Report No. 22070 Honduras Public Expenditure Management for Poverty Reduction and Fiscal Sustainability June 28, 2001 Poverty Reduction and Economic Sector Management Unit Latin America and the Caribbean Region Document of the World Bank CURRENCY EQUIVALENTS (Exchange Rate Effective April 10, 2001) Currency Unit = Honduran Lempira HNL. 15.29 = US$1 US$0.06 = I Honduran Lempiras FISCAL YEAR January 1 - December 31 Vice President: David de Ferranti Country Director: Donna Dowsett-Coirolo Sector Director: Ernesto May Lead Economist: Ian Bannon Task Manager: Yasuhiko Matsuda Honduras Public Expenditure Review Main Acronyms and Abbreviations AECO Asociación Educativa Comunitaria BCH Banco Central de Honduras CESAMO Centro de Salud Área Metropolitana CESAR Centro de SaludÁrea Rural CG Contraloría General CONATEL Comisión Nacional de Telecomunicaciones DGP Dirección General de Presupuesto DGSC Dirección General de Servicio Civil EAP Escuela Agrícola Panamericana ENA Escuela Nacional de Agricultura ENEE Empresa Nacional de Energía Eléctrica ENP Empresa Nacional Portuaria EPHPM Encuesta Permanente de Hogares de Propósitos Múltiples FHIS Fondo Hondureño de Inversión Social HIPC Heavily Indebted Poor Countries HONDUTEL Empresa Hondureña de Telecomunicaciones IDB Inter American Development Bank IHSS Instituto Hondureño de Seguridad Social IMF International Monetary Fund MTEF Medium Term Expenditure Framework NFPS Non-Financial Public Sector NPV Net Present Value OECD Organization for Economic Cooperation and Development PAHO Pan-American Health Organization PEM Public Expenditure Management PER Public Expenditure Review PETS Public Expenditure Tracking Survey PF Poverty Fund POA Programa Operativo Anual PRAF Programa de Asistencia Familiar PROHECO Programa Hondureño de Educación Comunitaria PRSP Poverty Reduction Strategy Paper PTR Pupil-Teacher Ratio SANAA Water and Sewerage Company SEFIIN Secretaría de Finanzas SEP Secretaría de Educación SER Social Expenditure Ratio SIAFI Sistema Integrado de Administración Financiera SISPU Public Investment Reporting System SOPTRAVI Secretaría de Obras Públicas, Transporte y Vivienda SS Secretaría de Salud STS Staff Tracking Survey UNAH Universidad Nacional Autónoma de Honduras UNAT Unidad de Apoyo Técnico UPEG Unidad de Planificación y Evaluación de UPN Universidad Pedagógica Nacional USAID United States Agency for International Development WDR World Development Report  Honduras Public Expenditure Review Table of Contents Executive Sum m ary ...................................................................................................................................... I Public Expenditure Management for Poverty Reduction and Fiscal Sustainability ................................... i Fiscal Sustainability.............. ........................................... i Public Sector Pay, Employment and Personnel Management ..................................... n Public Expenditure Management ................................................... In Selected Sectoral Policy Issues: Education............................................... vii Selected Sectoral Policy Issues: Infrastructure ................................. ..... vii Strategies for Strengthening the Institutional Capacity for Poverty Reduction ......... ........ viii CHAPTER 1 FISCAL SUSTAINABILITY IN THE HIPC CONTEXT.........................................1 Background ...............................................................I The Road to Becoming a HIPC .................................................1 The Road after HIPC ........................................................3 The Size of the Public Sector....................................................4 Economic Classification of Central Government Spending ........................ .......5 Functional Distribution of Central Government Spending .......................... ..... 8 Social Spending and the PRSP ..................................................9 CHAPTER 2 PUBLIC SECTOR PAY, EMPLOYMENT AND PERSONNEL MANAGEMENT .12 Public Sector Pay and Employment Trends .............................................. 12 Personnel Management......................................................15 Personnel Management in the Health and the Education Sectors: Findings of Staff Tracking Survey ....17 Conclusions and Recommendations ................................................ 26 CHAPTER 3 PUBLIC EXPENDITURE MANAGEMENT ......................................................... 28 Institutions and Budgetary Performance................................. ........... 28 Budget Process in Honduras ..................................................29 Institutional Roots of Performance ......................................... .....37 Conclusions and Recommendations ............................................... 43 CHAPTER 4 HEALTH........................................................................................................................... 46 Overview of the Sector ......................................................46 Health Expenditure in Honduras..................................................47 Equity and Access.........................................................49 Efficiency and Service Quality .......................................... .......51 The Crisis in the IHSS .......................................................55 Conclusion: Institutional Constraints to Improved Sector Performance and the Agenda for Reform......57 CHAPTER 5 EDUCATION ............................................................................................................60 Overview of the Sector ......................................................60 Education Expenditure in Honduras .................................... ......... 62 Equity of Access to Education and of Educational Outcomes................................. 65 Improving Equity, Efficiency, Quality and Coverage in Public Education .........................67 Regulatory Failure and the Agenda for Reform ......................................73 CHAPTER 6 INFRASTRUCTURE SECTORS ............................................................................... 75 Introduction ..............................................................75 Sector Structure and the Role of the Public Sector..................... ...............76 Main Characteristics of Infrastructure Performance in Honduras..........................77 Broad Patterns of Public Expenditures............................................80 Policy Issues and Priorities...................................................82 Conclusions ..............................................................90 Honduras Public Expenditure Review Tables Table 1.1: Consistent Debt-Deficit Limits .................................................4 Table 1.2: Average Annual Gross Disbursements 2000-15...................... ................4 Table 1.3: NFPS Public Spending.......................................................5 Table 1.4: Economic Distribution of Central Government Spending................................6 Table 1.5: Sustainable Central Government Wages and Salaries....................... .......6 Table 1.6: Functional Classification of Central Government Spending..............................9 Table 1.7: Social Spending Scenarios....................................................10 Table 1.8: Funding For Additional Social Spending................................ ......10 Table 2.1: Wage Expenditures By Employment Category........................... .......12 Table 2.2: Central Government Employment Composition.................................... .............13 Table 2.3: 1998 Private Sector Incomes (Monthly) ............................................... .............13 Table 2.4: Evolution Of Average Monthly Wages (In 1990 Ls.) .......................... .............14 Table 2.5: Health Staff Not Located To A Specific Workplace In The Central Database.............. .......19 Table 2.6: Hiring Processes And Contractual Status In The Health Secretariat........................21 Table 2.7: Hiring Processes And Contractual Status In The Education Secretariat................22 Table 2.8: Proportion Of Workforce That Has Migrated Since Their "Acuerdo" Was Issued (%) ..........23 Table 2.9: Other Paid Employments Of Health Staff - Survey Findings............................24 Table 2.10: Other Paid Employments Of Education Staff - Survey Findings..........................24 Table 2.11: Hours Of Work Of Health Employees......................... ...................25 Table 2.12: Hours Of Work Of Education Employees..........................................25 Table 3.1: Basic Elements Of Public Expenditure Management: "Three-Level" Analysis .................28 Table 3.2: In-Year Budget Re-Allocation - Ministry Of Health, 1994-1999..........................34 Table 3.3: Asignaciones Globales And Their Probable Agency Destination In 2000 ....................35 Table 3.4: Asignaciones Globales And Servicios Centralizados (% Of Total Executed Budget) ............36 Table 3.5: Revenue And Expenditure Shares Within The Non-Financial Public Sector (Nfps) .... ......40 Table 4.1: Health Outcomes 1970-95.................................................... 46 Table 4.2: Expansion of The Ss Network, 1990-99...........................................47 Table 4.3: Public Health Spending By Agency 1990-99.......................................48 Table 4.4: Structure of Ss Spending By Program, 1999, %..................................... 48 Table 4.5: Structure of Ss Budget By Expenditure Group, %.................................... 49 Table 4.6: Suppliers of Ambulatory Attention By income Quintile................................ 49 Table 4.7: Distribution of S Attentions, By Population Quintile.................................. 49 Table 4.8: Incidence of Health Problems and the Decision to Seek Attention......................... 51 Table 4.9: Incidence of Hospitalization (% Per Person/Year).................................... 51 Table 4.10: Staffing and Productivity in Health Posts.......................................... 52 Table 4.11: Average Number of Months of Closure Per Year In Rural Health Posts.......................... 52 Table 4.12: SS Hospital Productivity, 1991-1998 .............................................53 Table 4.13: Insured Population of IHSS................................................... 56 Table 4.14: Health Expenditure By Supplier and Quintile/1...................................... 56 Table 5.1: Trends In Educational Outcomes............................................... 60 Table 5.2: Education Outcomes Compared With Inteational Benchmarks........................... 60 Table 5.3: Educational Coverage, 1990-99................................................ 60 Table 5.4: Age-Specific Secondary School Attendance By Cycle, 1999, % Of Cohort................... 61 Table 5.5: Supply of Preschool, Primary And Secondary Education in Honduras...................... 61 Table 5.6: National Education Accounts By Funding Source and Financial Agency, 1998, L Million. .......... 62 Table 5.7: Public Education Spending By Agency 1990-99..................................... 63 Table 5.8: Structure of Sep Spending By Program, 1990-99, %................................... 64 Table 5.9: Structure of Sep Budget By Expenditure Group, %................................... 64 Table 5.10: Attendance at A Primary School, March 1999........................................ 65 Table 5.11: Trends in Primary Attainment, 1987-97.............................................65 Table 5.12: Age-Specific School Attendance By Household Income Quintile, 1999.....................66 Table 5.13: Educational Attainment Of 25 Year Olds By Income Quintile % ............... ...........67 Table 5.14: Impact Of Single Teacher Schools On Incomplete Grade Coverage........................67 Table 5.15: Dispersion Of Resource Assignment And Outcomes In Nacaoe........................................................ 68 Table 5.16: Incidence And Cause Of Closures In A Sample Of 24 Primary Schools, 1998.. .................68 Honduras Public Expenditure Review Table 5.17: Missed Days In Primary Schools: Household Survey Evidence ...........................68 Table 6.1: Principal Agencies Involved In Infrastructure Services In Honduras ..........................77 Table 6.2: Household Survey Data On Trends In Access To Water And Sanitation.....................78 Table 6.3: Honduran Infrastructure Performance Compared With Central America .... ...........................78 Table 6.4: Public Investment In Infrastructure, 1990-99..... ....................................... 80 Table 6.5: Distribution Of Public Expenditures And Investments By Sector................................... 81 Table 6.6: Infrastructure Spending By Financial Source % Of Total Spending - Current Plus Capital..............82 Table 6.7: Targeting Performance Of The Electricity Subsidy In Honduras ..................... 87 Table 6.8: Public Spending On Road Investment And Maintenance..................................... 89 Table 6.9: Estimated Maintenance Needs For The Roads System ....................................................................... 90 Figures Figure 1.1: Central Government Revenue and Expenditure 1960-98............................... 2 Figure 1.2: Non-financial Public Sector Deficit 1960-98...................................... 2 Figure 1.3: International Comparison of Central Government Spending... ......................... 5 Figure 1.4: Deficit and Public Investment................................................ 7 Figure 1.5: Public Investment and Disbursements............................................ 7 Figure 1.6: Cumulative Additional Resources............................................. 11 Figure 3.1: Credibility of the Honduran Budget - In-year Deviations by Agency..................... 38 Figure 3.2: Rigidity in the Central Government Budget....................................... 42 Figure 4.1: Correlation Between Health Spending and Per Capita Income at Departmental Level.............. 50 Figure 5.1: Correlation Between Education Spending and Per Capita Income at Departmental Level, 1999.. 65 Boxes Box 2.1: STATUTORY PAY AGREEMENTS ............................................... ............ 18 Box 3.1: COORDINATION BETWEEN CAPITAL AND RECURRENT EXPENDITURE PLANNING .. ................30 Box 3.2: GOVERNANCE AND TRANSPARENCY IN HONDURAS .................. ........................36 Box 3.3: SISPu - AN INSTRUMENT FOR AN EFFICIENT PUBLIC INVESTMENT PROCESS ..................................42 Box 4.1: MEANS-TESTING IN CHILE: THE FICHA CAS.................................................55 Box 5.1: DIAGNOSING THE CAUSES OF EDUCATIONAL FAILURE .................................69 Box 6.1: THE SAN PEDRO SULA WATER CONCESSION ................................................ 85 Annexes ANNEX 1: DEBT SUSTAINABILITY ANALYSIS - A METHODOLOGICAL NOTE .............. .................91 ANNEX 2: SELECTED STATISTICAL RESULTS OF THE STAFF TRACKING SURVEY ............. ...............93 ANNEX 3: STAFF TRACKING SURVEY METHODOLOGY.................................................99 ANNEX 4: HONDURAS: TRACKING POVERTY-REDUCING EXPENDITURE ASSESSMENT AND ACTION PLAN......... 108 ANNEX 5: AN ASSESSMENT OF BUDGETING SYSTEMS: HONDURAS ................... ................... 114 Acknowledgements This report was prepared based on the World Bank missions in March and May, 2000. The World Bank team was composed of Yasuhiko Matsuda, Humberto L6pez, Suzanne Dove, Jyoti Shukla, Steve Webb, Maria Albino, Josef Trommer, and George Guess. Ian Walker, Elmis Cardenas, and Marco Moncada from the ESA Consultores participated in the main mission in May, 2000. The peer reviewers were Allister Moon and Vicente Paqueo. Throughout the preparation of the report, the team benefited from close guidance from lan Bannon, (Lead Economist). Comments on various drafts were received from Helena Ribe, Maria-Luisa Escobar, Joel Reyes, David Warren, William Dorotinsky, Anand Rajaram, Santiago Herrera, and Norman Hicks. In carrying out the mission, the team received logistical support from Iris Medina and Carol Mejia at the Resident Mission in Tegucigalpa, and from Meseret Kebede and Patricia Mendez at headquarters.  Honduras Public Expenditure Review EXECUTIVE SUMMARY Public Expenditure Management for Poverty Reduction and Fiscal Sustainability 1. As a highly-indebted poor country (HIPC), Honduras faces a dual challenge of reducing poverty while ensuring medium-term sustainability of its public spending to avoid recurrence of over indebtedness. While Honduras should continue to promote private participation in productive and infrastructure sectors to enhance growth, the government still has a critical role to play in achieving the long-term poverty reduction goals. The government is now preparing a Poverty Reduction Strategy Paper (PRSP) to guide the country's medium-term policy priorities in poverty reduction. Besides identifying key policy programs that are expected to have significant impact on poverty reduction, the PRSP needs to be consistent with a realistic fiscal framework, and robust institutional arrangements and sound fiscal management that ensure good public expenditure outcomes. This Public Expenditure Review (PER) is intended to contribute to the government's overall poverty reduction efforts through efficient use of public resources and fiscally sustainable improvement in public services. It provides an assessment of Honduras' institutional capacity for good public expenditure management (PEM) and identifies key policy priorities in selected sectors (health, education and infrastructure). Fiscal Sustainability 2. Honduras appears to have overcome the problem of weak fiscal discipline that threatened the country's macroeconomic stability and growth until the early 1990s. Currently, the size of the Honduran public sector is not excessive in comparison to other Central American and HIPC countries. But in the absence of a medium-term fiscal strategy and institutional mechanisms to ensure that new policy decisions' fiscal implications are duly considered before their legislative passage, Honduras is still vulnerable to policy decisions that effectively imply not only present increases in public spending but also future government liabilities. Particularly worrisome are the increasing wages and salaries, which are currently on an unsustainable path due in part to the implementation of the so called Estatutos. The government faces a stark policy choice between increasing funding for new social programs on the one hand, and financing the rising wage bill on the other. Developing reliable estimates of the sustainability of public sector wages and salaries, and fiscal costs of desired social programs, therefore, is critical for helping the government make informed policy and expenditure decisions. 3. Sustainability ofpersonnel expenditures: Projections under different macroeconomic and fiscal scenarios would indicate that the projected wage bill of 9.5 percent of GDP for 2001 is well above the sustainability level (by 2.3-3.1 percent depending on growth scenarios). The government might maintain current salary levels by either increasing taxes by about 2.5 percent of GDP, or cutting other budget items. Since the simulations are based on both purchases of goods and services and transfers at prudent levels, the likely candidate would be capital investment cuts. However, the required cuts to bring fiscal sustainability back would take capital spending and net lending by the central government down to about 5.5 percent of GDP. With capital transfers and net lending at 4-5 percent of GDP, the central government fixed capital formation would be capped at about 1.5 percent of GDP, which is probably unrealistic in view of the country's reconstruction needs, and could endanger future growth. Further adjustment of the public sector wage bill seems unavoidable, though for the adjustment to be sustainable it should be part of a more comprehensive approach that would review the Estatutos, especially those of teachers and doctors. 4. Sustainability of social sector spending: The macroeconomic limits on deficits imply limits to the availability of funds for the social sectors, given the likely levels of tax effort. Under different growth scenarios ranging from 2 percent to 5 percent, by increasing the Social Expenditure Ratio (SER) from the 1990s' average level of 36 percent to 40 percent of the central government spending, the country would ensure a minimum of US$1 billion in additional resources for the social sectors. This amount is about the Honduras Public Expenditure Review ii same as the HIPC debt service relief that Honduras will benefit from over the coming years. Overall social spending levels could range between US$8.7 and US$13.7 billion depending on GDP growth and social allocations. However, it should be noted that just to hold the spending levels on the existing programs constant on a per capita basis, the social budget would have to increase to reach a total of US$10.4 billion (2000 prices) over the 2000-2015 period. Therefore, after taking into account the required additional needs implied by a growing population, the additional resources for new programs, or improvements in existing ones, could be as much as US$3.2 billion (5-percent growth/44-percent SER), or they could be negative US$1.7 billion (2-percent growth/36-percent SER). Thus, when the SER is fixed at 36 percent, unless growth averages more than 4 percent per year, Honduras would devote a decreasing amount of resources to the social sectors in real per capita terms. If we look at a median case (SER of 40 percent and 3.5-percent average growth), additional available resources would be about US$500 million spread over the next 15 years. If these additional resources are allocated exclusively to new programs, existing programs would have no financing (in real per capita terms) above the 1999 levels. 5. This analysis indicates that Honduras' ability to allocate increasing amounts of public resources to social programs depends critically on the country's growth prospects and the government's ability to allocate increasing shares of its spending to the social sectors. The preferential allocation to social sector programs, in turn, depends on the government's ability to contain the rising personnel expenditures, which are currently on an unsustainable path. This also highlights the importance of increasing efficiency of social spending. Public Sector Pay, Employment and Personnel Management 6. Effective management of personnel expenditures is critical for both maintaining aggregate fiscal control and ensuring efficient delivery of public services. Both fiscal discipline and efficient service delivery are essential for achieving the government's poverty reduction objectives. Increases in public sector wages and salaries, even in the social sectors (i.e., teachers, health sector professionals), are unlikely to result in improved social services, unless they are accompanied by incentives to perform better and mechanisms to hold staff accountable for their conduct and performance. The assessment of the personnel management practices carried out with a focus group and a staff tracking survey in the health and the education sectors reveals a variety of institutional weaknesses in personnel management, which the government should tackle as a high priority. 7. Personnel management - general assessment: A focus group of a cross section of civil servants revealed a number of weaknesses in the central government's personnel management, characterized by a variety of informal practices that circumvent either the spirit or the letter of the existing rules and regulations. Specific deficiencies identified include: * Inadequate identification of personnel needs by line ministries, and a resultant mismatch between the ministries' authorizations for hiring specific types of personnel and their real needs; * Discretionary recruitment procedures which invite politicization of the actual recruitment process; * Low public sector salaries, that encourage or even necessitate income supplement by holding second jobs - salary increases have typically been awarded based on non-professional criteria rather than on job performance, although this situation has been partially corrected with the recent salary adjustment; * Deduction of public employees' salaries (about 5 percent of monthly salaries) by political parties; and Honduras Public Expenditure Review iii * Selective use of personnel evaluation, which some civil servants perceive as an excuse for dismissal. 8. Staff tracking survey: The survey to track actual staff deployment in the health and education sectors provides additional evidence of the weak personnel management that hampers effective service delivery. The survey confirms that there are important weaknesses in the staffing information and payments administration systems in the Health and Education Secretariats, and warns that their modernization should take account of the need to build in rigorous auditing systems. The study indicates that there is a limited problem of "ghost" staff who do not appear to work anywhere but draw their salary regardless, but it finds that absenteeism is a major problem in the Health Secretariat. Multiple employment is also an important problem in health, reflecting "employee capture" of the system by senior doctors. In education, arguably the most serious problem is bureaucracy: almost 40 percent of staff is engaged in administrative functions even though the Education Secretariat carries out little sector-wide planning and has transferred the responsibility to manage the sector's capital investment programs (to FHIS). Finally, both ministries are plagued by the migration of posts between service production units and towards the administrative offices. This is due to a combination of employee capture (the post belongs to the individual) and administrative inflexibility (managers cannot get new posts where they need them so they connive at shuffling exiting posts between institutions). 9. Personnel management reforms: The government has carried out certain actions to improve government-wide personnel management, such as an initial reform of the civil service pay structure, and is currently working on a new civil service law. These efforts should be continued, but a more concerted approach at reform of the public personnel management would be needed, even though one of the most serious problems, the constraints imposed by the Estatutos, will be politically difficult to address. As a first step, the government should aim to develop a coherent and comprehensive public sector pay and employment policy that addresses simultaneously fiscal affordability and performance incentives in the public sector. The articulation of such a policy will not eliminate the ability of powerful and organized groups to push for further salary increases but would strengthen the government's hand in politically difficult negotiations and stand a better chance of mobilizing public opinion, especially when unreasonable demands pose a serious threat to fiscal stability or would crowd out other priority expenditures. As part of this effort, it would be advisable for the government to begin developing a medium-term fiscal envelope for personnel expenditures, which provides a top-down resource constraint on the size of the public sector employment. 10. At the same time, the government should develop bottom-up (i.e., initiated by line ministries) assessments of staffing needs, based on a functional review of each government agency's roles, functions, and organizational structure, so as to arrive at a more or less reasonable picture of each agency's medium- term staffing needs. These staffing needs should be costed out and be reconciled with the medium-term personnel expenditure framework. The government should also continue efforts to develop adequate capacity for monitoring and controlling public sector employment, for example, by developing a more reliable personnel data system, and an institutional mechanism for its maintenance and operation. These measures should be developed on a priority basis in the health and education sectors, where the weak personnel management seriously hamper the effectiveness and efficiency of key poverty reduction programs. The much more difficult challenge of performance management will have to be addressed once there is a minimum capacity for control, although there may be scope for greater use of community supervision as a substitute for administrative control and accountability, in areas such as primary schools and rural health posts. Public Expenditure Management 11. Aggregate fiscal discipline: In recent years, Honduras has maintained a reasonable degree of fiscal discipline, avoiding lax fiscal policy and financial management practices that weaken fiscal control Honduras Public Expenditure Review iv (e.g., weak commitment control that results in arrears). However, to the extent the government's capacity to maintain fiscal discipline has improved since the mid 1990s, it remains relatively fragile as it depends less on robust formal institutional mechanisms (e.g., quantitative fiscal targets based on reliable macroeconomic projections) than on a limited number of informal practices, particularly the strong centralizing role of SEFIN, the Budget Department in particular, in fiscal and financial management. SEFIN is the dominant actor in budgeting and financial management. A dominant ministry of finance is a key institutional element to strengthen fiscal discipline. However, the current state is still not very desirable from the point of view of both sustainable aggregate fiscal discipline and operational efficiency in public expenditure management. 12. Despite the relative effectiveness of fiscal control via SEFIN's institutional dominance vis-A-vis line ministries, the most serious threat to aggregate fiscal discipline, the large increases in personnel expenditures, comes from an arena that is effectively outside SEFIN's control. These increases were negotiated directly by the powerful labor unions (gremios) in the health and education sectors (sometimes even bypassing the sectoral ministry) and supported by the Congress. In light of the budget system's inability to produce binding policy and budget decisions, tight central control of financial management becomes a necessary second-best alternative to control aggregate spending levels. This method of fiscal discipline, however, also carries a high cost in terms of shifting de facto policy priorities during the year and reducing operational efficiency. 13. Prioritization in resource allocation: From the poverty reduction point of view, the system's capacity to allocate budgetary resources to those programs designed to address poverty, such as social programs, is of great concern. In the mid to late 1990s, when the overall public spending dropped due to fiscal adjustments, the level of social spending in Honduras stagnated as percent of GDP (6.8 percent in 1998 compared to 7.7 percent in 1990). However, the government protected social spending as a share of total public spending (33.2 percent in 1998 compared to 33.8 percent in 1990). Partly as a result, outcomes in primary health and primary education either improved or, at least, stayed at a reasonable level throughout the 1990s. 14. Despite this relatively favorable picture within a difficult environment of fiscal adjustment, Honduras still spends below the regional average in health. Further improvement in health outcomes may require some additional funding, as well as improvement in the efficiency and equity of health spending. In education, in contrast, Honduras spends above the regional average. While some progress in expanding primary and pre-school coverage took place, the education outcomes achieved during the 1990s do not correspond to what might be expected from the level of spending. Measures to improve efficiency are relatively more important in education, although there is some need for targeted increase in educational spending as well (e.g., to raise public provision of secondary education). Given the overall fiscal constraint, however, this would require reallocation of available resources either within the sector, or from other lower-priority sectors, rather than by increasing the total spending level. In infrastructure, improvement in sector performance has been more limited, although important steps have been taken in some of the sub-sectors (e.g., the airports concession, the establishment of the road maintenance fund). In all three sectors, lack of effective access especially among the rural poor is a continuing challenge, which requires more effective targeting of expenditures. 