiblv be suMo2rted. The FDC, has not yet made its first loan, but the basic idea is to give credit to rural groups, such as cooperatives or communities based on so-called "solidarity" guarantees, where each member of the groups cross-guarantees every other member. In the development of FDC it will be essential to assess the successes and failures of a multitude of NGO programs.

7.60 In the medium-term, credit activities in the private sector through rural credit cooperatives should be helied and strenothened. Such credit cooperatives and savings and loan institutions were relatively successful in helping small farmers with some marketable surplus prior to the hyperinfla- tion. Credit institutions dealing with small farmers are basically opening up a new market. The necessary training of credit officers and borrowers in the workings of small-scale rural credit is thus a positive externality. Private benefits are smaller than societal benefits as there is no guarantee that staff will stay with the institution, once the initial investment is made. Thus, is does appear that a case could -be made for public sector subsidization of training of credit coonerative stafr.

12 CEDEAGRO (an NGO) is implementing such schemes in the poor, high valleys of Cochabamba.

1i A recent evaluation sponsored jointly by several international financing agencies, including the bilateral cooperation organizations of the Netherlands, the Federal Republic of Germany and Switzerland, is very positive about the prospects for FDC. 116

Imorove Small-scale Irriaation

7.61 Access to water for irrigation could greatly stabilize food production and improve welfare especially for the farmers with no marketable surplus. The possibilities for small scale irrigation projects, based on traditional community organization and active inputs from the communities, should be explored. The ir,lementation of such a project would require not only the technical aspects of irrigation design but also radical changes in the laws relating to communal property and land titles. Support for small scale irrication irojects in the Hiahlands and Valleys, based on traditional social structures annears well iustified.

Linking the Poor to the Ma ket Economy

7.62 General development benefits will only "trickle down" if the rural poor, are linked to the modern market economy. This may occur in two ways, firstly through the labor market and secondly, in the case of those who have land, through expanded demand for agricultural products. At present in Bolivia the overall economy is at best growing slowly and the labor market does not appear to be expanding nor is it likely to expand sufficiently rapidly in the near future to have a major impact on rural poverty. Hence attemots to alleviate rural povertv for the maioritv of farmers should concentrate on linking the rural poor to the market economy on favorable terms. This basically means that efforts have to be directed towards providing the rural poor the means to profitably produce, process and market saleable products so as to increase their incomes.

7.63 The following major constraints should be dealt with to induce development of agricultural enterprises geared to meeting market demands:

a) The identification of those products that can be produced by the poor and that are in high demand;

b) Evaluation of the feasibility of producing them profitably;

c) Venture capital and technical support for pilot projects in production, processing and marketing to test commercial viability of enterprises;

d) Mechanisms for transfer of technology, credit and know-how to move from the pilot phase to the fully fledged expansion phase for the selected product lines;

The existing de facto social organization can provide the basis for the development of communally managed agricultural enterprises. Communities will require assistance in areas such as project management, technical support and credit.

7.64 The examples of government becoming involved in the management of agricultural enterprises clearly indicate that this should be left to other agencies. On the other hand, it is the role of the government agencies to provide policies and guidelines within which a multitude of private, public, 117 non-government and foreign aid agencies can complement each others activities. However, in order to define these policies, government officials need as base information objective analysis of the sector, potential areas for development, appraisal of different possible development strategies and practical examples of how small- or large-scale agro industries can be effectively promoted. This base information and the develorment of oilot scale new aaro induetries may beat be obtained outside the mainstream covernment apparatus, iwossiblv in a autonomous joint private/oublic foundation. In this scenario the government agencies would still have a vital role to play. They would be responsible for the setting of policies and coordination of activities and providing what are considered as public goods such as education, research, infrastructure, etc.

Migration

7.65 In the long-term, it seems unlikely that aariculture can form the basis for maintaining at a reasonable level the oopulationof the Hiahlands. The possibilities for profitable agriculture and horticulture are greater in the Valleys and hence local agricultural development may well provide for the needs of the population. The solution then seems to lie in migration of part of the population to the potentially more productive Lowlands. However, the social ties, the uncertainty and fear of moving to a new habitat and the lack of viable agricultural opportunities (with the exception of coca) make spontaneous migration unlikely.

7.66 The key to successful migration must be the existence of viable enter- prises in agriculture in the Lowlands. The development of such enterprises can be established on the lines laid down in the previous section. However, the situation is more complex as in the previous case we assumed availability of land. In this case, land would have to be made available. One of the impediments to migration is the close ties to the community in the Highlands. One possibility would be to allow the whole community to move. The new colony could maintain its links and privileges of being part of the original ayllu, and provide opportunities for seasonal or permanent migration. In this manner, the old traditional social structures would not be destroyed and could form the base units for profitable agricultural enterprises in the Lowlands. Suonort for micration to the Lowlands. based on orofitable aaricultural enterprises. and linked to the traditional social oroanizations of the Hiahlands shows promise. Past experiences with divided migration have been dismal in Bolivia. The necessary infrastructure and agricultural extension was not provided. The new migrants found themselves in a ecological area totally foreign to them. The result has been a continuation of poverty and substantial environmental degradation. Thus, in the short run the exoerience of past migrations should be assessed. UNICEF has recently proposed an interesting study of the results of past migrations from the Central Andean Region to the Lowlands.

Women's Issues

7.67 Rural development projects should direct the use of anorooriate technoloav eaually to women and men and specificallv take women's aaricultural role into account. In the Highlands, women have the principal responsibility for animal husbandry and should be taught how to improve livestock production 118 through improved husbandry and pasture management. Training programs ahould be established in order to improve the community's ability to organize and manage rural development projects and to solve communication problems. These programs should be designed for both community agricultural promoters and the community in general, with a special emphasis on women's participation.14 Radio should be used for community education in organization, production, health and nutrition, and appropriate technology, among others. Attention should also be given to the women's ability to communicate with formal insti- tutions because of the many problems related to legal rights, inheritance (land titles), and other issues which are difficult to solve without communi- cations skills and knowing how and where to apply them.

Strenathen National Policy and Planninq

7.68 Efforts to strengthen national policy and planning capacity within MACA have not led to positive action in poverty reduction because they have been largely divorced from the farm community. A first step would be suooort of the reaional develooment councils in each department and encouraging the participation of community organization representatives in such councils. A second step would be for MACA to actively pursue coordination of olannina efforts with RDCs and regresentatives of farmer organizations. This would include supporting efforts to collect regional data and analyzing the data jointly. To do this, however, farm communit' institutions must be strengthened, with a system of delegating authority and representation at the national level in order to voice their needs and concerns at the policy-making level. Third, MACA must increase its own analytical caoacitv at the central level. In addition to exploring means to increase public sector salaries to attract and retain qualified professionals, emphasis should also be placed on creating a small cadre of highly skilled experts in agricultural policy analysis and poverty issues in order to attain a degree of influence at the inter-ministerial and international levels.

14 CIPCA/Cochabamba provides a good model for administration, accounting, and marketing programs. iCBAPTR8: flNSPORTATIOM

XNTRODUCTION

8.1 Improvements in market access are essential to raising the income and productivity of the rural poor. Bolivia's economy has historically suffered from insufficient road infrastructure and the resulting burden of high transport costs. The transport bottlenecks have been particularly serious in rural areas where populations are dispersed and mountains and rivers act as natural barriers. The outline of this Chapter is as follows. First, the institutional framework is described. Second, the analysis focuses on Government policies and investments. Third, several specific sector constraints to poverty alleviation will be discussed, including poor policy setting and coordination, insufficient maintenance, and low salaries. Finally, a set of recommendations is presented including several relating to shifts in expenditure priorities and to improving the institutional structure.

INSTITUTIONAL FRAMEWORR

8.2 The transport sector incorporates several ministries and agencies. Given that the transportation subsector with the most direct impact on the rural poor is road transport, the analysis focuses primarily on institutions directly related to road construction and maintenance. These can be grouped into three broad categoriest

a) Global Plannina Level: the Ministry of Transport and Communica- tions (MTC).

b) Decentralized Plannina and Execution: The National Road Service (SENAC) is the principal institution responsible for the national road network in Bolivia. Although it is a decentralized agency of the MTC, it has both technical and administrative autonomy. Established in 1964 to construct, maintain and administer (and hold title to) the country's road network, it concentrates primarily on principal roads connecting departmental capitals. Secondary roads providing access to principal roads are managed largely by the RDCs. The RDCs also help local communities in financing the construction and maintenance of rural feeder roads.

c) Local Execution: ESP, NGOs, private contractors, and community or- ganizations.

8.3 In this organizational structure, the MTC holds the functions of con- trol, regulation and decision-making, whereas the decentralized agencies are generally responsible for execution. The one exception is the RDCs which, although responsible to the MTC, enjoy a certain autonomy in regional planning and project management. Road transport itself is in the hands of many individual operators for interprovincial services and interdepartmental truck- ing, and of small companies for international and interdepartmental bus ser- vices. The individual operators are generally organized in unions, although membership has not been mandatory since 1983. 120

GOVERNMENT POLICIES AND INVESTMENTS

8.4 Transportation policy and investment have focusad on economic growth potential. Following the rapid growth of the mining industry, most of the construction of the existing Andean rail lines started about 60 years ago. The construction of roads and rail lines in Santa Cruz began in the fifties mainly to connect the hydrocarbon industry with the rest of the country. Since 1985 the Government has reemphasized developing the road system in the Eastern Lowlands to promote agricultural exports. Rural roads have not been considered, even though they can have higher rates of return than several other projects, mainly because of the lack of influence of rural communities and the prevailing urban bias.1

Sector Strategy

8.5 The stated obiectives of the sector's investment are to: (a) maintain and rehabilitate existing road and rail lines, especially in the Cochabamba and Santa Cruz departments; (b) integrate the three main populated areas by paving the still unpaved portions of the La Paz-Cochabamba-Santa Cruz Highway; (c) upgrade the two main export corridors to the Atlantic and Pacific Oceans; and (d) construct penetration roads to northern and southern parts of the country by connecting the major population centers with the Lowlands and navigable rivers.

8.6 The strategy does not acknowledge the importance of developing rural road infrastructure to facilitate development in poorer areas. This has been left up to RDCs and local initiative. From a poverty alleviation point of view the strategy is flawed in its emphasis on construction of new roads connecting all the major cities in Bolivia (point (d) above), while disregarding the needs to integrate urban centers with nearby rural areas. Such an integration would allow rural areas to develop as market access would be improved. Some of the roads connecting the cities of the Lowlands show very low rates of return2 and could be substituted for well designed rural roads.

Sector Investment

8.7 Xnvestment plans reflect even less of an equity concern than the Government's strategy. Public investment in transport totaled US$118.1 mil- lion in 1988 (33.9% of total). Planned investment over the next few years will fluctuate between US$100-122 million per year. Although not as dominant as in 1988, future public investment in transport will account for about 26- 28% of total public investment. Examining the types of projects, public

I See Chapter 4.

