Honduras - First and Second Social Investment Fund Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13839-HO PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECTS I AND II (CREDITS 2212-HO AND 2401-HO) DECEMBER 30, 1994 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. ACRONYMS BMI - Bono Materno Infantil (Maternal Child Coupon Program) BMJF - Bono Mujer Jefe de Familia (Woman Head of Household Coupon Program) EDSAC - Energy Sector Adjustment Credit FHIS - Fondo Hondurefio de Inversi6n Social (Honduran Social Investment Fund) GNP - Gross National Product GOH - Government of Honduras ID - Institutional Development KfW - German Credit Institute for Reconstruction LSMS - Living Standards Measurement Surveys MHS - Honduras Multiple Purpose Permanent Household Survey MIS - Management Information System MOE - Ministry of Education MOH - Ministry of Health NGO - Non-Governmental Organization PASI - Programa de Asistencia al Sector Informal (Informal Sector Credit Program) PCR - Project Completion Report PRAF - Programa de Asignaci6n Familiar (Family Assistance Program) PROCATMER - Programa Credito y Asistencia Tecnica Microempresa Rural (Rural Micro-Enterprise Technical Assistance and Credit Program) RUTA - Regional Unit of Technical Assistance SAR - Staff Appraisal Report SECPLAN - Planning Ministry SIF - Social Investment Fund Project FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Dieator-Genwal Operadon Evaluation December 30, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Honduras Social Investment Fund Projects I and II (Credits 2212-HO and 2401-HO) Attached is the Performance Audit Report on Honduras- Social Investment Fund Projects I and II (Credits 2212-HO and 2401-HO) prepared by the Operations Evaluation Department. This report presents an audit of two IDA credits that provided support for several social programs, the most important of which is the Honduras Social Investment Fund. Started as an emergency program, the Fund, today, is best described as an institution that manages the process of allocating funds received to a large number of small projects meant to help the poor. In the typical case, municipalities prepare and propose projects which are executed by NGOs and private contractors under supervision by the Fund. Projects are approved if they meet conditions such as availability of a budget for each municipality based on a poverty map, type of project, employment generated, and commitment for future operations and maintenance. Implementation has been rapid and smooth, so much so that it has attracted growing donor support. There has, however, been less participation by local communities than is desirable, a shortcoming that the current management of the Fund is attempting to correct The project outcome is rated as satisfactory, institutional development as substantial, and sustainability as likely on the assumption that lessons learnt will induce significant changes in Fund practices to involve local communities. The Audit attributes the success of this Fund to the establishment of a special institution with carefully defined, monitorable goals and more latitude than typical line ministries have to determine how to achieve them; staff and leadership of high professional quality who are hard-working and well motivated; personal involvement of the President who has used his powers to promote and protect the organization; and substantial oversight by donors who have insisted on high performance standards and substantial openness and transparency. The Audit also warns against the tendency to ask the Fund to undertake tasks outside its primary area of expertise, the tendency of donors to allocate an increasing portion of their assistance to the Fund at the expense of line ministries, and the tendency by the Bank and other donors to attempt to establish similar institutions in other countries without careful consideration of the details necessary for success. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECTS I AND II (CREDITS 2212-HO AND 2401-HO) TABLE OF CONTENTS Page No. PREFACE ....................................................... i BASIC DATA SHEET ............................................... iii EVALUATION SUMMARY .......................................... vii 1. INTRODUCTION ............................................... 1 2. BACKGROUND ................................................ 1 3. THE PROJECTS ................................................ 3 4. FHIS ......................................................... 4 A. Design and Operating Procedures ............................... 4 B. Sources and Uses of Funds and Level of Activity..................... 6 C. R esults. ................................................... 6 5. OTHER COMPONENTS ......................................... 21 A. Family Assistance Program (PRAF) ............................. 21 B. Living Standards Measurement Surveys (LSMS) ................... 22 C. Social Sector Restructuring Program ............................ 23 D. Technical Assistance for an NGO Liaison Office ................... 23 6. ROLE OF THE BANK .......................................... 23 7. PARTICIPANT ASSESSMENT ..................................... 24 8. ISSUES AND LESSONS .......................................... 26 TABLES 1. Sources & Uses of FHIS Funds ...................................... 7 2. Project Implementation 1990-July 1994 ............................... 10 This report was prepared by Ronald G. Ridker. Pilar Barquero provided administrative support. This docuent has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. 3. Units Affected (built or renovated) by FHIS compared to National Inventory, 1990-July 1994 ............................... 11 4. FHIS Expenditures Compared to Government Expenditures and GDP ......... .12 5. Projects and Employment Generated by Sector ........................... 14 6. Investment according to Poverty Map for Municipalities .................... 16 7. FHIS Staff, 1990-August 1994 ...................................... 16 FIGURES 1. FHIS Accumulated Investments and Projects Approved ..................... 8 2. FHIS Disbursements: Investments and Adm. Costs ......................... 9 LIST OF REFERENCES ............................................ 29 1 PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECTS I AND II (CREDITS 2212-HO AND 2401-HO) PREFACE This is a Performance Audit Report (PAR) of the Social Investment Fund Project in Honduras. The project has been supported by two Bank credits, Credit 2212-HO for SDR 14.3 million (US$20 million equivalent) approved on February 28, 1991 and Credit 2401-HO for SDR 7.4 (US$10.2 million) approved July 29, 1992. The first credit closed on March 31, 1994, three months beyond the original schedule, and was fully disbursed by June 13, 1994. Funds from the second credit have been committed and should be fully disbursed by December 31, 1994, the scheduled closing date. This PAR is based on the Staff Appraisal Reports, Loan Agreements and Mid-Term Review documents for these projects, the Project Completion Report for Credit 2212, project files, and government documents (see reference list). Additional information was obtained through discussions with Bank staff, bilateral and other donors, government officials and participants involved in the project, the latter being contacted during a visit to Honduras in August 1994. The cooperation and assistance of these individuals -in particular the staff of FHIS-is gratefully acknowledged. The Audit found that the PCR gave an accurate account of the implementation of Credit 2212-HO and the operation of FHIS up to the spring of 1994. The PAR includes a review of Credit 2401-HO, includes later information on FHIS operations and on its impacts, to the extend available. Following standard OED procedures, copies of the draft PAR were sent to the Borrower for comments on October 26, 1994. No comments were received.  iii PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECT I (CREDIT 2212-HO) BASIC DATA SHEET KEY PROJECT DATA Item Appraisal Actual or Actual as % of Appraisal Expectation Current Estimate Estimate Total Project (US$ Million) 68.00 98.38 145 Credit Amount 20.00 20.00 100 Cancellation 0.03 Disbursed 20.00 19.97 100 Institutional Performance Substantial CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS (SDR Millions) IDA FY FY91 FY92 FY93 FY94 Appraisal Estimates 2.7 8.6 14.2 14.3 Actual Disbursements' 0.0 12.5 13.7 14.3 Actual as % of Appraisal 0 145 96.5 100 a. Undisbursed amount of SDR 18,448.39 was cancelled July 29, 1994. PROJECT DATES Processing Steps Date Planned Date Revised Date Actual Identification (EPS) 05/14/90 05/14/90 Prep./Preappraisal 06/15/90 07/15/90 07/26/90 Appraisal 08/15/90 10/21/90 Credit Negotiations 11/15/90 12/17/90 01/14/91 Board Approval 12/18/90 01/29/91 02/28/91 Credit Signature 01/15/91 03/14/91 Credit Effectiveness 03/30/91 08/15/91 06/26/91 Credit Closing 12/31/93 03/31/94 03/31/94 Credit Completion 06/15/94 06/30/94 iv STAFF INPUTS (STAFFWEEKS) Stage of Project Staffweeks Cycle HQ Field Total Through Appraisal 21.8 55.2 77.0 Appraisal through Board Approval 11.4 0.0 11.4 Supervision 22.0 50.2 72.2 PCR 2.4 6.4 8.8 Total 57.6 111.8 169.4 MISSION DATA Mission Month/year No. of SW in Specialty Performance Types of Persons Fiel Rating Problems Identification 05/90 5 5.0 A,C(2),D,E N/A Prep./Preappraisal 07/90 6 12.0 A,B,C(2),D.F N/A Appraisal 10/90 8 38.2 A(2),B, N/A C(3),D,H Supervision 1 04/91 1 1.0 C 1 N/A Supervision 2 06/91 1 1.0 A 1 N/A Supervision 3 07/91 2 4.1 A,G 1 N/A Supervision 4 11/91 6 30.0 A,B(2),D,E.G 1 N/A Supervision 5 02/92 2 2.0 A,C 1 N/A Supervision 6 11/92 4 12.0 A,B,E,G 1 N/A Supervision 7 07/93 S 1 1.0 I 1 N/A Supervision 8 10/93 ' 1 1.5 B 1 N/A Project Completion 02/94 2 4.0 B,E 1 N/A Total Supervision 56.6 a. A = Financial Analyst; B Human Resources Specialist; C = Human Resources Economist; D = Economist; E = Credit Specialist; F = Architect; G = Engineer, H = Lawyer; I = Procurement Analyst. b. 1 = problem free or minor problems; 2 = moderate problems; 3/4 = major problems. c. Jointly with Supervision Mission of Loans 2804-HO and 2401-HO. d. Jointly with Supervision Mission of Loan 2212-HO and Preparation Mission of Basic Education. C. Jointly with Supervision Mission of Loan 2452-HO and Preparation Mission of Basic Education. V PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECT II (CREDIT 2401-HO) BASIC DATA SHEET KEY PROJECT DATA Item Appraisal Actual or Actual as % of Appraisal Expectation Current Estimate Estimate Total Project (US$ Million) 67.5 67.5 100 Credit Amount 10.2 10.2 100 Cancellation Disbursed 10.2 8.06 77.6" Institutional Performance Substantial a. Disbursements as of 09/01/94. As of September 30, 1994, FHIS has committed the entire balance of credit funds assigned to FIllS projects, estimated at USS1.9 million, and full disbursement is expected by December 31, 1994. b. Due to differences in valuation, the total undisbursed+disbursed may not equal the approved amount. CUNIULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS IDA FY FY93 FY94 FY95 Appraisal Estimates 5.5 4.7 - Actual Disbursements* 4.7 3.0 .3 Actual as % of Appraisal 85 64 : DisburScmentL as of 09 01/94. PROJI'CT DATES Processing Date Date Date Steps Planned Revised Actual 1 Mdentification (EPS) 01/22/91 01/22/91 2. Prep./Preappraisal 01/22/91 01/22/91 3. Appraisal 02/20/92 02/20/92 4. Credit Negotiations 05/04/92 05/04/92 5. Board Approval 06/25/92 06/25/92 6. CrcJit Signature 07/29/92 07/29/92 7. Credit Effectiveness 11/25/92 11/25/92 S. Credit Closing 12/31/94 9. Credit Completion Vi STAFF INPUTS (STAFFWEEKS) Stage of Project Staffweeks Cycle HQ Field Total Through Appraisal 3.5 .2 3.7 Appraisal through Board Approval 12.7 1.5 14.2 Supervision 39.9 25.4 65.3 PCR - - - Total' 56.1 27.1 83.2 a. Data as of 06/30/94. MISSION DATA Mission Month/year No. of SW la Speciaky Performance Types of Persons Field Rating