15. The budgetary institutions' capacities to prioritize policies and allocate expenditures accordingly are quite weak. There is a tendency for expenditure items to be included in the budget without appropriate considerations for their strategic importance and/or for their longer-term fiscal affordability. There are some built-in institutional obstacles to reallocate resources to identified priorities, such as: (i) the lack of comprehensiveness of the budget; (ii) the structural rigidity of the budget, where more than 50 percent of the spending is either constitutionally and legally pre-committed or difficult to cut back for political and practical reasons, and the rising trend of the wage bill, which would further limit budget flexibility; and (iii) systematic use of non-transparent budget classifications (e.g., asignaciones globales Honduras Public Expenditure Review v and servicios centralizados). Besides, the significant deviations between approved and executed budgets that are observed year after year indicate that the formal budget, as approved by the Congress, is not fully credible as an instrument for resource allocation and program implementation. 16. Operational efficiency: HIPC assistance will allow Honduras to devote a larger share of public resources to key priority programs, if growth of non-priority spending is effectively controlled. In addition, Honduras may be able to assign additional resources to poverty reduction objectives by reallocating resources from lower to higher priorities. But many of the remaining challenges, such as improving educational outcomes of children already enrolled in school or low productivity in rural health posts, cannot easily be addressed by increased resources alone, and gains in program outcomes depend critically on improving the efficiency of resource management in the existing programs. 17. One of the obstacles to operational efficiency is tight expenditure control by SEFIN, combined with the excessively detailed line item budget that requires line ministries to spend much time reformulating the budget during the fiscal year. The reform of the government's financial management with the pilot implementation of the Integrated Financial Management System (SIAFI) is an important step in this direction. SIAFI now covers six main ministries representing 70 to 80 percent of the central government budget and financial transactions. But its further progress, especially as an instrument to facilitate gradual devolution of financial management decision-making to line ministries, would require a reform of the legal framework governing government financial management and reform of the organizational structure and functions of SEFIN, whose centralist tendency may be hampering progress of financial devolution within the central government. Gradual relaxation of the central control should be accompanied by strong efforts to strengthen the government's capacity to monitor both expenditure and program implementation. Among them is strengthening of the Expenditure Programming and Execution Units (UPEG) in SEFIN and in line ministries and of the Controller General's Office (CGR), which is currently too weak both institutionally (for lack of effective independence and limited budget) and technically (for lack of qualified auditors). 18. Another key obstacle that is found in many spheres of public administration is the weak personnel management, as discussed above, especially in the health and education sectors. Additional reform measures to advance financial devolution within the central government and strengthen personnel management are critical for improving efficiency in public expenditures, and therefore be top priorities in the government's poverty reduction strategy. 19. To compensate for the problems of weak incentives and capacity in the public sector, Honduras has attempted to complement the state's role in service delivery with participation of the private sector (mainly in infrastructure) and communities (most notably in PROHECO schools, but also in social investment programs by FHIS). The government might consider pursuing these approaches more widely and systematically in areas currently not addressed with a similar arrangement (e.g., primary/community health care). 20. Capacity to track public spending: Honduras has made significant progress improving the quality of the expenditure data with the gradual introduction of the SIAFI and the accompanying procedural and presentational improvements in its budget. But, there are still some shortcomings that obfuscate the tracking of spending programs, particularly poverty-related programs, including: * Crude macroeconomic and fiscal scenarios used in the annual budgeting exercises, which does not permit establishment of a medium-term baseline to estimate the additionality of pro-poor spending; Honduras Public Expenditure Review vi * The use of opaque budget categories (asignaciones globales, servicios centralizados), which obscure the actual allocations of significant portions of the budget and make monitoring difficult, and unreliable classification of the functional distribution of public expenditures; * The lack of comprehensiveness in the government budget, which omits the budgets of decentralized and deconcentrated units that are important for delivery of poverty reduction programs; * The weak audit mechanisms and unclear roles and functions of internal and external audits; * Incomplete coverage of the effective integrated financial management system (SIAFI); and * Weak government capacity for program evaluation. Selected Sectoral Policy Issues: Health 21. Sector performance: In spite of the country's continuing high level of poverty and low rate of economic growth, and comparatively lower allocation of public expenditures to the sector, health outcomes in Honduras have greatly improved over the last couple of decades. The advances (e.g., the significant decrease in the infant mortality rate, a corresponding increase in life expectancy at birth) are due to the relative effectiveness of the public sector's primary health programs in areas such as vaccination, control of diarrhea and respiratory infections, and a greater emphasis on integral mother and child health care, as well as to the expansion of the public sector health network during the last decade. With the expansion, there has also been a shift of ambulatory consultations away from hospital outpatient clinics (down 24 percent) toward lower-cost ambulatory facilities (up 28 percent), which suggests an improvement in network rationality. 22. Policy priorities: In recent years the main emphasis in health has been on expanding coverage in primary services. Although there are still important differences in access to clinical services, the majority of the population is now covered by basic services and the results in the improving health status of the population are clear. The sector now needs to enhance its efforts on the institutional reforms needed to embed quality and value-for-money into the public health system and to promote the mobilization of private resources for non-primary treatments. The necessary measures to be taken include: * Improve effective access of the rural poor to the health facilities - this would entail extending their effective operating hours, and improving their service quality, both of which would require better regulation from the center; * Strengthen the sector's planning and regulatory capacities; * Further improve efficiency in the sector's financial management through the development of modern management information systems, the linking of financial and physical programming and the decentralization of decision-making and responsibility, linked to performance targets; * Improve quality and efficiency through measures that might include the systematic use of NGOs and private providers for basic health care, and raise the productivity of the public sector health workers through the modernization of labor contracts and introducing more performance orientation to service delivery; * Promote private sector provision of health care by carefully defining the services that will be subsidized and establish full cost recovery for tertiary services (including from beneficiaries of the IHSS and private insurance) and introduce strict means testing for subsidized access to such Honduras Public Expenditure Review vii services. There is also a need to develop better regulation of the private insurance market in health in order to increase coverage in this area, which is presently insignificant; and * Reform the social security system by separating the three functions (health insurance, old-age pension, and health service provision), and re-establishing rational contribution levels that are sufficient to cover the efficient cost of providing a basic package of health insurance. Selected Sectoral Policy Issues: Education 23. Sector performance: Honduras has registered steady improvement in many key educational indicators over the last three decades, including the doubling of the average number of years' study and pre-school attendance, the fall in the illiteracy rate, and the attainment of a relatively high level of primary enrollment. While the urban-rural gap still remains large, the rural population benefited relatively more from these improvements. However, relative to its comparatively high funding level (7.1 percent of GDP, of which 4.4 percent are from public sources), Honduras' achievements do not fare particularly well vis- A-vis those of its neighbors such as El Salvador. In particular, the low secondary coverage stands out as one of the main weaknesses, which is partly a result of the limited public provision of secondary education. Public provision is also limited at the pre-primary level, where informal private sector suppliers have grown rapidly over the last decade. At the tertiary level, there is clear evidence of the "crowding out" effect, where the existence of low-cost public universities arrests the growth of private alternatives. 24. Policy priorities: The most important challenge facing Honduras' education sector is to improve efficiency in primary education rather than expanding physical supply of primary schools as such. Emphasis should be on getting the child into school at the right age and providing sufficient educational quality to reduce repetition and failure rates. This would require a better targeting of finangial and human resources to remote schools in areas with access deficiency, and more effective mechanisms for regulating teacher performance and for providing them with appropriate administrative and pedagogical support. Mainstreaming of PROHECO-type community-based schools should be considered as one of the options for improving the quality and efficiency of educational services. 25. The second major challenge is expanding secondary coverage overall, and in particular for the poor. There is a need for a considerable expansion of state provision in this area, while encouraging private provision (both formal and informal) and avoiding crowding out when state provision expands. The third major challenge is improving equity, efficiency and effectiveness in the University sector in order to reduce inequitable subsidies to the non-poor. Selected Sectoral Policy Issues: Infrastructure 26. Sector performance: Despite a number of important initiatives, and some advances in improving access in infrastructure since the mid 1990s, infrastructure performance remains an impediment to competitiveness and growth in Honduras. According to the World Micro-Economic Competitiveness Index, Honduras ranked 58 out of 59 countries. Even though public expenditures on the infrastructure sectors appear adequate, virtually all indicators of access, efficiency and service quality continue to be low in absolute terms and in comparison with other countries in the region. In large part this is due to the fact that sector structures and the related regulatory environment are not grounded in clear objectives of improving operational efficiency and quality of service to consumers. This applies equally to all infrastructure services whether publicly or privately managed. Significant levels of public spending for infrastructure and subsidies for infrastructure services have been provided predominantly to address the question of affordability rather than addressing the issue of greater access, especially to the poor. A notable exception is the social investment fund (FUHS) activities, which seem to have had a positive impact in improving access to water and sanitation services. The consequence of this approach has been Honduras Public Expenditure Review viii to support policies which favor consumers already connected to existing service networks rather than to expand access to those who are currently not covered by these services. On the issue of maintenance, Honduras has made some progress in the roads sub-sector with the establishment of the road maintenance fund and the use of community-based micro enterprises. Progress has been more limited in other sectors. 27. Policy priorities: Improving infrastructure services remains an important priority for Honduras to stimulate growth and reduce poverty. Levels of infrastructure and infrastructure services are crucial to growth and the poor tend to value highly access to key infrastructure services (clean water and sanitation, electricity and transportation). To improve performance, it is crucial that Honduras define a clear and coherent strategy for private sector participation and public spending for infrastructure services. Major policy recommendations include: * Given existing large infrastructure gaps, private participation in the infrastructure sectors will be essential for the future; * To maximize development gains, private sector participation arrangements should be structured to give greater weight and incentives to improving access, quality and efficiency in service delivery rather than maximizing short term fiscal gains; * The greatest impact on improving the quality and efficiency of infrastructure services is likely to come from better sector regulation, promoting competition wherever possible, and regulating with clear objectives of improving efficiency and quality of service; * Regulatory autonomy and effectiveness can be improved by merging the various regulatory entities into a single multi-sector regulatory body; * Government subsidies in infrastructure sectors need to be restructured, toward improving access to infrastructure services which is more likely to have an impact on reducing poverty, and away from improving affordability to those with existing network access; and * Expenditure patterns should place priority on maintenance activities. Strategies for Strengthening the Institutional Capacity for Poverty Reduction 28. A comprehensive approach: Despite some progress made with the state modernization program and other related efforts, there still remain multiple challenges to strengthen the Honduran state's institutional capacity for poverty reduction. The role of public expenditures is crucial. In addressing these challenges, it is important to recognize that a good institutional arrangement for effective public expenditure management requires achieving the three inter-related objectives of aggregate fiscal discipline, allocation of resources to strategic priorities, and operational efficiency. Therefore, isolated measures that improve performance at one level while negatively affecting another should be avoided. For example, a typical trade-off found in many developing countries, and also in Honduras, arises from the use of cash budgeting with tight ex ante expenditure control. While this practice aides in maintaining aggregate fiscal control, it usually has a strong negative impact on operational efficiency, and thus on achievements of program objectives in public service delivery. Similarly, a common short-cut to improving specific aspects of public sector operations is to create special programs or entities that are exempt from many of the budgetary control requirements (e.g., rules and regulations for public procurement, personnel management). Such measures may improve operational efficiency of those specific programs and entities in the short run, but experience from other countries shows that their sustainability is often suspect, and they may also introduce perverse effects in government-wide management of resources such as undesirable fragmentation of public expenditures. Honduras Public Expenditure Review ix 29. Strategic sequencing of reform measures: With Bank and other donors' assistance, Honduras has been making steady progress in improving its public expenditure management. Among the measures taken, the pilot implementation of the SIAFI seems to be progressing reasonably well, though there is much more to be done. These ongoing efforts should continue, including the passage of a new legal framework that governs the government budgeting and financial management procedures (the so-called SIAFI Law) and a thorough reform of the roles, functions and organization of SEFIN. The government is also working on a new civil service law, as a necessary element of a public sector personnel policy that is badly needed. Other necessary measures to strengthen public personnel management is development of reliable and comprehensive personnel data systems and strengthening of the Direcci6n General de Servicio Civil, which currently lacks the required institutional capacity to manage public sector human resources. The analysis presented in this review also suggests additional priorities, some of which can and should be tackled in the short run, while others may take longer to achieve. 30. Efforts will need to be made to improve budgetary performance at all three levels. But, priorities may be first to consolidate the emerging capacity for aggregate fiscal discipline, while simultaneously introducing measures to improve operational efficiency both at the government-wide level and at specific sectoral level. The capacity for allocating resources according to the strategic priorities may be a more difficult challenge that can only be addressed over the medium to long term. Advanced OECD countries have pursued this by devolving greater autonomy for policy making and implementation to line agencies within an agreed fiscal envelope and a medium-term expenditure framework. Honduras, like most developing countries, still has some way before reaching this stage because granting greater financial and operational autonomy to line agencies without first establishing reliable central mechanisms for monitoring and control would raise the risk of corruption, waste and other undesirable outcomes without necessarily achieving the desired goal of greater efficiency in both allocation and execution of budgetary resources. Therefore Honduras should pursue a more gradual path toward improved strategic resource allocation. Continued efforts to refine the Poverty Reduction Strategy and to develop accompanying sector strategies that would guide intra-sectoral prioritization in more concrete detail would be a necessary prerequisite that should be vigorously pursued. The implementation of the national public investment system (SISPU) that is currently under development should also aid in achieving better control over allocation of capital expenditures. Over time, the country should strengthen policy-making capacities both at the center (e.g., SEFIN) and line ministries that include adequate mechanisms to vet the fiscal affordability of new policies. Relaxing some of the earmarked expenditures (e.g., university) and developing institutional means to discipline the Congress' budgetary decision-making would be a longer- term agenda. Immediate steps 31. One of the problems in Honduras' public expenditure management that can be tackled immediately is the need for improvement in the budget classification system for better fiscal transparency. The government should eliminate use of the opaque spending categories, especially asignaciones globales and servicios centralizados at least in reporting the budget execution, and continue its efforts to improve presentational clarity of the budget document. The immediate elimination of asignaciones globales may not be feasible for entities such as the judiciary, the legislature, and the Public Ministry because of their independence from the Executive branch, but should be possible for the Ministries of Health and Education, which jointly accounted for more than 40 percent of the total asignaciones globales in 2000. Similarly, to the extent asignaciones globales are used to cover costs that are not properly estimated (about a fifth of the total of budget of the Ministries of Health and Education), the immediate elimination of this category in the budget without developing a capacity to estimate those costs accurately and attribute them to appropriate spending categories will probably result in another form of non-transparent budgeting. However, the government should at least be able to report how the resources originally allocated as asignaciones globales were actually spent, when reporting on the final budget execution for all public sector entities using this spending category. With the SIAFI implementation, the quality of Honduras Public Expenditure Review x expenditure reporting is expected to improve. The government should take full advantage of the improved capacity for expenditure monitoring to make the information on budget execution available to the public. SEFIN should also strive to consolidate budgetary data from the variety of decentralized entities, at least for their execution ex post so as to carry out comprehensive reporting on public expenditure management. Short-term measures 32. To consolidate fiscal control, some procedural reforms and technical improvements are needed. One of the first steps should be to develop a capacity and a procedure for utilizing reliable macro fiscal forecasting to guide fiscal policy decisions over the medium term. For each fiscal year, the budget should be formulated within a pre-established macro fiscal framework for that year. Once this is accomplished, the government should gradually extend the time horizon of the forecasting, first by covering the fiscal year and one outer year. Eventually, the forecasting might cover the budget year plus two or three additional years to become the macro fiscal framework for medium-term expenditure planning. 33. Besides the projection of the macro fiscal envelope, the government should also develop capacities to project some meaningful spending sub-categories. Given the importance of controlling aggregate personnel expenditures, the government might consider starting this exercise with forecasting the medium-term trend in spending on wages and salaries. This would require, inter alia, information on the current size of the public sector employment and the pay level, and the expected trend into the future given the various sectors' staffing needs and plans. Improving the government-wide and sector-specific personnel data system should be one of the top priorities for this purpose. Government commitment to formulate and execute the budget within these sub-category ceilings is essential for this approach to produce meaningful results. The ceilings should be established early in the budget formulation process (before SEFIN receives spending requests from line ministries). 34. To improve the capacity for strategic allocation of resources, it is urgent that the government develop concrete sector strategies in key sectors in which the bulk of pro-poor public expenditure programs are to be implemented, the government should gradually develop a capacity to evaluate key policy programs, as it is being contemplated under the Economic and Financial Management Project. Initially the efforts may be somewhat ad hoc, and it may take a while for them to evolve into a coherent evaluation system with a full feedback loop into the policy and budgetary decision-making processes. But even ad hoc and selective evaluation of priority programs should provide valuable information about program performance and serve as a guide for resource (re-) allocation. Even before sector ministries develop capacities to formulate their budgets fully on the basis of program evaluation and costing, it is important that the sectoral budgets be formulated within broad sectoral ceilings consistent with the medium-term macro framework and established early in the budget process by SEFIN. 35. To improve operational efficiency, one of the key measures would be to modify the budget classification, which is currently excessively detailed and is detached from the operational realities of the line ministries. Much time is wasted in budget modifications during the fiscal year. Adopting broader line item categories without eliminating the relatively effective input control should alleviate the high transaction cost of budget modifications. To make the matter worse, Congress required its own approval of the most minute spending modification requests during part of 2000. This practice has been abolished in the mid year, and should not be repeated. 36. There are also a variety of measures to be taken to improve the incentive framework in the whole public administration. However, this (e.g., professionalization of the civil service) would require systematic and concerted efforts for a number of years. In the meantime, the government should pursue supplementary institutional measures to improve governance at the sector level, including privatization and concessioning of infrastructure services, and community participation in social service delivery. The Honduras Public Expenditure Review xi successful PROHECO model of community school management should be mainstreamed in the rest of the public primary education, and a similar approach should be explored for other social service delivery including primary health care and community water management. Medium to long-term measures 37. One of the most important obstacles to sound public expenditure management in Honduras is the arbitrary policy-making that pays little attention to either fiscal affordability of a policy proposal/decision or technical rationale of the decision from the point of view of effectiveness and efficiency. In the recent past, this has taken the form of lax public investment decisions (e.g., in 1993), as well as the decision to award significant salary increases to health professionals and teachers without appropriate analysis of their fiscal implications. Over the medium-term, Honduras should aim at developing disciplined policy- making and budget processes that guide the country's scarce fiscal resources to priority programs without jeopardizing medium-term fiscal sustainability. Furthermore, these budgetary decisions should be made and implemented in such a way as to ensure appropriate levels of operational efficiency and accountability. A medium-term expenditure framework (MTEF), as practiced in a number of OECD and a few developing countries, is a good approach (though not sufficient by itself) which Honduras could aim to develop over time to achieve these multiple objectives. Developing a credible MTEF would require, at a minimum: * Capacity to estimate a medium-term fiscal envelope; * Capacity to estimate costs of existing government programs and of new policy/program proposals over the medium term; and * Commitment and an institutional mechanism to ensure that new policy decisions are made consistent with the medium-term fiscal envelope. 38. As suggested above, Honduras might begin by strengthening its capacity to estimate a medium- term macro fiscal envelope and some broad spending categories such as wages and salaries and interest payments. Once this is achieved, the efforts should shift to developing capacities of line ministries to estimate program costs and SEFIN to review their validity. A rational policy-making process would also require availability of reliable information on program effectiveness and efficiency, which systematic evaluations should provide. 39. Other medium-term priorities besides development of an MTEF would include: * A civil service reform to create the capacity both in line ministries and in SEFIN to control public sector staffing, further rationalize the pay levels and patterns in the public sector, and gradually introduce modern performance management techniques; and * Institutional reform of the CGR to establish effective independence and to improve technical qualifications and incentives of its staff. 