2 See the forthcoming Updating Economic Memorandum (UEM), 1990 (Report No. 8623-BO), for a review of the returns of the road projects in the investment program. 121 investment in the transport sector will not have any major poverty impact. In 1988, investment in road infrastructure made up 73% of total transport investment followed by air transport (16%). Investment in road infrastructure was concentrated in large projects. Out of 69 road projects, the three largest projects accounted for almost 50% of the public investment in transport in 1988. Regarding air transport infrastructure, nearly all (as much as US$18.6 million or 96%) of the investment in 1988 went to a single airport construction project in Cochabamba, which follows neither poverty nor national strategy objectives.

8.8 The plan for future public investment in transport shows improvement in redirecting projects toward the sector's stated objectives but continues to disregard the importance of rural roads. Most resources in road infrastruc- ture will be devoted to the maintenance and rehabilitation of the existing principal road network built among the economic and commercial centers of the country, construction of major penetration roads, and upgrading of certain segments of export corridors. Heavy concentration on large projects is likely to continue over the next years and is broadly warranted according to the strategy.

8.9 Because investment in rural roads has been mainly undertaken by RDCs, the transport development strategy adopted by RDCs as well as their planning, technical, and financial capabilities significantly affect the poverty focus of these transport projects. For example, the Santa Cruz RDC (CORDECRUZ) has put the highest priority on transport sector investment, spending 70% of its investment budget and allocating about 80% of its total transport investment to rural areas. Before the change in administration in August i989, its engineering department3 had a cadre of qualified engineers and prepared a medium-term plan for road infrastructure development. Especially, a series of rural road projects (caminos vecinales) have targeted the areas with inade- quate road network, such as the newly colonized areas and the southeastern parts of the department. In contrast, the Chuquisaca RDC (CORDECH), with its poorly staffed engineering unit, paid limited attention to investing in the transport sector in the past despite serious transport bottlenecks in Chu- quisaca. only three projects were implemented in 1988, of which one was for aviation (Monteagudo Airport), and less than US$0.09 million was budgeted for road investment in 1989. With a drastic shift in its development strategy, CORDECH now proposes to devote US$9.3 million for rural road development over the coming years. Although this change is highly laudable in principle, it is doubtful that CORDECH can manage and finance all the planned investments judging from its present capability. 8.10 ESF has provided funds to various organizations in the public and NGO sectors. As of mid-June 1989, US$14.35 million had been disbursed for 184 road projects (out of tnis, US$13.47 million were spent for 160 rural road

However, most of CORDECRUZ's future investment projects in transport are not included in the new public investment program and might change totally, as everyone who did not belong to any of the political parties in power was fired. This situation largely paralyzed the Corporation for a long time. 122 projects). ESP funded a number of small projects, focusing on development of rural roads. In 1988, ESF financed US$9 million for 119 projects mainly for the construction upgrading and maintenance of rural roads. In 1988, ESF investment in road infrastructure was particularly important in Potosi and Oruro, where only limited funds were provided by both the SENAC and RDCs. As is the case in agriculture, it was initially intended that FDR would take over ESF type projects. However, this appears unlikely now.

SECTOR CONSTRAINTS

Rural Roads

8.11 Although rural roads are of great importance to che rural poor, con- struction and maintenance of such have traditionally been neglected. A back- ground paper for this study4 looked at 14 rural road projects in the poorest areas of Bolivia. The paper concludes that rural roads are important but also that they should be carefully assessed. Simple rules of thumb can be deve- loped to assess whether a road makes sense from a socio-economic '?erspective. Projects often do not have a direct impact in terms of increased production but usually in increased sales of barley. Donkeys are sold as a substitution towards road transportation takes place and the barley they would have eaten is sold. This increases sales without increasing production. Rural roads are shown to be economically efficient investments, have a high involvement of the community and be generally cost effective. One issue is the lack of an institutional home for rural road construction and maintenance. Today, it is the responsibility of the RDCs but SENAC's local engineer is often the one with the local expertise. The Regional Development Fund (FDR) could potentially play an important role as coordinator and keeper of the technical expertise, but plays no such role today.

Weak Planning and Coordination

8.12 The three functions of the planning system--investment planning, project management, and annual investment budgeting--involve both ministries and decentralized agencies in the transport sector. Each agency has its own planning department that assesses projects and make recommendations. Rivalries among agencies have led to weak coordination of investment projects and a lack of progress in meeting the needs of rural people. To illustrate, assessments normally originate with the executing agencies (SENAC and RDCB), are passed up to their respective supervising ministries (MTC or MINPLAN), then to the National Committee for Projects, and finally to the Cabinet. The National Committee for Projects is run by the MINPLAN, but includes representatives of MOF and the initiating ministry of each project. RDCs are expected to be largely responsible for departmental transport infrastructure, yet most sector strategy proposals are formulated at the central level by the SENAC Planning Department. Consequently, no appropriate coordination system exists for incorporating the transport plans of the RDCs into SENAC's sector

4 "La Importancia de 1a Infraestructura Rural en el Alivio de la Pobreza," Rodrigo Cisneros, La Paz, January 1990. 123 strategy. To complicate matters, SENAC's District offices, which have primary responsibility for executing basic rural road projects, have virtually no say in the country's road planning. Although they participate somewhat in the design of the annual operational program for road maintenance by providing the central maintenance office with regional information, the needs of the rural poor are often overlooked because there is -irtually no means by which to communicate their priorities. Once the planning proposals have been made, the Ministry reviews and attempts to incorporate them in the National Transport Plan according to the availability of financial resources, a factor that can rarely be taken for granted (see next section).

8.13 SENAC's planning departments and the RDCs produce project concepts, make pre-feasibility and feasibility studies and engage consultants. While these are important steps in the planning process, the proposals are processed with little or no study of alternatives (alignment, standards, etc.), except when required by international lending agencies. As in other sectors, this tends to distort the internal priorities of the respective institutions and often creates longer term difficulties at the project implementation stage. Moreover, decisions on project implementation often appear to be determined by regional and institutional pressures, frequently without regard to the system of project planning. In most cases, the dominance of such pressures over a rational assessment of local needs will neglect the needs of the poor. Although some of the wealthier RDCs are very capable of undertaking studies, capacity varies widely among departments. Some RDCs have almost no technical or human resources to invest in roads at any stage of the project cycle. Departments are also meant to maintain project inventories, but often do not do so.

Scarcity of Ooerations and Maintenance Resources

8.14 Throughout the eighties, insufficient funding made it impossible for SENAC to meet all its objectives, particularly in relation to maintenance of existing roads and construction of rural roads. Although it is relatively well-organized administratively, SENAC is almost entirely dependent on TGN and external donor financing for road construction, improvement and maintenance. Consequently, its planning and execution practices have been distorted. Because the full amount of programmed TGN resources may not reach the sector, executed projects fall far short of what was originally programmed. Dependence on external financing for the remainder of its resources requires SENAC to follow priorities that are not always in line with internal and/or regional necessities. Consequently, stated objectives in the roads sector are rarely met from a global perspective. While certain regions may be receiving particularly careful attention and financing, others may be neglected altogether. This scarcity of resources has a profound effect on SENAC's physical and human resourcesas well.

Low Salaries

8.15 Although SENAC itself has a history of being staffed with highly qualified, experienced, and dedicated personnel, this is changing as many wealthier RDCs are attracting some of the best engineers away from central SENAC to their respective regional offices and the relative salaries of SENAC 124 continue to fall. While this is a positive move for those RDCs that can afford it, it exacerbates the existing regional inequities, and leaves the principal national agency considerably weaker from a human resource perspec- tive. Without adequate salaries, this resource drain is likely to continue.

RECOMMENDATIONS

Sector Priorities and Expenditures

8.16 The importance of rural roads should be recognized among the sector priorities. The basic idea in a poverty oriented transportation policy is to facilitate the development of local markets. Investments in rural roads should pass the same strict cost-benefit criteria as other roads. Among the principal roads presently in the investment program some show an internal rate of return of less than 12% while projects for rural roads exist that have a return much higher than 12%. For example, by substituting these high-return roads for the low-return principal roads, there is a possibility of improving both equity and efficiency of the investment program.

8.17 This can be accomplished, in part, through some of the community par- ticipation and labor-intensive means pursued by ESF. The transfer of ESF methodology to SENAC and the RDCe would ensure that the same types of poor communities would continue benefiting from such projects in the future. Given that SENAC is considering proposals for decentralizing all its secondary and tertiary road activities, special attention should be paid to the development of capacity to reach poor communities at the regional level. The future role of the Regional Development Fund (FDR) in promoting ESF style projects should be assessed and possibly promoted.

8.18 The economically efficient rural roads projects have not been undertaken because of lack of attention and pro-urban biases in the decision making. To overcome these obstacles, costs must be kent down, the settina of 2riorities improved, and technical skills strenathened. Costs can be kept down by: (1) focusing on spot improvements to drainage, bridges and other vulnerable sections; (2) using local labor as far as possible; (3) revising the procurement system for small projects, as the system is relatively more expensive for small than for large projects; and (4) ensuring that roads once improved are kept maintained. The experience of ESF is important in both the identification of high-return spot improvements, in the use of local labor, and in using a different procurement system.

8.19 To improve the setting of priorities, community involvement should be enhanced and planning should consider other activities in the project areas. First, the communities should be involved. This could be done by allocating a share of road investment funds to rural roads projects proposed by communities (see next paragraph). The communities would be required to provide counterpart funds (e.g., 20%) in the form of labor to assure a thorough selection by the community. The Government agency should contract out the works to a private contractor under the community's supervision, or to the community itself. This would help ensure the successful completion of the 125 projects. Second, ro4d planning should consider proposed and ongoing activities to improve agricultural production (irrigation, extension, etc.).

Institutional Recommendations

8.20 Chanee budgeting process. To ensure the consideration of rural roads in the investment plans a certain percentage of the national investment budget should be set aside for such projects. These funds would probably have to be distributed roughly among regions to avoid the more powerful regions using up all the funds. The funds should be used for matching grants to local or regional institutions proposing projects. Communities should be provided with technical assistance in the design of projects (see below), and be required to provide a certain share of the costs in the form of labor or local materials. The selection and financing of such projects would be an important role for SENAC or FDR to play during and after the decentralization process to help alleviate regional differences.

8.21 Establish clear lines of resDonsibilitv. A system of planning, coor- dination and delegation of responsibility should be established between SENAC, its District Offices, and the RDC Transport Units. The RDCs should be given both the responsibility and budget to concentrate on regional priorities. If this option is pursued, however, it would be important to analyze the specific roles that each institution should play, the capacity of the RDC Transport Units and the potential costs of transition. A preliminary analysis would suggest the importance of maintaining the existence of a national institution. Its principal functions would include: setting national objectives and norms; providing uniform monitoring and evaluation standards; retaining responsibili- ty for maintaining, rehabilitating, and, where necessary, constructing the principal national arteries/networks.