Problems Identification 02/92 1 .2 A N/A Prep./Preappraisal Appraisal 03/92 1 1.5 B N/A Supervision 1 03/93' 8 A(3),B,C,D,E,F 1 N/A Supervision 2 07/93' 2 1.0 A,G 1 N/A Supervision 3 10/93* 1 0.3 A 1 N/A Supervision 4 02/94' 1 0.1 A 1 N/A Project Completion Total Supervision 25.4 a. A = Human Resources Specialist; B - Financial Analyst; C - Engineer; D - Administrator; E - Credit Specialist; F = Economist; G - Procurement Analyst. b. 1 = problem free or minor problems; 2 = moderate problems; 3/4 - major problems. c. Jointly with Supervision Mission of Loans 2804-HO and 2452-HO. d. Jointly with Supervision Mission of Loan 2452-HO and 2804-HO. c. Jointly with Supervision Mission of Loan 2212-HO and Preparation Mission of Basic Education. f. Jointly with Supervision Mission of Loan 2452-HO, Project Completion Mission of Loan 2212-HO and Preparation Mission of Basic Education. vii PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECTS I AND II (CREDITS 2212-HO AND 2401-HO) EVALUATION SUMMARY 1. The plight of the poor in Honduras is es- IDA credit of $20 million, cofi-nancing of $33 pecially difficult because of the country's low per million, $8 million from the Government and $7 capita GNP ($580) and extremely unequal distri- million from project benefi-ciaries (individuals, bution of income and wealth. In 1990, the situa- municipalities, NGOs and other entities). In the tion for many in this group became even worse end, substantially more than these amounts were as a consequence of an economic stabili-zation mobilized, with the re-suit that total project cost and adjustment program. To cope with these increased to $98.4 million. consequences, the government launched, at about the same time, two innovative social safety 3. The principal objective of SIF-11 was to net programs, which the IDA helped fund. This provide additional funds for FHIS. Additionally, report reviews this experience and the use that a small amount was set aside to provide techni- was made of two of the IDA credits (Cred-its cal assistance to a NGO liaison office to improve 2212-HO and 2401-HO). NGOs' ability to plan, prepare and implement projects for FIS. The Projects FIS: Rationale, Concept and Operations 2. The objectives of the first project, SIF-I, were to mitigate the social costs of adjustment, 4. The rationale for FIS, originally pat- lay the basis for a targeted food assistance pro- terned after the Bolivia Emergency Social Fund, gram, and improve service delivery in the social was that line ministries were not capable of ra- ministries. The first component, which absorbed pidly and efficiently administering large emer- over 98 percent of the credit, provided support gency employment and social investment pro- for a social investment fund known as FHIS grams, that a new agency with more freedom in (Fondo Hondurefio de Inversi6n Social). The procurement, contracting and hiring of personnel se-cond component funded a pilot scheme to was necessary, at least for a time. It was esta- distri-bute food coupons through primary health blshed by special legislation as a semi-inde- cen-ters to poor pregnant and lactating mothers pendent government agency attached to the and children under the age of five. The Office of the President with a life of four years. remaining two components were to provide At the end of this period, new legislation extend- technical assis-tance to help design a program to ed its life for an additional four years. Today, it improve the efficiency of the Ministries of is best described as an agency that manages the Health and Educa-tion and to establish a process of allocating funds received to a large monitoring and evaluation system for the number of small projects meant to help the government's social programs through a Living poor. In the typical case, municipalities prepare Standards Measurement Study (LSMS). The and propose projects which are executed by total cost of the project was ini-tially estimated NGOs and private contractors under supervision at US$68 million equivalent to be funded by the by HIS. Projects are approved if they meet a viii set of conditions which include a budget for each 7. Projects. Over 5,500 projects have been municipality determined on the basis of a pover- approved and over 4,500 completed. FUIS's re- ty map, type of project, employment generated, habilitation projects have affected 32 percent or and commitment for future operation and main- public schools and 39 percent of public health tenance. For projects of $30,000 or less, which centers. In 1993, MIS expenditures accounted constitute roughly 2/3 of the dollar amount of its for 13.5 percent of public expenditures in the so- projects, FHIS contracts directly without compe- cial sectors. The project numbers also suggest a titive bidding, using as guides data bases of costs speed of operations that is unprecedented in and of past performance by contractors. Honduras. Both statistical data plus personal observations suggest that project quality is rela- 5. FHIS got off to a rapid start. Commit- tively good. ments have grown steadily until about Septem- ber 1993, thanks both to its ability to develop 8. The most promising of these projects are and implement projects and to growing donors some 56 that have provided credits to NGOs and support. Since September 1993, commitments community banks for onlending to micro- have fallen and (up to September 1994) failed to enterprises. Bad debts associated with credit recover, a consequence of uncertainties about projects located in urban areas have averaged whether FHIS would continue beyond the first less than Five percent, a result of careful selec- four years, and difficulties in obtaining bridge tion of NGOs, enforcement of detailed proce- financing until new funds for the second four dures, technical assistance provided when need- years arrived. During the remainder of 1993, ed, and substantial monitoring and supervision. disbursements increased dramatically, a conse- quence of Hurricane Gert and a desire to com- 9. Employment created per project and per plete projects before the end of FHIS-I. Equally dollar of project expenditure is higher than dramatic was the fall in disbursements in January country-wide averages and similar to other Latin and failure to recover during the last six months, American social investment funds, but not as a result of severe shortages of cement and elec- good as it might be because no attempt was tricity as well as the decline in commitment for made to use labor-intensive technologies. new projects. Administrative costs as a percent of total expenditures, which averaged less than 10. Targeting. Despite the use of a poverty ten percent during the first four years, have risen map to allocate available resources geographical- dramatically, a result of the failure to reduce ly, wealthier municipalities received more MIS temporary employment in tandem with the de- funds during the first year, simply because they cline in business. This situation should correct more quickly identified and prepared acceptable itself rapidly once new funding comes on stream project proposals. Concerted efforts to assist in the next few months. poorer municipalities have corrected this imba- lance; but there is still a ways to go before Results poorer areas are significantly favored. There are no good data to indicate what has happened 6. Evidence on results comes from FHIS re- within municipalities, but most projects are of a cords, central statistical data, site visits and dis- type that benet poor more than wealthy fami- cussions with FHIS staff, recipients of FHIS con- lies. tracts and participants in FHIS-supported pro- jects. All sources of information supported each 11. Institutional Development (ID) has been other. In particular, observations of participants substantial. Not only has FIS grown in size, were consistent with the picture derived from complexity and competence, but through its con- more centralized sources of information, tracts and the provision of technical assistance ax when needed it has contributed to the ID of mu- Role of the Bank nicipalities, NGOs and private contractors. Even line ministries have been positively affected by 15. While the Bank played very little role in having to deal with FHIS. the initial design of EIS and was remiss in not devoting more attention to the minor compo- 12. Sustainability. While FHIS was initially nents of SIF-I and II, it devoted considerable considered a temporary organization, a new divi- attention to FIS operations -39 staffweeks per sion of labor seems to be emerging in which the year (FY92, 93, and 94 for the two credits cor- MOE and MOH concentrate on policy planning, bined), two mid-term reviews and numerous norm setting and operations, leaving manage- studies by consultants. These efforts contributed ment of the investment cycle to FHIS. If this at least three ways to the success of FIS. (1) development continues, the government will Out of these investigations the Bank was able to need something like FHIS on a permanent basis make a number of useful suggestions for im- and would fund it, to the extent it could, even if provements in FIS operations. (2) The conti- foreign assistance to FHIS were substantially re- nuous, detailed monitoring, insistence on fre- duced. quent auditing and reporting, and numerous re- quests for detailed information and data on ope- 13. On the other hand, a number of FHIS rations, encouraged a high standard of perform- projects have run into trouble after completion ance and helped protect FHIS staff from de- because of lack of maintenance and sometimes mands for behavior that would have weakened operating budgets. In its first four years, FHIS its reputation. (3) This detailed monitoring gave dealt with this problem in its latrine projects by the Bank a solid basis for providing the seal of adding education and training activities (which approval needed to mobilize other donor re- proved to be effective) and in its water and sew- sources for FIS. age projects by promoting the creation and in- volvement of user committees (which have had Conclusions and Lessons mixed results). In its second four years, FHIS has institutionalized its efforts to promote com- 16. The two IDA credits reviewed in this re- munity participation for sustainability, but action port have been used for activities that demon- programs are still in the formative stage. How- strate that, under the right circumstances, social ever, a more basic problem is the limited finan- investment funds, microenterprise credit pro- cial capacity of poorer municipalities, a problem grams, and food-coupon programs can cost-ef- that is largely out of FHIS's control. fectively assist the poor in important ways. While there may be other ways to create the 14. Other Components. The pilot food cou- proper conditions for the successful operation of pon program demonstrated that this is a cost-ef- these programs, in the case of FHIS four factors fective way to provide food-cum-income supple- need emphasis: ments to the poor. But provisions for terminat- ing these programs, if they are truly meant as Carefully defined, monitorable goals with temporary safety nets, need to be built into them latitude to determine how to achieve at the beginning. The remaining components ac- them. complished very little, in part because too little attention was paid to them. This is unfortunate Quality of staff. Senior staff seem to be since they addressed important issues. Hopeful- professional, hard-working and well moti- ly, this oversight is being corrected in the vated. Good salaries, lack of tenure, Nutrition and Health Project and the currently carefully detailed operating procedures, being negotiated Education Project. and a first executive director who gave X priority to technical competence and ef- and to enhance their contribution to ficiency are probably the most critical project sustainability will have to be factors explaining