40. Introducing a medium-term expenditure framework: In a number of both OECD and developing countries, a medium-term expenditure framework (MTEF) has been introduced as a method to guide budgetary decision-making that balances medium-term fiscal affordability/sustainability and strategic priorities of policies and expenditure programs. A medium-term perspective is often necessary to bring public finance to a sustainable path, and more importantly, to develop necessary institutional mechanisms for making sustainable budgetary decisions, because lasting adjustments in expenditure allocations usually require efforts that go beyond the annual budgeting exercise. Possible sequencing of introduction of an MTEF in Honduras is presented here for illustrative purposes. The adoption of an Honduras Public Expenditure Review xii MTEF should be gradual, though involving some changes in decision-making early in the reform process. In this illustration, emphasis is placed on developing a medium-term framework for personnel expenditures, whose control is critical for fiscal sustainability, and the government's ability to fund other critical spending items for poverty reduction. Phase 1 * Revise budget cycle to include a budget framework, prepared by SEFIN, decided by the government three months before finalizing the annual budget. * Budget framework comprises: (i) macro framework (detailed first year, outline for outer years), (ii) wage bill proposals (indicative for outer years if possible), and (iii) proposed sector expenditure ceilings for next annual budget. * Framework made available to legislature. Phase 2 * Budget framework exercise repeated, beginning earlier in the cycle, process coordinated by SEFIN, but with significant input from sector working groups. * Framework comprises: (i) three-year macro framework; (ii) medium-term strategy on pay, staffing consistent with three year wage bill proposal; and (iii) sector expenditure ceilings over three years, linked to sector policy commitments, with particular emphasis on major resource shifts. * Budget framework published in some form prior to finalization of annual budget. Phase 3 * Process as previously, strengthen capacity of sector working groups, opening of sector working groups to civil society consultation. * Framework develops increasing sectoral depth, more explicit linkage between existing/new policy and sector ceilings, introduction of more program detail, gradual introduction of performance indicators, routine public reporting on performance against prior year framework targets. * Greater use of the published budget framework and forum to stimulate public debate, civil society actively involved in public forum. Honduras Public Expenditure Review CHAPTER 1 FISCAL SUSTAINABILITY IN THE HIPC CONTEXT Background 1.1 Honduras is at a difficult juncture in terms of its overall fiscal situation. On the one hand, there are enormous challenges ahead as a result not only of the impressive damage brought by hurricane Mitch in 1998 and the associated reconstruction needs, but also of the prevailing poverty levels in the country. According to the government's Interim Poverty Reduction Strategy Paper (PRSP) about two thirds of Honduras' families would be below the poverty line, including about 50 percent below the extreme poverty line. On the other hand, in July 2000 Honduras qualified for the Heavily Indebted Poor Countries (HIPC) Initiative, implying that past fiscal policies were unsustainable, and hence that notwithstanding the debt relief associated to the HIPC Initiative Honduras has to show fiscal restraint to avoid returning to a situation characterized by unsustainable debt. 1.2 This chapter reviews Honduras' fiscal policy over the last four decades and argues that whereas the country implemented sensible fiscal policies during the 1960s, 1970s, and mid and late 1990s, the disequilibria of the 1980s and early 1990s created a situation that took the country to an unsustainable fiscal position. It then explores sustainable levels of fiscal deficits that Honduras may afford in the future under the HIPC framework without returning to a situation characterized by unsustainable debt. It also addresses the size of Honduras' public sector, and concludes that within the Latin American context Honduras' public sector is about average, reviews the economic classification of the public expenditures, paying special attention to the two largest components: wages and salaries and public investment, and concludes with an analysis of the functional distribution of expenditures and the financial possibilities of the PRSP. The Road to Becoming a HIPC 1.3 During the 1960s, Honduras' Non-Financial Public Sector (NFPS) deficit fluctuated at about 1 percent of GDP, and public debt in 1970 was below 16 percent of GDP, of which 10 percent external and the remainder 6 percent domestic. During the 1970s and the 1980s, the state increased its direct participation in the economy and public investment grew sharply, peaking at 12 percent of GDP in the early 1980s during the construction of the hydroelectric plant of El Caj6n. These high levels of investment generated large fiscal deficits which reached 12 percent of GDP in the mid 1980s. As a consequence, the stock of public debt almost doubled during the 1970s, reaching 30 percent of GDP in 1980 (of which about two thirds was external), and it more than doubled again in the 1980s, reaching 66 percent of GDP in 1989 (of which approximately one-half was external). In the balance of payments, the internal disequilibria led to deficits in the current account, which tripled between the mid 1970s and the mid 1980s, when they reached 11 percent of GDP. External debt service, less than 2 percent of GDP during the 1970s, more than doubled to 4.3 percent of GDP during the 1980s. 1.4 Access to foreign assistance, provided generously while the civil conflict was at its peak in neighboring countries, allowed Honduras to stave off a major balance of payments crisis. But when regional political circumstances changed in the late 1980s, and external finance became scarce, the government had to cut expenditures. Despite fiscal tightening, Honduras continued to face high external debt service obligations (at about 40 percent of revenue in 1988), and the government resorted to external arrears to finance its deficit. In 1989 Honduras went into arrears with the multilateral institutions and was isolated from the international community. Honduras Public Expenditure Review 2 1.5 The Callejas administration (1990-94) Figure 1: Central Government Revenue and initiated a program of stabilization and adjustment Expenditure 1960-98 supported by international financial assistance. The 30 stabilization package included a sharp devaluation of 25 the Lempira (178 percent), and the external debt to 20 .-15 GDP ratio jumped from 37 percent in 1989 to 91 10 percent in 1990, while external debt service almost 5 tripled to 13 percent of GDP (about 80 percent of (0 0> t .- r- co co o 0) a central government revenues in 1990). In 1993, the o) C) 0) 0) a> a ) a> 0) outgoing administration reversed the course of its economic policy, and the central government's fiscal enue - Expenditure_ deficit almost doubled from the previous year, to 7.5 percent of GDP. Inheriting an unsustainable fiscal condition, the Reina administration (1994-98) had to take strong measures, including a cut in central government expenditures from 26 percent of GDP in 1993 to 22 percent in 1994, and some tax adjustments, which had their main impact in 1995 and beyond. 1.6 The Flores administration (1998-02) adopted a Figure 1.2. Non-financial Public Sector Deficit fiscal package to furthEer reduce the fiscal deficit, which 1960-98 had fallen to about 3 percent of GDP in 1997. The 14 _____________ n0 measures~~~~~-evnu inrExpenditureig,loeedinoe ax 1 To finance the increased spending and tax reductions, to 7. the sales tax rate was increased from 7 to 12 percent. o G t2prenryenta o994 the fstom e ajsmnts, whic 1998 thi 2anipc n19 n eod -- indicated that fiscal performance was better than 0 expected, and despite the damage brought by Hurricane -2 a> a> 0) a> M> a a a> a> a, a Mitch in November 1998, the fiscal deficit fell below 2 percent of GDP. 1.7 Over the 1995-98 period, the relatively tight fiscal policy kept an average deficit at 2 percent of GDP and contributed to slowing down in the accumulation of public debt This combined with growth rates averaging 4 percent a year, and an average annual real appreciation of the Lempira of 6.7 percent resulted in a steady decline of the stock of external debt from 95 percent of GDP at the end of 1994 to 66 percent of GDP at the end of 1998. Over the same period, external debt service obligations declined from Pr percent to about 7 percent of GDP. After Hurricane Mitch, the government revised its economic program for 1999-02, with measures to shift spending priorities toward emergency relief and to contain growth in non-emergency expenditures. Consequently, the fiscal deficit in 1999 remained at about 4.5 percent of GDP, due in part to slower than expected implementation of reconstruction projects. Debt service payments due for 1999 were about 5 percent of GDP. 1.8 In July 2000, Honduras' debt was declared unsustainable within the framework of the Enhanced raIPC Initiative and qualified for debt relief of Net Present Value (NPV) US$556 million, after the application of traditional debt relief mechanisms which would provide about NPV US$181 million. This assistance would lower the NPV of debt from US$3.3 billion to the sustainable position, as defined by the Enhanced HIPC Initiative, of 250 percent of NPV to government revenue or NPV US$2.6 billion. Thus, after the episodes of mismanagement and ill luck that made Honduras a highly indebted country, the debt relief program offers some room for the country to retake control of its public finances. Honduras Public Expenditure Review 3 The Road after IPC 1.9 Honduras' gross disbursements are expected to remain at high levels for the next two years to complete the post-Mitch reconstruction work. In the medium term, however, say after 2003, the country needs to ensure that its spending and borrowing are on a fiscally sustainable path and will have corresponding levels of low-cost financing (foreign concessional). This section looks first at the sustainable level of deficits and then at the likely levels of concessional financing, the two constraints that Honduras will face. 1.10 Despite the often-voiced concern with fiscal sustainability, the term does not have a single meaning. This report defines a fiscal policy path to be sustainable if (i) it is consistent with stable government debt dynamics at levels below those implied by the HIPC sustainability thresholds, and (ii) it can be financed with available external resources. We focus on external debt and financing, rather than domestic, because the government's medium-term macroeconomic framework assumes that the fiscal deficit is fully foreign-financed. 1.11 Annex I summarizes the theoretical framework used to assess consistency between government debt dynamics and fiscal deficits. Briefly, the principal determinants of medium term debt dynamics are given by the equation A Stock of foreign debt-Deficit -K* Stock of foreign debt (1) where the stock of foreign debt and the deficit are expressed as shares of GDP, and K(n+o-c)/(l+x+a), with n being the inflation rate, a the GDP growth rate, F the exchange rate depreciation rate, and A before a variable indicates rate of change. In simple economic terms, equation (1) tells us when the debt to GDP ratio will increase or decrease. It will increase when the deficit is higher than the starting debt to GDP ratio multiplied by the factor K. In the special case that the deficit is equal to the initial value of debt multiplied by K, the debt to GDP ratio will stabilize. As for the impact of changes in the macroeconomic environment, increases in the growth rate of nominal GDP will be associated, ceteris paribus, with lower debt to GDP ratios. In contrast, a sharp exchange rate depreciation will lead to higher debt to GDP ratios. 1.12 For equation (1) to be useful as a device to project sustainable fiscal deficits there are two requirements. First, there is a need to specify the macroeconomic scenario that will determine K, and second set the debt to GDP target that will stabilize the debt dynamics. The present analysis assumes that inflation falls in the medium term to 7 percent per year, and that the exchange rate evolves according to the purchasing power parity theory. As for GDP growth, the analysis simulates different scenarios with rates running from I percent to 6 percent per year. Finally, regarding the debt to GDP target, the analysis considers three alternative possibilities with debt targets at the HIPC implied sustainability level, and at 10 and 20 percent of GDP below the HIPC sustainability level. 1.13 To maintain a debt stock at the HIPC debt sustainability limit, the maximum non-financial public sector (NFPS) deficit depends critically on the projected growth rate. The deficit could be as high as 4.2 percent of GDP when GDP growth is projected at 6 percent per year, or as low as 1.7 percent with a sustained GDP growth projection of 1 percent. As a rule of thumb, a l percent reduction in the long-run sustained GDP growth rate would require a reduction in the sustainable fiscal deficit of about 0.5 percent of GDP. Under an intermediate GDP growth scenario (at 3-4 percent per year) and the same aim of remaining at the debt sustainability limit, the NFPS sustainable fiscal deficit would be about 3 percent of GDP. Under a low GDP growth scenario (2 percent per year) Honduras could not run deficits above 2.2 percent of GDP. Table 1.1 summarizes these results. Honduras Public Expenditure Review 4 1.14 The deficit targets in the first line are set so Table 1.1:Consistent Debt-Deficit Limits as to remain right at the debt sustainability limit, which in practice would severely constrain the ability GDP Growth High Intermediate Low of the government to react to unexpected negative 6% 5% 4% 3% 2% 1% shocks. If the debt target is set to allow some Debt Target (1,2) Overall Deficit (1) maneuvering room, say 10 or 20 percent of GDP Sustainability 4.2 3.8 3.2 2.7 2.2 1.7 below the HIPC threshold, then the maximum deficit would be lower, as shown in the bottom two lines of the table. All in all, the sustainability analysis -20% 2.8 2.5 2.2 1.8 1.5 1.1 indicates that as long as GDP growth averages at (1) as a percentage of GDP least 2 percent per year, fiscal deficits at about 2.5 (2) in deviations from the sustainability level percent of GDP would be sustainable. Growth rates below 2 percent per year would require fiscal deficits also below 2 percent of GDP. 1.15 The results are sensitive to deviations of the exchange rate from purchasing power parity, which, as noted above, is assumed in the baseline projection. A real exchange rate depreciation of I percent would require reductions in the deficit by about 0.5 percent of GDP. Conversely, a real exchange rate appreciation of 1 percent would allow higher deficits by 0.5 percent of GDP. These results neglect the effects of real depreciation or appreciation on inflation and growth, and hence should be taken cautiously since in practice what could be gained with a higher appreciation might be lost with lower growth. These results are not sensitive to the projection of inflation, because of the purchasing power parity assumption. 1.16 Since the sustainability of the debt and Table 1.2:Average Annual Gross deficit levels just discussed depends on their being Disbursements 2000-15 financed at concessional terms, the feasible level of deficits depends on the availability of financing on and g a n u t such terms. Table 1.2 shows the annual gross High intermediate Low disbursements that correspond to the deficits for 6% 5% 4% 3% 2% 1% fiscal sustainability shown in Table 1.1 (plus the Debt Target Gross disbursements scheduled repayment of debt). Based on current Sustainability 492 436 374 328 288 253 information, however, Honduras would get -10% 439 388 343 305 267 239 concessional financing of US$250 million or less -20% 386 346 311 276 251 225 per year on average for one decade after 2003. (1) Millions of constant 2000 US$ deficitsdepends n(2)easaiaapeliyeof fenncingDo Except for the lowest growth scenarios, this would (3) an d eeviati onsfr thG usaDbPiyee be a much tighter financing constraint than what would be fiscally sustainable. Looking at the issue from the side of financing availability, only about US$250 million per year on average, the deficit would need to average no more than 1.25 percent of GDP, with growth rates in the 3 to 5 percent range. Higher financing flows would allow for higher deficits, up to the fiscally sustainable limit. Similarly, lower flows of lending would require tighter fiscal policy. The Size of the Public Sector 1.17 Having reviewed the fiscal deficit that Honduras' NFPS could afford in the medium term, the next issue to address is the size of the government, or the level of public spending that can be considered sustainable over the medium and long run. With NFPS revenue projected at about 30 percent of GDP over the medium term, this would imply spending possibilities of about 31.3 percent. Table 1.3 reports spending levels for the NFPS disaggregated at the central government, rest of the general government and public enterprises levels for the period 1999-2002. The column with the heading Equil. would indicate the equilibrium level consistent with the sustainability analysis above. Honduras Public Expenditure Review 5 1.18 Within the Latin American context a Table 1.3: NFPS Public Spending central government this size would be about 1999 2000 2001 2002 Equil average. Figure 1.3 shows central government prel. proj. proj. expenditures for the other Central American Central gov't 23.7 22.9 25.1 24.6 22.25 countries (El Salvador, Costa Rica, Guatemala, Rest of general gov't 4.2 5.2 6.5 4.4 4.5 Panama, and Nicaragua), the other Latin Public enterprises 5.0 4.1 5.0 4.6 4.5 American HIPC (Guyana, and Bolivia) and the Total NFPS 32.9 32.2 36.6 33.6 31.2 median for the whole Latin American region over (1) as a percentage of GDP the period 1997-2000. Figure 1.3 suggests that Honduras' central government spending was similar to the regional median of about 22 percent of GDP in 2000. A comparison with the three other regional aPC, Guyana (42 percent of GDP in 2000), Nicaragua (32 percent of GDP in 2000), and Bolivia (29 percent of GDP in 2000) would suggest that Honduras' public sector is smaller than the other heavily indebted countries. As for the Central American countries, Honduras' spending is at about the same level as Costa Rica and Panama, two of the better-managed public sectors in Central America, and is larger than Guatemala and El Salvador (about 15 percent of GDP in 2000), which most observers consider to be lacking in public services. Figure 1.3: International Comparison of Central Government Spending Source: World Bank staff estimates. Economic Classification of Central Government Spending 45.0 - 40.0 - 35.0 3gg 1996 25.0 -- - - - ----- *1997 01998 20.0 mod- .01999 15.0 *2000 10.0 5.0 HONDURAS LAC (cg) Panama (cg) Nicaragua El Salvador Guatemala Guyana Bolivia (cg) Costa Rica (cg)* (cg) (cg) (cg) (cg)* (cg) 1.19 Table 1.4 presents the economic disaggregation of Honduras' central government spending as a share of GDP, and as a percentage of total expenditure since 1990. Total expenditure averaged 22.2 percent of GDP over the period 1990-2000, and is expected to average 24.9 percent over 2001-2002. Most of this increase (2.1 percent of GDP) will be associated with higher capital outlays whereas current expenditures are expected to increase by 0.6 percent of GDP (from 16.1 percent over the 1990-2000 period to 16.7 percent in 2001 and 2002). 1.20 Notwithstanding this small increase in current spending, it is worth noting the evolution of the wage bill in Honduras. Wages and salaries represented 6.1 percent of GDP in 1997, increased to 6.9, 8 and 8.9 percent of GDP for 1998, 1999 and 2000 respectively, and are expected to further increase to 9.5 percent of GDP in 2001 before falling to 9 percent of GDP in 2002. These increases have been possible in part by lower interest payments (the correlation coefficient between interest payments and wages and Honduras Public Expenditure Review 6 salaries for the period 1998-2002 is estimated at -.96 i.e. they almost move perfectly in step). But this raises questions about its long-term sustainability (assuming no further increases in revenue efforts) when interest payments to bilateral creditors resume after the end of the Paris Club debt deferral. Table 1.4: Economic Distribution of Central Government Spending 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 prel proj. proj. (% of GDP) Total expenditure 22.8 22.7 22.0 26.5 21.9 21.5 20.1 19.9 20.4 23.7 22.9 25.1 24.6 Current expenditures 17.2 18.5 16.6 17.7 14.9 14.8 15.4 15.0 14.6 16.6 16.4 16.7 16.7 Current primary exp. 14.5 12.5 12.7 14.3 10.9 11.0 12.0 11.6 11.9 14.6 14.7 15.3 14.8 Wages and salaries 8.6 7.2 7.4 7.2 7.0 6.7 6.6 6.1 6.9 8.0 8.9 9.5 9.0 Goods and services 3.7 3.1 3.0 3.0 1.8 2.0 2.3 2.5 2.0 2.2 2.3 2.4 2.6 Current transfers 2.2 2.2 2.4 4.1 2.0 2.3 3.1 3.0 2.9 4.4 3.5 3.4 3.2 Interest Payments 2.6 6.0 3.8 3.4 3.9 3.8 3.5 3.4 2.8 2.0 1.7 1.4 1.9 Capital expenditures 5.6 4.3 5.4 8.8 7.0 6.7 4.7 4.9 5.8 7.1 6.5 8.4 7.9 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 prel proj. proj. (% of total) Total expenditure 100 100 100 100 100 100 100 100 100 100 100 100 100 Current expenditures 75.5 81.2 75.4 66.8 68.0 68.8 76.6 75.4 71.6 70.0 71.6 66.5 67.9 Current primary exp. 63.9 54.9 58.0 53.8 49.9 51.3 59.4 58.4 58.0 61.6 64.2 61.0 60.2 Wages and salaries 37.9 31.5 33.6 27.1 32.1 31.2 32.9 30.5 33.8 33.8 38.9 37.8 36.6 Goods and services 16.3 13.5 13.4 11.2 8.4 9.3 11.3 12.6 9.9 9.3 10.0 9.6 10.6 Current transfers 9.7 9.8 10.9 15.6 9.4 10.7 15.3 15.3 14.3 18.6 15.3 13.5 13.0 Interest Payments 11.6 26.4 17.5 13.0 18.1 17.5 17.2 17.0 13.7 8.4 7.4 5.6 7.7 Capital expenditures 24.5 18.8 24.6 33.2 32.0 31.2 23.4 24.6 28.4 30.0 28.4 33.5 32.1 Source: SEFIN 1.21 As for the share of each item in total expenditure, Table 1.3 indicates that during 1990-2000 Honduras directed more than one-fourth (27 percent) of its total spending toward capital spending, highlighting the importance traditionally given to the public investment program. This level of effort is about double the Latin American average where capital expenditure would represent about 14 percent of total expenditure. A question that arises is whether such high levels of investment would be justified. The analysis of infrastructure investment (Chapter 6) indicates that despite the high level of public investment, there remain significant gaps in infrastructure access, especially in rural areas, and there are also problems of inefficient resource use. All in all, the analysis of the public expenditures by economic classification suggests an evolution of wages and salaries that might not be consistent with long-term sustainability and very high levels of government investment, at least within the regional context. These two issues are addressed next. The broader issues of public pay and employment and personnel management are addressed in more detail in Chapter 2. 1.22 Wages and Salaries: Given the sustainable level Table 1.5: Sustainable Central Government of central government spending computed above (22.3 Wages and Salaries percent of GDP), and projections for the other broad (percent of GDP) economic categories, it is possible to compute a sustainable GDP Growth level of wages and salaries. The projections used here are: Deficit 6% 5% 4% 3% 2% 1% central- government capital spending (including net lending) (% of GDP) and current transfers are capped at 8 and 3.5 percent of 1.25 7.6 7.5 7.4 7.3 7.0 6.6 GDP respectively over the long run and purchases of goods and services at 2.5 percent of GDP. Table 1.5 reports the results for the different growth scenarios with Honduras Public Expenditure Review 7 an NFPS fiscal deficit of 1.3 percent of GDP, which imply sustainable personnel costs ranging from 6.6 percent of GDP under the low growth scenarios to 7.6 percent under the high growth scenario. The additional room for wages and salaries under high growth would come from the lower debt service payments (as a percentage of GDP) associated with the high growth scenario. In every case, however, the results suggest that the projected wage bill of 9.5 percent of GDP for 2001 seems well above the sustainability level (between 2.9 and 1.9 percent depending on growth scenarios). To give an idea of the magnitude of the problem, the nominal adjustment that public wages and salaries would require in 2001 to return to its sustainability level would be approximately 25 percent. 1.23 The government might maintain current salary levels by either increasing revenues by about 2.5 percent of GDP, or cutting other budget items. Since both purchases of goods and services and transfers have already been projected at prudent levels, the likely candidate for adjustment would be capital investment cuts. However, the required cuts would take capital spending and central government net lending down to about 5.5 percent of GDP. With capital transfers and net lending at 4 percent of GDP, the central government fixed capital formation would be capped at about 1.5 percent of GDP, which is probably unrealistic in view of the reconstruction needs of the country, and could endanger future growth. 1.24 Public Investment: As noted above, the share of Honduras' central government investment in total expenditure is about double the Latin American average. Capital expenditure (including net lending) of the Non Financial Public Sector at almost 9 percent of GDP in 2000 and estimated at above 11 percent in 2001 is also high, and has traditionally been behind Honduras' already noted high deficits and build-up of external debt (Figures 1.4 and 1.5). Figure 1.4: Deficit and Public Investment Figure 1.5: Public Investment and Disbursements 10% IT_ __ _ _ __ _ _ 0 0 a- 18'u CI g Pttic Instmt a Dicit * Riusm EDOtuserls I 1.25 Public investment seems to have played two opposing roles in the Honduran economy. On the one hand, there is the positive impact created by the provision of infrastructure services. On the other hand, the high levels of public investment have contributed to the macroeconomic instability of the 1980s and early 1990s. In principle, public investment should be welcomed when it increases the profitability of private investment (i.e., when public investment crowds-in private investment). Though the examination of microeconomic data reveals that some components of public investment are not well targeted or implemented efficiently, historically Honduras' public investment seems to have had a positive impact on the profitability of private investment. 1.26 Analysis undertaken for this PER' suggests that every additional percentage of GDP of public investment would crowd in 0.2 percent of GDP in private investment. A possible explanation for this finding could be the drastic redefinition of the role of the state in the economy that has taken place over 1 The Impact ofPublic Investment on Private Investment: The Case ofHonduras, LCC2C Mimeo. Honduras Public Expenditure Review 8 the last two decades. The process, initiated in 1985, privatized over 50 enterprises that had been financed by the bankrupted National Investment Corporation and some facilities owned by decentralized public institutions. In the late 1990s the only public enterprises were the National Electricity Company (ENEE), the Water and Sewerage Company (SANAA), the Telecommunications Company (HONDUTEL), the Ports (ENP), and the Airports. 1.27 Therefore and notwithstanding documented cases of deficient projects, a large share of public investment, has been directed to sectors that have a positive impact on the economy, like infrastructure and communications, in addition to areas where private participation is inherently limited and public goods content is high (e.g., education and health). It is worth noting that during the 1990s, approximately half of Honduras' public investment has been implemented by five agencies focusing on transportation and infrastructure services: the four public enterprises and SOPTRAVI have usually been responsible for more than 25 percent of public investment (40 percent in 1993). 1.28 However, this is not the end of the story, since the analysis mentioned above also suggests that the positive impact of public investment on private investment only holds for moderate levels of the former (at below 9 percent of GDP). Above 9 percent of GDP, public investment would crowd out private investment and hence, the negative effects created by additional capital spending would out-weigh the positive effects of additional infrastructure (both human and physical). This result therefore calls for moderation in the levels of public investment, and could be added to the previous results calling for moderation of public wages and salaries, and the level of total spending. Functional Distribution of Central Government Spending 1.29 The functional classification of public spending permits assessment of the influence of economic policy in the budget formulation and execution stages. In principle, lack of strategic resource allocation would result in volatility, showing up as unpredictable and major allocation changes. This would be an indication that a government does not have stable medium-term policy priorities, and the policy shifts due to unstable resource allocations would likely impede pursuit of medium-term development objectives. Table 1.6 reports Honduras central government spending by functional classification for the period 1990- 2000 both as a share of GDP and as a share of total expenditure. 1.30 As Table 1.6 shows, the budget shares allocated to the social sectors have been fairly stable through most of the 1990s both as a share of GDP and as a share of total spending. When the fluctuations (as measured by the standard deviation) are corrected by the average level of spending, the social expenditures component of spending would be the one with the lowest volatility both as a percentage of GDP and as share of total spending. It is followed by the environment and general administration components, and finally by the defense and infrastructure components. The relative stability of social expenditures in the context of fluctuating total budgets (as percentage of GDP) reflects primarily the relatively high share of wages and salaries (for teachers and health professionals), which tend to be less volatile than other spending items. But it could also be taken as evidence that the government has tried to protect the social sectors when total expenditures decreased. Honduras Public Expenditure Review 9 Table 1.6: Functional Classification of Central Government Spending 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 (% of GDP) Total expenditure 22.8 22.7 22.0 26.5 21.9 21.5 20.1 19.9 20.4 23.7 22.9 General administration 5.4 3.2 3.0 4.0 2.9 3.6 3.1 3.6 4.8 5.3 4.8 Defense and public safety 2.5 2.1 1.6 1.4 1.2 1.1 1.0 1.1 0.8 0.9 1.0 Social spending 8.1 8.0 7.9 8.7 7.0 7.0 7.8 7.2 7.3 8.8 9.8 Interest 2.7 6.2 3.9 3.5 3.6 3.6 3.5 3.4 2.8 1.8 1.7 Infrastructure 2.7 2.3 4.0 7.0 3.0 2.8 2.9 2.8 2.0 1.8 2.4 Environment 1.4 0.8 0.7 1.0 1.3 1.5 0.9 0.8 1.2 1.2 1.4 Other 0.0 0.1 0.9 0.9 2.9 1.9 1.0 1.1 1.5 3.9 1.9 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 (% of total) Total expenditure 100 100 100 100 100 100 100 100 100 100 100 General administration 23.7 14.3 13.4 15.3 13.2 16.9 15.4 17.8 23.5 22.3 20.8 Defense and public safety 10.8 9.2 7.5 5.3 5.3 5.1 5.0 5.4 4.1 3.9 4.4 Social spending 35.4 35.1 35.7 32.8 32.0 32.8 38.6 36.3 35.6 37.2 42.8 Interest 12.0 27.2 17.7 13.1 16.4 16.7 17.3 17.3 13.5 7.5 7.4 Infrastructure 11.9 10.0 18.1 26.4 13.8 13.2 14.4 13.9 10.0 7.7 10.5 Environment 6.2 3.6 3.4 4.0 6.1 6.8 4.3 3.9 6.0 5.0 6.0 Other 0.0 0.7 4.2 3.2 13.3 8.5 5.0 5.5 7.4 16.4 8.2 Source: Ministry of Finance, IMF and WB estimates. Social Spending and the PRSP 1.31 The macroeconomic limits on the overall level of public spending have obvious effects on the overall funds that can be devoted to the social sectors. Since one of the objectives of the PRSP is to present a strategy that is fiscally responsible and realistic, this section simulates different envelopes that might be available for the social sectors. These simulations also consider the government's focus of spending in the relevant sectors. 1.32 One such measure is the Social Expenditure Ratio (SER), defined as the percent of central government spending assigned to social services in health, education, and safety net programs, such as the Honduran Social Investment Fund (FHIS) and the Family Assistance Program (PRAF). Since 1990 Honduras' SER has averaged 36 percent, with a maximum of almost 43 percent in 2000 and a minimum of 32 percent in 1994. If we further take into account UNDP's recommendation for a SER of 40 percent, it is possible to conclude that whereas Honduras was below target during most of the decade, the increase in 2000 would have brought the country to a comfortable position. It should be noted, however, that a large part of the increase in the SER since 1998 is due to the salary increases granted to teachers and doctors, and hence that the increase in the SER is not completely associated with an increase in social sectors service delivery. 