8.22 SENAC should provide both consultative and trainino services in regional road planning, project design, implementation and evaluation. Particular emphasis should be placed on training in cost-effective methods. To comple- ment these training efforts, it would be useful for the weaker RDCs to take advantage of the experience gained in other RDCs by instituting a cross- departmental training program. Where necessary, specific measures could be considered by a central institution such as the FDR to provide the necessary financial resources to attract additional professionals to the RDCs where road expertise is the weakest. In sum, all relevant road institutions should receive appropriate technical assistance not only to guide the process of decentralization, but also to strengthen the functional roles of each institution once decentralization has been achieved. 126

CHAPTER 9: EFFICIENCY I88UES

INTRODUCTION

9.1 Since investment resources are scarce, the opportunity costs of an investment strategy oriented to poverty alleviation must be assessed. It is often assumed a priori that investments in poor regions are less cost-effec- tive and have lower benefits. This assumption is based on the fact that productivity is low due to isolation from markets and low density of benefi- ciaries. This chapter reviews how investment costs vary between poor and non- poor regions and the degree of efficiency or impact of investments in poor areas. The Chapter concludes that it is possible to make cost-efficient investments in poor areas if appropriate technology is used and communities are involved in the implementation. Recommendationa are made for improving the return on investments, and the possibility of leveraging investment resources in poor regions is explored.

COSTS OF INVESTMENTS IN POOR AREAS

9.2 The experience of the ESF is used to determine whether the cost of, for example, building a school or maintaining a road varies between regions. Through its financing of three thousand small projects in infrastructure and social services, the ESF has amassed a data base on investment costs. This information falls into two categories: first, expected costs based on market prices, and second, actual implementation experience.

9.3 To control their contracting procedure, the Index(La Paz 1) ESF established a data 1.4 base of the costs of labor 1.2- and major construction ma- 1 terials around the coun- try. These prices are for 0. urban areas only. In 0.6 rural areas prices vary 0.4 R too greatly between speci- fic locales and times 0.2oof the year to be used as awi general index. Using nine O' ?oct!flbo c5tAo%, ca 16AP~ types of standard pro- jects, the ESF developed Construction Empedrade Enimosodo RoadUclnlenance estimated costs per unit rojectOpes. Sb of output for eight urban toad Oinpus t obsur d rencesCn areas (excludingthe Empedrodo and Enlosela4doare forms of street- areas (excludingthe ~payingwtfh stne w cobbIeaon.s. respeditvety Departments of Beni and Source, ESF Pando because of the wide I variation in costs). Figure 9.1: 3SF Standard Costs in major cities Figure 9.1 shows these unit costs for eight of the project types. Since these are benchmark prices based only on the cost of standardized inputs, they obscure differences In 127 topography, access, construction style, and the density of beneficiaries served by the infrastructure.

9.4 Most of the project types in Figure 9.1 vary by less than ten percent from the La Paz base price. Santa Cruz has consistently higher prices, approaching a forty percent differential between road projects carried out in Santa Cruz versus those in Potosi or Oruro. In fact, the poorest cities tend to have the lowest investment costs, as witnessed by the fact that all of the projects when costed for the cities of Oruro and Potosi are below the La Paz index price, and those in El Alto are either equal to or below the index. Therefore, looking only at urban areas, costs of focusing infrastructure investments on poverty regions should be expected to be lower.

9.5 To judge whether this expectation is fulfilled in implementation, the actual costs of sixteen types of projects were analyzed. The data were broken down between rural and urban areas by poverty area.I The poorest provinces (Area 5) are characterized by a predominance of dispersed, rural populations that live at a subsistence level. The wealthiest provinces (Area 1) include provinces that contain department capitals. These projects are not standar- dized and therefore often not fully comparable, but do reflect regional contexts. This information aggregates different types of technologies and does not consider donated local materials or labor in the cost of the project.

9.6 Table 9.1 provides - indicators of technical COSTAWITBa POVERTYAEMS F ESF PIWCTS efficiency as measured by Poverty Ares 1 Poverty Area 4 PovertyArea S cost per unit for projects Urban Rural Urban Rural Urban Rural in water supply, sanita- tion, roads, school con- Water Supply struction and health faci- Cost/Meter 18.62 14.54 N/A 12.54 7.97 26.41 lities. The information Seitation on water supply reveals Cost/Meter 19.45 10.48 18.23 16.60 12.40 20.40 wide variations in costs, with the lowest cost per Cost/Km 3,?96 1,965 N/A 4,692 "/A 1,692 meter installed in urban centers in the poorest Schoot Construction provinces. Unit costs of Cost/School 40,239 62,559 40,880 18,706 38,065 25,238 sanitation projects in ealthContructicn rural areas increase with Cost/Center 37,813 27,056 25,641 7,669 109,49112,892 the level of poverty, while costs of urban pro- oure. 8SF jects decrease with the W level of poverty. This is Table 9.1: Cost/Unit Variations by Poverty Areas probably because sanita- tion technology can be more easily made adequate for local needs, for urban systems than for rural systems. The urban systems in Area 1 often need to include more expensive treatment facilities, whereas such facilities are not necessary (or at least not used) in the small towns in Area 4 and 5. Road maintenance project cost vary widely, though costs are comparable between the

1 See Annex II. 128 poorest and wealthiest regions in rural areas. School construction is least expensive in poor rural areas, undoubtedly reflecting smaller scale. Health construction follows a similar trend. The lowest costs are in poor rural re ions, accounted for by a predominance of health posts in rural areas and hospitals in urban areas. The conclusions from Table 9.1 suggest that in simple cost per unit terms, poor rural areas often out-perform other regions when technology is adapted to local needs.

9.7 Table 9.2 uses cost . -. -...... per beneficiary as an C/UMEICJARY VARIATIOSY POVERTYAMA additional measure of l U$/8enflcary) efficiency. The dispersed Poverty Are* 1 Poerty Area 4 Poverty Are" 5 nature of poor rural areas Urbn Rwal Urban Rural Urban Rurat of Bolivia would be expected to give these itater Syply 54 U/A N/A 16 43 20 regions a decided aoeitan 26 33 88 58 67 55 disadvantage in terms of this measure. construc- Roo& 7 32 N/A 28 2 6 tion of schools and health Sdo lCotr. 3? 96 115 82 9 218 facilities, show the higher costs of providing aeetIh Contr. 3 2 I3 12 68 10 a standardized service to dispersed populations. SOI:ES? However, the evidence shows that for water -- showsy, thats forper Table 9.2: Cost/Beneficiary Variations by Poverty supply, costs per ue beneficiary are about A three times lower in poor rural regions than in the wealthier urban areas. The poorest provinces also out-perform the rest of the country in cost per beneficiary of road main- tenance projects. Sanitation projects show higher costs in poor areas, but rural projects are competitive with urban projects on a cost per beneficiary basis.

9.8 The data presented in Tables 9.1 and 9.2 show that infrastructure investments in poor regions, and particularly poor rural regions, can be competitive on a cost per unit or cost per beneficiary basis. This is particularly true in types of projects where a technological package can be developed that in appropriate for dispersed rural populations. One example is in water supply where costs per beneficiary can be controlled by switching from the urban model of centralized systems to a decentralized system of individual pumps which correspond to the more dispersed population. In more standardized types of investments, especially investments in social services like primary education, mechanisms must be sought to reduce the inherent inefficiency of dispersed populations. For instance, the unit cost of building schools in poor rural areas is generally lower because of the smaller size and availability of donated labor or local materials. However, the cost per beneficiary of the investment is much higher in poor areas in general and poor rural areas in particular. 129

EFFICIENCY AND IMPACT OF INVESTMENTS IN POOR AREAS

9.9 This section presents a summary of the experience of investments in poor rural areas of Bolivia. As presented in the preceding section, investment costs vary greatly between regions and poverty levels. To gauge efficiency, these costs must be matched against output. Unfortunately, little systematic data exist on economic or social return of projects in poor areas. Table 9.2 presented one indicator, cost per beneficiary, for a group of projects carried out by ESF. This section summarizes the existing information on efficiency and impact for roads, irrigation, water supply, health facilities and schools.

Roads

9.10 Because of difficult topography and low levels of agricultural productivity, it is often assumed that the construction of roads in very poor areas of Bolivia is not economically justified. A study commissioned for this report analyzed the economic impact of roads in poor, rural areas of Bolivia. Of seven roads2 identified and analyzed, only 1 generated an economic return below 12%. The positive economic impact of road opening and upgrading includest (i) the sale of pack animals and excess forage as peasants sub- stitute forms of transportation, (ii) an increase in agricultural production primarily in zones with under-utilized capacity, and (iii) an income effect from lowered transport costs, reduced prices of goods imported into the community, and expanded opportunities for seasonal migration of labor. The report found that the level of economic benefits depends on the structure of the local transportation market, the level of community organization, the present level of production, and the possibilities for expanding production.

9.11 The structure of the transport market is important in determining how benefits are distributed. In one case, the benefit to local residents in terms of lowered transportation costs was only an absence of an annual increase, not an actual reduction in transportation costs. Transport was controlled by a monopoly trucking firm that absorbed most of the benefits and did not expand services. One project had no benefit at all due to lack of community organization. The project was located in an area with 13 communities spread over two departments and three provinces. Both Aymaras and Quechuas live in the area and religiously there is a strong division between catholics and evangelicals. Thus the communities could not agree on a joint sale of production after harvest that would lower transportation costs, and could not put any pressure on the trucking company to lower tariffs.

9.12 The study found that roads are beneficial if production is not techni- cally constrained. If other technological and market constraints exist, investments in roads will have little impact. Therefore, new investments in rural roads should: (i) be oriented to regions with under-exploited productive capacity, and (ii) be complemented by programs to organize collective transport of products and provide information and technical inputs for commercializing local production.

2 The seven roads had benefitted from a total of fourteen individual projects. 130

Irrigation

9.13 Irrigation investments have had a mixed record in Bolivia. Recent assessments have shown the following factors as being important for the success of irrigation projects. First, small scale is essential, no large scale irrigation scheme has succeeded. second, community involvement is essential. Third, follow-up assistance in improved farming methods and marketing can substantially increase the net returns. Fourth, net benefits vary negatively with the altitude of the area.

9.14 The National Service for Community Development (SNDC) has implemented thirty-nine small-scale irrigation projects in the Highland and Valley regions. The costs per hectare range from $18 to $229,407 per hectare and from $2 to $376 per beneficiary depending on the system applied. Detailed ex- post evaluations carried out on two of these projects highlight some of the factors that influence the impact of the investment.

9.15 The first project is located in a community of 250 people on the road from Potosi to Sucre. A $17,000 investment in irrigation increased the irrigated land in production by 152% and the annual yield by 203%. However, family income was only 32% higher in 1989 than when the project was initiated in 1982, far short of project goals. Increases in transport costs and lower prices offered by the intermediaries reduced the benefits of the surplus generated. In addition, because of the limited size of the local economy, production increases forced down prices fetched in the local markets. The project had not considered the potentially negative effects of commercia- lization and market absorption.

9.16 The second project is located in an isolated community in the department of Oruro that survives on a very limited base of potatoes and livestock, mostly llamas. Due to weak technical implementation and miscalculation of available water resources, production increased only 26%. Family income remains unchanged due to unchanged low productivity, and few market oppor- tunities. No crop diversification resulted since the community did not receive any technical assistance or training during project execution. The community could, therefore, not fully utilize the expanded opportunities afforded by irrigation.