this result. designed and implemented. Personal involvement of the President * While FIS and the food coupon who used his powers to promote and scheme can ameliorate some of the ef- protect the organization. fects of poverty, they can do little or no- thing about its underlying causes: high * Substantial oversight by donors who in- population growth rates coupled with sisted on a high degree of openness and low economic growth rates plus unequal transparency in FHIS operations. distribution of land and capital, including human capital. So long as these forces 17. It is also important to identify what social dominate, such programs will be needed investment funds like FHIS cannot do and in the on a long-term basis and will be running process sound several cautionary notes. hard just to keep up with ever-expanding needs. * In its present form, FHIS does not deli- ver education and health services; it only * FIS is being pulled in many new direc- provides the construction component of tions-for example, technical assistance, the full package of inputs necessary for training and community mobilization as this purpose. It would require quite a well as project management-which it different organization, to provide the may be far less well-equipped to handle. services themselves. One solution for this is to hive off spe- cialized functions to new FMlS-type or- * Just because of FHIS success, there has ganizations, as is happening with its cre- been a tendency on the part of donors dit programs. Another is to contract out to allocate an increasing portion of their additional functions, as is happening with assistance to FHIS at the expense of line some of its training programs. A third is ministries. This may be leading to ex- to improve the capacity of the line minis- cessive allocations to capital expenditures tries so they can undertake these func- and exacerbating the shortage of funds tions, as is being planned in the Public for operations and maintenance. New Sector Management Project currently approaches to involve local communities being prepared PERFORMANCE AUDIT REPORT HONDURAS SOCIAL INVESTMENT FUND PROJECTS I AND I (CREDITS 2212-HO AND 2401-HO) 1. INTRODUCTION 1.1 This report presents a review of the activities supported by two IDA credits, the Social Investment Fund I and II (Credits 2212-HO and 2401-HO). The focus is on the most important of its components, support for the Fondo Hondurefio de Inversion Social (FHIS) which absorbed over 98 percent of the funds. FHIS was established by special legislation giving it a life of four years starting April 1, 1990. In early 1994 its life was extended by an additional four years. These two phases of FHIS are known in Honduras as FHIS-I and FHIS-II. Credit 2212 was approved in February 1991 and nearly all its funds were committed by the end of 1991, far more rapidly than assumed at appraisal. At that point the Government requested additional IDA funding to continue the activities through March 1994. The second credit was approved in July 1992 and nearly all of its funds committed by the end of FHIS-I. The first credit was closed March 31, 1994, three months later than originally scheduled; the second credit is scheduled to close December 31, 1994. A third credit to fund new projects under FHIS-II is being discussed. 2. BACKGROUND 2.1 Honduras is one of the poorest countries in the Western Hemisphere. In 1992 GNP per capita was US$580; maternal mortality was 221 per 100,000 live births and infant mortality was over 50 per 1,000 live births, with nutritional deficiencies being a major contributing factor in 60 percent of the cases. Nearly 40 percent of children under five and over half of women attending health centers suffer from at least moderate malnutrition. Over 40 percent of all households lack adequate sanitary facilities; over a third are without potable water connections; over 40 percent have inadequate access to primary health care facilities. While primary school enrollment rates indicate good access, educational quality is low and repetition and dropout rates are high. On average, the population has no more than 3.9 years of schooling and roughly a third is classified as illiterate. 2.2 These national averages mask tremendous disparities between families and across regions. In 1989, the wealthiest ten percent of the population received 48 percent of the income while the poorest 20 percent received 2.7 percent Extreme poverty affects over 50 percent of the population nationally and nearly 80 percent of the rural population. In the three worst departments, where doctors, teachers, health clinics and educational facilities are especially scarce, maternal and infant mortality rates are nearly three times the national average.' 1. Data for this and the previous paragraph come from the World Development Indicators for 1994 and the World Bank's Country Economic Memorandum/Poverty Assessment for Honduras, 1994. 2 2.3 The principal reasons for this state of affairs are chronically low rates of economic growth coupled with very high rates of population growth, inequality in ownership of land and assets, and low productivity of labor resulting from inadequate investment in health and education. 2.4 In March 1990, the Government launched a comprehensive economic stabilization and adjustment program designed to correct distortions resulting from past mismanagement of the economy and to resume economic growth. This program included exchange rate and tariff reforms, liberalization of the financial sector, administrative decontrol of agricultural pricing and marketing, and public sector reforms to reduce the fiscal deficit. 2.5 The immediate effects of these policies varied across economic groups. On the one hand, exporters and farmers were benefitted. On the other, the urban population and net consumers of food in rural areas were faced with higher prices for food and imported commodities and fewer jobs in the public and import-competing sectors. Thus, for many of Honduras's poor, a poor situation threatened to become worse. 2.6 The potential seriousness of this situation induced the Government to adopt both a short-term and a medium-term strategy. The short-term strategy involved the introduction in 1990 of two innovative social safety net programs: the FHIS, a mechanism for rapidly funding a large number of small scale, labor-intensive infrastructure, informal sector and social service projects, and the Family Assistance Program (Programa de Asignaci6n Familiar - PRAF), which delivers subsidies in the form of food coupons to poor primary school children of female headed households,' poor children under five and pregnant and nursing mothers. These programs were viewed as temporary in nature, to allow a rapid response to a critical situation until line ministries' capacity to cope with more fundamental poverty problems could be built up. The medium term strategy was to develop and implement a social sector reform program that would improve policy formulation, increase efficiency and equity in social sector programs, and strengthen the institutional and managerial capacity of social sector line ministries. 2.7 Both the structural adjustment and the safety net programs have received substantial external assistance. In 1991, Honduras received Official Development Assistance of US$52 per capita, amounting to 9.1 percent of GNP. The Bank and IDA have supported the Government's economic restructuring efforts through a Second Structural Adjustment Loan of US$90 million approved in September 1990, a Structural Adjustment Credit of SDR 14.3 million in January 1991, an Energy Sector Adjustment Credit (EDSAC) of SDR 37.95 million in October 1991, and a supplemental credit attached to the EDSAC of SDR 23.8 million. In addition to the two credits reviewed in this report, IDA has provided a credit of US$25 million for a Nutrition and Health Project, approved in 1993. Other donors in the aggregate have provided substantially larger amounts for social sector programs. 2. Female-headed-household as an eligibility condition was dropped as of 1993 and replaced by poverty indicators, e.g. high incidence of malnutrition. 3 3. THE PROJECTS 3.1 The objectives of SIF-I, according to the Staff Appraisal Report (SAR) were to "(i) mitigate the social costs of adjustment and allow economic reform measures to be put into place rapidly and make them acceptable to the population, (ii) lay the basis for a targeted food assistance program to support the poor and malnourished, and (iii) improve service delivery in the social ministries" (para 3.2). For these purposes, the project was to finance four components: (1) FHIS, (2) the food coupon program, to be implemented by PRAF initially on a pilot basis, (3) technical assistance to help design a program to improve the efficiency of the Ministry of Health (MOH) and the Ministry of Education (MOE), and (4) the establishment of a monitoring and evaluation system for the Government's social programs using Living Standards Measurement Surveys (LSMS). 3.2 The total cost of the project was initially estimated at US$68 million equivalent to be funded by the IDA credit of US$20 million, cofinancing of US$33 million, US$8 million from the Government and US$7 million from project beneficiaries (individuals, municipalities, NGOs and other entities). Of the US$33 million, the German Credit Institute for Reconstruction (KfW) was to contribute US$12 million and USAID US$8 million. In the end, substantially more than these amounts were mobilized, with the result that total project cost increased to US$98.4 million. 3.3 After finding that disbursement from SIF-I was proceeded more rapidly than anticipated, the government requested additional funding for FHIS. An appraisal mission was fielded in February 1992 and SIF-Il became effective in October of that year. The objective of this second credit was to help expand the operations of FHIS and provide technical assistance to a NGO liaison office to improve the NGO's ability to plan, prepare and implement projects. 3.4 Conditionality in both credits was limited mainly to procedural issues, for example, requirements that an operating manual be prepared that was acceptable to the Bank, that procurement procedures (which differed from standard government procedures) be acceptable to the Bank, that external audits, which included physical as well as financial reviews, be undertaken quarterly, and that recommendations developed in mid-term reviews would be quickly implemented. Nearly all these conditions were met without undue delays. 3.5 Supervision was heavy. In addition to normal supervision inputs averaging 35 staff weeks per year (for both credits over the period FY92-FY94), two intensive mid-term reviews were held, the first in November 1991' and the second in April 1993. These reviews generated a number of suggestions for improvements in operations, some of which are discussed below. In general these credits have been relatively problem-free. 3. While the credit had been agreed to only a few months before, FHIS-I had been in operation for almost a year and a half of its four- year life. 4 4. FHIS A. Design and Operating Procedures 4.1 In essence, FHIS is an agency that manages the process of allocating funds received to a large number of small projects meant to help the poor. In the typical case, municipalities prepare and propose projects and private contractors and NGOs receive contracts from FHIS for their execution, although there are numerous variations on this theme. 4.2 The idea for such a fund, originally based on the successful Bolivian Emergency Social Fund, was embraced by President Callejas when he assumed office in early 1990. The notion was that the line ministries concerned with the social sectors-mainly the MOE and MOH-were not at the time capable of rapid and efficient administration of a large emergency employment program, that a new agency with more freedom in procurement, contracting and hiring and paying personnel was necessary, at least for a time, until these ministries' capacities could be built up and/or the emergency period successfully completed. 