1.33 These elements allow us to compute different budget envelopes for the social sectors, four of which are shown in Table 1.7 for different scenarios of GDP growth, ranging from 2 to 5 percent per year. For the period 2001-02 total central government spending is given by the government's medium-term macroeconomic framework, regardless of the GDP growth rate. After 2002, we assume the equilibrium central government deficit as implied by a NFPS deficit of 1.25 percent of GDP, central government revenue of 19 percent, and a surplus for the rest of the NFPS of 2 percent of GDP. In other words, the central government would have 22.3 percent of GDP to allocate. Honduras Public Expenditure Review 10 1.34 Regarding the government's spending allocation Table 1.7: Social Spending decisions, three scenarios for the SER take values of 36, 40 Scenarios and 44 percent. Thus, all scenarios would devote at least the Cumulative resources over 2000-15, same share of total spending to the social sectors as in 1990s: in US$ million at constant 2000 prices 36 percent. This choice is consistent with additional efforts Growth SER aimed at poverty reduction, implicit in the Poverty Reduction Percent 36 40 44 Strategy. To facilitate comparisons for the Poverty 5 11,327 12,521 13,714 Reduction Strategy costing exercise, all reported data are in 10,371 11,459 12,547 constant 2000 US$, and for each scenario, we report the 3 9,508 10,500 11,492 cumulative social spending and cumulative priority social 2 8,72 9,633 10,539 spending budgets over the period 2000-15. Table 1.7 shows1. pend how total resources for social spending could vary between US$8.7 billion (low growth-low SER) and US$13.7 billion (high growth-high SER). The average case (SER at 40 percent, growth at between 3 and 4 percent) would imply resources of about US$11 billion over the 2000-15 period, or about US$687 million per year. Under all four growth scenarios simulated here, increasing the SER to 40 percent, from the decade average of 36 percent, the country would ensure about US$1 billion in additional resources for the social sectors. This amount is about the same as the HIPC debt service relief that Honduras will benefit from over the coming years. In other words, a minimum increase in the SER by four percentage points would be consistent with the notion that the HIPC debt relief is devoted to the social sectors. 2 1.35 The analysis above shows the overall resources available for social sectors, including existing programs, such as running existing schools, hospitals, etc. This cost will surely increase over the coming years due to the expected increase in the population. To hold real per capita social spending constant at the 1999 level, the social budget would have to increase to reach a cumulative total of US$10.4 billion (2000 prices) over the 2000-2015 period. This computation would correct for the salary increases of 2000 and those expected for 2001 that would increase the cost of service delivery but would not add additional services. It also assumes that the relevant salaries remain constant over the period 2002-15. If one deducts that figure from the projections of available resources in Table 1.7, the remaining amount would be available for additional programs and increases of per capita spending in current programs. With this calculation, Table 1.8 shows the additional component of resources in the same growth and SER scenario as in the previous table. 1.36 After deducting the cost of running existing programs Table 1.8: Funding for for a growing population, the additional resources for new Additional programs, or improvements in existing ones, could be as much Sial Spending Cumulative over 2000-2015 as US$3.2 billion (high growth-high SER), or they could in US$.million at constant 2000 rices negative US$1.7 billion (low growth-low SER). Thus, when the is_dvote t SER is fixed at 36 percent, unless growth averages more than 4 Growth SER percent per year, Honduras would devote a decreasing amount wil_suel 36 40 44 of resources to the social sectors in real per capita terms. If we 5 901 2,095 3,289 look at a median case (SER of 40 and 3.5 percent averge 4 -55 1,033 2,121 growth), additional available resources would be about US$500 re-918 74 1ld million spread over the next 15 years. If these additional 2 -1,698 -792 113 resources are allocated exclusively to new programs, existing programs would have no financing (in real per capita terms) above the 1999 levels. Therefore, this estimate should be taken cautiously as an upper limit on resources available for additional programs, because if improvements to existing ones are also planned, the resources for new programs would be lower. Figure 1.6 reports the cumulative additional 2 In practice, not all pro-poor spending is in the social sectors, and thus it is possible for the government to increase the social spending by less tn 4 percentage points, and still allocate all of the HIPC debt relief to pro-poor spending, as identified in the Poverty Reduction Strategy. Honduras Public Expenditure Review 11 resources that could be used in the social sectors under different SER and the salary profiles described above. 1.37 It is worth noting that should public Figure 1.6: Cumulative Additional Resources* wages and salaries have been constant in real terms in 2000 (a nominal increase of 10 3000 - -------- percent) and 2001 (similar nominal increase), 20 the additional funding for the social sectors 1000 would range from negative US$691 million . 0 under the low growth-low SER scenario to -1000 -Scenano Sacen=rpo3 US$4.3 billion under the high growth-high -2000 SER scenario, with a median case (SER 40 ER 40 SER44 percent and growth at 3.5 percent) of US$1.5 billion. In other words, the cumulative cost *scenario 1 assumes constant salaries in real terms over 1999-2015 of the salary increases of 2000 and 2001, Scenario 2 assumes constant salaries in real terms over 2002-15 under the assumption of real wage Scenario 3 assumes a I percent increase in real terms over 2002-15 inflexibility, would be about US$1 billion. Finally, should salaries increase by 1 percent per year in real terms over the 2002-2015 period, under the median case there would be no additional resources for the social sectors beyond maintaining spending constant on a real per-.capita. basis. This clearly highlights a policy choice: higher salaries in the public sector or additional social services. Honduras Public Expenditure Review CHAPTER 2 PUBLIC SECTOR PAY, EMPLOYMENT AND PERSONNEL MANAGEMENT Public Sector Pay and Employment Trends 2.1 Trends in expenditures: One notable trend in the composition of public expenditures is the fast growth of wages and salaries in the last few years, due to increases in both the number of certain types of public employees (primarily teachers) and salary levels. The most significant impact on the wage bill stems from growing salary expenditures in the health and education sectors since 1997. Total real wage expenditures' remained roughly constant until the last part of the 1990s, when significant pay raises legislated under separate employment regimes (Estatutos) took effect (Table 2.1). Since 1996 there has been a 48.6 percent increase in total real wage expenditures (from 3.9 percent to 5.0 percent of GDP). Of the three main public employee categories-central government civil servants, teachers and health sector employees-the latter two have seen their total wage bills increase by more than 60 percent since 1996, with a more modest 12.4 percent increase for central government civil servants. Teachers and health sector employees accounted for 64 percent of the total wage bill in 1990, which is projected to rise to close to 80 percent in 2000. Unless these trends can be controlled, the government will face major problems in maintaining fiscal discipline, ensuring a minimum technical and managerial capacity in the public sector, and preventing a crowding out of non-salary expenditures. Appropriate allocations for non- salary expenditures are critical to ensure adequate operations and maintenance of existing facilities and to improve the quality of public services. Table 2.1: Wage Expenditures by Employment Category (in million of 1990 Ls) 1990 1992 1994 1996 1998 1999 2000 % Change (prel.) (proj.) 1990-94 1994-96 1996-00 Civil Servants 252.9 233.0 218.4 179.5 188.7 218.6 201.8 -13.6 -17.8 12.4 (% of total) 36% 35% 31% 28% 27% 28% 21% Top-level 7.0 5.9 5.3 6.1 8.2 7.8 7.2 -24.3 15.0 18.0 Middle-level 102.5 92.9 83.8 77.4 85.8 98.2 93.9 -18.0 -7.6 21.3 Support 143.4 134.2 129.2 95.9 94.6 112.6 100.8 -9.9 -25.7 5.1 Teachers 379.7 367.6 408.7 378.8 412.7 416.4 613.7 7.6 -7.3 62.0 (% of total) 53.9% 54.7% 58.1% 59.0% 58.9% 53.9% 64.3% Health Sector 71.2 71.3 76.4 84.1 99.0 137.0 138.9 7.3 10.1 65.2 (% of total) 10.1% 10.6% 10.9% 13.1% 14.1% 17.7% 14.6% Doctors 38.4 31.4 32.8 36.0 47.5 67.6 73.1 -14.6 9.7 103.1 Nurses 2.8 3.9 7.3 8.4 9.4 14.3 15.3 160.7 15.1 82.1 Other health 3.9 3.9 3.9 3.3 5.3 5.0 5.2 0 -15.4 57.6 Aux. Nurses 26.1 32.1 32.4 36.5 36.8 50.1 45.4 24.1 12.7 24.4 Total Central 703.8 672.0 703.6 642.4 700.4 772.1 954.5 -0.0 -8.7 48.6 Government 100% 100% 100% 100% 100% 100% 100% Sources: SEFIN, Escalaf6n, World Bank calculations Note: For 1990-94, wage costs for teachers include only primary and secondary. These figures include 13th and 14 month4 2.2 Trends in public sector employment: Throughout the 1990s there has been an upward trend in the total number of employees in the central government. The number of central government civil servants fell 5 percent during the first four years of the decade, and then dropped sharply by almost 15 percent between 1994 and 1996. Since 1996, their number has remained constant, although falling as proportion of total central government employment (Table 2.2). The steady growth in the number of These figures are calculated using base salaries including 13th month, and l4ff month after 1995. They do not include pension contributions, seniority, and other colaterales. 4 Since 1995, all government annual salaries include a 14th month (decimocuarto mes). Honduras Public Expenditure Review 13 teachers has offset the effect of civil service reductions and resulted in a larger total public sector employment by the end of the decade. Table 2.2: Central Government Employment Composition (selected years) % Change 1990 1992 1994 1996 1998 2000 est 1990-94 1990-96 1996-00 Civil Servants 26,399 26,314 25,045 21,482 21,324 21,202 -5.0 -14.0% -1.0% (% of total) 43% 41% 38% 31% 29% 28% o/w: Top-level 146 127 136 118 136 135 -6.8 -13.0 14.0 Middle-level 7,183 7,222 6,905 6,867 6,960 6,760 -3.9 -0.5 -1.6 Support staff 19,070 18,965 18,004 14,497 14,228 14,307 -5.6 -19.5 -1.3 Teachers 30,176 32,829 35,298 40,923 45,025 48,086 17.0 16.0 18.0 (% of total) 48.8% 50.8% 53.0% 59.1% 61.1% 62.6% Health sector 5,260 5,466 6,268 6,847 7,366 7,581 19.0 9.0 11.0 (% of total) 8.5% 8.5% 9.4% 9.9% 10.0% 9.9% o/w: Professionals 1,825 1,857 2,283 2,498 2,826 2,756 25.0 9.4 10.3 Auxiliary nurses 3,435 3,609 3,985 4,349 4,540 4,825 16.0 9.1 10.9 Total Central 61,835 64,609 66,611 69,252 73,715 76,869 8.0 4.0 11.0 Government 100% 100% 100% 100% 100% 100%1 Sources: SEFIN, World Bank calculations. Note: 1990-94 data only include primary and secondary teachers; 1994-00 includes pre-school, adult education, and teacher-administrators. 2.3 Trends in the level of remuneration: While Table 2.3: 1998 Private Sector Incomes (monthly) total wage expenditures remained relatively stable until the late 1990s, employment numbers were rising. Top Middle Support As a result, there was a substantial erosion in average Executive Manager Staff salaries for all categories of public employment Maximum 127,196 36,105 5,298 (teachers, health personnel, and central government civil servants) during the first half of the decade. The income gap between equivalent positions in the public Minimum 125 505 15 and the private sectors was the largest at the middle management level (Table 2.3). In the second half, Source: Consorcio Cincorp-Hay, 1998 most public employees recovered what they had lost in previous years. The Estatutos gave health sector employees, especially doctors, large gains, as their average real wages rose significantly between 1996 and 2000 (more than 30 percent for nurses and "other" health sector employees, such as dentists and pharmacists, and more than 100 percent for doctors). Auxiliary nurses' wages have risen the least among health sector employees over the past four years, and their unions are now engaged in an energetic campaign to negotiate an Estatuto of their own. Teachers also negotiated an Estatuto in 1997, which stipulated a 69 percent increase in the base salary for all primary and most secondary school teachers' over a four year period (10 percent in 1998, 19 percent in 1999, 20 percent in 2000 and 20 percent in 2001). Although it could be argued that the Estatutos corrected in some sense for earlier real wage losses, it should be stressed that the salary increases granted over the period 1998-2001 took into account neither the fiscal possibilities of the country nor the medium term economic framework, and especially the declining inflation path. Additionally, the Estatutos introduced a certain level of inequality among public SOnly a few groups (docentes, who work as instructors in public sector agencies to train other employees, nurses, and auxiliary nurses) experienced increases in average wages between 1990 and 1994. 6 Secondary school teachers with an education degree were included in the 69 percent increase. Under the Estatuto, secondary school teachers who are in the process of obtaining an education degree and have completed at least half the coursework will receive half of the 69 percent increase; when they are awarded their degree, they will receive the full increase. Over half the secondary school teachers in Honduras have a teaching degree. Honduras Public Expenditure Review 14 employees since whereas some groups enjoyed relevant wage adjustments others had to face salary increases more in line with the fiscal reality. 2.4 Central government civil servants' wages fell steadily for most of these employees. Between 1990 and 1996, for example, the average wage for a ticnico had fallen 20 percent, and for an administrativo 24 percent. In addition, the early 1990s brought a situation characterized by a salary compression that made difficult to maintain qualified staff. The government has been implementing a salary rationalization program (nivelacidn salarial) in order to rectify the most egregious cases of underpayment by bringing technical employees' salaries in line with the newly developed salary scale.' The program's original goals were met for both the technical and support-level staff. As a result, the average base salary for an individual in this group rose sufficiently to regain the level at the beginning of the decade in real terms. Table 2.4: Evolution of Average Monthly Wages (in 1990 Ls.) 1999 2000 % Change 1990 1992 1994 1996 1998 (prel.) (proj.) 19 1990-941994-96 1996-00 Top-level 3710 3558 3026 3723 4329 4138 3791 -18.4 23.0 1.8 Superior 4839 5033 4242 5954 6779 6023 5454 -12.3 40.4 -8.4 Ejecutivo 3406 3196 2651 3028 3513 3452 3140 -22.2 14.2 3.7 Middle-level 1098 990 933 806 881 972 992 -15.0 -13.6 23.1 Prof. Univ. 1794 1576 1402 1258 1682 1763 1661 -21.9 -10.3 32.0 Tdcnico 830 758 748 657 684 893 842 -9.9 -12.2 28.2 Docente 818 820 845 740 735 754 829 3.3 -12.4 12.0 Support staff 578 544 552 473 475 558 503 -4.5 -14.3 6.3 Admin. 725 658 668 550 557 651 587 -7.9 -17.7 6.7 De obra 547 525 530 483 481 551 497 -3.1 -8.9 2.9 De servicio 440 430 434 399 405 481 433 -1.4 -8.1 8.5 Teachers 968 862 891 687 662 668 962 -8.0 -22.9 40.0 Doctors 2266 1863 1862 1628 2048 3018 3301 -17.8 -12.6 102.8 Nurses 898 1034 959 947 887 1247 1312 6.8 -1.3 38.5 Other health 1118 1002 937 964 1096 1329 1265 -16.2 2.9 31.2 Aux. Nurses 585 685 626 599 578 753 672 7.0 -4.3 12.2 Sources: SEFIN, Escalaf6n, World Bank calculations Notes: Teachers' average wage is calculated as average of primary and secondary school earnings. Teachers, doctors, nurses and other health professionals were not included in the nivelaci6n salarial. Their salary increases are legislated in the Estatutos. 2.5 The sharp rise in the wage bill, to a budgeted 9 percent of GDP in 2000, represents a potentially serious constraint on the government's future ability to maintain fiscal discipline, its objective of further rationalizing the public sector salary structure, and its ability to improve the quality and poverty-impact of public services. Personnel expenditures are very difficult to adjust, especially in the short run, and hence severely constrain fiscal flexibility. In spite of some progress made with the nivelaci6n salarial among middle-level officials, it is likely that further salary decompression is needed to have a major impact on the government's ability to attract and retain qualified staff. However, given the already high wage bill, a second stage of salary reform will only be possible if coupled with downsizing efforts targeted on the (still large) groups of lower-level staff. 2.6 The commitments acquired in the Estatutos weigh heavily on the government's flexibility in allocating social spending. Increased wage expenditures in the social sectors are not necessarily the best 7 Employees whose earnings were determined to be higher than they should be under the new salary scale were given only a minimal increase. Honduras Public Expenditure Review 15 target in a poverty reduction strategy. With a high and growing share of social spending going to cover the payroll, there is little left for complementary inputs (e.g., teaching materials, medicines) that are needed to improve service quality. As discussed in the next section on personnel management, compounding the problem is the government's weak control over performance and the geographic distribution of public employees who deliver social services. There is also considerable evidence that the productivity of social sector employees is quite low, so that higher spending on teachers and health workers (through increased numbers of employees, higher wages, or both) does not necessarily provide a corresponding increase in service output and quality where it is most needed. Personnel Management 2.7 While this report highlights weaknesses in the human resource management in the health and the education sectors because of their more direct impact on efficiency in the government's poverty reduction efforts, weak personnel management is a problem that pervades Honduras' public sector and affects policy implementation at all levels. To gain a better understanding of public personnel management issues, we conducted a focus group with seven middle-level public officials from five Secretariats' and probed different aspects of human resources management, covering both existing formal regulations as well as the degree to which these regulations are enforced. While the information gathered here is inevitably subjective and does not allow us to assess the frequency and the magnitude of the problems reported, the reported patterns of informality in personnel management are not uncommon in developing countries with weak institutional capacity and cultures of patronage. Some of the practices reported, such as non-competitive recruitment, are corroborated with the findings of the staff tracking survey reported in the subsequent section. 2.8 Recruitment and selection: The first step in recruiting new staff is identification of personnel needs by the hiring unit. In Honduras, this process is made quite inefficient by rigid budget rules that constrain agencies' ability to hire needed personnel. During the annual budget formulation process in Honduras, SEFIN's Budget Directorate assigns each entity a certain number of posts with corresponding salaries. Line managers complain that these assignments are often made arbitrarily by SEFIN, without sufficient communication with spending agencies. If the line ministry does not use all the posts and salaries that have been budgeted for it, these accumulate over time and are drawn on by SEFIN to respond to unexpected demands that arise in other parts of the public sector.' The line ministry cannot hire someone in a post that has not been budgeted for. For example, if a Secretariat needed to hire an engineer but had only been assigned vacancies and corresponding salaries for secretarial posts, the engineer had to be hired at the same wage as a secretary and would appear on the personnel lists as a secretary. Practices such as these result in a lack of transparency in personnel information, demotivate staff, and prevent agencies from attracting the most qualified candidates. 2.9 This way of assigning posts with little regard to the hiring agency's needs directly affects the recruitment process. The Civil Service Law stipulates that vacancies be announced to the public by the DGSC. In practice, most recruitment is done by direct invitation by the hiring manager because informal practices such as hiring of an engineer into a secretary's position make it impractical for the agency to announce a vacancy, as it would likely invite applications from secretaries (the announced position) rather than engineers (the actual skill need). It is expected, however, that the job reclassification that was implemented in 1999 has had an effect on reducing the scope of discretion. 2.10 There was widespread agreement among officials with whom we spoke that personnel offices in public entities are politicized; personnel directors themselves speak proudly of their loyalty to the political 8 The participants' time as public sector employees ranged from 11 to 30 years, and most had held their current post for around five years. 9 One public official explained that personnel assignments are used by SEFIN as a "cushion." Honduras Public Expenditure Review 16 appointee heading their agency. One personnel director stated: "the person in charge of the Personnel Office has to be one of the agency head's closest positions 'de conflanza."' From the applicant's standpoint, without a political recommendation it is very difficult to penetrate the public sector job market. Given the non-transparent way in which most recruitment appears to be conducted, applicants without special connections are unlikely to have access to real information about vacancies. In the education sector, the Education Ministry received pressures from legislators regarding teacher hiring, until hiring was decentralized in 1996. Today, according to a top official at the Education Ministry, legislators have even easier access to the Departmental Directors in their region, and exert influence over the hiring committees that select new teachers each year. Anecdotal evidence indicates that, until very recently, political contacts were needed in order even to apply for a teaching position. 2.11 Once the hiring manager chooses a preferred candidate for the position, the DGSC then assesses the candidate. The public officials we spoke to in Honduras, at every level of involvement in personnel management, complained without exception of the outdated selection processes used by the DGSC (e.g., outdated questions included in the examinations and interviews, DGSC officials' lack of qualification to evaluate candidates' performance for technical posts) and the heavy influence of political criteria in selecting personnel."o 2.12 Evaluation and dismissal: The 1968 Civil Service Law requires that the DGSC work with public agencies to create a performance evaluation system, in order to evaluate personnel at regular intervals. The Law suggests that employees be evaluated based on efficiency, initiative, personality, behavior, and skills. Based on evaluation results, the DGSC is charged with designing professional development plans for public employee. 2.13 According to officials at the DGSC, annual evaluations were conducted systematically until the early 1980s, when they were suspended because they had become too subjective and were conducted merely as a formality. Focus group participants told us that evaluations are still carried out on a very selective basis, often as an excuse to dismiss people when the hiring manager wants to give the post to someone else. A negative performance evaluation is a way to create "evidence" of dismissal with cause, in case the employee appeals the dismissal and the DGSC investigates the decision. 2.14 Turnover in the Honduran public sector does not seem to be as high as in some Latin American countries, especially among middle-level staff. This may be partly explained by the fact that the labor pool from which to select technically qualified middle-level candidates is quite limited. When there is a change in government, most of the turnover is concentrated in the top-level posts and somewhat less among lower-level staff. In addition, all political appointees (of which there may be 10 or 20 in each Secretariat) and their staff (personal secretary, advisers, etc.) are replaced. 2.15 Nevertheless, focus group participants (all of whom were middle-level employees) did express some uneasiness about job stability. There were anecdotes of technical staff being reassigned to secretarial jobs or simply marginalized from the daily functioning of the office, in hopes that they would eventually quit out of frustration. If the Secretariat decides to dismiss someone without cause, they must 10 Focus group participants told us that in some regions of Honduras, "caciques" act as intermediaries between job seekers and politicians. "Caciques" negotiate with politicians to arrange employment for people who need jobs, and in exchange they guarantee the community's votes to the politician who needs electoral support. Political influence in personnel management also takes the form of forced monetary contributions to political parties, which use public sector salaries to help finance their activities, particularly during electoral campaigns. In most public agencies, around 5 percent of public employees' monthly salaries are deducted by the ruling political party. Based on some rough estimates, we can calculate that this would result in a total of about 19 million Ls/month (around US$1.3 million) for the political party. These contributions are not voluntary; employees who resist may lose their jobs. Honduras Public Expenditure Review 17 pay the person an indemnization. The dismissal can be appealed, although it appears that there are limited cases of successful appeals. 2.16 Under the new proposal for a revised Civil Service Law, each post will be assigned a fixed base salary, as well as the possibility of earning a variable remuneration on an annual basis (i.e., not a permanent salary increase). The amount that each agency can pay its staff in variable remuneration will be assigned by SEFIN at the beginning of the fiscal year, according to availability of funds. Decisions about how much variable pay to award each employee will be taken by personnel officers at the agency level, in conjunction with the DGSC (or a new personnel office). These decisions are supposed to be based on annual performance appraisal results. However, the subjective and selective use of evaluation, as reported by the focus group participants, raises some concern about the fairness with which the new evaluation system will be applied to determine the level of variable pay. 2.17 The government has been developing measures to control these informal practices, and strengthen the personnel management system. Some of the measures already taken, such as the job reclassification, are expected to have reduced the level of discretion. The salary adjustment would also have contributed to lessening some of the incentives to abuse the recruitment and post assignment practices as a means to allow better pays than what was permitted in the formal scheme. The government has also been preparing a new civil service law, to replace the current law, which is outdated. However, having a new law as such will be hardly sufficient to improve public personnel management. For the new law to be effectively implemented, concerted efforts to strengthen relevant entities in charge of personnel management, especially DGSC, will be essential. Personnel Management in the Health and the Education Sectors: Findings of the Staff Tracking Survey 2.18 The picture that emerges from the foregoing analysis is a weak personnel system without a minimum capacity to plan and deploy staff according to actual needs, without sufficient monetary and non-monetary incentives supported by a robust performance management system. Such weaknesses are supposed to be pervasive across the public sector, and thus affect areas that are key to the government's poverty reduction goals, such as the health and education sectors, where efficiency in service delivery depends critically on staff performance. The results of the staff tracking survey (reported below) confirm the findings that emerged in the focus group and one-on-one interviews with personnel managers in public agencies." 2.19 Several additional factors complicate personnel management in these sectors. For example, the establishment and maintenance of employee information and payroll systems are particularly challenging because the highly decentralized nature of line positions in these sectors makes it difficult to monitor hiring and payment practices and to keep track of employees' attendance, location and job performance. Likewise, the existence of laws (Estatutos, covering doctors, nurses, teachers, and some other medical professionals), which allow public workers to engage in certain practices (such as holding multiple contracts or changing to a new work destination relatively easily) hamper the government's ability to control its human resources and ensure their productivity. Overcoming these problems is not easy, but it is essential for the. two sectors to improve their operational efficiency and achieve the stated poverty reduction goals, particularly in reducing the observed service gaps in rural and remote areas. " The following section reports the key findings of the staff tracking survey. Tables that report various findings are found in the annex. Honduras Public Expenditure Review 18 Box 2.1: Statutory Pay Agreements The impact of inflation on public service salaries during the stabilization program in the early 1990s was severe, as tough fiscal programs were introduced while the exchange rate was liberalized. However, in the medium term, this turned out to be a false economy. Frustrated by their inability to secure offsets for inflation through negotiations with the line ministries and Finance Ministry, the well- organized professional groups ("gremios") turned to the National Congress for redress and found populist politicians ready to help. The result was a series of statutes to regulate employment conditions for doctors, professional nurses and teachers, leaving the sector authorities with no freedom to negotiate either on remuneration or on productivity. The Ley del Midico Empleado was originally passed in 1985. Among its provisions are the following: * It sets a minimum salary for qualified doctors (applicable in private or public employment) * Over 90 percent of the doctors employed in any hospital or clinic must be Honduran by birth * The working day is set at six hours * Doctors are allowed to hold two jobs from the same employer * Any doctor contracted to work two hours or more is entitled to the benefits given by the law. Although the minimum salary for doctors was adjusted slightly in 1993 and 1994, it fell behind inflation during the first halfof the 1990s. By the start of the election year of 1997, the doctors' minimum salary was L. 3,500, the equivalent of 3.5 minimum wages. In that year, following intensive lobbying and strike threats by the Colegio Midico, the Ley del Midico Empleado was amended to provide for a phased program to increase the basic salary by 400 percent over a four-year period. The increases are guaranteed in real terms by being tied to the minimum wage. At the end of the adjustment period, a specialist's minimum salary will be equivalent to 14 minimum wages and that of a general physician to 12 minimum wages. The Estatuto also provides obligations on the employees, for instance, stipulating that a general physician should provide 36 consultations in a working day, and a specialist, 24. However, such requirements too often are honored more in the breach than in the observance. Although the Ministry has attempted to regulate hours at w%ork through the control of entny and exit, this is made difficult in practice by the fact that the personnel administrators are usually more junior than the medical staff. Unsurprisingly, other groups of employees ha%e been quick to attempt to emulate the doctors. In 1997, the teachers achieved the implementation of an Estatuto del Docenie and in July 1999, the professional nurses union finally managed to get their own Estaturo passed, including in each case a significant inflation-prootled salary rise. In 2000. the auxiliarn nurses were pushing hard for a similar provision. Although the Ministry negotiated to concede their specific demands without these being cast in law, the lobby of the National Congress continued, and the beginning of the "electoral season" increases the risk that further laws of this type might be passed bN politicians hoping to enhance their popularity. Similar to the cases in the health sector. the teacher statute raises concern on both fiscal and performance grounds because of the comparative shorter teaching time in Honduras -the average teacher works 160 daN sa %ear compared to 200 in Costa Rica and cver 180 in El Salvador. I hese Esturos are a major factor underlN ing the upw%ard trend of "real" social sector spending in the late 1990s through an increase in personnel costs. How%ever, contrary to the effects of an expansion in the number of positions. salary increases usuallk do not contribute to increases in the real production of Honduras Public Expenditure Review 19 2.20 Personnel information systems: Proper establishment control depends on a reliable information system that records and tracks actual staff deployment. There are weaknesses in this area in both sectors, though the system deficiency appears to be more serious in the education sector. 