9.17 A recent evaluation3 of 558 mi-ro-irrigation projects in La Paz, Oruro, and Potosi Departments, on the other hand, reached very positive conclusions. The report stresses the importance of community involvement throughout the project cycle starting with identification. Communities often identified projects, participated in construction, and organized themselves to maintain the projects. The technology chosen was problematic in some cases. There is a strong need to balance the limited knowledge of the communities in project implementation with the need for hydrologically sound investments. Most of

"Informe de la Kisi6n de Evaluacidn de Proyectos de Desarrol.o Agropecuario y Microriego en los Departamentos de La Paz, Oruro, y Potosi;" MINPLAN, MACA, UNDP, and Swiss and Netherlands' Technical cooperation; May 1989. 131 the irrigation pro4ects did not change farming methods in the affected areas, due to lack of follow up with extension and other services--similar to the individual experiences of SNDC cited above. Even without changes in technology the benefits were substantial in reducing the variability of the amount harvested. Some examples show double bxnefits with extension and irrigation, compared to irrigation alone. The cost benefit ratio calculated at a discount factor of 12% p.a. showed positive net benefits for 93% of the projects in terms of area irrigated. The net benefits vary inversely with the altitude of the project area.4

9.18 The lessons from past projects show that to increase the efficiency of irrigation investments, more attention needs to be placed on the commer- cialization of production and other post-harvest activities, the development of integrated packages including agricultural extension and credit informa- tion, and on community involvement. Technical problems in many small-scale irrigation projects point to the nee. to refine site analysis procedures and types of technology selected.

Water Supplv

9.19 To merit investment in dispersed rural populations, water supply technology must be carefully chosen. As the figures from the ESF show, costs per beneficiary of water supply projects in rural areas generally out-perform those of more sophisticated urban projects. This conclusion is tempered by the fact that urban water systems, while expensive to install, often are managed by much better institutions that more easily could apply tariffs5 that theoretically will allow recuperation of the investment over its lifetime.

9.20 Substantial knowledge about rural water systems and appropriate tech- nological packages is available in Bolivia. UNICEF, CARE, bilateral programs and some RDCs have been active in implementing these systems. The World Bank is planning a pilot project for low-income communities in Oruro and Potosi based on several years of testing. From these experiences, some efficiency guidelines can be established. For instance, UNICEF calculates that in very dispersed populations (under 200 people) hand pumps are most economical at an average cost of $200 per unit and US$8.00 per beneficiary. Depending on topography, spring-fed systems also are attractive at a construction cost of $150 and a cost per beneficiary of about US$7.00. Populations with a size between 200 and 800 people need a more sophisticated system of gravity pumped water that costs approximately US$11,900 per unit and US$30 per beneficiary. If we assume that the benefits from access to water are equal, it is more attractive to invest in more dispersed populations.

9.21 The return on these water investments can be maximized through two simple strategies. The first is to organize project implementation in such a

4 From gross benefits of US$200/ha at 4,000 m and higher to 3500$/ha in areas below 3,000m.

Such water companies can also often practice cross-subsidization from richer to poorer customers. 132 way as to ensure continuity of maintenance after the initial investment. This can be done by using tariffs to create the cashflow necessary to cover maintenance expenses. or, in extremely low-income communities, training community members in routine maintenance during the construction period will help to increase the effective lifetime of the investments. The second strategy involves education efforts to allow beneficiaries to derive the maximum benefits offered by access to water. Often opportunities for improv- ing both food production and the health status of the family remain unrealized because no training in adapting to the new systems is provided. The costs of education activities in water projects are typically less than 2% of the total project costs. This represents a very low-cost method of improving the return on these investments.

Health Care

9.22 The experience with ESF-financed projects has proven that developing standard packages for basic equipment, controlling construction costs, and focusing only on essential infrastructure can bring down costs to an acceptable range for introducing health services to rural populations. However, efficiency depends on the appropriateness of scaie, location, quality of services and community extension activities.

9.23 Rural hospitals are generally too large and sophisticated, and under- staffed for the population it is intended to serve. Rural Bolivia is littered with multi-bed facilities lacking mid-level staff (nurses and auxiliary nurses), operational budget and patients. An optimistic estimate by the Ministry of Health puts the average occupancy rate for rural hospitals in most departments at less than 20%.

9.24 In recent years, more attention has been focused on the need to build less sophisticated infrastructure to provide preventive health care and limited medical attention. The ESF has financed requests for over 250 basic health posts, most located in rural areas. The basic model, developed by the NSF and by UNICEF, includes infrastructure and equipment for a minimum level of service that costs US$10,000-US$12,000. Even with pared down investment costs, the lack of resources to cover recurrent costs of salaries and medical supplies limits the possibility of extending this model to increase the provision of health services to the rural poor.

9.25 While health posts have lowered the cost per beneficiary, their potential impact on improving the health status of the population has yet to be realized. One of the principal problems is low utilization. In a study of 11 rural health projects executed by non-governmental organizations, patient visits to health posts averaged 20-63 per post per month (para. 5.33). At these levels of utilization, infrastructure alone is obviously not the answer. Health facilities face significant obstacles to efficient service delivery. These obstacles include cultural beliefs and customs among indigenous groups that reduce demand for modern health services, the strong presence of traditional medical practitioners, economic constraints on paying for consultations and medicines, and a lack of rural infrastructure, such as roads, to facilitate use of health services. 133

9.26 Large efficiency gains are available from existing infrastructure. Investments should be focused on increasing tht utilization of existing infrastructure. New facilities should only be built in areas with a suffi- ciently large population that can be integrated into servi,. networks radiating from existing under-utilized facilities such as hospitals. Gains can also be made by closing down infrastructure that is duplicative, a problem in several rural areas with a high presence of NGO and government infrastruc- ture. Finally, developing outreach programs, appropriate educational material and service approaches that attract the confidence and support of the indigenous population is essential.

9.27 Besides increasing efficiency of existing facilities, a poverty-oriented investment strategy should seek to expand coverage to the unserved. Improved efficiency of the present system will free up resources to allow this expansion. One promising strategy is to expand the effective coverage of infrastructure with investments in mobile teams using a health post as a base. This avoids the cost of additional health posts to serve very dispersed areas. However, calculations must be made carefully to avoid spending more money on expensive transport modes than would be spent alternatively building basic fixed infrastructure. In the Amazon region of Bolivia, this idea of mobile teams can be translated into a river-based system of health boats to serve the dispersed population living along these rivers.

Education

9.28 Investment costs per primary school do not vary greatly among regions. indeed, rural communities may require less financial outlays because of the potential for community donations of materials and labor to construct schools. Unit costs of material inputs like textbooks and desks do not tend to vary significantly among regions either. For instance, based on ESF figures, the cost of transportation of textbooks and desks is a low percentage of the total costs of production, reaiultingin fairly uniform costs of provision around the country. A standard educational package would have roughly equivalent prices independent of region. However, while most rural communities have access to a primary level school, they usually lack the trained teachers and materials to deliver education.

9.29 To raise the economic efficiency of rural primary education, mechanisms must be found for raising student-teacher ratios. Two approaches bear support.6 UNICEF has been developing a program in multi-grade instruction. Instead of striving for one teacher for each class level, rural teachers would be trained and provided with materials that would allow them to teach multiple grades within the same classroom. This program requires investments in training, materials and some redesign of schools, with almost no incremental recurrent costs since the teachers are already receiving salaries. This approach will have little budgetary impact since external financing is usually available for investment but not for recurrent costs.

6 See also Chapter 6. 134

9.30 A further strategy for increasing both the efficiency and coverage of rural primary education is the establishment of rural boarding schools. Boarding facilities would be built onto existing rural primary schools to allow access for children who live in neighboring areas too dispersed to justify a separate school. The advantage, again, is that no new budget positions are needed. In fact, the existing teachers would teach more children, raising the efficiency of the system. There are isolated examples of such schools, usually run by religious organizations. An NG0 in the poor province of Cordillera in Santa Cruz plans to adopt this system, covering the costs of boarding through a combination of food donations and food gardens implemented by the school.

9.31 Because of their small size and difficult access, there are many isolated communities for which it is hard to justify investments. In education, experimental programs have been implemented in primary education and adult literacy using radios coupled with the distribution of written material. However, these programs should be viewed as a second-best solution since the quality of the primary education received is usually lower than in a school setting.

LEVERAGINGINVESTMENTS

9.32 As has been shown, making essential infrastructure and social service investmente in poor regions can be justified on efficiency grounds. There is no a priori reason to exclude investments in poor areas. In addition, there is room to increase the efficiency of these investments through better selection of technologies and delivery mechanisms. These conclusions support the case for shifting a portion of the public investment program to these poor rural regions. However, in Bolivia's very centralized system, resources are slow to trickle to poor rural areas. Resources will need to be generated within these regions themselves to ensure investments, at least to provide counterpart funds for nationally sponsored projects. The issue becomes to what extent central government or donor money can be leveraged through local counterpart financing.

Financial Resources

9.33 The availability of counterpart financial resources comi-g from public and private agencies implementing investments in poor areas ia usually considered quite limited. Tax bases are weak and the investment budgets of agencies constrained. To get an idea of the distribution of available comple- mentary financing, Table 9.3 presents figures from the ESF on counterpart financing of projects by region and poverty level.

9.34 Urban-based projects provided 11% of total project costs, with rural areas slightly les at 8%. In terms of poverty levels, as expected, the region of highest degree of poverty provided the least financial counterpart, averaging only 4.75%. However, $2.3 million in financing was generated in rural areas designated poverty level 4, which is significant. 135

9.35 These counterparts [come from central government CTERPARTFIAWCIAL C0NTRITUTI?WS tO PVERTYLEVEL agencies,agncmesfromuncnpral municipa- ESFPRQJECTS SY REGION, AND Counterpart lities, NGOs, and others. Comitment Counterpart Total As X ofTotat In the present context of c(s) 80 extremely limited public Poverty Area 1 mechanisms Rural 7,405,965 449,777 7,855,742 5.73 ssectoremelyurces, eector resources, mechanisms Urbn 59,247.401 9,171,211 68,418,612 13.40 for leveraging complementary Total 66,653,366 9,620,988 76,274,354 12.61 expenditures by agencies working in poor areas must Poverty Area 2 be eploed.Table 9.4 Rural 16.496,123 2,231.105 18,727,228 11.91 be explored. 9.4 Ruran 17bler:,723,618 1,317,173 19,040,791 6.92 shows that the highest over- Total 34,219.741 3,548,278 37,768,019 9.39 all shares of counterpart financing for ESF projects Poverty Area 3 have been provided by RUral 18,3854514 547,737 18,933,251 2.89 have ~~~~~ ~ ~~~Urban3,677,239 178,760 3,855,999 4.64 religious organizations, Totat 22,062,753 726,497 22,789,250 3.19 cooperatives, grassroots organizations, and municipa- Poverty Area4 lities. Central government Rurat 20,249,8212,351,704 22,601,525 10.41 Urban 1,700,420 40,775 1,741,195 2.34 agencies have provided very totat 21,950,241 2,392.479 24,342,720 9.83 little in the way of finan- cial support, again reflec- PoWertyArea 5 ting severely limited UrbanRurat 5,426,799784,468 308,2021,747 5,735,001786,215 0.225.37 investment budgets particu- Total 6,211,267 309,949 6,521,216 4.75 larly for the type of decentralized, small-scale Totat Rural 67,964,222 5,888,525 73,852,747 7.97 projects presented to ESF. Urban 83,133,146 10,709,666 93,842,812 11.41 Muricipalities have a Total 151,097,368 16,598,191 167,69s,559 9.90 certain ability to generate their own resources and NGO tSurce: ESf often have channels of external financing. The Tale .3: C F Contribut fact that cooperatives and Table 9*3: CounterpartFinancial Contributions grassroots organizations were able to generate close to $2.4 million to support projects is encouraging.