4.3 Project Selection. Project selection was to be demand driven, any agency being eligible to propose projects provided certain conditions were met. A budget was established for each municipality depending on its poverty index and its population size (see discussion of targeting below). Eligible projects included small scale economic and social infrastructure (rehabilitation and expansion of health posts, schools, water and sewage facilities, drainages and roads), basic needs projects (provision of health, nutrition and training services, support to the elderly and street children, construction of school desks, establishment of family vegetable gardens and small livestock farms), and small-scale credits to informal sector workers and microenterprises. Evaluation was based on labor and materials content, counterpart financing, unit cost estimates, technical viability and satisfactory provision for operations and maintenance. No rate of return or cost-effectiveness methodology was utilized. 4.4 At the outset, there was a shortage of good proposals and most came from more affluent municipalities which typically were better informed and equipped to identify projects and prepare proposals. To correct this situation, FIS began promoting its services and offering technical assistance to poorer municipalities. FHIS-Il has institutionalized these procedures by establishing a special promotion office for these purposes. Over time, the number of proposals received has grown rapidly. By now, more proposals are received from each municipality than is budgeted, so that poverty targets should be easier to achieve. 4.5 Procurement and bidding procedures. To speed up the project cycle and ensure a wide distribution of small projects, both the Government and IDA agreed to allow FHIS to utilize different procurement procedures. (1) For projects of US$30,000 or less, FHIS is permitted to contract directly without competitive bidding. For this purpose, two rosters are maintained and continuously updated, one of experience with contractors and another of prices. The price data base is broken down on a regional basis for all prices typically used in estimating project costs. The procedure for gathering and utilizing these prices to assess the sole-source bids seems very efficient. (2) For projects between US$30,000 and US$70,000, a small number of contractors are invited to bid. (3) For projects between US$70,000 and US$500,000, publicly announced local competitive bidding procedures are 5 utilized. (4) For projects of US$500,000 or greater, international competitive bidding procedures must be used (to date, however, no projects have been of this size). While IDA accepted these procedures (for reasons discussed in paragraph 3.21 of the SAR), it placed a limit on the dollar amount that could be funded by direct contracting and requested monitoring procedures be established to ensure that contracts were not divided into smaller packages than appropriate and that the same contractor was not awarded an excessive number of contracts. 4.6 The limitation placed on direct contracting has not proved binding since the additional donor funds acquired by FHIS-I after the Bank-funded project started do not have this restriction. As a consequence, over two-thirds of the dollar amount of FHIS projects are contracted directly. There may have been some division of projects into smaller packages - for example, a school room rehabilitation project separated from a fencing project for the same school -but sometimes this occurred because budget limitations required such projects to be sequenced. No systematic information is available to indicate the prevalence of this practice or to determine whether contract awards were excessively concentrated, an oversight that should be corrected. However, no one interviewed believed that such practices were frequent or serious; and all praised the speed and geographic distribution resulting from the use of these procurement procedures. 4.7 Monitoring and Supervision. FHIS activities are heavily monitored, by donors, line ministries, and internal and external auditors. The latter conducted financial-cum-physical audits quarterly based on a rotating sample of projects under implementation. The auditors summarized this experience to the mission by indicating that in more than 95 percent of the cases, no serious problems were found; and when exceptions were reported, corrective actions were implemented fairly quickly. The consequence has been a significant drop in exceptions in 1993 compared to the 1991-92 period (Table 6.C of PCR). However, the contract with the external auditor expired at the end of FHIS-I and has not been renewed or replaced with another one. This may not be a serious problem as yet because of the presence of an effective internal auditing system and the low level of business during the last six months; but it should not be allowed to continue." 4.8 FHIS, in turn, monitors and supervises its projects fairly heavily. Required reports and procedures are spelled out in detail in its operations manual. The 1991 mid-term review found that the system was becoming overburdened because of the rapid expansion in the number of projects and recommended employment of additional supervisors. This was done, in part by contracting out some of the supervisory work to qualified individuals and NGOs. The system worked well until the last few months of FHIS-I, when morale and discipline became a problem. This led to the replacement of many of the staff of the supervision directorate when FHIS-Il started. The system seems to be back on track again, but it has not yet been adequately tested because of the low work loads since the inception of FHIS-II (see below). 4. Invitations to submit proposals for a new contract were issued by FHIS in September 1994 and the contract is expected to be signed in November 1994. 6 4.9 Monitoring is facilitated by computerized systems that are capable of rapidly producing up-to-date reports in a variety of formats, a capability observed by this mission after requesting tables for this report. However, these systems have not been fully integrated-reporting on physical progress, for example, has not been linked to financial reporting. FIS-II intends to correct this situation in the near future.' B. Sources and Uses of Funds and Level of Activity. 4.10 FHIS got off to a rapid start and has grown steadily up to about September 1993, thanks both to its ability to develop and implement projects and to donors' support. Table 1 indicates how rapidly donor commitments to FHIS have increased over time; it also demonstrates the uses that have been made of these funds. By now, FHIS has become an important actor in the social sectors. In 1993, it accounted for 13.5 percent of all public expenditures on these sectors. 4.11 Since September 1993, commitments have fallen sharply (Figure 1). Three factors appear to be responsible: uncertainties about whether FHIS would continue beyond its first four years, failure of FHIS-I to plan for continuation of the organization by seeking new sources of funds, and failure of FHIS-II to quickly obtain bridge financing until new sources of funds begin to arrive. Instead of developing new projects, FHIS-I staff focused on trying to complete projects already under way. Hurricane Gert, which hit Honduras in September 1993 and the Presidential elections of November 1993 as well as the imminent end of FHIS-I probably influenced this decision. The result was a dramatic increase in disbursements during the last four months of 1993 (Figure 2). The equally dramatic decline in January and the failure to recover thereafter is a result of severe shortages of cement and electricity that developed in the last months of 1993, as well as the decline in commitments for new projects. C. Results. 4.12 Projects - Numbers and Quality. From its inception through June 30, 1994, over 5,500 projects have been approved and over 4,500 completed (Table 2). These are relatively large numbers in this small country. They include, for example, rehabilitation projects affecting 32 percent of the public schools and 39 percent of the public health centers (see Tables 3 and 4 for these and other indicators of magnitude). The average size has been small -US$20,000- though the informal (credit) projects have been several times larger; and average implementation period has been about 6-9 months. Eighty percent of these projects have involved social and economic infrastructure. 5. Interestingly, FHIS started by borrowing and then building on the system used in Bolivia. The El Salvador social investment fund then borrowed and built upon FHIS's system. FHIS may now borrow the new El Salvador system and build upon it. Table 1. Sources & Uses of FHIS Funds (US$ 000) 1990 1991 1992 1993 1994 Total SOURCES - TOTAL 6,799.4 19,360.3 31,627.3 43,830.6 11,003.3 112,621.0 Percent of Total Government 27.7 10.8 25.4 14.7 0.1 16.4 World Bank 0.0 51.9 31.6 13.3 11.7 24.1 KFW 0.0 8.9 23.7 17.9 18.7 17.0 IDB 0.0 0.0 11.7 44.4 50.5 25.5 USAID 53.8 23.2 0.8 6.4 12.0 11.1 Other 18.5 5.1 6.7 3.2 7.0 5.8 USES - TOTAL 4,046.1 20,390.3 26,187.1 46,579.1 8,018.1 105,220.6 Percent of Total Investment 70.2 92.7 90.3 94.3 84.3 91.3 Advance Payments* 10.9 14.0 0.7 6.6 29.5 4.0 Institutional Strengthening 0.0 0.6 0.9 1.0 1.8 0.6 Infrastructure 52.0 61.4 68.1 68.4 89.2 68.0 Basic needs 4.9 10.9 17.0 18.0 28.4 16.6 Informal sector 2.5 5.8 3.7 0.5 2.1 2.2 Administration 20.8 6.7 8.3 5.3 14.5 7.6 Changes in assets 8.9 0.6 1.3 0.4 1.2 1.1 Changes In reserves 2,753.3 1,030 5,440.2 2,748.5 2,985.2 7,400.4 *Not clamified by pmgram. FIGURE 1 FHIS ACCUMULATED INVESTMENTS AND PROJECTS APPROVED LEMPIRAS NUMBER OF MILLION PROJECTS 700 6000 600 - 00AMOUNT -5000 500 --PROJECTS 4000 400 3000 00 300 2000 200 1000 100 UAPJuly 1994 1991 1992 1993 1994 FIGURE 2 FHIS DISBURSEMENTS: INVESTMENTS AND ADM. COSTS (MILLION LEMPIRAS) 35 30 _ _ _ _ _ 2 5 ADM. COSTS 20-- 15 - 10- 5 0 - tr4 t Source : UAP July 1994 1991 1992 1993 1994 Table 2. Project Implementation 1990-July 1994 1990 1991 1992 1993 1994 Total No of Amount No of Amount No of Amount No of Amount No of Amount N' of Amount Proj. ($000) Proj. ($000) Proj. ($000) Proj. ($000) Proj. ($000) Proj. ($000) Projects Status Received' 14,604b Evaluated 6,806 - Approved 1,043 8,710 1,615 29,850 1,344 35,370 1,476 38,410 48 1,000 5,526 113,340 Under Implementation 19 50 62 1,780 213 8,500 671 19,410 48 1,000 1,013 30,730 Completed 1,024 8,670 1,553 28,060 1,131 26,870 805 19,000 0 0 4,513 82,600 Projects Approved by Sector Infrastructure 983 7,730 1,322 22,670 1,086 28,300 1,092 26,360 42 630 4,525 85,670 Basic Needs 57 400 274 5,150 246 5,490 360 9,240 0 0 937 20,280 Informal 3 580 19 2,030 12 1,590 24 2,810 6 370 64 7,380 Total 1,043 8,710 1,615 29,850 1,344 35,370 1,476 38,410 48 1,000 5,526 113,340 Noter Amounts in thousands of USS (estimate), accumulated from average yearly exchange rates. a. The nature of database does not permit the projects received and evaluated to be broken down by year sub-totals. b. The number of projects received in relation to the number of projects evaluated is slightly skewed due to fact that 5,662 projects were received in the last six months. Source- FHIS/UAP, July 1994. 11 Table 3. Units Affected (built or renovated) by FHIS compared to National Inventory, 1990-July 1994 Selected Categories National FHIS Percent of Inventory Affected Inventory Education Schools 7,709 2,478 32.1 Classrooms 22,429 4,238 18.9 School Desks 492,543 335,526 68.1 Health Health Centers 841 331 39.4 Latrines Build (90-93) 250,424 78,451 31.3 Water Systems Population with access 3,862,000 185,030 4.8 Population without access 1,350,000 Sewer Systems Population with access 3,797,000 569,186 15.0 Population without access 1,415,000 Training Midwifes 6,632 4,632 70.0 Data (Preliminary): Units and Persons Source- MEP, MSP, FHIS/UAP July 1994. 