2.21 Health: Of the total of 15,309 Table 2.5: Health Staff not Located to a Specific Workplace in people employed by the SS, the Ministry's the Central Database SIGAF database does not identify a workplace for 422 staff (2.7 percent of the Number % of staff Average Total total), including 53 physicians (3.4 percent Type of employee of that earnings, earnings, of the physicians) (Table 2.5). The salaries tpe Lo L8/mo5th of these staff total L.2.1 million a month or General physician 28 4.2 17,000 476,000 L.32 million a year; and those of the Prof Nurse 12 1.5 7,000 84,000 physicians alone total L.1 million a month. Auxiliary nurse 60 1.2 3,600 216,000 SS staff interviewed about the whereabouts Promoters 30 7.0 3,000 90,000 Technicians 9 5.4 7,000 63,000 of these employees said that they were Non medical 258 3.5 2,615 674,670 mostly new entrants to the SS who had not Total employees 422 2.7 5,013 2,141,170 yet been assigned to a workplace, or were in process of transfer from one place to another. They were confident that these are not, in general, fictitious posts. Among the staff properly registered in the system, there is a heavy concentration of ministry staff in Tegucigalpa. The metropolitan region (which has 13.4 percent of the population) has 37.3 percent of all ministry staff and 49 percent of all specialist physicians. 2.22 The Education Secretariat (SEP) administers teaching and non-teaching staff separately. The 8,351 nonteaching staff, including school directors, sub directors, secretaries and messengers, together with staff whose contract is under Civil Service arrangements, are administered through and paid by the SEP's personnel office (Subgerencia de Recursos Humanos). Of these, the Civil Service law covers 51 percent, the Estatuto del Docente covers 48.5 percent and 0.5 percent (32 people) are "empleados de confianza" (high level executives or other political appointees cf the Minister). The teaching staff, totaling 46,020 (of which 31,954 or 69 percent, and 14,066 or 31 percent are primary and secondary, respectively) is administered through the Oficina de Escalf6n. The figure for secondary education refers to a total number of contracts, not of individuals because in secondary education the system defines assignments of class hours, rather than full time jobs. As such, one teacher could be holding more than one contract legitimately, and the system is unable to track the exact number of teachers working in the public sector. 2.23 The personnel data for Central Ministry staff, for Civil Service terms staff and teachers with administrative duties are not in a proper database within the Secretariat, but are handled in lists in spreadsheets.'" The information system in the Escalaf6n is in database form but has the following weaknesses: 1 . There are separate databases for preschool, primary, secondary and adult education, making it hard to trace multiple assignments in jobs that are supposed to be undertaken in the same hours. 2. The information system does not include a permanent code (such as the national Identification Number) for the individual identification of the teachers. Although there is a code for each individual in the Escalaf6n, this can change if they get re-graded. This makes it impossible to analyze, for example, the incidence of employees who simultaneously draw checks from both the Civil Service and the Escal6f36n. yet been asindto a worplae,oroer It is possible thaere is database in the Civil Service office with this information, but if oit is ntapparently shared with the Education Secretariat. Honduras Public Expenditure Review 20 3. The codes used in Escalaf6n for the schools are different from those used by the SEP's Statistics Department, which makes it difficult to analyze the relationship between staff assignments and payments and education outcomes. 4. The information for Departmental Offices staffing is incomplete. These are new entities, which do not have full staff establishments authorized in the central budget; so many of their employees are on loan from a secondary school. This is done with official authorization but leads to underreporting of Departmental Office staff in the information systems. 2.24 Evidence on the real assignment of staff* Given the low reliability of the personnel information systems, one of the key questions about personnel management in Honduras' social sectors is: what proportion of staff, assigned to different entities and to different functions, really work in the place of employment indicated in the central personnel databases? Staff deployment is a critical factor in ensuring that public services are actually provided. Any doubt about actual presence of the personnel who are presumed to be working in particular areas needs to be clarified, if public expenditures allocated to specific poverty reduction purposes, be it provision of primary education or health, is to bring about expected benefits. 2.25 Health: Among the surveyed workforce of 14,454 people whose supposed place of work was recorded in the central database, the study found that 9.3 percent (1,344 people) did not work in the surveyed workplace.13 Of these, 1.7 percent had retired or were permanently incapacitated; 5.2 percent had been transferred elsewhere but were still recorded as being hired in this place, and 2.4 percent (347 people) were totally unknown (i.e., pure "ghosts" who are drawing salaries from the SS but are not apparently working anywhere in the system, "earning" the estimated L.1.74 million a month)" (Table 2.2.1 in Annex 2). Physicians (8.3 percent of the GPs and 5.1 percent of the specialists) registered by far the highest proportion of persons who are not known in their supposed workplace. These findings suggest that many physicians who draw salaries from the SS either simply do not work for the system or their place of work is incorrectly identified in the central information systems, making it impossible to monitor compliance with their contract. 2.26 Education: The survey found a lower proportion of "missing persons" in the education sample than in that for health. Among the surveyed workforce of 44,048 people, the study estimates that 5.9 percent (2,615 people) did not work in the surveyed workplace. Of these, 3.0 percent had been transferred elsewhere but were still recorded as being hired in this place, and 2.9 percent (1,280 people) were totally unknown. This is the apparent number of pure "ghosts," whose estimated monthly "earnings" are L.6.4 million. Primary school teachers registered the highest proportion of persons who are not known in their supposed workplace (4.6 percent). The highest proportion of transfers was registered for non-teaching staff (8.5 percent) and for secondary teachers in Institutos Polivalientes (7.9 percent) (Table 2.2.2 in Annex 2). 2.27 Correspondence of real function with training and with nominalfunction: As discussed in the section on government-wide personnel management above, misalignment of staffs skills and their nominal assignments appears to be a problem that causes inefficiency. The survey confirmed that particularly in the health sector, there is significant numbers of trained and highly paid medical professionals (19 percent of the specialist physicians, 12 percent of the GPs, 27 percent of the professional nurses, and 27 percent of the technicians) are assigned to administrative posts, and that 13 Those found to be actually present include some people who could not be interviewed because they were absent from work on the day, or, if they were working, they refused to cooperate, were out in the field, worked a different shift etc. These totaled 16 percent of the sampled workforce. In these cases the survey staff verified their status with colleagues. 14 This is in addition to the 413 people whose apparent workplace could not be identified from the central database, some proportion of whom are also possibly fictitious. Honduras Public Expenditure Review 21 around 20 percent of employees are not in posts that correspond with their formal education or training. This problem is especially marked among nurses, promoters and educators and "others." These relatively low-paid posts might be expected to be particularly prone to the problem of politically motivated assignment of posts. In contrast, most physicians' posts are held by people with appropriate education (96 percent in the case of specialists, and 91 percent of general physicians). This is the result of the formalization of the selection process under the Ley del Midico Empleado, which makes it difficult for someone with the wrong qualifications to get nominated to these posts. 2.28 In the education sector, the mismatch of staff skills and their actual assignments seems to be less of a problem. Overall, 90 percent of the Education Secretariat's staff does a job that corresponds to their nominal position; the main exceptions to this are the Normal Schools (58 percent) and the non-teaching staff (80 percent). Eighty three percent do a job that matches their qualifications; once again, the lowest ratios are registered in the Normal Schools (60 percent) and the non-teaching staff (70 percent). 2.29 The recruitment process and contractual status: The staff tracking survey provides some corroborating evidence on the use of non-competitive recruitment process. 2.30 Health: When asked how they got their present job, 39 percent of SS staff answered that they got it in a competitive hiring process ("concurso"); 32 percent said they were hired directly, 16 percent were "recommended", 8 percent had been transferred from another post, 2.3 percent started by covering someone else's absence and 2.4 percent began by doing volunteer work (Table 2.6). The typical strategy is to enter the SS on a temporary contract and then to lobby for the creation of an acuerdo to make their employment permanent Only 55 percent of all SS employees had an acuerdo from the start of their employment in the SS, and for the case of GPs this falls to 43 percent. Those who started on a temporary contract had waited, on average, 18.5 months before getting their acuerdo. At the time of interview, 96 percent had an acuerdo and 4 percent were on contract. Among GPs, the proportion without an acuerdo is higher, at 11 percent. (Table 2.6). Table 2.6: Hiring Processes and Contractual Status in the Health Secretariat Spec Prof. Aux. Promoters & ialist GP nurse nurse educators Technical Others Total How staff were hired Percentages Competition ("concurso") 62.9 64.4 44.2 45.4 63.0 36.3 23.1 38.7 Hired directly 25.7 21.3 25.4 35.3 22.9 30.7 34.7 32.3 Recommended 5.0 5.8 7.7 7.5 4.9 12.3 30.7 16.3 Transferred 0.0 2.2 9.7 7.2 8.3 9.8 8.5 7.7 To cover someone else 0.0 1.1 3.1 2.5 0.0 1.5 3.0 2.3 After volunteering 2.5 5.3 9.8 2.0 0.0 9.3 0.0 2.4 Other 3.9 0.0 0.1 0.1 0.9 0.1 0.0 0.3 Contractual status Percentages Have acuerdo now 95.5 89.0 96.9 93.7 98.9 95.7 98.6 96.0 Had acuerdo from the start 57.9 42.5 53.8 65.0 52.5 52.9 46.4 55.1 Months Time on temp. contract (av.) /1 22.7 20.1 16.9 15.8 22.9 17.3 18.8 18.5 Note: 1/ For those hired originally on temporary contract 2.31 Education: The vast majority of SEP employees have an acuerdo (99.6 percent) and 84 percent were originally hired with an acuerdo. Only 14 percent entered the Secretariat's employ originally on a short-term contract (Table 2.7). However, not everyone who is hired on an acuerdo achieves it in the stipulated competitive manner (through a public concurso). Some 60 percent were hired through a concurso, but over 20 percent were hired directly or "recommended." Around 10 percent had entered the SEP covering for someone on leave on a short-term contract, and had later negotiated an acuerdo when a Honduras Public Expenditure Review 22 permanent post became available. For those who entered the SEP in this way, the average time spend on a short-term contract was 25 months. Table 2.7: Hiring Processes and Contractual Status in the Education Secretariat Non teaching (Ministry, Primary Secondary teachers Total Depts, teacher Basic Commerce & Poliva- Bach Acad./ Directors) (7-9) computing lientes Normales How staff were hired Percentages Competition ("concurso") 60.2 65.3 43.6 55.4 49.3 55.7 60.1 Hired directly 28.0 13.6 17.3 22.3 25.0 30.7 18.8 Recommended 0.9 2.1 1.5 0 18.7 0 2.7 Transferred 5.6 3.9 25.8 7.4 0 5.9 6.6 To cover someone on leave 5.3 13.2 6.2 13.9 0 6.8 9.7 After volunteering 0 1.8 0 0 5.7 0 1.3 Started in another post 0 0 5.6 1 1.2 0.8 0.7 Total 100 100 100 100 100 100 100 Contractual status Percentages Have acuerdo now 100 99.5 99.6 97.8 100 100 99.6 Had acuerdo from the start 93.9 81.2 77.3 92.6 81.7 70.4 84.1 Months Time on temp. contract (av.) /1 41 23 27 38 5 27 25 Note: 1/ For those hired originally on temporary contract 2.32 The migration of posts: Another problem which the survey confirmed is the migration of posts. Given the heightened importance of geographic targeting to address the rural access problems, the government's ability to assign staff and ensure they perform their duties in the assigned locations is critical for poverty reduction. The survey findings suggest, however, that migration of posts is considerable in both sectors. 2.33 Health: Overall, 5.2 percent of the workforce has been transferred since the SIGAF data on workplaces were generated (in early 2000). Promoters and educators (21.7 percent) and GPs (10.2 percent) were the groups with the highest proportions of transfers. This is reinforced by the survey data on employment histories, which confirm that 40 percent of SS employees have moved their place of work since their acuerdo was originally issued (Table 2.2.13 in Annex 2)." The highest migration rates are registered among general physicians (60 percent of whom have moved) and technical staff (56 percent). The lowest rates are registered among auxiliary nurses (37 percent) and "others" (35 percent). Close to 50 percent of the present day staff of regional offices, national hospitals, CESAMOS and CESARs have migrated since their post was confirmed. Much lower proportions of migrants are registered in the central ministry and regional hospitals. The very high figure of 74 percent for the CLIPERs (peripheral clinics) reflects their recent establishment, using staff transferred from other hospitals. Most migration is from lower to higher levels of the system, and few people migrate downwards. However, a considerable number move between institutions of the same level. For example, of the 49 percent of the present national hospital staff that are migrants, fully half (24.8 percent) came from other national hospitals. 15 This is very close to the figure of 38 percent reported by the 1985 staff census, indicating that little has changed in this regard. Honduras Public Expenditure Review 23 2.34 Migration between posts is even Table 2.8: Proportion of Workforce that has Migrated greater in the SEP. Fully 60 percent of Since their "Acuerdo" was Issued education staff have moved to another post since their acuerdo was issued. The Where they work now: migratory patterns are very clear - from Central Departments Secondary Primary Total lower to higher levels of the system, and Originally Migrants as % of all employees in their present within the same level, but almost no assigned in: workplace, by original workplace migration downwards. Many secondary Central 0 0 0 0 0 teachers move up to occupy posts in the Departments 0 0 0 0 0 Ministry. Such migrants make up nine Secondary 9 0 22 0 6 percent of Central Ministry staff at present. Primary 0 50 15 50 42 Primary teachers move up into the Unknown 0 0 18 11 13 Departments: 50 percent of the present staff60 ofe me perent ofd thaereen sff Note: Based on the 84%/ of staff who initially entered the SEP with an of the Departments had transferred fromacuerdo. primary schools. Fifty-five percent of staff in secondary schools had moved to their present post from elsewhere, 22 percent of them from other secondary schools and 15 percent from primary schools (with 18 percent unknown). In the primary sector, 62 percent of teachers had transferred, 50 percent from other primary schools and I1I percent from somewhere unknown. 2.35 Multiple employment - analysis of the payroll database: In both sectors, significant portions of the staff hold more than one job. This is likely to be a major cause of short working hours (discussed below) and inefficiency in health and education facilities. 2.36 Health: The Estatuto del Midico Empleado allows physicians to hold up to two posts within the public sector and many take full advantage of this. Unfortunately, the institutions concerned are not well equipped to ensure that both the jobs are carried out satisfactorily. Analysis of the national SIGAF database shows that the holding of multiple employment within the SS is observed mainly among physicians, especially among specialists."6 The most common case of dual employment in the SS (112 cases) is that of specialist physicians who also hold on-call posts (guardia), which do not require them to be present at the workplace. This combination of posts leads to a virtual doubling of their salaries without any measurable increase in the amount of time worked for the Ministry. Since in total there are 317 specialists posts and 288 posts for specialists on call, this implies that people who are already hired by the Ministry as full-time specialists also hold 42 percent of the on-call posts. On average, in August 2000 a specialist earned L.2 1,500 (US$1,433) per month for a 6 hour work shift; but if he or she also held an on- call post, this nearly doubled to L.42,000 (US$2,800). This costs the ministry L.34.4 million (US$2.3 million) a year, a figure that will rise in real terms during 2001 as the salary hikes generated by the 1997 reform of the Estatuto del Midico Empleado continue to work through the system. In comparison, a general physician was earning, on average, L.15,861 (US$1,05 7). 2.37 Analysis of the personnel databases of the National University (UNAH) and the Social Security Institute (111SS) together with that of the SS reveals that a significant proportion of SS staff also hold posts in these institutions. Seventy-five SS physicians also hold teaching posts at the University Faculty of Medicine in Tegucigalpa and 66 of these are at the top range of the University pay scale (Profesor Titular III), earning L. 14,269 per month on average Ojust under a thousand dollars). Thirty-eight of these staff work for the SS in the national teaching hospital (Hospital Escuela) or the adjacent Hospital Materno Infantil. Since the Faculty of Medicine is sited in the teaching hospital, it is very easy for staff to rIn all, 182 physicians (12 percent of all physicians in the ministry) hold more than one ministry post. In 51 cases, general physicians (GPs) hold a combination of limited hours posts (e.g., a 2 hour GP's post in one place and a 6 hour GP's post in another), which gives little cause for concern. In all, 214 (14 percent of the ministry's physicians) are on part time contracts that require less than the statutory 6 hours (for a specialist) or 8 hours (for a general physician) associated with full time contracts. Honduras Public Expenditure Review 24 give classes that might take place during their formal work shift under the SS contract without anyone noticing. No one monitors this. A further ten are hired by the SS to work at the nearby Hospital San Felipe. More strangely, there are four physicians on the Tegucigalpa-based UNAH faculty whose SS post is in the Mario Catarino Rivas hospital in San Pedro Sula, a four hour journey away. A further three are assigned to La Paz, an hour and a half's drive away. It seems likely in these cases that someone might be getting short changed. Twenty-six SS staff also have contracts in the IHSS, 22 of them holding full-time specialist physicians' positions in the IHSS and four of them, general physicians' posts. Overall, these physicians earn on average L.16,182 in the SS and L.19,542 a month in the IHSS. Three quarters of these physicians are Tegucigalpa-based and 25 percent of them are in San Pedro Sula." 2.38 But first prize for negotiating multiple jobs in the Honduran public health sector must surely go to the doctor who holds a full-time specialist physician's post in the IHSS, a four hour specialist's post in the SS (worth two thirds of a full salary) and a teaching post at the Medicine Faculty in UNAH (naturally, on the Profesor Titular III grade), with total monthly earnings of L.45,000 (US$3,000). 2.39 Survey evidence on multiple employment: The survey data confirms the picture of multiple employment of professional SS staff already given by the foregoing analysis of the payroll database. Although only 11.2 percent of all SS staff have other paid employment, for specialist physicians the figure is 54 percent and for GPs, 33 percent. Most physicians' second jobs are in their private clinics, but 23 percent of specialists' second jobs are in the SS. Only 10.5 percent of professional nurses have second jobs, but 55 percent of these are in the SS. Those who have second jobs earn from them, on average, L.6,512 a month, which is 31.1 percent more than they earn from the SS. Specialists earn on average from their other posts L.17,563 a month (equal to 90.2 percent of their declared SS earnings); general physicians earn L.10,153 (59.5 percent). Table 2.9: Other Paid Employments of Health Staff - Survey Findings Promoters Specialist GP Prof. Nurse Aux. nurse & educators Technical Others Total %with another paid job 53.6 33.4 13.5 6.9 15.6 20.6 4.8 11.2 Earnings from second job (L./month) 17,563 10,153 9,106 2,203 3,578 5,801 1,586 6.512 2.40 In the education sector, the proportion of SEP staff with another paid post is 23.1 percent. The incidence of dual employment is particularly high among secondary teachers, reaching close to 50 percent among staff in basic schools teaching seventh through ninth grade and in the Escuelas Normales (teacher training colleges) and Bachillerato Acadimico programs. This reflects the fact that the SEP contracts of such staff are for class hours, not full time posts. It is normal for teachers to seek a second contract to top up their salary. Among primary teachers just under 20 percent have another paid post, and the figure is 16 percent for non-teaching staff. Table 2.10: Other Paid Employments of Education Staff- Survey Findings Non teaching (Ministry, Primary Secondary teachers Total Depts, teacher Basic Commerce & Poliva- Bach Acad./ Directors) (7-9) computing lientes Normales % with another paid job 16.1 19.7 48.2 41 15.5 49 23.1 Earnings from second job (L./month) 3,709 5,605 3,365 4,297 5,890 5,378 4,652 1A further 28 physicians hold posts in both IHSS and the Faculty of Medicine. Honduras Public Expenditure Review 25 2.41 Almost all the second jobs are similar in nature to their post in the SEP (93 percent) and the overwhelming majority of second jobs are in other SEP schools (65 percent) or in a public university (1.8 percent) or public technical institutes such as INFOP (1.7 percent). Only 18 percent are in the private sector. As in the health sector, earnings from second jobs represent an important proportion of the income of those who hold them. On average, they total L.4,652 per month, 93 percent of the average earnings from the SEP post that was sampled. 2.42 Attendance and work hours: Anecdotes indicated that staff absenteeism and short working hours are serious problems in both schools and health posts. Therefore, the survey probed this question in the sampled facilities, and confirmed the suspicion that absenteeism (either licensed or unlicensed) and short-working hours are a generic problem in the system, leading to low productivity and poor value for money in service delivery. 2.43 Health: Among the people really employed at the site in the health sector, the average attendance rate over the last five days was 73 percent, with the GPs showing the lowest average attendance rate (61 percent). No group had an attendance rate above 76 percent (Table 2.2.3 in Annex 2). Vacations account for 21 percent and illness for 12 percent of the absences. Both these rates are reasonable: they represent on average 12.5 days per employee year in vacations and 7.5 days per year on average on sick leave. In hospitals of all types, the granting of compensatory time off to offset extra hours worked is a major cause of absences, accounting for 16 percent of all absences in the system (about eight days per employee year on average). But the most disturbing finding is that for 39 percent of absences, no apparent justification was given. This represents, on average, 23 days per employee year in unaccounted for absences, which is roughly 10 percent of the total days the staff are supposed to work. 2.44 The holding of second Table 2.11: Hours of Work of Health Employees posts by specialist physicians is made feasible by the short Spec- Prof. Aux. Promoters Tech- working hours required under the ialist GP Nurse nurse & educators nical Others Total Estatuto del Midico Empleado. Average number of hours worked on the previous day The (self reported) hours of work Administrative 1.4 1.6 2.7 1.0 4.8 3.4 4.4 2.9 of specialists show that they stick Health services 4.6 5.2 5.2 6.9 3.1 4.0 2.7 4.5 strictly to the statutory limit, Other 0.1 0.1 0.3 0.1 0.2 0.3 1.2 0.5 normally working from 7:00 AM to 1:00 PM in a single straightToa6. 68 8. 80 81 76 82 79 shift, allowing them considerable leeway to work in the afternoon and evening in other posts (Table 2.11). GPs report working on average a little more than the statutory six hours, perhaps because fewer of them have second jobs to go to in the afternoon. On average, the self-reported hours of work for SS staff were 7.9 hours and the figure is very similar for the different sorts of employee. 2.45 Education: In the education Table 2.12: Hours of Work of Education Employees sector, the average attendance rate ________________________________ among the people really employed atT thesieverthlstivadysiasNo GPrysecndrey tedcaors nclOhr oa the ste oer te las fiv day wasteaching teacher Basic Commerce Poliva- Bach AcadToa 86 percent, which is considerably (7-9) computing lientes Normales bstervedhn thathecatgor of Average number of hours worked on the previous day Administrative 0.6 0.3 0.9 0.4 0.5 0.9 0.5 staff that had the lowest attendance Educational rates was secondary teachers in the services 4.9 5.2 5.1 5.1 5.7 4.2 5.1 teacher training schools (Normales), Other 0 0 0 0 0 0.2 0 with 63 percent. Vacations account Total 5.5 5.6 6.0 5.5 6.2 5.4 5.6 for 23 percent and illness for 18 percent. As in the case of the SS, both these rates are reasonable: they represent on average eight days per Honduras Public Expenditure Review 26 employee year in vacations during term time" and 6.5 days per year on average in sick leave Strikes accounted for 12 percent of absences, and unpaid leave, for 15 percent. The proportion of unaccounted- for absences is 15 percent of the total (Table 2.2.6 in Annex 2). As for the working hours, on average, the self-reported hours of work for SEP staff were 5.6 hours and the figure is very similar for the different sorts of school, rising above six hours a day only for secondary teachers in Basic schools and Polivalientes (Table 2.12). Conclusions 2.46 This chapter shows that effective management of public sector personnel is key to both aggregate fiscal discipline and efficiency in implementing government programs. The constraints imposed by the Estatutos will be politically difficult to address. As a first step, the government should aim to develop a coherent and comprehensive public sector pay and employment policy that addresses simultaneously fiscal affordability and performance incentives in the public sector. The articulation of such a policy will not eliminate the ability of powerful and organized groups to push for further salary increases but would strengthen the government's hand in politically difficult negotiations and stand a better chance of mobilizing public opinion, especially when unreasonable demands pose a serious threat to fiscal stability or would crowd out other priority expenditures. At the same time, the government should continue efforts to develop adequate capacity for monitoring and controlling public sector employment. The much more difficult challenge of performance management will have to be addressed once there is a minimum capacity for control, although there may be scope for greater use of community supervision as a substitute for administrative control and accountability, in areas such as primary schools and rural health posts. 2.47 The analyses of public expenditure management in the health and education sectors (Chapters 4 and 5) show that, broadly speaking, public expenditures are not egregiously misallocated within each sector. While some adjustments are still needed (e.g., increase in public provision of secondary education), more important challenges are in improving the efficiency with which these resources are utilized. In this regard, one key bottleneck is the weakness of public sector personnel management. The government should address strengthening of personnel management in these two sectors as top priorities within its poverty reduction strategy. 