Communitv Contributions

9.36 The financial costs of carrying out investments in poor areas can otten be lowered through the contribution of donated labor and local materials by beneficiary communities. While this system tends to complicate project execution and requires more technical supervision, the benefits in terms of leveraging scarce investment dollars and increasing community participation are significant.

9.37 There is very little systematic information on the value of community contributions in investment projects. The most rounded experience comes from the National Service for Community Development (SNDC). The agency requires that all unskilled labor and local materials be provided by the community in the execution of its rural infrastructure projects. Of their projects in flood control (defensive walls), roads, water supply, and small-scale 136

irrigation, the imputed value of community contributions ranged from 16% to 79% of total COUNTERPARTWNTtRIOlNS TO 8SF project costs, with most communitiee able to provide between 30 and 40%. In irrigation PUOJECTSBY SOLICITINOAGSAO projects, the SNDC leveraged S1.3 million in total Comerpt as investments with an expenditure of a little over AsnKv ShArssof YLCs $SOO,000. The micro-irrigation study found a community contribution of on average one third. RurosiCU 6rgm9zatfon: tIrban 20.80% Total 17.102 CONCLUSIONS AND RECOMMENDATIONS Cooprative& Gressroots OruriationS: 9.38 Based on the existing evidence, there is no Urba 14.17% reason to suggest a priori that investments in poor Total 15.15% regions, particularly rural areas, can not be 1c1palltless competitive. While differences exist in labor and Urbanural 12.97%18.61% material costs, they do not systematically Total 14.6S% influence the costs of implementation of specific SocutarWOOs: projects both among regions and areas of different Rural 13.9M poverty levels. Of far more importance to the Utalbn 3 112 economic viability of a project is the scale and ReionalDveloqpwt Corp. type of technology implemented and the attention Rural 8.67% paid to issues of market structure and commer- Urban P.M cialization possibilities. TUtetr-A _. , *s~~~~~~~~~~~~~~ohrGovernmucApmciz (CMOIS,OPSSP, SENAC ate.) 9.39 For roads, construction of rural roads should lural 1.95% take place under the following conditions: A Urban 7.26% sufficiently large pre-project production per km. Total 4.7 of road constructed, an expected increase in Source:ISf production, or a generated saving independent of the increase in production. The level of benefits _ _ depends on the degree of community organization and Table 9.4: Counterpart the level of agricultural potential in the area. Contributions to ES? The level of benefits to the poor also depends on the degree of competition in the local trucking industry. The recommendation is to improve existino rural roads. and construct new rural roads in areas with some existino groduction and a oood community oroanizat_ion.

9.40 Irrigation shows good potential in poor areas. Irrioation oroiects should be followed uD with substantial technolooy transfer and assistance ln marketina. Again community organization is key. In water supply, technology needs to be adapted to local needs. Investments in water should follow the ouidelines established by UNICEF. Training in maintenance or establishment of water cooperatives are key to the sustainability of these projects.

9.41 For projects that offer social services such as health care and educa- tion, dispersed populations are decidedly at an efficiency disadvantage. Health investments, in particular, have been badly planned to take into account the dispersed nature of rural populations in Bolivia. However, large efficiency gains are possible in both health and education by rethinking approaches to service delivery. In health. there is a need to build low cost 137 basic fagilitiesand asneially to Mmroeh use gf isINa aiiis Outreach programs, education, and comunity boad sorvice delivery will all be important ln ralisng the effiliency of eristLiug health lntrastructure. In education. focusshoul be gn umnaeina bafclte.atter.reMir of strUctursnd constructio of bhOana sghgools cTherec M r o nd tions pre- sented ln Chapter 6 for reform of the rural system would all belp in uelng exlitlng facilitlesmore efflicently. Thes rLoanuendatios lncludes (a) introductionof multi-gradeteaching; (b) pilot programs ln rural boarding schoolsg (c) improvingthe provisLonof textbooksand other materLialtand (d) improvingcommunity control and partLcLpatLon. 138

CHAPTER10: RECOMM3NDATXISOAND ACTION PLAN

INTOD=UCTXON

10.1 This Chapter presents the recommendations of the previous chapters as an outline of an action plan. Since the report is focused on sectoral policies and institutional issues, the action plan is too. First, the chapter reviews the causes of poverty d.-scuesedin Chapter 2, summarizing the actions recommended to attack each cause. Second, a set of key cross-sectoral criteria for successful interventions is proposed. These criteria may be used as a yard-stick against which to judge any proposed action. Finally, an outline of an action plan is presented. To fill out the outline, a series of specific projects and programs must be agreed upon.1

THE CAUSES OF POVERTY REVISITED

10.2 In Chapter 2, the causes of poverty were divided into five broad categories:

a) general development problems; b) poorly developed human capital; c) poor asset base; d) lack of public goods; and e) discrimination.

The following paragraphs summarize how each of these problem has been attacked in the outline of the action plan.

10.3 It is almost a tautology to say that general develoument problems are solved through overall development of the economy. In the context of poverty alleviation, it was found that the most important role for macroeconomic policy is to avoid the distortions that lead to a higher return to the already privileged. That is, the best overall development philosophy is to maintain the free-market economic policies, combined witn direct interventions, to benefit the poor through sector policies and specific projects.

10.4 The problem of poorly developed human caoital should be attacked from several fronts. The efficiency of education should be improved though improvements in productivity and salaries of teachers, provision of more education materials and increased involvement of NGOs in management of schools. Basic education, instead of inequitable university spending, should be supported. The health status of the population should be improved through more focus on primary health care, involvement of beneficiary groups, integration of traditional and modern medicine, and increased involvement of NGOs. Nutrition should be improved through dissemination of improved production techniques for staples and iodation of salt across the country, supplemented by closely focused nutrition interventions for children.

I Annex 1 contains a preliminary list of ideas for projects and studies. 139

Especially for women, literacy campaigns using NGO experiences, reproductive health services and culturally sensitive reproductive education should be pursued.

10.5 The poor asset base in rural areas is mainly related to secure access to good land. Thus, the action plan recommendss recognition of communal property rights in the formal title system and streamlining of the process for obtain- ing titles. Rural credit should be channeled through local private or mixed institutions such as credit cooperatives, etc. In the medium-term, the Peasant Development Fund should be strengthened. Technology for subsistence crop improvement should be developed within the private sector and distributed through peasant organizations and other private sector groups. In urban areas, the main problem is related to lack of cheap credit at appropriate times. Ongoing, sustainable ZGO and private sector activities should be supported to this end.

10.6 The provision of public services should be demand driven. In health and education, the Social Investment Fund should be supported. In transport, the successful experience of the Emergency Social Fund should be absorbed by the Regional Development Fund and Regional Development Corporations (RDCs). The priorities of the transportation sector should be shifted toward maintenance and enabling market access for rural producers. Rural roads should be built where cohesive community organizations exist, where they will increase access to markets, and where pre-project production is high enough to justify it. Irrigation should be pursued on a major scale through many small projects. The RDCs should play a key role in coordination and control. Communities should be required to share in the cost of provision of the services through the donation of labor. The provision of infrastructure without full cost recovery should be the main form of subsidization of poor rural communities.

10.7 Discrimination is often regarded as a general development problem. It is true that much discrimination is socially or culturally determined and thus slow to remove. As a minimum, laws should be changed to remove discriminatory clauses. Such changes should include the Labor Code and the Agrarian Reform Law. Bolivia is too poor to be able to afford the luxury of limiting the productivity of large segments of its population through discrimination.

CRITERIA FOR SUCCESSFUL POVERTY ALLEVIATION INTERVENTIONS

10.8 Bncourage a multitude of activities, programs, solutions--do not strive for centralized, national, global solutions. Large national schemes have generally not taken the concerns of the poor into account. They have been perverted by the interests of the rich and powerful and have in only few cases reached the poor. An exception to this rule is the national vaccination campaigns and some other national health campaigns.

10.9 Interventions should be small-scale. Small projects have a much better track record in terms of poverty alleviation and efficient implementation than large multi-billion dollar projects. Such small projects include irrigation, local road projects, and simple water supply schemes. 140

10.10 Interventionsmust take the cultural heritage of conumitis Auto account. Community structures can become very efficient partners ln developmentif projects are developed with beneficiaries. Many projects have failed when cultural norms were not considered. The indigenouscultures have practiced sustainable agriculture for centuries, the need is to build on these traditions, not to discard them as inferior.

10.11Projects should be demand driven. One way to lnvolve communities and take their cultural heritage into considerationis to react to demand for interventions,and not rely excluaivelyon a centralizedtop-down planning process. A demand-drivensystem must be modified to include substantial outreach activities. Communitieswithout the necessarycapacities for developingproject proposalsmust be helped. The modified demand-driven system presentlyin place in the Social InvestmentFund could serveas a model.

10.12 Implementprojects with local private organisations. Locally based NGOs, communities,neighborhood councils, mothers' clubs, etc., could all play an importantrole. At times communityorganizers might be needed. This could be an NGO expandingits geographicalor sectoralscope.

10.13 The "bottomline" in a strategybased on such interventionsis to help the poor help themselves. Only if this route is followed will interventions be effective,efficient and sustainable. ACTIONPLAN Lont Term ObJectives Medium Term Obfectlves Shor Term Oblectives

Limlt direct public in- Deepen economic reforms Implementstructural re- terferencein economlc forms in investment,ml- decisions ning and hydrocarbons

Further limit state in- Streamline licensing and terference registrationrequire- ments for businesses

Continue free trade and Maintain auction for ex- Continuepresent system, exchange rate regime change rate strive towardsmore com- petitive level

Improvecustoms collec- Streamlinecustoms ser- tions vice along the lines of the implementationof the tax reform.

Maintain fiscal stabi- Implementfully the tax Improve collections, lity reform targeting large tax- payers.

Implementland tax for Collect tax from large large landholders,in- landholders,exempt crease rates for large smallholders. landholders

Llmit current Reallocateexpenditures expendituresto levels towards agricultureand of revenues social sectorswithin overall flscal constraint.

Limit defense expendituresto 201 of central government spending.

141 Lon# Term Oblectives Medium Tgm Obeive Short Term Objectlves

InstitutionalRecommendations

Increase focus on de- Strengthenlocal insti- Find institutionswho mand-drivenapproach tutions could help set up local organizations(could in- clude NGOs, neighboring municipalities,or suc- cessful grassroots groups from other areas)

Support institutional Allow communitiesa say Streamlinesystem for pluralism in who should run their such approvals health, education,ex- tensionor other system (NGOs,private or pub- lic)

Foster beneficiarypar- Send representativesof ticipation public institutionsto local farmers'markets to listen to local wishes

ImplementSIF outreach program, assess how Pea- sant DevelopmentFund could involvebeneficia- ries.