12 Table 4. FHIS Expenditures Compared to Total Government Erpenditures and GDP 1990 1991 1992 1993 Projected 1994 A. Million Lempiras FHIS (LPS mill) Education 7.1 86.0 112.0 134.0 71.0 Health 22.1 45.0 53.0 70.0 48.0 Other 17.4 31.0 313 46.0 69.3 Total 46.6 162.0 196.3 250.0 188.3 Public Social Expenditures (incl. FHIS) Education 522.0 620.4 754.3 889.0 1,081.4 Health 344.2 392.0 467.0 523.3 555.3 Other 91.1 107.2 146.2 181.0 90.0 Total 957.3 1,119.6 1,367.5 1,593.3 1,726.7 GDP 11,156.0 13,975.0 16,103.0 18,554.0 21,145.0 B. FHIS as a Percent Public Social Expenditures Education 1.4 13.9 15.0 15.1 6.6 Health 6.4 11.5 11.3 13.4 8.7 Other 19.1 29.0 21.4 25.4 77.0 Total 4.9 14.4 14.4 15.7 10.9 GDP 0.42 1.15 1.22 1.35 0.89 Source: SECPLAN, FHIS/UAP July 1994. 13 4.13 Problem projects seem few and far between. According to FHIS, cost overruns have been experienced by about five percent of its projects,' and time overruns were a negligible problem prior to September 1993. Since then, problems have increased because of Hurricane Gert which struck in September, the shortages of cement and electricity which became serious later in the year, and some deterioration in operations in the supervision department which occurred during the last months of FHIS-I. All told, there have been 62 problem projects -projects turned over to the Legal Department for action. Thirty five of these resulted from giving contracts to municipalities for project execution during the first two years, a practice that has nearly ceased. Most of the remainder occurred during the last months of FHIS-I. Over half of the 62 have been satisfactorily resolved and the remainder are still being negotiated. 4.14 Evidence on the quality of workmanship and training is spotty and a lot must be based on anecdotal reports in the absence of more than a few systematic reviews. Taken together, however, a fairly systematic picture of respectable quality emerges. Observations by this mission, which visited eight project sites and interviewed a number of beneficiaries, municipal officials and project supervisors, support this picture. A survey of completed projects is currently under way that should provide better evidence (see paragraph 4.38). 4.15 Evidence on the importance of individual projects for sectoral goals is also weak. For example, while there is little doubt about the need for rehabilitation of classrooms and health centers, most of these projects involved some expansion of capacity as well; yet there is no school or health center mapping studies to identify where expansion is most needed, where schools and centers can be consolidated, etc. This is a failing of the planning process, not of FHIS; but the speed with which FHIS is operating exacerbates this problem. 4.16 Employment Generated. The data on employment in Table 5 has been derived from FIS project appraisal documents, not from what actually happened. They may be too high because of efforts to promote proposed projects; but they are too low in that they ignore multiplier effects of these projects on jobs in the rest of the economy.' 4.17 Compared to the economically active population and to estimates of unemployment plus underemployment, these figures seem small -an impression that may reflect more the magnitude of the problem than ineffectiveness on FHIS's part. The investment cost per job created" is lower than country-wide averages and is similar to what is found in other Latin American social investment programs. However, these projects use standard technologies and pay market wages. While it may not be possible on political grounds to pay less than market wages (even though it can be justified because of the high underemployment figures) it should 6. Prior to 1994, cost overruns were permitted up to 30 percent of initial cost if FHIS approved work order changes. During that period, about five percent of the projects experienced such overruns. Since the beginning of 1994, no cost overruns have been permitted. 7. A school map is planned for 1995 under the proposed Basic Education Project. 8. The figures for 1990 and 1994 are atypical and should be ignored. In 1990, the low cost per job created resulted from an emphasis on pure job creation. In 1994, the low figure results from the fact that credit projects, which generate more employment per unit of investment, were a larger proportion of an abnormally small total-only 48 new projects were started during the first half of 1994. Table 5. Pjects and Employment Generated by Sector 1990 1991 1992 1993 1994 To~a A~om P &lcse ~pom A ~om rlct miye A ~am reei Emply~ A~-•• P~le Empi~ A~mne Pr~|ct ep~ A~am Pr~al E ms. ($m-B) (#s) (pes/uns) (hulD) (#s) (para/fas) (hulB) (#s) (ps/muos) (1fmill) (#s) (pmss/mos) (Suill) (#s) (pe/nms) (SmalB) (#u) (pem/uns) Education 1.3 467 41,030 15.8 966 69,485 20.2 963 73,645 20.6 1,038 56,725 0.3 18 815 58.2 3,452 244,742 Health, Water & Sanitatiom 4.1 95 14,610 7.3 227 63,865 8.8 197 33,315 10.4 263 18,065 0.3 21 890 31.0 803 131,480 Food & Nutrition 0.1 6 9,105 1.0 29 745 0.8 36 455 0.3 17 225 0.0 0 0 2.1 88 10,614 Soca & Economic Infrastructure 2.5 448 146,555 3.6 360 68,815 3.9 130 14,255 3.5 86 22,480 0.1 3 380 13.4 1,027 253,075 Attentio to SpedalGroup 0.0 3 195 0.1 5 270 0.1 2 315 0.7 39 80 0.0 0 0 1.0 49 907 Environent 0.2 21 6,060 0.1 10 1,060 0.1 4 430 0.2 10 1,305 0.0 0 0 0.6 45 8,879 Informal Sector 0.6 3 120 2.0 18 400 1.6 12 336 2.7 23 666 0.4 6 116 7.1 62 1,638 Total 8.8 1,043 217,675 29.9 1,615 204,640 35.5 1,344 122,751 38.4 1,476 99,546 1.1 48 2,201 113.4 5,526 651,335 Total (Employm. in peron-year) 18,140 17,053 10,229 8,2% 183 54,278 Employment - percent Economically active population 1.2 1.4 0.9 1.0 n.d. n.d. Underemployed and unempioyed 3.3 4.0 2.4 .4. n.d. n.d. Unemployed 24.2 30.2 19.2 n.d. .d. n.d. Noe on Emp1ojanmt Fipaa Figurs derived from project appraal docemmenta. Employment in informal wedor i a asumed to be prmunet. Hecce, numbers prided are multipled by 12 to obtain peron-moth. Projects in other sectr are asumed to pr~de employment only during implementation period. Hence, numbers prmded are multipi by average ipe tation period for esch type f project. 15 be possible to find and utilize more labor intensive technologies without giving up a great deal in efficiency. This would help but will hardly solve the problem. 4.18 Targeting. The poor are targeted in two ways, first by allocating more investment funds to poorer regions, and second by selecting projects of most value to the poor. 4.19 Geographic targeting is accomplished by assigning a poverty index weighted by population to each department, and within each department to each municipality. For example, the department of Intibucd which has a poverty index of 60, is given a budget of 36 percent more than would be its share strictly on a population basis whereas the department of Corts with an index of 18.4 is given a budget of 24 percent less. This procedure, which by now is generally accepted and understood, provides a simple, transparent way to allocate limited funds and a good defence against political pressures to favor one area over another. 4.20 This form of targeting has not worked as well as originally expected. The actual amount received by each municipality depends not only on its budget but also on the number of acceptable proposals it puts forward. In the first year of operation, the inevitable bias towards those who reacted quickly favored wealthier municipalities. FHIS has made a concerted effort to correct this through education and promotion campaigns and assistance to poorer municipalities in identifying and preparing projects. As Table 6 indicates, that effort has helped eliminate this bias; but (ignoring the very small sample in 1994) there is still a ways to go before poorer areas are significantly favored. Moreover, some knowledgeable observers believe that within municipalities, urban and wealthier areas have been favored. Unfortunately there is no statistical evidence to investigate this possibility. 4.21 There are, however several reasons to believe that geographic targeting will improve in the future. First, as noted above, the number of proposals has grown to exceed budgets for each municipality. Second, FHIS-II plans to put greater emphasis on TA and community participation in the formulation of projects. And third, FHIS has demonstrated its capacity to reach remote communities, something that the MOE and MOH have not done in very many cases. All three together should have a positive effect. 4.22 Equally important as a targeting mechanism is the nature of the projects. Wherever placed, most of them have been of a type that the poor make most use of. This is true of the latrine, potable water and sewage projects, most of the primary health care projects and even many of the primary school rehabilitation projects (since wealthier groups so often use private facilities). Thus, even if it is true that wealthier areas have been favored, the poor within these areas are still the most likely beneficiaries. 4.23 The geographic targeting procedures can be improved by obtaining data for smaller geographic units, reconsidering the indexes themselves and updating the data (the poverty map is still based on 1989 data). But in the end, a proper investigation of targeting requires information on beneficiaries, not on geographic areas. That was supposed to be provided by the LSMS component of SIF-I, but it seems not to have happened. 16 Table 6. Investment accoring to Poverty Map for Municipalities Levels of Poverty Population Pop/ 1990 1991 1992 1993 1994 Average Total Very Poor 895,167 0.20 0.08 0.19 0.21 0.20 0.26 0.19 Poor 941,040 0.21 0.12 0.17 0.25 0.24 0.12 0.22 Deficient 1,090,253 0.24 0.28 0.21 0.24 0.22 0.18 0.22 Regular 1,191,770 0.26 0.44 0.35 0.26 0.25 0.23 0.29 Acceptable 390,521 0.09 0.03 0.08 0.04 0.07 0.07 0.06 Various Levels 0.06 0.00 0.00 0.02 0.13 0.02 Total 4,508,751 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Sourcr FHIS/UAP, July 1994. Table 7. FHIS Staff 1990-August 1994 1990 1991 1992 1993 1994 (Mt) Permanent Staff Professional 33 68 86 80 82 Other 24 61 70 94 87 Total 57 129 156 174 169 Contract Employees n.a n.a n.a n.a 118 Total - - - - 287 Souc* Peronnel Office. August 23, 1994 17 4.24 Operational Efficiency. FHIS has acquired a reputation within the government and amongst donors for speed and efficiency. While line ministries sometimes take up to a year to complete the bidding process, FHIS has been able to do so for the same type of project within a month. According to the Midterm Review of SIF-I (para 22), the MOE staff itself estimated that FHIS can build, for about the same cost, many times more classrooms in a given year than the Ministry can. During the last year this has led donors and the MOE and MOH to shift some construction work that would normally be done by these ministries to FIS. One of the reasons for this difference in speed is that ministries must go through a central contracting agency which can add six months to the procurement process. This requirement is scheduled to be abolished in the near future. Other reasons given include greater institutional capacity, more capable and motivated staff, more standardized products, and more direct contracting (see paragraph 4.5). 4.25 These results have been achieved at reasonable administrative cost-between six and ten percent during 1991-1993. For the first half of 1994, these costs have increased to 15 percent because the members of employees was not reduced commensurately with the decline in commitments and disbursements during this period (Table 7).' 