2.48 The staff tracking survey confirms that there are important weaknesses in the staffing information and payments administration systems in the Health and Education Secretariats, and warns that the modernization of the latter should take account of the need to build in rigorous auditing systems. The study indicates that there is a limited problem of "ghost" staff who do not appear to work anywhere but draw their salary regardless, but it finds that absenteeism is a major problem in the Health Secretariat. Both ministries are also plagued by the migration of posts between service production units toward the administrative offices. This is due to a combination of employee capture (the post belongs to the individual) and administrative inflexibility (managers cannot get new posts where they need them so they connive at shuffling exiting posts between institutions). Uncontrolled migration of posts undermines any government effort to target specific geographic areas in increasing supplies of health and education services. A better system to track staff deployment, including improved personnel data systems and practical mechanisms for managing and supervising staff assignments (e.g., community supervision) needs to be developed. Since it is unlikely that the government will be able to re-allocate these migrated staff back to their original places of assignment, there will be a need to formalize this informal situation by allowing these individuals to change their official posts to where they actually are. This one-time adjustment would allow the government to establish a more accurate personnel database on actual staff deployment. Simultaneously, the current system where budgeted positions move with individual staff should be re-examined. 18 In addition to this leave taken during term time, teachers have long leave in December and January when the schools are closed. Honduras Public Expenditure Review 27 2.49 Multiple employment is also an important problem in health, reflecting "employee capture" of the system by senior doctors. This problem will not be easy to unravel, in the face of the statutory limitations in the Estatuto del Mddico Empleado. However, as a start, the Health Secretariat should consider stopping hiring expensive specialists for administrative posts, using instead university professionals with training in administration, and not assigning on call posts to specialists that are already hired on full-time contracts unless there is no one else who could occupy the post and it is indispensable to have it filled. In education, multiple employment is even more widespread than in health, but is concentrated among secondary teachers, whose contracts are for class hours, rather than being full-time posts. Nevertheless, the information systems should be strengthened to allow the managers to identify the total hours contracted to each individual, in order to ensure that they are programmed consistently with the teacher's ability to deliver the class. In education, arguably the most serious problem is bureaucracy: almost 40 percent of staff is engaged in administrative functions. Given the shortage of teachers in some areas, the possibility of reassigning these teachers in administrative functions to teaching positions should be explored. Honduras Public Expenditure Review CHAPTER 3 PUBLIC EXPENDITURE MANAGEMENT Institutions and Budgetary Performance 3.1 As discussed in the previous two chapters, Honduras has regained aggregate fiscal discipline in the latter half of the 1990s and has made attempts to protect social sector spending. These are both important efforts for promoting economic growth and poverty reduction. There are questions, however, on the fiscal sustainability of the current spending patterns, and efficiency of resource use in the social sectors. This chapter assesses the quality of the government institutions for public expenditure management (PEM), especially with respect to their capacity to maintain aggregate fiscal discipline, allocate resources to policy priorities, and implement expenditure programs efficiently. 3.2 Effectiveness of public expenditure management is determined to a significant degree by the quality of institutions in the government. Institutions, understood as sets of formal and informal rules, influence behaviors of public officials involved in government activities, in this case, the process of planning, allocating and spending public money for the purpose of implementing government policies and programs. These rules determine what roles each actor (individuals or organizational units) play, how incentives of individual officials are structured, and how information is captured and flow among them. 3.3 A good PEM institutional arrangement is one which aligns formal organizational roles, incentives, and information in such a way as to maximize effective coordination among different actors involved in the budget and financial management process and to produce desired budgetary outcomes. governments, especially those in developing countries that suffer from resource constraints, need to maximize the efficiency of resource use both in terms of their allocation and the "value for money" they get from executing the allocated resources, while not jeopardizing macroeconomic stability and sustainability of the expenditure programs. Good budgetary institutions would allow governments to balance the three inter-related objectives of aggregate fiscal discipline (level 1), efficient allocation of resources according to the government's policy priorities (level 2), and efficient use of resources to achieve program objectives (level 3). Simultaneous achievement of these three levels of budgetary outcomes is a key goal of any PEM institutional reform. In addition, as a HIPC country, Honduras is expected to demonstrate a minimum capacity to track public resources that are allocated to poverty reduction. This largely depends on the extent of overall transparency in the budget and financial management systems, and the robustness of the existing accounting and reporting mechanisms. Table 3.1: Basic Elements of Public Expenditure Management: "Three-level" Analysis Aggregate Fiscal Budget totals should be the result of explicit, enforced decisions; they should not Discipline merely accommodate spending demands. These totals should be set before individual spending decisions are made, and should be sustainable over the medium-term and beyond. Allocation to Strategic Expenditures should be based on government priorities and on effectiveness of public Priorities programs. The budget system should spur reallocation from lesser to higher priorities (Allocative Efficiency) and from less to more effective programs. Operational Efficiency Agencies should produce goods and services at a cost that achieves ongoing efficiency gains and (to the extent appropriate) is competitive with market prices. Source: Allen Schick (1998). 3.4 The way in which the budget process is organized has significant impact on the performance of the public expenditure system. Different aspects of the budget process can affect one or more of the three levels of budgetary performance. This section provides a summary analysis of Honduras' budget process, highlighting institutional and technical weaknesses that give rise to observed budgetary performance in Honduras Public Expenditure Review 29 the country. The following section summarizes the budgetary performance at the three levels and discusses some of the key institutional roots of the observed performance. Budget Process in Honduras 3.5 Budget preparation: The budget preparation process is characterized by: (i) incrementalism, whereby allocations are decided mostly as semi-automatic adjustments to the previous year's allocation; (ii) limited use of performance information such as the physical and financial goals in POAs; (iii) reliance on formalistic legal analysis (dictamen); and (iv) scant legislative review of the Executive's budget proposal. The result of these practices is a budget that does not adequately reflect the government's policy priorities, or the spending ministries' real financial needs, a typical problem in a number of developing countries. 3.6 Preparation of instructions and guidelines: The first step in the budget process is preparation of instructions to spending ministries. These should include tentative ceilings based on estimates of macroeconomic variables such as growth and inflation. The basis for the estimates should be transparent and available to ministries for discussion prior to the determination of sectoral ceilings by SEFIN. The Honduran Budget Law requires that evaluations and forecasts begin in March based on results of the previous year's outlays. The results of the macroeconomic exercise are: (i) forecasts of aggregate revenue, expenditure, and fiscal deficits; (ii) estimated inflation; (iii) GDP growth; and (iv) aggregate and sectoral ceilings. The Economic Cabinet, the Budget Department (DGP), and spending ministry representatives meet to develop: (i) a tentative overall public expenditure ceiling, estimates of earmarked funds for the fiscal year, and preliminary revenue/expenditure estimates; and (ii) sectoral ceilings based on the government's policy priorities. 3.7 Development of needs assessment/program planning: From July to August, the spending ministries develop program strategies (lineamientos y estrategias) to include in their proposals to SEFIN (anteproyectos). These are based in principle on both evaluations of activities (e.g., 1999 Informe Anual de las Actividades en Salud) and operating plans for the year (POAs). In principle, ministries develop physical and financial goals for each department/activity within the context of tentative ceilings provided by DGP. Ministry and DGP advisors review proposals and negotiate a final set of programs and objectives in August. Except for capital projects, spending plans are annual only. 3.8 In practice, development of program strategies at the ministry level are driven by SEFIN revenue ceilings, government policy directives, and constitutional/legal earmarks of expenditures (e.g. five percent of gross receipts transferred to local governments). Most annual budget presentations are largely based on incremental increases/decreases from last year's base with adjustments for inflation. Requests are not made on the basis of performance reports indicating needed allocational changes in activities. Key public investment projects, such as those carried out by FHIS or the decentralized institutions (public enterprises) are not within SEFIN scrutiny or part of the computerized public investment reporting system (SISPU) due to the lack of comprehensiveness of the central government budget. Even those capital expenditures included in the central government budget are based largely on demand for new facilities and crude estimates of existing facility condition. Honduras Public Expenditure Review 30 Box 3.1: Coordination between Capital and Recurrent Expenditure Planning9 A typical %%eakness in public expenditure planning in developing countries is poor coordination between capital and recurrent expenditures. Historically many countries have maintained dual budgeting practices, whereby capital investment planning is not integrated with the budgeting exercises. While Honduras abolished its Secretariat of Planning several Nears ago, in an effort to overcome the institutional separation between investment planning and budgeting, there is a continued problem of capital-recurrent coordination, especially between the social fund (FHIS) and social sector ministries. In the education sector, after the Secretariat of Public Education (SEP) surrendered its school building functions in 1996, FHIS has spent over half its resources on building new schools and rehabilitating and replacing existing classrooms. In 1995-98. FHIS constructed or upgraded 15 percent of the national stock of primary classrooms, taking advantage of its freedom from the bureaucratic constraints, which govern contracting in the line ministries, and using private contractors. In the same period. FHIS decisions about investments in educational capital arose from the local-level consultations with communities about their priorities, within the FHIS' menu of basic social infrastructure. Although SEP had to sign off on the proposed in%estment and although the FHIS follow%ed general SEP guidelines on the design of the works, there was little SEP input on priorities from a sector planning perspective. This led to concerns that the resulting resource distribution pattern may not reflect the country's (as opposed to specific communities') priority needs. It also led to complaints by the SEP property di%ision that it had insufficient say on the specification and quality of the works. Reliance on local'demands for determining allocations of investment resources, when local actors are not required to provide for operations and maintenance (O&M) of facilities, is conducive to a bias in favor of investment at the cost of rational recurrent expenditure planning and allocations. The national government may be forced to pro% ide O&M of facilities which it might not have built on its own, ifithad a rational system of investment planning nation-wide. If such patterns cumulate, they could negatively affect sustainability of specific in'estments as well as fiscal balance of the sector budget. These caveats notwithstanding. in fact. the correspondence between the pattern of FHIS' education investments and the pattern of apparent needs was relatively good, with an r2 of 0.43. In contrast- there was almost zero correlation between the allocation of new teachers (controlled by SEP) and the distribution of the coverage deficit. This hardly suggests that the SEP would have been likely to allocate capital resources more closely in line with apparent needs. In fact, it is fair to say that SEP currently lacks a credible plan for physical investments and for recurrent expenditure (i.e.. mainly personnel). Partly due to SEP institutional weakness. FHIS has been drawn in to play the key role in articulating local demands. While FHIS has played an important role in promoting community participation in investment planning, its dominance in social sector investment planning carries the above- mentioned danger of capital-recurrent imbalance. Besides. there remain serious problems of access to primary education in remote areas. and reaching those remote communities within a tight fiscal resource envelope would require an ability to determine relative importance of resource distribution from a national perspective that is duly informed bN an accurate sense of local needs Making that determination should not be a role for FHIS. but for SEP. FHIS can and should contribute by capturing local demands. but simultaneous efforts at strengthening SEP's planning capacity, including physical investment and staff deployment plans based on sharpened targeting criteria, should be pursued as one of the most important challenges for Honduras' education sector. Over time, it is also necessary to de%,elop municipal governments' capacitN to coordinate local investment needs. following broad polic, guidelines from the national go%ernment. '9 The specific information here draws from "Informe Final: Evaluacion Ex-Post del Fondo Hondureno de Inversion Social (FHIS 2)," July, 1999, ESA Consultores. Honduras Public Expenditure Review 31 3.9 Review of ministry requests: After receiving the ministry requests in late August, SEFIN performs a legal/financial analysis of spending requests (via dictdmenes). The dictamen is a formalistic legal analysis in which recommended expenditure levels for a particular project do not follow directly from transparent cost-benefit reviews. In addition, SEFIN's Expenditure Analysis Unit performs a series of ongoing program reviews, comparing planned objectives and expenditures with outlays, and attempting to explain the differences. Its report (Andlisis de Presupuesto Para Ejercicio 2000) provides relatively crude performance information to SEFIN and accompanies the final SEFIN budget to Congress. Before congressional submission, ministry finance/planning teams are permitted to appeal final SEFIN recommendations based on dictdmenes before submission to Congress. Typically SEFIN cuts about 15 percent of the ministries' additional requests. Beyond these cuts, SEFIN makes few changes to the composition of the recommended requests. 3.10 Legislative review and approval: The Executive's proposal is presented to the Budget Committee of the Congress by September 15. The proposal is often not reviewed until the very end of the fiscal year, despite SEFIN's effort to present it to the Congress on time.20 In the end, the Committee makes few inputs and serves largely to rubber stamp proposals made elsewhere within Congress and the Executive. Differences between the legislative and executive versions are by law negotiated by January 1 at which time the fiscal year begins. But in few fiscal years are there major differences. This does not mean, however, that the Congress is entirely powerless in budgetary decision making. Quite the contrary, the Congress has, in fact, passed legislation with significant fiscal and budgetary implications (e.g., Estatutos, the budget management rule in 2000 that every budget modification, however small the amount, be submitted for congressional approval). The likely reason for the Congress' scant attention to the Executive's proposal has more to do with its ability to override the budget law with few institutional restraints. 3.11 Budget execution: There are a number of technical and institutional weaknesses during budget execution in Honduras. Different accounting methods are used by the Budget Department, the Accounting Department, and the Central Bank, which makes effective monitoring of budget execution unnecessarily cumbersome, and in some cases may even result in payments arrears that go undetected for a while. Monitoring of physical execution is weak and incomplete (e.g., FHIS projects not covered by SISPU). The bureaucratic processes for processing commitments and payments are quite cumbersome, and slows down the operational efficiency of spending units. This is reinforced by the lack of credibility in the approved budget, which necessitates a large number of in-year transactions for modifying budget allocations. This appears to result from a combination of: (i) weak capacities of spending units to estimate their real spending needs; (ii) excessive details in the budget classification which force spending units to assign estimates to very minute line items, combined with their limited discretion for reassigning allocated funds across line items; and (iii) incremental budgeting by SEFIN, which exacerbates the disconnect between spending units' real needs and budget allocation. 3.12 Typically major methods of ensuring control during execution are: (i) periodic reporting of actual outlays against the budget plan; (ii) quarterly apportionment/allotment of funds by the ministry of finance; (iii) enforcement of transparent transfer and reprogramming policies that prevent changing the government's or legislative purposes but allow for management flexibility to meet changing realities on the ground; (iv) control of personnel positions to ensure linkage of the departmental establishment list, approved rates and positions, and the treasury payroll system; (v) monitoring revenue forecasts and collection rates and capacity to impose cash limits or aggregate budget adjustments without substantial impairment to planned allocations and program operations; (vi) monitoring physical progress toward service objectives via comparison of norms/standards with workloads and results; and (vii) internal 20 Congress did not review the Executive's 2000 budget proposal until three days before the close of the fiscal year. Congress then debated the proposal without prior establishment of norms or criteria for debate. Honduras Public Expenditure Review 32 control at the ministry of proposed commitments and outlays for legality and funding sufficiency. In Honduras, there are problems with the design and application of all such methods. 3.13 Quarterly allotments and cash management: After congressional approval, SEFIN exercises general control by allotting funds to ministries in quarterly quotas. The Cash Flow Committee composed of the Central Bank (BCH), and SEFIN's Budget, Public Credit, and Accounting Departments meets every Thursday to monitor cash flow and decide on releases for the next week and (at some of the meetings) on the next quarterly allotment. However, the Committee does not utilize modem financial programming techniques to allot quarterly quotas according to the spending units' spending schedules. Instead, cash releases are based on availability of funding in a given quarter (i.e., cash budgeting). This practice, commonly found in many developing countries, has contributed to avoiding overspending, but undermines operational efficiency of line ministries. In fact, these quotas are effectively quarterly budgets in that ministries have to go back to SEFIN within 15 days of the next quarter and newly request funds. 3.14 The flow of budgetary information for decision-making (allotments, reporting, accounting) during execution is slowed by rules and procedures over which line officials have little discretion. There are at least eight laws affecting budget execution which had to be examined by ministry officials prior to pay approval in the past (e.g., Ley de Presupuesto, Ley de Contabilidad, Ley de Contrataci6n del Estado, and Ley de Control Interno). There are multiple levels of often-intrusive input controls over all budget transactions. For example, ministry purchases of maintenance parts require compliance with the budget law (pay orders), law of contracts (purchasing) and internal control (for outlays). Each law requires separate forms and approvals which are time-consuming and resource-intensive. Some spending activities, such as maintenance, may be deferred to avoid the trdmite involved. In addition, the governing structures of ministries ensure that decisions must be approved at the highest levels even for relatively minor line-level transactions. In the wake of Hurricane Mitch in 1998, which required rapid releases of funds for reconstruction and rescue projects, a budget delegates cadre (preintervenci6n) was created to facilitate payments by providing one-stop approvals. 3.15 Payments process: In spite of the one-stop approval process mentioned above, sectoral ministries must still request purchase and pay orders (form F-1) for all supplies and equipment from SEFIN's Budget Department. The budget director gives final approval to all of them except those for the Secretariats of Education and Health, which are piloting the integrated financial management system (SIAFI) and some procedural simplifications. Checks are issued to vendors and employees by the Treasury Department. While the Health Secretariat must actually send their directors of personnel to retrieve their monthly checks in Tegucigalpa, the Education Secretariat is served by pagadurias especiales from the Treasury. There is a similar degree of centralization with respect to disbursement of funds originating in external loans. Although the vast majority of these external credits finance investments in sector ministries, all the disbursement decisions of the approved credits are centralized in SEFIN's Public Credit Department. 3.16 With the introduction of SIAFI which seeks deregulation of operations/compliance and decentralization of decision-making authority, the days required to approve pay orders has decreased from 56 to six and the documents required by vendors for issuance of a check have been reduced from 20 to only six on average. To date, however, "decentralization" would mean in practice increased reporting requirements rather than increased managerial discretion over budgets and staff. Further advances in financial devolution within the central government would require both a change in the legal framework for government financial management and a reform of SEFIN's functions and organizational structure. 3.17 Accounting and reporting: Approved purchase and pay orders permit ministries to execute planned activities consistent with budget categories. Ministries receive quarterly allotments of funds from SEFIN. Ministries then report quarterly outlays against planned quarterly budget targets to the Honduras Public Expenditure Review 33 Accounting Department (Contaduria) and to the Budget Department which record them respectively according to the chart of accounts and budget codes. One problem which could affect the quarterly balance is that BCH accounts are recorded on a cash basis while Treasury accounts are on an accrual basis. In May 2000 (when the PER mission took place), there was a L.758 million difference attributable to outstanding approved pay orders and checks in process. This problem also occurs in public enterprise and decentralized institution accounts where the basis is some mixture of modified accrual and cash. The resultant deficit and financing depends on the accounting basis, meaning that financial accounts contain heterogeneous data. This has led to arrears and a floating debt in some cases. This sort of problem could be alleviated by adoption of a unified fiscal account at BCH, which would eliminate the need for bureaucratic processes associated with issuing checks and recording payments. It would add transparency to the allocation of resources, eliminate idle funds and facilitate the efficient use of available liquid resources to finance budget activities. 3.18 With the introduction of new accounting codes for SIAFI, there have been problems reconciling accounting codes with budget categories. Thus, accounts are still posted on a cash basis while budgets are recorded on a mixed-accrual basis (accrued spending and cash revenues). Detailed line-items for improved control of balances also mean that ministries must request modification approval from their senior staff levels/SEFIN/Congress for even the most minor shift of funds. In addition, there is still no monthly reconciliation of budget and accounting transactions at the commitments level which may be creating cash management and expenditure control problems. Accounting maintains cash accounts; the Budget Department records transactions at the commitments level. SIAFI will eventually include commitments data and permit regular reconciliation. Accounting prepares and publishes quarterly financial cash flow statements for all general and special funds (Memoria: Actividades Realizadas). 3.19 While outlays for personnel expenditures are also reported by ministry, they are not linked by the Budget Department to personnel plans in workload or tasks or compared to actual payroll outlays by position. As yet, POAs do not attempt to link activities to personnel resource inputs, nor does the Expenditure Analysis Unit of SEFIN focus as yet on personnel productivity. Because of the line ministries' weak capacities to administer their payrolls, to this date, salary payments (especially teachers') are handled directly by SEFIN's Budget Department, even though the series of financial management reforms clearly envision the Budget Department to play less operational roles and instead play the roles of norm-setting, oversight, coordination and evaluation of public expenditure management. 3.20 Budget modification: While the Budget Department's quarterly cash flow forecasts are relatively accurate, institutional and regulatory factors intervene to change planned vs. actual ministry allotments. Budget outturn data indicates that many planned and unplanned modifications are occurring without seriously affecting the aggregate level of fiscal discipline (see analysis of budget credibility below). There are at least three types of modifications: (i) congressionally approved packages of modification requests from ministries; (ii) reprogramming by SEFIN to accord with changes in policy priority; and (iii) diversions by the executive and/or congressional earmarks for pet projects. According to SEFIN, about 98 percent of the modification requests are due to cost estimation problems on existing or new projects and needed counterpart funds. The other two percent can be attributed to legislative projects added to the budget during the fiscal year. 3.21 However, data from the Health Secretariat indicates that there may be an "institutionalized" pattern of budget modifications due to the mechanical incremental budget formulation. Table 3.2 shows budget allocations to ambulatory and hospital programs. In each of the six years reported here, there is a significant re-allocation of the budget from the hospital programs to ambulatory programs through budget modifications. One plausible explanation for this repeated phenomenon is that the Secretariat of Health does not take into account the previous year's budget execution level as a base line for formulating a given year's budget proposal. Rather the budget proposal appears to be an incremental adjustment to the previous year's approved budget (or possibly the ministry's budget proposal to SEFIN). The Health Honduras Public Expenditure Review 34 Secretariat also reported that it requested 460 modifications in 1999, which took about one and a half months for SEFIN approval. This is clear evidence that line ministries do not see the approved budget as authoritative means for budgetary resource allocation. Table 3.2: In-year Budget Re-allocation - Ministry of Health, 1994-1999 Approved Modified Executed 1994 102 Enfermed. Trans. y c. ambi 22,900,000 80,556,279 79,767,609 103 Atenci6n medic. Hospitalar 51,652,076 9,480,073 9198,143 1995 102 Enfermed. Trans. y c. ambi 32,949,496 117,533,382 115.124,122 103 Atenci6n medic. Hospitalar 70,790,205 9,420,806 9,310,320 1996 102 Enfermed. Trans. y c. ambi 42,858,000 140,204,827 139,658,659 103 Atenci6n medic. Hospitalar 95,757,912 26,857,854 26,814,872 1997 102 Enfermed. Trans. y c. ambi 57,533,232 168,178,486 160,469,006 103 Atenci6n medic. Hospitalar 111,333,484 9,155,593 9,119,535 1998 102 Enfermed. Trans. y c. ambi 64,497,377 184,450,930 180,580,542 103 Atenci6n medic. Hospitalar 136,112,974 9,852,650 9,417,506 1999 11 Atenci6n ambulatoria 70,545,265 190,327,038 188,731,047 13 Atenci6n hospitalaria 160,541,933 32,774,016 31,633,535 Source: Secretaria de Salud 3.22 Expenditure monitoring: Expenditure monitoring is the responsibility of the Budget Department, which monitors financial performance throughout the year. The SIAFI system now covering six ministries permits relatively integrated entries into its budget, accounting, and treasury modules that allow tracking of budget progress and facilitate cash management and control. Purchases are tracked and controlled in the accounts, but not monitored at the storeroom or inventory level. FIS public works projects are tracked internally by FHIS. Spending ministries make regular internal reports on their own physical program performance, but SEFIN's expenditure analysis unit lacks staff for a thorough analysis of budget programs. The government is also developing a system (SISPU) to track financial and physical execution of public investment projects. 3.23 Evaluation and audit: The final step in the budget process consists of feedback on physical program performance into next year's budget formulation phase. Lessons from a review of the strengths and weaknesses of program design and implementation should influence decisions on the next budget. 3.24 Honduras is no different from most countries in this regard. Not much feedback occurs, or if it does, the information is not used to reallocate resources the next year. The Management Planning and Evaluation Unit in SEFIN (UPEG) is responsible for this function for both program results and management performance. While rather high-quality evaluations are performed by some spending ministries of their program performance in relation to norms and standards (e.g., health), UPEG and the rest of SEFIN focus on macro-level concerns such as: fiscal deficits, exchange rate policies, GDP, balance of payments, debt management, and exceptional programs such as those related to Hurricane Mitch. 