Decentralizeadministra- Use phased approach Begin decentralization tioii in a few sectors.

Limit regional differen- Establishmechanism to ces. redistributerevenues between regions.

Assess the role the Re- gional Development Fund could play in such redi- stribution.

Provide technicalre- Establishmunicipal and sources to regions. regionalgovernment training institute.

142 Lomt Term Gbfectives Medium Term Obfectives Shrt Tem Gb fe_tive_

Decide degree of local appointmentof offi- cials.

Improvepolicy, planning Strengthenthe policy Restructure and coordination role of Sub-secretariat subsecretariatto for Social Policy. includepractical planners and policy analysis

Strengthenpolicy and Create attractiveposi- planning units of mini- tions in policy units stries throughdownsizing of administrativestaff

Focus ministrieson pol- Abolish implementing icy and planning,not units of ministries implementation

Raise salaries of civil ImplementPublic Sector Implementuniform salary service staff Management (PSM) Scheme scale for counterpart as planned in Economic financedpositions, as ManagementStrengthening agreed among donors. Operation

Identifykey positions in poverty alleviation to benefit from scheme

Establishgoals of PSM fund

Reduce number of public Set limits on admini- sector employees. strative staff's share of all positions

Limit politicalpatro- Establishcivil service Agree on limitednumber nage career path for profes- of politicalappointees sional development (e.g., down to the level of Sub-secretary)

Reach agreementamong politicalparties.

Increase term for local elected officials

143 Lo,ntr Term Ob1 ecelves um Term Oh ect:ves Short Term Gb Lecti e

Donor CoordL=aion Improve donor coordina- Strengthen monitoring All donors should work tion and planning role of Mi- through HINPLAN. nistry of Planning Improve inter-donorcon- Establishlead agency in sultations each sector Establishsub-groups for each sector chairedby lead agency

Continue regular donor coordinationmeetings in La Paz, maintain Consul- tative Group process

Health Policies

Implementexisting good Review investmentpro- ImplementMaternal policies gram to make it conform Health and Child Survi- to policies val Program (MHCSP)and the IntegratedHealth Model in Cochabamba, Santa Cruz, La Paz, and El Alto

Emphasize primary health Focus on mothers and ImplementMHCSP care children.

Shift expenditures to- Assess financing of new wards primaryhealth hospitals. E.g., scale cars. back planned Bracamonte hospital in Potosi

Involve communitiesin ImplementSocial Invest- planimng and design of ment Fund Program facilities within uni- form standards

Increase role and coor- Coordinateactivities of Review agreements(con- dination of NGOs NGOs venLos) with NGOs to en- sure better coverage of country

144 Ter7 O?blectlves ,Medlum Term Ob fectives ,Short Term Ob lect

Improve local integra- Let NGO networks par- tion of NGO activities ticipate in sectoralDe- partmental Technical Councils.

Integrate traditional Use tested traditional Train doctors and nurses and modern medical sys- proceduresin modern me- in the use of herbal me- tems dicine dicines

Integratetraditional Train midwives and com- practitionersinto munity leaders in pre- health care system natal care, nutrition etc.

Improve equity of spend- Focus expendituresin Shift expendituresto Ing rural and peri-urban better reflectpopula- areas. tion distribution between urban and rural areas.

Focus on maintenanceand recurrentcosts.

Develop budget for main- tenance and recurrent costs based on existing infrastructure

Develop database to as- Carry out an inventory sess needs and resources of existing of each area infrastructure and staff.

Support SIF's mapping of poverty areas.

Women's Issues Diminish abortionpro- Implementagreements of (above and beyond those blem recent conferenceon mentionedabove) abortion.

Develop reproductive Assess NGO and church educationprograms, sen- experiencein the field sitive to cultural and religiousbeliefs

145 Lon# Term Oblectives Medlum Term ObJectives Short Term Obectilves

Nutrition Monitor and coordinate Coordinatenational nutritioninterventions nutritionsurveillance system (SVEN)with effort by National Instituteof Nutrition, establishnorms in Mini- stry of Health.

Expand iodine program; Includeblock salt in developprogram to alle- the iodine program viate iron deficiency

Assess community-based Asbess NGO experiences nutritionprograms, agree on standard limits on length of program and substitutionof local foods for donated foods Educatin

Improve equity of spend- Increase share of expen- Limit universityspend- ing dituresfor basic educa- ing to current share of tion educationspending.

Increase spendingon ma- Implementefficient pro- terialsand other non- grams in SIF to increase salary items educationsector invest- ment

Unify rural and urban Improverural education Implementmulti-grade systems. system teaching accompanied with improved teacher salaries in rural areas.

Test pilot projects of rural boarding schools for dispersedpopula- tions.

Introducepilot schemes for multi-cultural teaching

Implementregionalized school year adaptingto agriculturalcalendar

Improveefficiency of Improve teachers'sala- Remove non-teachingtea- system ries chers from payroll 146 Long Tre OblJectiv Medium Tem Obctlves e ObI

Introducebonuses for more hours worked, more students taught, train- ing received,managerial positions

Abolish seniorityas on- ly base Zor pay increa- ses

Improvequality of tea- Establishtraining pro- chers grams for teachers.

Implementregular test- Ing of all teachers.

Shift resources towards Introduce bidding for materials production of text books to reduce cost. Use NGO experiencewith dif- ferent text books.

Improveallocation of Develop national infor- teachingresources mation system indicating schools,number of stu- dents and number of tea- chers.

Increase community par- Involve parents and tea- ticipation chers in mnJagement of schools. Create elected council of parents for each school. Give council supervisory power.

Allow communityto "con- Approve more applica- tract' out its school tions for having school system to NGOs. districtsmanaged by NGOs with good track re- cords hagiculture

Provide poor communities Base investmentson ex- Transfer ESF experience with infrastructure pressed needs of commu- to regional governments nity and Regional Development Fund

147 ont Term Oblectlves IaShort TermObfectlves

Do not require full cost recovery of investments

Improve planning and po- Refocus Ministry toward Make planning department licy setting the need of the farming more powerful community

Involve local rural or- Implementadvisory coun- ganizationsin planning cils in rural areas with private sector partici- pation

Improve titling Recognizecommunal land- Move to approve already holdings as legitimate developedrequest for claims land titles of in- digenous group in the Lowlands.

Assess other claims by indigenousgroups on a case by case basis, dev- elopinggeneral methodo- logy in the process

Streamlinesystem of ob- Introducereform plans tainingtitles for approval

Improve irrigation Design projectswith Develop projectbank of communities,taking wo- well functioningproject men's roles into ac- ideas count. Implementrecommenda- tions of recent donor- supportedreport on small scale irrigation

Improve technologydeve- ReorganizeIBTA Focus on fewer crops lopment and transfer Improve incentives for IBTA staff

Contract extension out Assess ongoing extension to NGOs, RDCs, or produ- programs cers' associations

Improve agricultural Assess and improvePea- Support sustainable credit sant Development Fund private and NGO (FDC) includinga initiatives

148 ILn Term Obfecives edim Te0rm OblectIveas Short er Ob1ectyes

revisionof interest rates Liquidateor privatize BolivianAgricultural Bank

Link poor with market- Identifysuitable pro- able surplus to market ducts and necessarypro- cessing

Identify small scale agro-industrialenter- prises

Establishprivate/public Study other countries' foundationto identify experiencein this and fund such enter- prises

Endorse community-based, Developprograms allow- Assess existingexpe- voluntarymigration ing whole communitiesto riences and recommend migrate to areas of sus- actions for improve- tainableagriculture in ments. the lowlands

TransDortation

Integrateurban markets Increaseemphasis on Develop projectbank of with nearby rural produ- maintenanceof network, rural roads projects cers rural roads, and feeder roads

Scale down existing airportprojects. Stop building subsidized airports for the next 5 years

Decentralize management Let RDCs learn from ESF of roads experience in road upgrading. Efficiency

Improve efficiencyof Developappropriate For water, use UNICEF investmentsbenefitting technologicalpackage guidelines; the poor

149 Long TermaOblectives Sim n Short Term Oblectlyes

For roads, involvecom- munity and use ESF guidelines; For health and educa- tion, focus on using existing infrastructure more efficiently;

For irrigation,add ex- tension and technology transfer to projects.

Involve community in de- Require community to sign and implementation provide labor or local materials.

150 151

ANNE I Page 1 of 5

Ideas for Poverty-FocusedProjects and Studies

1. Multisector

* Institutionaldevelopment of centralministries--provide MINPLAN with technicalassistance to strengthenits capacity in macroeconomicmanagement, budgeting and monitoringof public expenditures,public investmentplanning, and coordinationof external assistance;support should be provided to Subsecretariat for Social Policy in developinga well-articulatedsocial developmentstrategy, translating it into practicalactions, and mobilizingdonor resources;

* Support to Public Sector ManagementProgram--provide the Governmentwith technicalassistance to implementa civil service reform; efforts should be made in rationalizingthe existing uneven wage structureto attract qualifiedpersonnel to the public sector;

• Support to decentralization--provideboth centralministries and regionalauthorities with technicalassistance to strengthentheir administrative,financial, and technicalcapabilities in line with renewed responsibilitiesunder decentralization;central ministriesshould be streamlinedc.nd reinforced to focus on policy planning and prioritization,while delegatingimplementation functionsto other local agencies;special attentionshould be paid to the existingregional disparitiesin institutional capacity;

2. Agriculture

o Institutionaldevelopment of MACA and related agencies--provide MACA and related agencieswith technicalassistance to strengthen their capacity in policy formulation,public investmentplanning, and project coordination;for the poor regions, the applied research system (i.e.,IBTA) should be upgraded to improve the technologicalbase for promotinggrowth;

* A study of land titling--developa methodologyfor land titling, based on a comprehensivestudy of the present land use and tenure situation;the study should also assess viabilityof a land consolidationprogram in Altiplanoand Valleys and a resettlement policy to serve as a basis for rationalizationof the present system;

* A study of a possibilityof developmentfor small farmers in the Highlands--assessnatural resources,potential products (cropsand livestock),technologies (irrigation and productionsystems), supportingservices (credit,marketing, and transport),and socio 152

Page 2 of 5

-culturalaspects to determinea developmentstrategy for poor farmers. Analyze the feasibilityof migrationto Lowlands;

• Small-scaleintegrated rural developmentin the Highlandsand Valleys--implementwell-focused micro-projects, with such componentsas agriculturalextension, productive infrastructure, rural feeder roads, health and sanitation,education; special attentionshould be paid to encouragecommunity participation in planning and implementation;

* Developmentof micro-irrigationsystems in the Highlandsand Valleys--supportsmall-scale irrigation projects to help the poor produce food for home consumption;projects should take into accountmanagement of land and water resourcesand traditional communitystructures;

* Support to rural credits--providetechnical assistance to rural credit cooperativesin the areas of planningand evaluation, promotionactivities, development of guaranteemechanisms; Peasant DevelopmentFund should be strengthened; o Formationof producers'associations--provide technical assistance to communitiesto establish/strengthencore organizationsto prepare and implementvarious developmentprograms to benefit small farmers;major effortsshould be made in increasing administrativeand financialmanagement capabilities of such associations;