4.26 However, there may be some other costs associated with this emphasis on speed: less involvement of communities in selection and design of projects, less adaptation to local conditions, less emphasis on use of local manpower and materials, fewer projects located in remote areas, and less concern about future operations and maintenance are all possibilities. FHIS-II plans to improve its operations in all these areas by, among other ways, promoting greater community and municipal participation.' While efforts to do so are likely to slow down the project cycle, and increase administrative costs, these are small costs to pay for the likely benefits. 4.27 Institutional Development (ID). ID has occurred at three levels. First, FHIS itself has grown in size, complexity, and competence. By 1993, it was a smoothly and efficiently functioning organization capable of efficiently handling a large volume of business. Some of the experienced senior staff and well-honed procedures were lost in March/April 1994 after the new government assumed power, but it remains an impressive organization. It is so treated by line ministries and donors alike. 4.28 Second, FHIS has contributed to the institutional development of the municipalities, NGOs and private enterprises through its contracting and supervisory procedures and by sometimes providing them with technical assistance and training." In FHIS-I, TA and 9. In August 1994, FHIS had 169 permanent staff and 118 contract employees. It is not dear why so many contract employees were necessary given the low level of activity at that time. 10. A particularly telling manifestation of this is an agreement just reached between FHIS, the National Association of Municipalities and the President's Office that municipalities will share in the investment costs, richer communities paying 50 percent middle income communities 33 percent and poor communities 25 percent. 11. For example, the PCR for SIF-I notes that 80 of the 291 municipalities now have engineers as permanent staff to plan and supervise small investment projects. 18 training were provided as the need arose; in FHIS-II, these functions are being institutionalized. 4.29 Third, line ministries, particularly in the social sectors, are having to adjust to the presence of FHIS and its style of operations. First, FHIS and these agencies are obliged to cooperate with each other to avoid duplication and ensure that FHIS's projects are consistent with ministry programs and norms. This requirement is encouraging line ministries to more clearly and carefully specify what is acceptable to them. Second, in the interest of ensuring sustainability, FHIS insists on commitments to provide staff and supplies for the schools and health centers it rehabilitates or expands. Third, FHIS presence has introduced a degree of competition for donor funds not before faced by these ministries, which will force them to improve the efficiency of their operations. 4.30 Microenterprise Credit Programs. FHIS includes two programs that provide funds for NGOs and community banks that specialize in lending to microenterprises (enterprises with less than ten employees and less than US$2,000 of capital). A special comment is in order because, per dollar of investment, these programs create more sustainable employment and institutional development than any other FHIS program. 4.31 PASI (fPograma deAsistencia al Sector Informal) which assists urban microenterprises, was started in October 1990 with government funds. The World Bank provided US$1.3 million (90 percent provided as a credit for on-lending and the remainder as a grant for administrative and TA operations)as part of SI-I. This magnitude was a compromise between the government's much larger request and Bank staff who were initially reluctant to include this component in the credit. This reluctance was apparently associated with poor repayment experiences of similar schemes in other countries and the belief that they require banking skills quite different from those required for emergency employment generation schemes (the initial image of what FIS would focus on). The credit agreement included limits on the extent to which IDA funds could be used for this purpose. A few months later the Netherlands Government provided an additional US$7 million (again with ten percent as a grant for administration and TA). 4.32 As of July 1, 1994, PASI has approved 56 credits totalling Lps 40 millions; 39 are still in operation. These credits have in turn financed 12,000 microenterprises out of a population of 250,000. The loan repayment rate has been excellent: in contrast to private banking where failure to repay typically averages eight percent, the PASI has experienced less than five percent. This record is attributable to clear, detailed and rigidly enforced regulations (including the elimination from the program of any NGO with more than a five percent arrears record), a high level of supervision, and TA provided to the NGOs and through them to the microenterprises.' 4.33 PROCATMER (Programa de CrdditoyAsistencia T&nica Microempresa Rural) which provides funds to NGOs operating in rural areas, was started in February 1993 with a loan 12. Individual microenterprises are not left alone for very long. Loans are provided in tranches, scheduled repayments are frequent, financial records are required and inspected frequently with rapid follow-up to encourage corrective actions if cash is not being accumulated fast enough for scheduled repayments. 19 of US$12 million from the European Community. Operating procedures are very similar to those of PASI, the main differences being that PROCATMER operates independently of FHIS management' and has a codirector appointed and paid for by the European Community. To date, some 20 credits totalling Lps 177 million have been provided to NGOs. Most of the projects have started too recently to judge whether the repayment experience will be as good as that of PASI (since it is operating in a more difficult environment, it probably will not be). 4.34 These programs are among the most promising in the FHIS family but cannot be expanded rapidly without lowering quality." Discussions have started to involve commercial banks in PASI operations. An in-depth international evaluation of both programs is currently under way. 4.35 Sustainability. In this case, sustainability has two aspects: sustainability of FHIS, and sustainability of its projects. 4.36 FHIS was initially conceived to be a temporary organization that would remain in existence until the MOE and MOH built up their capacity to take over the functions it has been serving. This no longer appears likely. Instead, a new division of labor seems to be developing, in which these ministries concentrate more and more on policy planning and norm setting and leave management of the project cycle to FHIS. An indication of this is the fact that, with donor encouragement, a number of ministries have begun arranging for FHIS to implement some of their projects. If this development continues, the government will need something like FHIS on a permanent basis and would fund it, to the extent it could, even if foreign assistance to FHIS were substantially reduced.' 4.37 On the other hand, a number of FHIS projects have run into trouble after completion because of lack of maintenance and sometimes operating budgets. In FHIS-I, the principal way of coping with this was to require statements in project proposals that adequate plans for operations and maintenance had been developed. In two important instances, however, it went further by developing specific programs. One case involves the latrine construction program which initiated an education and training activities after finding that only a small number of recipients were properly using and maintaining them; according to an evaluation study, the program was very successful in raising proper use and maintenance rates from 2 to 88 percent in a brief period of time. The other case involves the introduction of user committees for water and sanitation projects, to work with local authorities in establishing maintenance routines, charges and operating procedures, an effort which has helped in some but not all cases. FHIS-II has institutionalized its efforts to promote community participation and sustainability by establishing a new division for this purpose. This division's action program is still in the formative stage. 13. The same will be true of PASI by 1995. 14. The available capital stock will also limit expansion. Due to the rapid devaluation of the Lempira and the practice of granting loans to NGOs on a fixed interest rate basis. The original capital of PASI has depreciated by about 30 percent during the past three years. PASI is now planning a series of measures to correct this problem. 15. An amendment to the FHIS law, approved by Congress in November 1994, extends the life of the fund for an additional 12 years. 20 4.38 Such operations and maintenance problems are not believed to be severe. But so far the evidence is all anecdotal. At the behest of the Bank, the Regional Unit for Technical Assistance is undertaking a survey of all completed projects, asking a few simple questions about operations, use, maintenance and environmental impacts, which is to be completed and assessed before approval of a third IDA credit to FHIS. 4.39 There is, however, a limit to what FHIS can do. The more fundamental factor explaining inadequate operations and maintenance is the limited financial capacity especially of the poorer municipalities on which FHIS attempts to focus, a problem that is largely out of FHIS's control. Still, FHIS could play a useful role if its TA to municipalities helped improve their capacity to collect revenue (cadastral surveys, property taxes, water rates, etc.). 4.40 Political Interference. Political interference with technical decisions can distort project selection and location, contractor selection, staffing and costs of operations. If the distortions were serious and blatant, donor interest in funding could be eroded. The SARs for both SIF-I and SIF-I identified this (in more diplomatic language) as a major risk' 4.41 There is no evidence that serious problems of this sort have been present in FHIS. Some excessive turnover and hiring of staff seems to have occurred at the beginning of FHIS- II and this may account for some of the increase in administrative costs, but decision-making with respect to projects does not appear to have been affected. 4.42 Five factors appear to account for this commendable record. * Strong commitment and deliberate restraint by political authorities, enforced by the President. * Executive Directors who are selected for their proven commitment to the principles involved, are already wealthy, and have track records as successful managers. * Professional, dedicated staff, encouraged by good salaries and lack of tenure (contracts with staff cannot be longer than the legislative life of FHIS, which has been limited to four years, and are often much shorter). * Frequent, intensive monitoring by donors,"' internal and external auditors, ministries, and recipients, facilitated by a good MIS system and a policy of open access to information. 16. The SAR for SIF-I noted that "...there is the risk that the FIS may be tempted to respond too quickly to demand, which could result in poor use of funds or their diversion to uses not envisaged in the project." 17. Several of the staff are cpatriates whose salaries are paid by donors. In the case of the credit programs, a condition of one grant was the assignment of joint directors, one paid for by the donor, to run the projects. 21 Resource allocation procedures and parameters that are agreed to with donors and the government, and which are easily understood and monitored. 4.43 Other factors may have also been involved at the outset of FHIS. It undoubtedly helped that FIS was a new organization established by special legislation; as such it started with a nearly clean slate of rules, regulations, expectations and prior obligations. It may also have helped that it was started as a small, temporary organization to help resolve an emergency and that at the outset it very quickly committed its funds. These factors are no longer available to protect the organization. As it has grown, attracted more donor funds and begun to look like it may be permanent, it has undoubtedly become a more attractive target for rent seekers. 