3.25 Similarly, the Contraloria General (CG) is responsible for ex-post audits and in many countries this includes program performance reviews. In Honduras, the audits are largely financial. Due to lack of independence and budget resources, UPEG and the Contraloria do not focus on managerial and program effectiveness questions that would satisfy public demands for accountability. The CG also suffers from mixing internal and external audit functions, which confuses objectives of internal control (management improvement) and external audit (due diligence). Its budgetary and human resources are limited, and it has never conducted an audit of a final annual report of the public sector prepared by the Accountant General's Office. Instead, the CG focuses on specific audits of particular public sector entities or programs, some of which are requested by the Congress. Many of these special audit reports are not published. The mission was shown, for example, an audit report of a public sector entity in the education Honduras Public Expenditure Review 35 sector with a number of problematic observations. The audit was conducted several years ago, but the CG had not been given the authorization to publish the report. 3.26 Capacity for expenditure tracking: Honduras has made significant progress in improving the quality of expenditure data with the gradual introduction of SIAFI and the accompanying procedural and presentational improvements in its budget. It now produces very detailed line-item budget data including broad functional and economic classifications as well as broad breakdowns by programs. The budget is available on CD-ROM and on the Ministry of Finance's web site. In addition, and starting with the 2001 budget: (i) there will be some important presentational improvements, including comparisons with expenditure outcomes from previous years, clear classification of HIPC-related expenditure items and three-year projections; (ii) the costing of poverty programs included in the current version of the PRSP will be taken directly from the detailed three-year projections presented in the draft 2001 budget; (iii) the budget will be immediately published on the internet and available on CD-ROM once approved by Congress; (iv) SEFIN now presents to the Congress quarterly reports on budget execution; and (v) once the SIAFI system is more operational, monthly fiscal accounts will be posted on the internet within two months of execution, with some written analysis. 3.27 Notwithstanding this progress, there are still some shortcomings that complicate the tracking of spending programs, particularly poverty-related programs. The use of opaque budget categories which make it difficult to understand the actual allocations of significant portions of the budget make monitoring difficult. The most obvious is the extensive use of global allocations (asignaciones globales) as a spending category in the budget (about 14 percent of the 2000 budget and 20 percent of the budgets of the ministries of education and health). Even in the executed budget global allocations accounted consistently for more than 10 percent of total expenditures (except 1999), which is inexplicable if the main reason for having these allocations is a way of accommodating unexpected expenditures that may arise during the year. Eventually, much of these global allocations are spent on wages, suggesting a very weak public sector control and lack of precise knowledge of the required allocation for the wage bill. Moreover, some of the entities that receive these allocations (such as the legislative and the judiciary) are not required to report how they were spent. While these spending categories give the government considerable flexibility, this practice seriously undermines the transparency and accountability of the budget process (Table 3.3). Another opaque classification is that of officialized services for unspecified capital and current transfers (including FHIS). As in the case of global allocations, the purpose of this budget category is unclear, and it prevents an easy verification of the actual use (both intended and actual) of an important portion of the budget. To exacerbate these problems, a significant part of the global allocations are assigned to these centralized services (Table 3.4). Table 3.3: Asignaciones Globales and their Probable Agency Destination in 2000 Agency Asignaciones Globales as % of Share of total Asignaciones agency budget Globales Secretaria de Educaci6n 19.0 28.1 Secretaria de Salud 20.0 14.9 Poder Judicial 100.0 13.2 Poder Legislativo 100.0 10.2 Ministerio Pblico 100.0 5.2 Presidencia de la Repfblica 86.8 5.0 Organismo Electoral 100.0 3.9 Procuraduria General 100.0 0.9 Servicios Centralizados 12.0 15.9 Other --- 2.8 Total 14.4 100.0 Source: Secretaria de Finanzas. Honduras Public Expenditure Review 36 Table 3.4: Asignaciones Globales and Servicios Centralizados (% of total executed budget) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Asignaciones Globales 13.7 13.0 10.9 11.4 10.8 10.5 12.7 13.8 13.6 8.9 Servicios Centralizados 14.3 16.9 12.6 15.5 10.3 12.8 20.5 21.6 29.8 24.8* Source: Secretaria de Finanzas * Preliminary 3.28 On the institutional mechanisms for expenditure monitoring, SIAFI allows real-time tracking of budget execution in specific ministries (it has been installed in several ministries, including Finance, Education and Health). However, as long as the approved budget continues to suffer significant modifications during the fiscal year, it is a poor guide for ex ante identification of how the budget will be executed. Establishing credibility of approved budgets should be one of the priority efforts for effective expenditure tracking. 3.29 Beyond the relatively simple task of expenditure tracking/monitoring, a greater challenge is to evaluate impacts of the public expenditures. The government does not yet have a reliable system for measuring and evaluating the poverty impact of public expenditures, for which institutional strengthening of the Expenditure Programming and Evaluation Units (UPEG) is key for this purpose. In turn, audits by the Controller General's Office focus on evaluating the legality of public expenditure execution as opposed to its impact, and its capacity is, in any case, limited. Box,3.2: Governance and Transparency in Honduras Since the transition to democracy in 1982. Honduras has made a concerted effort to improve go%ernance and transparenc*. especially in curbing the corrupt practices prevalent during the military dictatorships when political elites overtly helped themselves to public funds, made deals with the private sector. and intervened personally in the allocation of budgetary resources and jobs. Although Honduras is making steady progress in improving governance and transparency, much remains to be done. While this PER does not aim to provide a comprehensive analsis of governance issues in Honduras, the report does point out a number of areas in public sector management where the adoption or deepening of reforms N%ould substantially contribute to impro%ed go%ernance, accountability and transparency. Despite recent ad%ances in state modernization, control mechanisms are still incomplete and those that do exist are rarely respected. For example, the 1976 civil service law requires that public agencies hire their emploNees based on merit considerations through a competitive hiring process. How%ever, a 2000 survey of health ministr-, emplo)ees showed that over 50 percent were hired without going through the required competitive selection. instead. they were hired -directly" or "by recommendation." The same survey found that approximately 30 percent of education ministry employees were hired outside a competitive hiring process. Anecdotal evidence suggests that political connections are often necessary even to gain access to competitive hiring. Focus group members and interviewees reported that local political bosses frequently control access to information about upcoming selection processes. There hate been a number of attempts during the 1990s to address issues of governance (e.g.. anti- corruption commissions) but they were not sustained or were viewed as political. The large increase in aid in response to Mitch underscored the need to refocus on governance. not only to assure donors that their support would be handled transparentl). but also as a key longer-term development challenge. In recent vears, and as part of the public sector modernization program, the government has begun to take important steps to improve governance and transparency. These include ongoing efforts to: strengthen the Controller General's Office. %ith IDB support impro%e procurement systems. develop and extend SIAFI. carr out a transparent privatization process, develop appropriate regulaiory agencies for pri,,atized enterprises and sectors. e%aluate compliance with the IMF's ROSC and develop and action plan to impro%e compliance and the government's ability to track spending on poerty. Honduras Public Expenditure Review 37 In addition, and going beyond public sector management, the government has recently launched a program aimed at designing a participatory anticorruption strategy. The program, coordinated by a steering committee which includes members of government, civil society and business community, consists of three main components. First, the compilation and analysis of a survey of households, businesses, and public officials to identify key problems of transparency and governance in the public and private sector. Second and on the basis of the surveys, the elaboration by the steering committee of an anti-corruption strategy. This anti-corruption strategy will include both long-term and short-term draft action plans that will establish the timing and responsibilities for each action. Third, the work of the steering committee will be discussed in a national workshop, to which all major stakeholders, including civil society, business, and government representatives will be invited to participate, In addition to raising awareness among the general public on the economic and social costs of corruption, the main objective of the workshop will be to discuss the draft action plans prepared by the steering committee. Different focus groups working in specific areas will be created to review the proposed strategies, and a final action plan to fight corruption in the public sector will be drafted. Institutional Roots of Performance 3.30 Aggregate fiscal discipline: As shown in Chapter 1, Honduras' level 1 performance has been relatively satisfactory since the mid 1990s. However, examination of the budget process and related institutional arrangements gives rise for concern. Budgeting relies neither on quantitative spending and deficit targets nor on a medium-term perspective even at the level of macroeconomic projection. It is evident that a reliable and binding macroeconomic projection is not being developed to inform budget formulation. The macroeconomic estimates of the Budget Department are often late in the calendar and are not made public. For its part BCH believes it cannot develop reliable macroeconomic estimates or budget ceilings since the closing budget statements from the past year have not yet been presented. Obviously not having a macroeconomic framework at the beginning of the budget formulation process prevents the government from developing a budget proposal that is solidly grounded in a realistic estimate of its own resource and expenditure envelopes. 3.31 To the extent the government's capacity to maintain fiscal discipline has improved since the early 1990s, it remains relatively fragile as it depends less on robust formal institutional mechanisms than on a limited number of informal practices, particularly the strong centralizing role of SEFIN, the Budget Department in particular, in fiscal and financial management. SEFIN is the dominant actor in budgeting and financial management. It sets budgetary ceilings for ministries, approves (or disapproves) creation or increases in administrative posts (by deciding or not to fund these posts), and controls actual release of cash during budget execution. Even within the entity's approved cash limits, the most minute spending decisions require SEFIN's authorizations (except in the Secretariats of Health and Education to which some amount of financial management decisions have been deconcentrated). Furthermore, some of the key government transactions such as supply of medicines and educational materials to the final service delivery points, take place as a transfer of centrally-purchased goods, thus limiting the possibility of overspending by these entities. 3.32 A dominant ministry of finance is a key institutional element to strengthen fiscal discipline.21 However, the current state is still not very desirable from the point of view of both sustainable aggregate fiscal discipline and operational efficiency in public expenditure management. The informal control by SEFIN to maintain fiscal discipline has proven relatively effective in recent years, but its limits are already quite obvious from the ways in which some of the recent threats to fiscal discipline have come about. Presently the most serious threat to aggregate fiscal discipline, the large increases in personnel expenditures, comes from an arena that is effectively outside SEFIN's control. These increases were 21 Ed Campos and Sanjay Pradhan, "Budgetary institutions and expenditure outcomes : binding governments to fiscal performance," World Bank Policy Research Working Paper 1646, 1996. Honduras Public Expenditure Review 38 negotiated directly by the powerful labor unions (gremios) in the health and education sectors (sometimes even bypassing the sectoral ministry) and supported by the Congress. Honduras lacks formal institutional mechanisms that could restrain this sort of behavior, such as the requirement that any new policy proposal that carries additional fiscal cost be accompanied by a proposal for reducing expenditures elsewhere. The Congress is simply able to raise the revenue estimate, even without a solid technical justification. Besides, without a medium-term expenditure framework, it is difficult to properly assess medium-term fiscal implications of such a decision, and the government is not in a position to negotiate on the basis of a reliable and credible analysis of the fiscal implications of their demands. Given the success of these groups in securing higher salaries and better working conditions, it is likely that other groups (e.g., nurses' assistants) and the same groups that have already won significant concessions will continue making their demands via passage or renewal of estatutos in coming years. In light of the budget system's inability to produce binding policy and budget decisions, tight central control of financial management becomes a necessary second-best alternative to control aggregate spending levels. This method of fiscal discipline, however, also carries a high cost in terms of shifting de-facto policy priorities during the year and reducing operational efficiency as explained in a later section. 3.33 Strategic allocation of budgetary resources: A typical symptom of weak institutional capacity to make binding decisions on strategic resource allocations is the significant in-year deviations between approved and executed budgets. Some degree of in-year adjustments are expected because of legitimate contingencies such as external shocks to the economy that result in lower-than-projected revenue collection. However, when the extent of in-year deviations at the agency level far surpasses the extent of adjustment in the aggregate, it is a clear indication that the budget is being re-allocated among agencies even in the absence of a macro-level adjustment. Figure 3.1 shows that while the deviation of the total budget has tended to be relatively limited (i.e., close to 100 percent execution ratio), several entities, including the Presidency, the Legislature, the electoral organ, and SOPTRAVI, have tended to overspend in their final execution vis-h-vis their original budgetary allocations. Figure 3.1: Credibility of the Honduran Budget - In-year Deviations by Agency (percent of the executed over the approved budget) 200 0- ~00 0 0 0 2 50 0- 2 0 0 0 0 5 04 o) 0 0 miso 1991 01992 011993 M1994 mi995 m1996 CM1997 1998 g1 3.34 The fiscal impact of overspending may not always be substantial (e.g., the Secretariat of the Presidency has consumed on average only one percent of the total budget during the 1990s). It is nonetheless a clear indication that the approved budget is not a binding document that guides the government's true resource allocation and that, to an extent, the government's policy intentions expressed in the budget do not reflect its true priorities. The cost of these practices is the lack of credibility of the Honduras Public Expenditure Review 39 approved budget as an authoritative mechanism of resource allocation, and thus policy implementation. Assuming this translates into lack of predictability in available funding for ministries, it also affects their ability to implement programs efficiently (operational efficiency). 3.35 Establishing budget credibility is no easy task. It requires a high-level government commitment to use the budget as a reliable and authoritative instrument for the government to express its policy priorities and allocate resources accordingly. However, there are various practices that systematically undermine the budget credibility, including inaccurate original budget estimates by the line ministries, and the tendency of external donors to offer assistance without regard to the integrity of the budget cycle.22 Further, even with the strongest of commitments, the government would need a range of policy and institutional instruments to turn the budget into a credible mechanism with limited in-year deviation. These are: (i) realistic sector strategies to establish relative priorities among sectoral expenditure programs; (ii) use of accurate information on budget execution of the previous year and program performance; and (iii) institutional features of budgeting (comprehensiveness, transparency, and contestability) to ensure all competing claims for limited resources are duly considered for funding and to enable re-allocation of resources once changes in priorities are agreed upon. 3.36 Sector strategies: No informed decision can be made about intra-and inter-sectoral prioritization unless there are macro and sectoral strategies of the government that guide such decisions, and there is accurate information on the costs of various policies and programs that compete for the limited room in the budget. This assumes adequate incentives and capacity to cost out the existing and proposed programs, and evaluate their effectiveness and efficiency as necessary input into making budgetary decisions. There must also be incentives for sectoral ministries to invest in developing a realistic sectoral strategy that will serve as a basis for formulating their sectoral budget plans, and a capacity at the center of the government (e.g., cabinet) to make binding decisions on necessary expenditure trade-offs. 3.37 In most sectors, Honduras has yet to develop adequate sectoral strategies on which to base the difficult decisions on intra- and inter-sectoral prioritization, though in some sectors efforts have been made to begin developing such a strategic vision for the whole sector. In the education sector, for example, the recent effort to develop a vision for the sector's "transformation" promises to provide a basis for such a sectoral strategy, but the vision as it is currently formulated still remains too general and too encompassing to guide decisions on intra-sectoral prioritization both in terms of policy goals (e.g., improvement of quality at the primary level vs. expansion of secondary coverage) and instruments (e.g., hiring more teachers vs. training existing teachers more effectively). In the near future, the Poverty Reduction Strategy Paper (PRSP) is expected to play this role. 3.38 Use of performance information: Despite the existence of POAs and some basic analysis by SEFIN's expenditure analysis unit, there is no evidence that SEFIN (or especially Congress) allocates resources on the basis of sectoral program performance and cost. SEFIN's expenditure analysis tends to focus on formal legal aspects (via dictamen) rather than cost-benefit analysis or even simple costing of the proposed programs. SEFIN's Expenditure Analysis Unit also performs a series of ongoing program reviews, comparing planned objectives and expenditures with outlays, and attempting to explain the differences, but such analyses do not seem to inform decisions on the budget allocations. In some cases, spending units receive less than programmed due to non-trivial discrepancies in actual revenue collection and the projection used as a basis for the financial programming. This can result in spending units over- committing first, and then correcting for reduced cash release from the SEFIN by running arrears. 22 Typically when the budget is presented to Congress SEFIN is aware of future deviations related to donor funded programs being negotiated. When these programs are financed with loans they cannot be included in the budget because of the need for Congressional approval to contract external debt. Honduras Public Expenditure Review 40 3.39 Within ministries, there is no evidence that inputs are related to outputs in the form of expenditure reallocations or transfers during the year. Although some performance information is generated by sectoral ministries such as health and education, mainly for the POAs and other sector- specific documents (e.g., Evaluacidn Anual by Health Regions), the budget is still developed on the basis of inputs. In measuring outputs, the government uses physical input norms such as the 35 student/teacher ratio instead of national financial norms, (e.g. cost/student). The same problem is reproduced at the lower level of a ministry's organizational hierarchy. Thus, a Departmental office of the Education Secretariat would complain that irrespective of their expressed needs (e.g., for facility repair, transport), annual allocations hardly vary from those of the previous years (i.e., incremental budgeting as discussed above). 3.40 Comprehensiveness: Limited coverage of the budget weakens the government's capacity to prioritize policies and expenditures. A sizable portion of public expenditures, including more than 80 percent of public investments, are carried out by decentralized entities whose budgets are presented individually to Congress without being integrated into the Central government budget that SEFIN coordinates. While budget fragmentation is neither unique to Honduras nor does it seem particularly egregious, it nonetheless limits the Executive's ability to adjust expenditures. One oft-mentioned example of a similar problem is the coordination problem between FHIS and sectoral ministries. While FHIS is officially not a decentralized entity with statutory autonomy, its programs are largely defined by external donors who provide financing and by needs expressed by communities. In spite of the FHIS management's claim that there is adequate coordination with sectoral ministries, it appears that there is further room for improvement. Even if the coordination between FHIS and the relevant ministries were effective, the fact that the FHIS budget does not appear in detail in the budget document makes it effectively an off-budget program that does not go through the same scrutiny as other budget items (FHIS' budget is included in servicios centralizados). Table 3.5: Revenue and Expenditure Shares within the Non-financial Public Sector (NFPS) (as percent of GDP) 1996 1997 Central Decen- Local NFPS Central Decen- Local NFPS _gov't tralized gov't gov't tralized gov't Income 16.6 15.4 2.4 34.4 16.9 14.8 2.6 34.3 Tax 14.6 0.6 15.2 14.1 0.5 14.6 Own 11.6 11.6 10.6 10.6 Transf from pub sec 0.5 1.4 1.2 3.2 0.5 2.1 1.5 4.0 Others 1.5 2.3 0.6 4.4 2.3 2.1 0.6 5.0 Expenditure 20.4 14.5 3.2 38.1 19.9 11.6 3.3 34.8 Current 16.0 9.6 1.2 26.8 16.4 8.6 1.2 26.2 Wages & salaries 6.4 2.3 0.5 9.2 6.3 2.2 0.5 9.1 Transfer to pub sec. 1.9 1.9 2.2 0.5 2.7 Other 7.7 7.3 0.6 15.7 7.9 5.9 0.7 14.4 Capital 4.4 4.9 2.0 11.3 3.4 3.1 2.1 8.6 Direct inv. 2.5 4.4 7.0 1.4 2.4 3.8 Transfer to pub sec. 1.8 1.8 2.0 2.0 Balance -3.8 0.9 -0.7 -3.7 -2.9 3.1 -0.7 -0.5 Source: Central Bank 3.41 In an aid dependent country like Honduras, it is often the case that many country priorities are effectively driven by external donors. We have not been able to compile reliable and comprehensive data on donor-funded public expenditures that lie outside the government budget. Despite indications that the magnitude of off-budget donor-funded resources is relatively small, it would be important to capture this information on a regular basis and consolidate it with the government budget data so that the aggregate impact of public spending can be assessed, taking into account budgetary allocations and allocations of Honduras Public Expenditure Review 41 extrabudgetary donor resources. It may be that after taking into account the extrabudgetary spending, the balance of total resource allocations among policy areas and programs may be different from the stated priorities of the government as reflected in the budgetary allocations. 3.42 Contestability: To the extent powerful interest groups can alter government expenditure priorities, the government will be forced to cut other expenditure items which may have been higher priorities. While interest group pressure is normal and expected in any democratic setting, it should also be kept in mind that the way in which this is permitted (i.e., institutional rules about allowing external interest groups to influence the budgetary process and outcomes) has implications for the quality of public expenditure management. Expenditure decisions made outside the context of a regular review and negotiations could bring about undesirable fiscal impact because such decisions are less likely to be based on sound analysis of fiscal costs and expected efficiency. The phenomenon of "officializing" secondary vocational schools would be an apt example of this sort of decision, whose cost effectiveness or simply fiscal affordability is suspect. Other decisions may imply significant medium-term commitment by the government to fund a program that may not be within the state's fiscal capacity to finance without seriously affecting other expenditure items. The best recent example is the decision to award the salary increases in the education and health sectors, when arguably the resources could have been spent more effectively on other purposes. Other examples of expenditure decisions that benefit relatively narrow segments of society and are difficult to reverse are the subsidies for urban transport and electricity. Many of these priority-altering decisions are likely to take place during budget execution. This is a common phenomenon in any country but more prevalent in countries with weak financial control. Finally, lack of transparency in any form would limit the contestability of budgeting because the non-transparent spending items cannot be subjected to the same degree of collective scrutiny to judge their relative merits vis-A-vis other competing spending claims. 3.43 Flexibility: Finally, the governments' ability to re-allocate budgetary resources according to policy priorities is further constrained by a high degree of structural rigidity in the budget due to legally or constitutionally mandated transfers, earmarking of revenues for specific government functions (or even off-budget expenditures) and other commitments (e.g., interest payments, personnel expenditures). In extreme cases such as Brazil, budgetary rigidity can reach 80 percent or more of the government budget.23 While budgetary rigidity in Honduras does not reach such a high level, given the overall resource scarcity against the country needs, it is still of some concern that at least 50 percent of the Central government budget is pre-committed to legally/constitutionally mandated transfers (e.g., UNAH, municipalities), a few subsidies that are politically difficult to reduce or terminate (i.e., urban transport, electricity), public sector wages, and interest payments.24 One source of budget rigidity, debt service, is expected to be reduced in coming years thanks to the HIPC Initiative. However, the expected increase in wages and salaries are likely to replace the reduced debt service and constrain the government's ability to re-allocate limited public resources to high priority programs. As shown in Chapter 1, the government would have to commit to allocate much larger shares of its budget to social spending in order to offset the large wage bill in the health and the education sectors, which means reduction in expenditure in other areas that are not directly relevant for poverty reduction. 23 The degrees of budgetary rigidity in Brazil were estimated to range between 76 percent and 83 percent of the Federal Government budget in 1995-1998. The breakdown of the sources of the rigidity is as follows: constitutional transfers (15-16 percent), personnel expenditures (26-32 percent), social security benefits (28-30 percent), and subsidies and other expenditures that are difficult to discontinue (5-7 percent). 24 The effect of the rigidity can be particularly perverse in some sectors such as education, in which as much as 15 percent of the sector budget has been allocated to the National Autonomous University in spite of the clear evidence of inequity (as the UNAH's enrollment is represented predominantly by students from relatively well-off segments of the society) and inefficiency and poor quality. Because of UNAH's statutory autonomy guaranteed by the Constitution, the government's ability to shift the scarce resources in the education sector toward areas of higher priorities such as the expansion of secondary coverage or the improvement of primary teacher training is severely limited. Honduras Public Expenditure Review 42 Figure 3.2: Rigidity in the Central Government Budget mT ran sport 100.0% subsidies WnElectricity 90 0% ,-------- --------- ------------ su bsidies =In st. N irez 80 o- - Farn ilia 70 0% - - - ---Inst. Nal. Agrario 60.0% -U. Pedagogica 50.0% mInterests 40.0% .WW ages 30.0% OD El (2.5%) 20.0% WlM unicipalities (4 10.0% UJudiciary (3%) 0.0% NUNAH (6%) 1995 1996 1997 1998 1999 2000 Source: SEFIN Box 3.3: SISPU - An Instrument for an Efficient Public Investment Process SEFIN is currently developing a system for managing the public investment process more efficiently. The Sistema de Inversiones del Sector Pblico (SISPU) will comprise a comprehensive database of both ongoing and planned investment projects in the entire public sector (including decentralized entities/public enterprises, which account for a large portion of public investments, but excluding municipalities, which execute relatively limited amount of the public sector capital spending), and will be linked to other government financial management systems (SIAFI, the evaluation system SINEG. and sector-specific management information systems such as the Health Secretariat's SIGADE). SISPU will include information on projected operations and maintenance costs of each investment project. With the introduction of SISPU, the government plans to have SEFIN's Public Investment Department screen all investment projects for their technical and financial viability. 