* Trainingprograms--provide local communitieswith trainingon technologyand environmentalproblems (i.e., crop diversification, improved seeds, land use and ecology, rural water systems, credits);training should be undertakenin indigenouslanguages to enhance communityparticipation;

* Trainingprograms for rural women--trainrural women in appropriatetechnology such as handicraftproduction (i.e., design, sewing, knittingand marketing)and animal husbandryto increasetheir income-generatingcapacity;

* Developmentof community-basedagricultural enterprises- -provide technicalassistance and infrastructuralsupport to communally. managed agriculturalenterprises; a phased approach (startingfrom pilot schemes) should be taken, and assistanceshould cover identificationof profitableproducts, examination of the production,processing, marketing mechanisms and commercial viability,provision of infrastructure,and servicesneeded; 153

ANNEX X Page 3 of 5

3. Transport

* Institutionaldevelopment of SENAC--provideSENAC with technical assistanceto strengthenits capacity in policy formulation, planning,monitoring and evaluation,and project coordination; technicalassistance should be provided to streamlineits funictionsand promote delegationof regionalplanning to decentralizedagencies;

* Developmentof rural roads--improveroad infrastructureof rural communitiesthrough cost-effective construction of rural feeder roads (i.e.,use of local materials,participation of local labor); attentionshould be paid to make such infrastructural supportconsistent with regional/sub-regionaldevelopment objectives;

* Naintenanceof existingroads--improve the existingroad network (primary,secondary, and rural) to improve communities'links to markets;

4. Health

* Institutionaldevelopment of MPSSP--providetechnical assistance to MPSSP to strengthenits capacity in planning,financial management,budgeting and human resource developmentto improve effectivenessof the public health system. Decentralizationof planning and budget control should be promoted;an inventoryof health facilitiesand personnelshould be developed;

* Primaryhealth care programs in rural areas--expandprimary health care services in rural areas, targetingmothers and infants; services should cover immunizationcampaigns, oral rehydration, and nutritionprograms;

* Child nutritionprograms--child nutrition centers should be supportedto prepare and distributemeals to childrenwith severe malnutrition;such nutritionprograms need to include improved targetingof beneficiarygroups, increaseduse of local foodstuffs,and stimulationof local food production;

* Provisionof health infrastructurein rural areas--construct, repair, and expand health posts in rural areas, with provisionof basic equipments,supplies and iedicines;location of health posts should be decidedbased on cou.rehensiveassessment of community needs in terms of their health status and accessibilityof basic services; 154

Page 4 of 5

* Provisionof sanitationfacilities--construct basic sewage dispo- posal infrastructure,including small-scale sewerage systems or latrines in communitiesin marginal urban and rural areas sufferingfrom high incidenceof contagiousgastro-intestinal diseases;

• Training of health professionals--traindoctors and nurses in applicationand integrationof local medical practicesand traditionalmedicines into modern health care; auxiliarynurses and traditionalmidwives shouldbe trained in prenatal care and nutrition.

* Health trainingprograms for communitymembers--under leadership of health promoters,provide community members with basic education in health,hygiene, and nutrition;

5. Education

* Policy dialogueto promotea reform in the primary educational system--assistMEC in embarkingon a major educationalreform, including:(a) shiftingeducation expenditures toward primary education; (b) improvingefficiency of allocationof human and material resources (i.e., increasedteachers' salaries, emphasis on materials and supplies);and (c) uniting the urban and rural systems (i.e.,administration, teacher classificationsystem, etc.). Donors would be able to link such reform to lendingopera- tions. In the unificationof rural and urban systems, a phased approach would be recommended,starting from improvementin rural primary educationto induce teachers'incentive to serve in rural areas;

* Institutionaldevelopment of MEC--provideMEC with technical assistanceto strengthenits capacity in policy planning,finan- cial management,information, and manpower planning to realize more effectiveallocation of human and financialresource; decen- tralizationof planning and budget control should be promoted;a comprehensiveinventory system should be establishedto make an accurate needs assessment;

* Provisionof educationalinfrastructure in isolated rural areas-- construct,expand, or repair primary school buildings in isolated rural areas, with provisionof equipment,furniture, and teaching materials;

* Literacy trainingprograms--support training programs for adults with less than secondaryeducation; special attentionshould be paid to improvingthe female literacyrate; these programs should have technical/vocationaland health/nutritioncomponents; 155

ANNIE Page 5 of 5

* Interculturaland bilingualeducation programs--develop and imple- ment bilingualeducation for the Quechua, Aymara, and Guarani populationsin indigenousareas; primary school curriculumshould be developedto reflect the country'scultural and linguistic diversity;

* Nultigrade school programs in rural areas--supportmultigrade schools in rural areas with low populationdensity; teachers should be provided in-servicetraining on multigrademethodology, and didacticmaterials and guides be produced and distributed;

* Girls' scholarshipprograms--develop pilot project for a scholarshipprogram for rural girls to increase their educational opportunities;programs could includesupport to both primary and technicalschools; 156 ANNEX II VARIOAPOVERTY PA" OF SOL'IVIA Page 1 of 5

Title Indicators Nethodology Results

RA. ort., '904MrllO y * 19 eI ciO-economic * For each of the 19 Indicators, the * The 21 poorest provinces Pabree an 96elWI," (1091) Indicators (health, MxiaD value of Ott provinces is are concentrated in Pandoa eiction, housirn, used as steranrd. ChuquisacaCepts. lotomnt, a * The poverty level of eachprovince duigraphy) are Is calculated based on the deviation selected. from this standard value. * Orifinal source: 1976 Natfonal Census...... UNICEF,MExtrm Pdor on * ICGV(irncx for * ICNVis caleulated by taking * The 20 poorest provinces btivila, 3 (IWS) quslity of tife) difference betweenmaxfns and are concentrated in Is calculated as minimiun values of indicators. Chuquisaca,Potos1, & major powerty * The poverty level Is determined Coch Depts. Indicea. beasedon the deviation from this ICIV. * IOEVfI basedon: e On the margin, other infant ortality. indicators (housing, migration, ltfe expetancy, & sanitation, etc.) are used. illiteracy rates. * Originat source: 1976 National Censs.

5..._.__._._._...... _.___...... _...... IP, Uroustas pareU*s - Infant mortality & * To rank the poverty tevel, the * The 10 poorest provinces estraee d _earrele de illiteracy rates In ftolowing critical tines are used: are concentrated in Chuquiswa Ses In (191) tord areas are used (a) Infant mortalfty rate--below 150, 6 PotosSDe.ts. a major poverty 150-179, 180 or above; Cb) illtiteracy fndices, rate -501. * Origirna source Is * Percentageof rural populatfon is tha 1976 National also used to supplementthe analysfs. Census,but dat are mdffled by including estimated demographic hdumnesIn 19?6-85. .__...... _...... UDW, 'qgp de to Extr_ * NoIsing conditions * Nousig conditfons are groupedinto * The poorest provinces Pebezanem tUiv$e.8(19) (i.e., san1itry four cateorfes, In term of are identiffed for both facilities) re used vaflbility of water, sanitary urban end rural ares. as the vajor poverty facilities and housing type. * Urban areas: l.Ci.rtf & indietor. * on the mrgin, avaflability of Yaeparaez(CHU), Aberda * Original source: electriefty IS elso used to rank the (ORU),erd Chayanta(PTS). 1976Nationsl Census.poverty level of each province. * Rural areas: G.Romena Padre de Dios (PAN), Itene2 (BE), and Iturralde (1PZ)...... OAM"O,Mltrsn etbre en * Infant mrtality * National average of each Inrdicator * The 21 poorest provinces SlIvie," (INS rate, mgration, is used as reference value, are concentrated in illiteracy rate, a * The poverty level is determined Chuquisacs,La Paz, houing conftdlons b the deviatfon from this reference Cochabrbe a Potosi Dapts. (aaltability of value. drinking water A lectricity) re ued as the major poerty indicators. * Origin source: 1976 Natiorat Census. 157 ANNEX II Page 2 of 3

IBRD 21869

BOLIVIA DISTRIBUTIONOF THE POORESTPROVINCES

* PoorestProvinces A DE ; I ~~~~~~ProvinceBoundaies B 7 88 -RAZIL IntemationalBoundaries

PERU BRAZIL 25 93)

0 71>vv10 294

Ki o et r ______>__ _

4~~~~~~~~~~~~~~~~~~ETME 1198

PARAGUAYIN

0 1CO 200 ar92ta vSPMR8 158 ANNEX II Page 3 of 5

TEM POOREST PROVINCES IN BOLIVIA

Order Number Map) Province DeRartment

1 34 Cochabamba 2 39 Tapacari Cochabamba 3 59 A. Ibanez Potosi 4 58 Nor Chichas Potosi 5 62 Sud Lipez Potosi 6 07 Nor Cinti Chuquisaca 7 41 Mizque Cochabamba 6 56 Chayanta Potosi 9 31 Cochabamba 10 57 Charcas Potosi 11 32 E. Arce Cochabamba 12 53 t. Frias Potosi 13 01 Oropeza Chuquisaca 14 55 C. Saavedra Potosi 15 03 Zudanez Chuquisaca 16 54 Bustillos Potosi 17 09 Sud Cinti Chuquisaca 18 65 G. Bilbao Potosi 19 72 Mendez Tarija 20 08 B. Boeto Chuquisaca 21 49 P. Dalence Oruro 22 06 Yamparaez Chuquisaca 23 04 Tomina Chuquisaca 24 02 Azurdy Chuquisaca 25 98 Abuna Pando 26 99 Gral. F. Roman Pando 27 50 L. Cabrera Oruro 28 05 H. Siles Chuquisaca 29 97 Hadre de Dios Pando 30 96 Manuripi Pando 159 ANNEX II Page 4 of 5

IBRD22289

BOLIVIA DISTRIBUTIONOF PROVINCESACCORDING TO LEVELOF POVERTY

L~LQF POVFRTYIBY PROpVlNC:ES

L Level2 ,_- DepartmentBoundaries 3

96 - IntermationalBoundaries \\ ve 1

PERU = evelLet2 s5 25 ~~~~~~~~~~~~~~~Areaswithin Boliviaonly.