4.44 Other impacts. Several programs have been of special importance for women. Basic needs programs such as the midwife training program, the small-scale farm and family garden program, and the maternal and child care program. At least half of the microenterprise credits administered by NGOs have gone to women, and a number of community banks have been established for and operated by women. FHIS has attempted to take environmental concerns into account through its water, sanitary land fill and latrine and septic tank projects and by avoiding adverse impacts in other projects. In FHIS-II this program will be expanded by cooperating with the Ministry of Environment to add projects specifically devoted to environmental improvement 5. OTHER COMPONENTS A. Family Assistance Program (PRAF) 5.1 The second of the government's safety net programs, started at the same time as FHIS, was a food coupon program for single mothers (Bono Mujer Jefe de Familia - BMJF) distributed through primary schools in poor regions. The Bank helped improve the design of this program through its Regional Unit of Technical Assistance (RUTA) located in San Jos6, Costa Rica, and recommended that a similar program for pregnant or lactating women and for children aged 0-5 with coupons distributed through the primary health system (Bono Materno Infantil - BMI) be initiated on a pilot basis using US$420,000 in funds from SIF-I. This was accepted and the pilot program started in December 1990. Both programs are administered by PRAF. These programs provide a monthly coupon worth Lps 20 (US$3.50 in 1992) which can be exchanged for food or for cash. At the time, this amount contributed approximately 20 percent of a poor individual's basic consumption needs. While the value has been eroded by inflation - it now covers about ten percent of requirements - an increase is planned for 1995. At the present time approximately 300,000 persons out of a total population of 5 million benefit from this program. 18. Agreements about the percent of budget that can be used in different programs and use of the poverty map for making allocations to municipalities leave much less room for influencing decisions. But what happens below the municipality level is far less clear. Centrally prepared school and health center maps would help at that level. See discussion of targeting. 22 5.2 A 1992 evaluation of these programs concluded that both programs were effective mechanisms for transferring income to large numbers of poor families. Approximately 80 percent of the additional income was spent on food and much of the remainder on other necessities like medicines. Administrative costs were substantially lower than food distribution programs and PRAF audits report that the coupons were honestly distributed. Both programs were found to have the side benefit of increasing enrollment/use of the schools and health centers distributing the coupons compared to institutions not in the program. The BMI was found to be especially effective in that it is self-targeting (since only poor families go to health centers). The BMJF was found to need more transparent and uniform eligibility criteria and better monitoring." 5.3 The evaluation concluded that both programs should be expanded. This was done as part of the Health and Nutrition project approved in January 1993, which provides US$6 million for them and is currently being implemented. 5.4 These programs only make sense as temporary safety nets; they do nothing to address the underlying causes of chronic poverty and malnutrition. But they have proven to be popular and will be difficult to terminate or scale down. Questions need to be raised about whether, and for how long, the Bank, as a development institution, should be involved in their support. B. Living Standards Measurement Surveys (LSMS) 5.5 This component was supposed to establish a mechanism for periodically monitoring, measuring and reporting on the impact of FHIS and PRAF as well as other GOH social policy programs. This was to be done by adding three components to the Honduras Multiple Purpose Permanent Household Survey (MHS): a consumption module to be used as a proxy for income in measuring living standards; a survey of social services, market prices and other aspects of the socioeconomic environment in areas identified by FHIS as being of high priority; and a survey of FHIS beneficiaries. SIF-I provided US$320,000 for this purpose. 5.6 These funds were used to obtain the assistance of an expatriate consultant to guide this enterprise, to provide training and equipment to the staff of the General Statistical Office, to develop the new modules and to conduct three rounds of surveys using these modules. Despite this, little usable results emerged. The sampling frame, which focused attention on the formal sector, was never adjusted to meet FHIS needs for a close look at the informal sector. Conceptual problems in measuring non-monetary consumption and income were never resolved. Most of the questions addressed to FHIS beneficiaries pertained to attitudes towards and participation in FHIS programs and were inadequately linked to information on their income and consumption. The few reports produced were of little value in evaluating the impact of FHIS. 5.7 Careful thought was not given during the design stage of this component to what was required. By simply assuming the need could be met by grafting onto an existing survey, the 19. This has been done in 1994 by the new administration of PRAF. 23 component was probably left with too little funds to do the job properly. Then, during implementation, while the Bank recognized that these problems were not being adequately addressed, it did not take corrective actions quickly enough to produce usable results before the budget was exhausted. An additional factor was the low priority assigned by the Planning Ministry (SECPLAN) to the task and the economy weak institutional support. C. Social Sector Restructuring Program 5.8 This component was to prepare a detailed social sector development strategy to address the weaknesses of the Ministries of Health and Education so they could eventually take over from FHIS. As softer funds became available, the government requested that this component be financed by a Japanese Technical Assistance grant rather than by the IDA credit. The activity was assigned to the secretariat of a newly created Social Cabinet and the funds were used to support its activities. 5.9 Nothing of any concrete value came of this component, which was largely ignored by Bank supervision missions. This was unfortunate as it could have laid the basis for a social sector restructuring project, talked about at the time but never developed.* Grant funds were redirected to the preparation of the Nutrition and Health and the Basic Education projects. Hopefully these projects will more effectively address the weaknesses of these ministries. D. Technical Assistance for an NGO Liaison Office 5.10 This component, for which US$200,000 was set aside in SIF-II, made sense as a way to improve NGOs' capacity to assist FHIS with its activities. One of the more useful things it could have done would have been to draft legislation to improve the legal status of NGOs. Some of the funds were used to support the operating budget of this liaison office without any output of lasting value to show for them; undisbursed funds were reallocated to FHIS subprojects. 6. ROLE OF THE BANK 6.1 While the Bank played very little role in the initial design of FHIS, and was remiss in not devoting more attention to the minor components of SIF-I and H, it devoted considerable attention to FHIS operations-57 staffweeks of supervision for SIF-I alone, length, two intensive mid-term reviews in the course of three years, plus numerous studies by consultants. These efforts contributed in at least three ways to the success of FHIS. 6.2 First, out of these investigations the Bank was able to make a number of useful suggestions for improvements in FHIS operations. (e.g. quarterly audit, improved financial projections and monitoring, attention to maintenance). 20. The GOH requested that IDA abandon the idea of a Social Sector Adjustment operation in favor of a Nutrition/Health investment operation. The energies of the secretariat focused on helping prepare the latter. The argument was that complex, long-standing ministries require in-depth study for reform to be effective. 24 6.3 Second, the continuous, detailed monitoring, insistence on frequent auditing and reporting, and numerous requests for detailed information and data on operations, encouraged a high standard of performance and helped protect FHIS staff from demands for behavior that would have weakened its reputation. 6.4 Third and closely related, this detailed monitoring gave the Bank a solid basis for providing the seal of approval needed to mobilize other donor resources for FHIS. 7. PARTICIPANT ASSESSMENT 7.1. The audit mission talked to contractors, local government officials and other participants in FHIS projects to determine if their impressions were consistent with more centrally provided information. These discussions also served to add meaning and significance to more aggregate information provided by secondary sources. This section reports on some of those discussions." 7.2. Discussion with midwife trainees and their teachers. This project, which started in 1993, intends to train over 4,600 midwives (80 percent of the total) using ten NGOs operating under contract with FHIS in different parts of the country. Each midwife is provided with one week of training, but then there are follow up sessions once per month. The NGO representative at the one site visited indicated that the follow up activities were proving to be more important than the formal training since 60-70 percent of the women are illiterate. Equally important in terms of improving service quality was the provision of supplies to the midwives. Here, there was a problem since the MOH often ran short. To improve the situation, FHIS agreed to one-time provision of six months of supplies to each center at the outset of this program. 7.3. Most trainees interviewed (including 2 men among the 12 women) had served as midwives before this training was provided, learning what they could from their peers on the job. None of them were paid a salary for their services; but they frequently received Lps 35- 40 for out-of-pocket costs from mothers. An important improvement they saw as a consequence of the program was the use of better materials (sharp knife, antiseptics, measuring tape) provided to them at the time of training; previously, supplies were often short and they had to buy them themselves. The only criticism of the program came from some of the better educated women who wanted to go into more depth on complications, so they would be better prepared for a wider range of emergencies. The NGO representative plans to propose a follow on course to FHIS. 21. There are three ways to acquire an indication of beneficiaries' attitudes towards a project: interviews with beneficiaries, observations on beneficiaries' behavior relevant to this question, and observations of political behavior relevant to this question. The first is the most direct but, unless based on a representative sample could be very misleading. The second provides the most convincing evaluation evidence -if people freely choose to make use of services offered, they must be assessing those services positively-but cannot answer many questions that might suggest ways to improve the provision of those services. The third takes advantage of politicians' tendency to promote programs that their constituents like, but again cannot provide detailed information about beneficiaries attitudes. Previous sections of this report have provided information on behavior-employment generated, number of persons served, percent of bad debts, etc. -to the extent it was available. This section reports on impressions acquired during site visits and discussions with local officials and beneficiaries. It also considers the evidence from the political arena. 25 7.4. There is no statistical evidence on impacts on maternal or child deaths, but it is thought to be significant because the midwives have begun referring more problem cases to health centers where there is a doctor. An impact evaluation is scheduled to take place before the end of 1994. 7.5. School rehabilitation projects. Four school rehabilitation sites were visited. Typically, they involved refurbishing classrooms; sometimes they included adding classrooms, paving paths and putting up fences around the school. In some cases the work had been completed, in others it was in progress, in one - a problem project - the work had been suspended for a time. Discussions that took place at these sites were with contractors, field supervisors (directly or indirectly working for FHIS), or teachers. 7.6. The construction work seemed to be making a significant difference in the quality of the school building, for example, substantially improving lighting by increasing the number of windows and providing better protection from the weather by repairing leaking roofs. Both the field supervisors and the contractors reported no major problems in their work or in relations with FHIS, other than some minor delays. Employment was supposed to involve mostly local labor, but sometimes did not if the contractor was not from that locality. Teachers seemed proud of their refurbished building but knew nothing about arrangements for maintenance. 7.7. Sheetmetal microenterprise. An association of 18 sheetmetal enterprises was visited. It provides services to individual members which include selling produce, buying raw materials in bulk, providing technical advice and keeping financial records. It began operating two years ago with loans from regional banks and received some training from UNIDO. Six months ago it received the first tranche (half) of a loan of Lps 500,000 from PASI, which it prefers because the banks required guarantees that only a few of its members could provide. This loan, which has a term of two years and an interest rate of 20 percent, comes with a grant of ten percent to be used for institutional strengthening (mainly, the salary of a bookkeeper) and some technical assistance. 7.8. The association member interviewed indicated that they were quite pleased with this package and high level of supervision from PASI and that it would allow them to expand their operation when conditions permitted. They have not been able to do so yet - and indeed have had some problems filling existing orders -because of electricity shortages that have plagued the country during 1994. 7.9. It is difficult to properly estimate the number of beneficiaries resulting from this loan. While there are 100-120 employees/owners involved plus perhaps five family members per employee, incremental employment is much less. 7.10. The ANNAH Artisan Group has been in existence for 20 years and has a membership of some 500 artisans. For the past year all their funding has come from PASI. Previously, their members got loans on their own. The package provided by PASI to the Group and from the Group to its members is similar to those provided in the case of the sheetmetal association. The PASI loan has permitted the Group to expand its business, including displaying and selling abroad. The individual artisans visited all seemed to have as much 26 business as they could comfortably handle. Again, however, there was no way to obtain a good estimate of the incremental benefits of the PASI credit. 7.11. Latrine Project. As Table 3 indicates, FHIS has been responsible for the construction of over 78,400 latrines (outhouses) during the 1990-93 period. These distinctive structures were seen scattered amongst the cottages of nearly every village visited. The recipients with whom the mission spoke in one village understood the reason they were provided - to reduce the risk of disease -and believed that most of their neighbors were diligent in their use. This was not always the case, however; before discussions with FHIS representatives, some of their neighbors used these structures as storage areas. Few do that today. 7.12. Visit with the Mayor and his staff in one municipality. This visit provided a clearer picture of how FHIS budgeting system works. By this time, the Mayor and his staff know approximately how much will be allocated to them each year and how to cope with the paper work of applying for the funds. FHIS has been a benefit to them in part because they receive a larger budget now (but not that much larger because MOE and MOH allocations have declined) but more important because they get what has been committed and receive it in a timely fashion. Previously, structures might sit for years half finished; now at least, something is finished and becomes available for use. Maintenance is a problem, however; the municipality can spare little money for it and have trouble obtaining funds from the ministries. How do you decide on priorities? He decides after going around and talking to his constituents? Is the assistance he receives from FEIS adequate? Yes in terms of technical assistance, no in terms of budget. 7.13. Political Viability of FHIS. Judging from the way FHIS has been treated in the political arena, it is very popular with the electorate. The strongest evidence of this is that even though the party in power during the FHIS-I period lost the presidential election of 1993, the new government extended the life of FHIS for an additional four years. It was too popular and too useful to terminate, as often happens to initiatives of a previous regime when a new party takes over. While a distinction is made by calling it FHIS-HI and a new director took over its leadership, it is essentially the same organization. 8. ISSUES AND LESSONS 8.1. The two IDA credits reviewed in this report have been used for activities that demonstrate several important points: * Social investment funds, under the right circumstances, can work to produce a large number of small, widely disbursed projects targeted to the poor and especially underpriveledged groups such as women and children. * Credit programs for microenterprises can be made to work well. * When successful, SIFs can also serve as powerful mechanisms to mobilize and coordinate donor assistance for the social sectors. These functions, particularly that of donor coordination, were unexpected but important bonuses. 27 * Distribution of food coupons is a cost-effective way to provide food-cum-income supplements to the poor, certainly more cost effective than direct food distribution programs. But provisions for terminating these programs, if they are truly meant as temporary safety nets, need to be built into them at the beginning. 8.2. Four factors seem most important in explaining FHIS's success: * Carefully defined, monitorable goals with latitude to determine how to achieve them. It is doubtful if a line ministry, whose goals are much broader and constraints more binding, could have accomplished the same thing. * Professionally competent, hard working and well motivated staff, particularly at senior levels. Good salaries, lack of tenure,' carefully detailed operating procedures, and a first executive director who gave priority to technical competence and efficiency are probably the most critical factors explaining this. * Personal involvement of the President who used his powers to promote and protect the organization. * Substantial oversight by donors who insisted on a high degree of openness and transparency in FHIS operations. 8.3. It is also important to identify what social investment funds like FHIS cannot do and in the process sound several cautionary notes. * In its present form, FHIS does not deliver education and health services; it only provides the construction component of the full package of inputs necessary for this purpose. It would require quite a different organization, which is unlikely to be as successful, to provide the services themselves. * Both the government and donors should keep this point in mind when allocating resources. There has been a tendency on the part of donors to allocate an increasing portion of their assistance to FHIS at the expense of line ministries. This may be leading to excessive allocations to capital expenditures and exacerbating the shortage of funds for operations and maintenance and delayed much needed quality improvements in basic health and education services. In this connection, it is unfortunate that the effort to develop an action plan for social sector programs came to so little. * While FHIS and PRAF can ameliorate some of the effects of poverty, they can do little or nothing about its underlying causes: high population growth rates coupled with low economic growth rates plus unequal distribution of land and capital, including human capital. So long as these forces dominate, such 22. Contracts with staff cannot be longer than the legislative life of FHIS, which as been limited to four years, and are often much shorter. 28 programs will be needed on a long-term basis and will be running hard just to keep up with ever-expanding needs. FHIS is being pulled in many new directions -for example, technical assistance, training and community mobilization as well as project management -which it may be far less well-equipped to handle. One solution for this is to hive off specialized functions to new FHIS-type organizations, as is happening with its credit programs. Another is to contract out additional functions, as is happening with some of its training programs. A third is to improve the capacity of line ministries so they can undertake these functions, as is being planned in the Public Sector Management Project currently being prepared. 29 LIST OF REFERENCES World Bank. 1990. Project Performance Audit Report "Bolivia: Emergency Social Fund Project (Credit 1829-BO). Report No. 8449. Washington, D.C. -------. 1991. "Honduras: Social Investment Fund Project (Credit 2212-HO)." Latin America and the Caribbean Regional Office. Staff Appraisal Report No. 9148-HO. Washington, D.C 1991. "Honduras: Social Investment Fund Project (Credit 2212-HO)." President Report No. P-5432-HO. Washington, D.C -------. 1991. "Honduras: Social Investment Fund Project." Development Credit Agreement 2212- HO. March 14. Washington, D.C -------. 1992. "Honduras: Second Social Investment Fund Project (Credit 2401-HO)." Latin America and the Caribbean Regional Office. Staff Appraisal Report No. 10451-HO. Washington, D.C -------. 1992. "Honduras: Second Social Investment Fund Project (Credit 2401-HO)." President Report No. P-5746-HO. Washington, D.C -------. 1992. "Honduras: Second Social Investment Fund Project." Development Credit Agreement 2401-HO. July 29. Washington, D.C -------. 1992. "Honduras: Social Investment Fund Project (Credit 2212-HO)." Latin America and the Caribbean Regional Office. Mid-Term Review. Report No. 10281-HO. Washington, D.C. -------. 1994a. World Development Report 1994. Infrastructure for Development. World Development Indicators. New York: Oxford University. -------. 1994b. "Honduras Country Economic Memorandum/Poverty Assessment." Latin America and the Caribbean Region. Report No. 13317-HO. Washington, D.C. -------. 1994c. Project Completion Report. "Honduras: Social Investment Fund Project (Credit 2212- HO)." Report No. 13573. Washington, D.C. -------. 1994. "Poverty Alleviation and Social Investment Funds. The Latin American Experience." World Bank Discussion Paper No. 261. Washington, D.C.