CHILE , V~~~~~BRZI

' 7 ~Areaof Msap

oGmtr010,0 200 300.;a BLV MllesX [ ':f F_X- 0 100 200;-<

Source: CONAPO(1988) APRILt990 160 CODWICACIONGEOGRAFICA DI 9OLSvxA ANNEXII Page 5 of 5

DEPARTAMENTO DI DEPARTAMENTODg DIPARTAMZNTO DI CHUQUISACA oRsuto SANTA CRUZ 01 OropZas 48 C.ond@ 74 Audrie tIbt:z 02 AVa?duV 44 Avuo 70 Wames 03 ZudiAo: 45 CazaIU 76 VoIio 04 Tomna 4 Sajam, 77'eo 06 Hmando 82.0 47 Utoal 76 Chlqitos 06 Yamp&?Aoz 46 PooP6 7T sab 07 NoT Cinti 49 Pantak6a Da10nce 60 Cordars Os 36104o Boeto 00 Ladssl.o Cabreoa 61 Vallerande 09 S&d Cinti SI Atabuatipa 82 Florida 0o TruisCalvo 52 Smear 83 Obiao Saati.otban 84 *Uflo de CbLvo: 65 Ansel Sandoval 86 ManuelMaria Cab. Unro DEPARTAMENTO DE DEPARTAMENTO DR DZPARTAMVNTODEL LA PAZ POTOSI BENI 11 Murllo 5I luaF 87 Cbucdo 12 Omaauoo 14 Bwtlos 8J Vau Dis 18 Pacales 5b Corndlo Saavedra 89 Orel. BamvlS 14 Camcho a Chayant 90 Yacua, is MUAOCaa 57 Chawas 91 Mo:. 16 LAecala 58 Not Cbieba 92 muAh 17 Franz Tamao G9 Alono de IbfiSo 98 MaMoAt is Inavi 60 Sud Cbieba 94 t6aoZ 19 Loayxa 61 No, Lip. 20 Iquidvi 62 Oud Lives DEPARTAMENTO DZ 21 Sud Yg6 QWo PANDO 22 Loa Andes Qo 23 Asoma 65 Gul. lb.ao 24 Not Yunsas 66 DanAelCampo 95 Nicoldssuito 25 Abel Itumlde 67 ModestoOUnMl 96 manur 26 BSautif Suvedaa 97 Mad" de Dios 27 Manco iCapo 98 Abum 26 Gualberto VWnoel 99 GOl. F. buom

DEPARTAMENTODI DEPARTAMENTODE COCHABAMBA TARIJA

29 Cewcado 68 Ctoedo 30 Campero 69 Am 1 CHUQUISACA 31 Ayopaya 70 GOn Chao 2 LA PAZ 32 Eaoban Arco 71 Avlez 3 COCHABAMBA 33 Anni 72 M6nd* 4 ORaUO 34 Arque 780 Consc 6 POTOSI 35 S TARUA 36 Jodin 7 SANTA CRUZ 37 QuWaoflo 8 BENI 39 Chbape 9 PANDO 39 Tapacat 40 Coamo 41 MIXQu* 42 Punats 161 ANNEX II

BIBLIOGRAPHY

Bank Documents:

Background Papers:

Cisneros, Rodrigo, The Importance of Rural Infrastructurefor Poverty Alleviation (in Spanish), January 1990

Cock, J. and Forno, D., Poverty Report: Agriculture, March 19, 1990

Cowi Consult, Poverty in the Lowlands of Bolivia, February 1990

Hedlund, Malene, Annotated Bibliogranhv on Poverty in Bolivia, January 1990

McGinnis, Linda, InstitutionalAsgects of Poverty Alleviation, January 1990

Mejia, Christina, Non-GovernmentalInstitutions in Core Poverty Areas of Bolivia (in Spanish), January 1990

Newlon, Dan, Education Reform and Poverty, February 23, 1990

Ohno, Izumi, The Government's Strategy and Investment Program and Poverty Alleviation, October 6, 1989

Ohno, Izumi, Donor Agtivities in Poverty Alleviation, March 1990

Webb, Anna, Women and Rural Poverty in Bolivia, January 24, 1990

Other Bank Documents on Bolivia:

Memorandum on Financing Requirements and Public Investment (BO 86-4), December 1986

Updating Economic Memorandum (Report No. 6455-BO), December 1986

Transport Sector Strategy Paper (Report No. 6882-BO), July 1987

Population, Health and Nutrition Sector Memorandum (Report No. 6965-BO), February 1988

Regional Development Strategy for the Eastern Lowlands (Report No. 7158- BO), April 1988

Updating Economic Memorandum (Report No. 7278-BO), June 1988

A Review of the Public Investment Program and Financing Requirements, 1987-90 (Report No. 7248-BO), June 1988 162 AMNNX IIr

Export Corridors (Report No. 7298-BO), February 1989

Staff Appraisal Report, IntegratedHealth Development Project (Report No. 8001-BO), December 20, 1989

Staff Appraisal Report, Social Investment Fund Project (Report No. 8248- BO), March 26, 1990

Country Economic Memorandum (Report No. 7645-BO), September 15, 1989

Public Sector ExpenditureReview With A Special Emphasis on the Social Sectors (Report No. 7746-BO), September 15, 1989

Other Bank Reports:

World Development Report 1990, Poverty (Report No. 8509), April 1990

Social Spending in Latin America: The Story of the 1980s (Report No. 8450-IAC),March 1989 -- Yellow Cover

Other Documents:

ACLO, Un Sistema de Promocion Liberadora, 1981

Arteaga, Vivian, La Muier Pobre En La Crisis Economica, 1988.

Asamblea Permanente de los Derechos Humanos, Estudio Sobre el Valor Adquisitivo del Salario d,e los Mineros, 1978.

Blanes, Jose, Asnectos Sociales en el DesarrolloRegional de Chuquisaca, 1987

Carafa, Carlos, Luz v Sombra de la Vida: Mortalidady Fecunidad en Bolivia, UNFPA, 1983

CEA.C,Estudio de Diamnosticodel Estado Nutricional de la Poblacion del Proyecto Materno-Infantil,Montero, 1988

CEDLA, Informalidade Ilegalidad:Una Falsa Identidad, 1989

CEDLA, El Sector InLformalUrbano en Bolivia, 1988

CEDLA, E13';ectorInformal en Bolivia

Centro de Estudios del Trabajo (CET), La Relocalizacion,1988

CET, Sector Informal v Novimiento Obrero, 1987

CET, iMercadode Trabajo, 1987 163 AM= III

Centre de Estudios de la Realidad Economica y Social, Catalogo ediciones CERES

Flores, Gonzalo/C.E.S.Y.M.,Los Proyectos de Desarrollo Rural IntegZado: ExDeriencias del Ouinguenio 1984-1989, 1989

Chudy, John Paul; Jones, James; Kraljevic, Ivo; Lubina, Allan, An Evaluation of the Bolivia PL 480 Title II Food for Work Program, 1988

Daza, Giovanni; Sev the Children: Estudio Nutricional de la Provincia Inguisivi. con enfasis en la deficiencia de Vitamina A, 1987 de Canedo, Teresa Delfin; de Baizeras, Susana Donoso; Provecto Proandes: Diagnostico SocioeconomicoRegion Interandina Central: Norte de Potosi X Sur de Cochabamba, 1989 de Cordoba, Miguel, Urioste Fernandez, El Estado Anticampesino, 1984 del Prado, Arturo Nur-!z,Economias de Viabilidad Dificil: Una Opcion por Examinar, 1988

Dornbusch, Rudiger, Short-Term Macroeconomic Policies for Stabilization and Growth, 1989

EMSO-MACA, Incentivos a la Exportacion, 1989

EMSO-MACA, Provectos Agrooecuarios, 1989

E14SO-MACA,Recursos Humanos y CaRacitacion, 1989

ENSO-MACA/RossM.A./Eduardo Alfaro, Honitoria y Evaluacion, 1989

Estes, Valerie Anne, Factories and Families: Urban Working Women in La Paz, 1984

Finot, Ivan, EvaluacionMacroeconomica de Politicas de Necesidades Basicas, 1989

Finot, Ivan, Redistribuciondel Ingreso Y Necesidades Basicas, 1989

FSE, Diagnostico de la CoXwntura Economica X Socisl. en Bolivia', 1987

IBRD/IDA/WHO,Bolivia: Water SuaRly and Sewerage Sector Study (Volumes I & II), 1974

INE/UNFPA/DTCD,de Baixares, Susana Donoso, Encuesta Nacional dte Poblacion y Vivienda, 1988

ISNAR, Fortalecimientodel Sistema de Investigacionv Transferencla de Tecnologia Agronecuaria en Bolivia, 1989 164 AMEX II

Laure, Joseph, Los Campesinos y La Crisis, 1988

Lazo, Manuel Lajo, SEMINARIO: Necesiades Basicas v Desarrollo, 1989

Lenz, Tomas, Una Imagen Campesina del Extremo Norte de Potosi, 1988

Machicado, Carlos, Estudio Diagnostico Debate,

Mann, Arthur J., The Political Economy of Tax Reform in Bolivia, 1988

Ministerio de Asuntos Campesinosy Agropecuarios, Tipologia de la Economia en Bolivia, 1988

McFarren, Wendy, Economic Crisis and the Politics of Adaption: Survival Strategies of Displaced Bolivian Mining Households, 1989

MINPLAN/MACA/PNUD/FIDA,Estrategia Rural de Desarrollo de Base Campesina, 1987

Montero, Vivian Arteaga, Recesion Economica. Estrategias de Vida y el Rol de la Muier en Bolivia, 1988

Morales, Juan Antonio, Impacto de los Aiustes Estructurales en la Produccion y los Niveles de Vida de los Campesinos de Bolivia, 1989

ODI; Addison, Tony; Demery, Lionel, Macro-Economic Stabilization.Income Distribution and Poverty: A Preliminay ]Study

PNUD, Documento Tecnico Y Declaracion Re ional Sobre la Pobreza, 1988

PREALC, EmDleos de Emergencia, 1988

Proandes, Plan de Desarrollo Alternativo 1990-1994. Volumen III Provincia Esteban Arze, 1989

Proandes, Plan de Desarrollo Alternativo 1990-1994 Volumen IV Provincia Mizgue, 1989

Proandes, Plan de Desarrollo Alternativo 1990-1994. Volumen VI Provincia Bolivar, 1989

Proandes, Plan de Desarrolllo Alternativo 1990-1994. Volumen V Provincia Tapacari, 1989

Proandes, Plan 0terativo 1990-1994 Propuesta, 1989

Proandes, Plan Operativo 1990-1994 Propuesta. Volumen II. Desarrollo Social, 1989

Proand&s, Plan Operativo 1990-1994 Propuesta.VolumenIII, Desarrollo Agrouecuario, 1989 165 ANNEX III

Proandes, Plan Operativo 1990-1994 Propuesta. Volumen IV. Desarrollo Pecuario v de Servicios, 1989

Programa Nacional de Desarrollo Porcino, Resumen Eiecutivo y Consolidado i ~~~~Naci-onal,1989

The Resources for Child Health Project, SuDervivencia Infantil en Bolivia: Situacion Actual y Prioridades Dara la Accion, 1987

Revista de la Universidad Mayor de San Andres, Contacto 27, 1988

Sandoval, Godofredo Z., Organizaciones no Gubernamentalesde Desarrollo, 1987

Seminario Taller, Lucha Contra el Aborto, 1989

UNDP, The Impact of Macroeconomic Policies on the Rural Poor, 1989

UNFPA, Salto al Futuro, 1984

UNFPA, Migraciones y Ocupacion del Territorio, 1985

UNICEF, Muier. Clase y discriminacion Social, 1960

UNICEF, Los Barrios PoRulares, 1985

UNICEF, Grupos Postergados de Bolivia, 1986

UNITAS, El Alto Desde El Alto, 1988

Urioste, Miguel, Segunda Reforma Agraria, 1988

Velasco, Antonio Peres; Sainz, Roberto Casanovas; de Pabon, Silvia Escobar; Cordova, Hernando Larrazabal, Informalidade Ile&alidad: Una Falsa Identidad, 1989