3.44 Operational efficiency: Developing an efficient, results-oriented public sector is one of the most difficult institutional development objectives faced by developing countries. Very few have made significant inroads. Many are struggling to implement a long-term process of institutional development and cultural transformation in the public sector. In recent years, well-functioning bureaucracies in a relatively small number of developed countries have shown possible ways of enhancing operational efficiency with a strong focus on efficiency and results. Like many developing countries, Honduras' public sector modernization is moving in the right direction, but has a way to go in approximating best practice standards on operational efficiency. Effective implementation of SIAFI would be a key step in making financial management processes more transparent and efficient. 3.45 Centralization of financial management: The current modernization of the state effort in decentralization is limited mostly to devolution of reporting requirements to strengthen central controls further. Within ministries, particularly education and health, program results are dependent on payment of purchase/pay orders. Rules and procedures for their authorization are still centralized despite instant recognition of approval/rejection in SIAFI. For example, hard copies are still hand-carried to senior ministry officials for approval when the SIAFI system readily available throughout SEFIN shows the Honduras Public Expenditure Review 43 status of the pay order. Similarly, centralization of decision-making still impedes transfer and reprogramming of funds where needed to achieve program goals. Regional level health officials indicated that they lack funding and authority to shift funds for vehicle maintenance which damages program results. Ideally, they recognize the function of vehicle maintenance itself should be contracted- out, but they lack authority to make such a decision. 3.46 Performance information: Performance information is generated by some sectoral ministries, but not systematically used to guide resource allocation, or monitor efficiency and effectiveness in program implementation. Regional units program supplies (e.g. books) to schools on the basis of demand and enrollment figures, but output measures (e.g., the number of graduates/student enrolled or cost/graduate) are rarely used to program expenditures by ministries. Within ministries, there is no evidence that inputs are related to outputs in the form of expenditure reallocations or transfers during the year. At the central level, SEFIN does not estimate the volume of services required to meet health and educational needs in a comprehensive fashion. Despite the existence of POAs and some basic analysis by SEFIN's expenditure analysis unit, the budget is still developed on the basis of inputs. Local units of ministries, such as schools and clinics solicit resources based primarily on demand estimates. Furthermore, programs are not costed to allow ministries to measure efficiency in their implementation. Neither is there a mechanism to measure the cost of service delivery at the ministry level (i.e., running cost estimates). 3.47 Weak control of human resources: Honduras, like many developing countries, suffers from inadequate supply of qualified professionals to staff the public sector, but the problem is compounded by weak personnel management. Health and education services suffer from low quality of professional staff, while there is evidence that moonlighting and absenteeism further reduce quality and access. Recent salary increases have not been accompanied by corresponding improvements in performance management and accountability mechanisms. 3.48 Weak external audit capacity: The Controller General is nominally independent, but is badly resourced. It has never issued an audit report on the financial statement of the entire government expenditures, and instead has mostly focused on a limited number of audits of specific programs. Many of the audits, however, have yet to be made public. It is likely that a number of the reports are obsolete by the time they are finally published. Conclusions and Recommendations 3.49 With donor assistance, Honduras has been making steady progress in improving public expenditure management. The pilot implementation of SIAFI seems to be progressing reasonably well, though there is much more to be done. The analysis presented here also suggests additional priorities, some of which can and should be tackled in the short run, while others may take longer. 3.50 Budgetary performance needs to improve at all three levels. Priorities may be first to consolidate the capacity for aggregate fiscal discipline, while simultaneously introducing measures to improve operational efficiency both at the government-wide level and at specific sectoral level. The capacity for allocating resources according to the strategic priorities may be a more difficult challenge that can only be addressed over the medium to long term. Some OECD countries have pursued this by devolving greater autonomy for policy making and implementation to line agencies within an agreed fiscal envelope and a medium-term expenditure framework. Honduras, like most developing countries, still has some way before reaching this stage and it should therefore pursue a more gradual path toward improved strategic resource allocation. Continued efforts to refine the Poverty Reduction Strategy and to develop adequate sector strategies that would guide intra-sectoral prioritization would be a necessary prerequisite that should be vigorously pursued. Over time, the country should strengthen policy-making capacities both at the center (e.g., SEFIN) and line ministries that include adequate mechanisms for vetting the fiscal affordability of new policies. Among the immediately possible actions, the government should abolish Honduras Public Expenditure Review 44 asignaciones globales and servicios centralizados, as part of a broader effort to make the budget more transparent and comprehensive. 3.51 In order to ensure fiscal discipline, it is important to develop a more robust institutional mechanism to check policy proposals that may imply government commitment to increased spending, including: (i) developing a capacity to project at least the aggregate budget envelope for the next 3-4 years; (ii) developing a capacity to estimate (even roughly) the fiscal costs of policy/program proposals; (iii) introducing procedural requirements that any new policy/program proposal must clear SEFIN's fiscal impact analysis; and (iv) developing and implementing a coherent public sector employment and pay policy that ensures the affordability of the public sector wage bill over the medium term. 3.52 To improve operational efficiency, SIAFI implementation should be extended beyond the pilot ministries, while gradually relaxing some aspects of the remaining central control. One option would be to broaden the excessively detailed line item budget classification, and grant ministries limited flexibility to reallocate resources within broader bands of spending categories. Honduras does not yet have a sufficiently robust monitoring and reporting mechanism to ensure ministerial accountability for results. Gradual relaxation of central spending control should be coupled with strengthening of the capacity for establishment control both by SEFIN and line ministries. Especially in the social sectors, controlling staff numbers and their performance would be critical for both aggregate fiscal discipline and for more effective and efficient service delivery. Dramatic improvement in public sector efficiency cannot be achieved unless the civil service is professionalized and adequately compensated. This challenge is particularly difficult for a country such as Honduras with a limited supply of qualified staff and limited fiscal scope to offer attractive public sector pay. In cases where citizen and user voices can be utilized to regulate public sector operations (as in PROHECO), those options should be vigorously explored. 3.53 Possible sequencing of introduction of a medium-term expenditure framework (MTEF) is presented here for illustrative purposes. The adoption of an MTEF should be gradual, though involving some changes in decision-making early in the reform process. In this illustration, an emphasis is given to develop a medium-term framework for personnel expenditures, whose control is critical for fiscal sustainability, and the government's ability to fund other critical spending items for poverty reduction. Phase 1 * revise budget cycle to include budget framework, prepared by SEFIN, decided by government 3 months before finalizing annual budget, * budget framework comprises: (1) macro framework (detailed year 1, outline for outer years), (2) wage bill proposals (indicative for outer years if possible) (3) proposed sector expenditure ceilings for next annual budget, and * framework made available to legislature. Phase 2 * budget framework exercise repeated, beginning earlier in the cycle, process coordinated by SEFIN, but with significant input from sector working groups, * framework comprises: (1) three year macro framework, (2) medium term strategy on pay, staffing consistent with three year wage bill proposal, (3) sector expenditure ceilings over three years, linked to sector policy commitments, with particular emphasis on major resource shifts, and * budget framework published in some form prior to finalization of annual budget. Phase 3 * process as previously, strengthen capacity of sector working groups, opening of sector working groups to civil society consultation, Honduras Public Expenditure Review 45 * framework develops increasing sectoral depth, more explicit linkage between existing/new policy and sector ceilings, introduction of more program detail, gradual introduction of performance indicators, routine public reporting on performance against prior year framework targets, and * greater use of the published budget framework and forum to stimulate public debate, civil society actively involved in public forum. Honduras Public Expenditure Review CHAPTER 4 HEALTH Overview of the Sector 4.1 Health outcomes and Table 4.1: Health Outcomes 1970-95 their causes: In spite of the A LA M. country's continuing high level of Honduras incme, LACJ poverty and low rate of economic 1970 1985 1990 1995 1980 1997 growth, health outcomes in Outcomes Honduras have greatly improved Life expectancy at birth, yrs. 54 64 n.a. 70 n.a 69 over the last couple of decades Infant mortality, per 1,000 live births 110 54 45 36 60 32 (Table 4.1). The advances include Matemal mortality (per 10,000 births) n.a. n.a. 182 108 n.a 185/2 the significant decrease in the % of children entering I' grade who infant mortality rate, and a are chronically undernourished n.a. n.a. 39.4 38.5 na. n.a corresponding increase in life % of malnutrition in children under 5 n.a. n.a. n.a. 18 n.a. 15/2 expectancy at birth. These Causes improvements are not the result of Vaccination cover BCG, children<5 n.a. 85% 90% 94% n.a n.a a particularly high level of public Global fertility rate 7.5 5.6 5.2 4.9 4.1 2.7 spending, which averaged 2.8 % of pregnant women attending percent of GDP in 1990-97, only controls n.a. 65 73 84 n.a. n.a slightly above the regional average % of births taking place in hospitals 37 54 n.a. n.a of 2.6 percent for Latin America. % of pop. with access to safe water 43 n.a. 50 77 73 75 Rather, the advances in health Public health spending % of GDP n.a. n.a. n.a. 2.8 n.a. 2.6 results in Honduras are due to the Sources: World Bank, 1998 & 2000, SS 1999 and UNAT, 1999. Data are for the nearest available date in each case. relative effectiveness of the public Notes: 1. Source for comparators: ADR 2000. 2. Average for Central America, sector's primary health programs 1990-97, from the WDR 2000. in vaccination, control of diarrhea and respiratory infections, and an increased emphasis on integral mother and child health programs. 4.2 Principal services and types ofprovider4 The principal public agencies involved in health care in Honduras are the Health Secretariat (SS) and the Honduran Social Security Institute (IHSS).5 The SS operates a network of 28 hospitals and over 1,100 health centers with programs of extension and promotion, which has achieved a high level of coverage nationwide. The IHSS operates three hospitals, 20 clinics and four emergency clinics, providing coverage mainly in the principal urban centers of Tegucigalpa and San Pedro Sula. In addition, there are some 56 private hospitals operating in Honduras.2 4.3 In terms of market share, the SS accounts for 48 percent of ambulatory consultations, 67 percent of preventative consultations and 70 percent of hospital in-patient treatments. The 3.5 provides, respectively, seven percent, seven percent, and 16 percent. Facilities run by NGOs provide only five percent of ambulatory consultations. The commercial private sector provides 39 percent of ambulatory and 23 percent of preventive consultations. In urban areas, the private sector is the single biggest provider of ambulatory consultations (49 percent of the total), reflecting a high willingness to pay, at least among the relatively well-off, for the comfort and convenience of a shorter queue, for more flexible opening hours and longer consultation times. In contrast, the private sector's share of hospital inpatient treatment is noticeably low (14 percent of the total), due to the low levels of sophistication in their emergency 2% The present study does not cover the Water and Sanitation sector. 26 Other public agencies involved in the health sector include the National Autonomous University (UNAH), where the schools or medicine, dentistry and nursing are located; the Honduran Social Investment Fund (FIS), which undertakes small scale capital works including the building of health posts and occasionally, hospitals; and the Family Allowance Program (PRAF), which provides both demand-side and supply-side incentives to increase take up and improve performance in rural health posts. Honduras Public Expenditure Review 47 rooms and their facilities for specialized surgery and the treatment of chronic illness. This reflects the availability of highly subsidized alternatives in the SS and IHSS hospitals, coupled with the low coverage of private insurance. 4.4 The Expansion of the SS Table 4.2: Expansion of the SS Network, 1990-99 network: During 1990-99, there were 1990 1999 Growth increases in the number of hospitals (47 National hospitals 6 6 0% percent), urban health posts (36 percent) Regional hospitals 6 6 0% and rural health posts (68 percent). Also Area hospitals 7 16 129% 15 new mother and child clinics were Sub total, hospitals 19 28 47% added to the network. The network Urban health post (CESAMO) 177 241 36% expansion has led to a 22 percent increase Rural health post (CESAR) 516 867 68% in the number of consultations per-capita Mother and child clinics 0 15 n.d. during the decade. There has also been a CLIPERs (Peripherical clinics) 0 3 n.d.. shift of ambulatory consultations away Sub total, ambulatory facilities 693 1,126 62% from hospital outpatient clinics (down 24 Total 712 1,154 62% percent) toward lower-cost ambulatory 323 244 -24% facilities (up 28 percent), which suggests Aobulta ct 693 888 28% an improvement in network rationality. Abl toalte 96 1,13 22% Health Expenditure in Honduras 4.5 National health accounts: According to a National Health Accounts study,27 total health spending in Honduras in 1998, including public and spending of all sorts, was the equivalent of 6.2 percent of GDP.2 In terms of service producer, 65 percent of the total was managed by public entities and 35 percent by private clinics, hospitals and pharmacies. The SS is the most important agency, itself accounting for 39 percent of all the health spending and a further three percent through FHIS for capital works on behalf of the S. The IHSS accounted for just six percent of health spending in 1998. Services provided by private sector doctors and clinics accounted for 19 percent of the national total and pharmacies, for 15 percent. The low share of private agencies reflects the low use of private hospitals for inpatient treatment, reported above, which is suggestive of crowding out of private hospital treatment by subsidized SS services. 4.6 On the financing side, the government financed 33 percent of all health spending with national funds and a further 20 percent was funded with external resources from international public agencies (multilateral and bilateral). However, private voluntary spending by households and firms (apart from taxes paid to the government) is a very important component of the total (47 percent). This includes households' and firms' contributions to the IHSS, co-payments by households to the SS for use of the Ministry facilities and payments by households to SANAA for water services. Within the total of private finance for health, private households paid for 43 percent and firms for four percent. This underlines the importance for sector policy of the mobilization of private sector resources for health through measures 27 Health accounts look at spending from two points of view: (i) how services are divided between the different service producers; and (ii) how funding is divided between the diverse financial sources and financial agencies that pay for the services. A 1995 World Bank study estimated the total at 7.2 percent of GDP. The main differences between these two estimates (apart from the different years) are: (i) the 1995 data included PRAF but excluded SANAA; (ii) the 1995 estimate for private consumption of medicines was higher than that in 1998 by a factor of three. The difference in the two estimates for medicine spending is to 1.7 percent of GDP. The 1998 estimate is based on the Central Bank income and expenditure survey, while the earlier estimate was imputed from national accounts data. Honduras Public Expenditure Review 48 such as means tested co-payments for the use of some of the SS facilities (especially at the tertiary level), and the strengthening of health insurance for formal sector employees. 4.7 Trends in public spending. Public Table 4.3: Public Health Spending by Agency 1990-99 spending in the health sector declined over the 1990s in real terms. In particular, the SS 1990 1991 1992 1993 1994 1995 1996 1997 1998 budget saw a steady decline in the second SS 100 82 87 86 92 103 83 69 69 half of the 1990s until 1999, when the FHIS n.a. 100 158 87 100 71 113 106 103 PRAF n.a. 100 67 333 33 200 233 133 67 increase in the sector's wage expenditures PAnS 100 6 158 87 100 7 13 1061 reversed the declining trend. In terms of Totl 100 79 113 98 98 102 102 87 83 allocations by program, 42 percent of SS resources were spent on hospital care, 29 percent on ambulatory consultation and control of transmissible disease and 13 percent on capital transfers to SANAA in 1999 (Table 4.4). The shares of hospitals and of administrative spending have increased sharply since 1995. This is the result of wage increases for doctors. Most doctors work in the hospitals, and most SS administrators are doctors. Table 4.4: Structure of SS Spending by Program, 1999, % - 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Environmental health & promotion 5.7 5.6 6.7 5.7 5.1 4.1 5.2 5.5 5.3 4.6 Ambulatory consultation and control of transmissible disease 18.9 26.7 23.7 23.8 24.5 22.4 27.0 32.3 31.8 29.4 Hospitals 40.1 34.0 33.0 34.7 34.4 28.5 36.4 36.8 41.5 42.0 Nutrition programs 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 Central administration 3.0 3.1 7.4 6.4 4.3 2.7 4.5 5.8 4.1 n.a. Normative services 2.1 2.6 2.6 2.1 2.1 1.5 1.9 2.6 2.9 n.a. Current transfers 1.3 1.3 2.0 3.1 2.1 2.2 2.6 3.5 4.4 2.9 Studies and construction of buildings 5.1 5.3 6.3 3.2 10.5 10.3 7.7 4.5 1.8 4.6 Equipment and other capital purchases 0.0 0.0 0.6 0.5 0.3 0.6 1.6 0.0 0.0 0.0 Capital transfers (SANAA) 23.6 21.3 17.6 20.4 16.4 27.7 13.2 9.0 8.2 13.2 Central activities /1 3.4 Total 100 100 100 100 100 100 100 100 100 100 Source: Executed budget, 1999. Note: 1/ Central activities in 1999 includes administration and normative services and which have not been unbundled in the 1999 statement of executed spending. 4.8 During the inflationary 1990s, the salary component of SS spending suffered large year-on-year changes. The share of personnel spending fell from 40 percent in 1990 to 31 percent in 1995, as nominal wages were held back in the face of inflation, and then shot back up to 50 percent in 1998 when the Congress legislated to increase doctors' wages. In 1999 the salary bill was 47 percent of the total (Table 4.5).29 The compensating adjustments were in the shares of construction and capital transfers (to SANAA). The purchase of materials and supplies (including medicines) remained relatively steady during the decade, and stood at 25 percent of the total in 1999. Over the decade, SS salary expenditure rose 32 percent in real terms and materials and supplies rose by 12 percent, while capital transfers were reduced by 40 percent. 29 SS's budget does not include payments for the 13'b and 14 month bonus given to all public employees, which are treated as a central item in the Finance Ministry's budget. They are worth 16 percent on top of ordinary salaries. Taking account of this, in 1999, SS salary spending would be 51 percent of its budget, and not 47 percent as reported in Table 4.5. Honduras Public Expenditure Review 49 Table 4.5: Structure of SS Budget by Expenditure Group, % 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Personal services 40.1 40.6 39.7 41.6 40.6 30.9 38.6 45.4 50.0 47.0 Other services 3.2 3.6 7.3 6.2 4.9 3.7 5.3 6.7 4.3 5.2 Materials and supplies 25.1 26.2 25.9 24.4 28.2 24.0 30.5 31.0 31.1 24.9 Machinery and equipment 0.3 0.1 0.9 0.5 0.9 0.6 1.6 0.5 0.4 1.8 Construction 5.2 6.0 5.5 3.0 3.3 10.1 7.6 3.7 0.7 4.6 Real estate purchase 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 Current transfers 2.0 2.1 2.2 3.1 5.1 2.2 2.6 3.5 4.5 2.9 Capital transfers 23.6 21.3 17.6 20.4 16.4 27.7 13.2 9.0 8.2 13.2 Global assignment 0.5 0.0 1.0 0.9 0.7 0.9 0.7 0.0 0.8 0.5 Total 100 100 100 100 100 100 100 100 100 100 Equity and Access 4.9 The analysis presented here on equity and access to health services shows that the expansion of SS primary programs has resulted in a relatively high proportion of resources reaching the poor. However, it also shows that there remain important differences in effective access to health services between the poor and non-poor, which may be due to the low quality of most of the public health services that are used by the poor. 4.10 How many SS resources reach the Table 4.6: Suppliers of Ambulatory Attention by poor?: The expansion of SS primary services has Income Quintile led to significant benefits for poor households. 1 2 3 4 5 The SS is the main supplier of ambulatory Supplirs: Percentage attended by type of'supplier services for poor households, attending to almost SS 68 69 50 41 18 70 percent in the bottom two income quintiles, IHSS :2 1 8 8 12 compared with only 18 percent in the top quintile Private 30 29 42 51 70 (Table 4.6). Furthermore, the distribution of SS - Commercial 22 24 36 45 63 attentions is remarkably even across income - NGO 8 5 5 4 5 levels, partly because the better-off segments of - Traditional 0 1 2 2 2 the population have the option of paying for Source: NHES 1995 better-quality, private medical services, which the poor may not be able to afford. In contrast, the IHSS is concentrated in the top two quintiles (the urban formal sector). Commercial private suppliers concentrate over 60 percent of their ambulatory attentions in the top quintile. 4.11 Table 4.7 shows how SS Table 4.7: Distribution of SS Attentions, by Population Quintile attentions of different sorts are divided Ambulatory and between population income quintiles. Quin All emergency consultation Hospitalization Twenty-two percent of all attentions -tile attentions Cesar Other are received by the poorest 20 percent Cesanio Hospital Total Maternity causeToa of the population, and 49 percent of all 1 22.2 21.3 24.1 22.4 16.6 17.1 16.9 attentions are received by the poorest 2 27.4 33.6 19.0 27.8 13.2 19.2 17.1 40 percent. The richest 40 percent 3 21.8 22.0 21.3 21.7 24.4 23.9 24.1 receives only 28 percent and the top 4 19.1 15.3 24.6 19.0 28.6 18.4 22.0 20 percent receives just nine percent 5 9.2 7.4 10.9 8.8 17.1 20.9 19.6 of the total. The bottom two quintiles Source: NNES 1995 consume 55 percent of ambulatory consultations in primary facilities (CESARs and CESAMOs) and 43 percent of those given in hospital facilities. In each case, the top quintile has a very low share (seven and 11 percent, respectively). Honduras Public Expenditure Review 50 4.12 For hospitalizations the picture is somewhat different, with a lower share for the poorest 40 percent and a correspondingly higher share for the third and fourth quintiles. But even here, the distribution of services is remarkably even, with the bottom two quintiles each Figure 4,1: Correlation Between Health Spending and per getting 17 percent of total capita Income at Departmental Level hospitalizations and the top quintile Frnrr7An) getting only its pro-rata share (20 percent).o There is considerable variation in per-capita spending at departmental 14 level, ranging from L.315 in Francisco Morazin (where the national hospitals are located) to L.31 in Santa Barbara. I However, the differences between F departments are not strongly correlated 60 with income. The correlation between per-capita income and per-capita health 20 spending has an r2 of only 0.25 and when 1 Francisco MorazAn is removed (to omit the effect of the concentration of national hospitals in that department) the rv falls to 0. 10. These data do not suggest that resources are being systematically channeled to the richer areas of the country. 4.13 Access: signfiicant differences remain: Evidence presented above shows that the Honduran public health system assigns a large proportion of its resources to the poor and that the most important basic services (such as vaccinations) now reach virtually all of the population. A recent evaluation of the FMIS program showed that the construction of new health posts has led to a significant increase in effective uptake of health services by the local populations (ESA Consultores 1999). Nevertheless, NHES data show that there are still significant differences in effective access to clinical health services by geographical location and income group. For example, while 53 percent of those with a health problem living in metropolitan areas seek medical help, only 34 percent in rural areas do so. And while 22 percent in the bottom quintile seek help, 53 percent of those in the top quintile do so. 4.14 The main factor affecting access is problems leaving home or work, which might reflect inflexible opening hours in the health center (mentioned by 13 percent of those who did not seek help). Four percent mentioned transport problems and nine percent said the health center was closed. As a result, the proportion of the bottom quintile that receives ambulatory attention each month is seven percent, compared with 16 percent in the top quintile. Therefore, improving the effective access of the rural poor to basic health services should be one of the priorities for the sector, and should go hand in hand with measures to upgrade service quality. 30 Estimates of subsidy distribution that take account of the unit costs of different types of service also show a positive pattenm of resource distribution. in 1995, the bottom quintile received an estimated 21 percent of SS resources and the second quintile, 25 percent. The top quintile received 12 percent (World Bank, 1998). However, this was due in part to a temporary and unsustainable repression of doctors' wages in the inflationary period of the early 1990s. Since 1995, large increases in doctors' pay have increased the cost of hospital services, worsening the distributive pattern. Based on data from the 1995 NHES, it is estimated that the doctors' wage rises will increase the cost of hospital outpatient consultations relative to the cost of a consultation in a CESAR by 122 percent, that of hospital emergency consultations by 83 percent, hospitalized naterities by 54 percent and other hospitalization by 64 percent. Honduras Public Expenditure Review 51 Table 4.8: Incidence of Health Problems and the Decision to Seek Attention Income quintile Location Sex (1= poorest) Total Metro Other Urb. Rural M F 1 2 3 4 5 % of popln. With a problem last month 32 31 37 30 29 34 32 34 34 30 31 % with problem that sought attention 37 53 32 34 36 38 22 35 36 48 53 % of popln that sought attention 12 16 12 10 11 13 7 12 12 14 16 % of population that got attention 12 16 12 10 11 12 7 12 12 13 16 Reasons why people did not seek attention when they had a health problem Not necessary (self perceived) 86 86 94 82 89 84 75 88 87 86 94 Problem to leave home or work 9 12 3 11 7 10 13 8 9 11 4 Lack of money 0 0 0 0 0 0 0 0 0 0 1 Transport problem 1 0 0 2 2 1 4 1 1 0 1 Problem in the health facility/I 3 2 2 5 3 4 9 3 3 3 1 Total 100 100 100 100 100 100 100 100 100 100 100 Source: NHES, 1995. Note: 1/ Includes absence of the nurse of doctor, temporary closure etc. 4.15 Effective access to hospital services is also Table 4.9: Incidence of Hospitalization unevenly distributed mainly due to the lower incidence of (% per person/year) hospitalized births in the rural areas (Table 4.9). In 1995, overall, 4.9 percent of the population spent some time in Qu ntile Total Urban Rural hospital. But for rural dwellers and those in the bottom 1 3.5 6.1 3.0 quintile, the average was 3.5 percent compared with 6.6 2 4.2 6.4 3.3 percent for the urban population and for those in the top 5.5 6.4 4.3 Quintile. Toa Urba Rural. 5 6.6 7.6 3.3 Efficiency and Service Quality Total 4.9 6.6 3.5 Source: NHES 1995 4.16 Following the rapid expansion of network coverage and improved access over the last few years, the SS now faces the challenge of how to ensure efficiency, effectiveness and quality in its services. One indicator of efficiency is the number of attentions given per nurse-day and per physician-day in different types of facility. 4.17 Health centers: According to SS statistics, in 1999 the average productivity of rural health centers (CESARs) (including vaccinations plus ambulatory consultations) was just 10 consultations per post/day (compared with 9.4 in 1997) and 8.3 per nurse/day (Table 4.10). Urban posts (CESAMOS) registered a total of 37.3 consultations per post/day (compared with 34.4 in 1997) and 29 patients per doctor/day." Taking account of all clinical staff members in urban health posts (on average 5.8 per post, not including vector monitors and lab technicians), the average number of consultations per clinical employee day was only 6.4. 3 The statutory benchmark for doctors' productivity is 36 patients per day. Honduras Public Expenditure Review 52 4.18 In some cases, low utilization results Table 4.10: Staffing and Productivity in Health Posts inevitably from the need to provide services in CESAMO CESAR CMI areas of low population density and difficult Staff numbers, 1998 / physical access. Under-utilization sometimes Auxiliary nurses 2.8 1.2 6.2 results from inappropriate health center General Physician 1.3 0 0.6 locations, too near to hospitals which users Specialist 0.07 0 0 regard as offering better services. Another Professional nurse 0.5 0.02 0.6 problem is staff shortages, due to the difficulty Dentist 0.5 0.02 0 in retaining personnel in remote areas. On Vector disease monitor 0.9 0.07 0 average CESARs have 1.2 nurses, compared Lab technician 0.5 0 0 aeaesTotal 7.2 2.3 7.4 with the norm of 2. This leads to temporary Consultations per day, 1999/2 closures during vacations and due to staff Per health post training and sick leave. Nationwide, the Per clinical staff member 3.3 10 n.d. average CESAR was closed for 0.7 months in Sources: 1. USAID Andlisis Situacional de laSecretaria de Salud. 1999. A small survey of 18 rural health posts, Based on a sample of 40 CESARs and 41 CESAMOS and 5GMI in undertaken for the 1998 FIIIS evaluation, found Regions 1,2 and 3, 1998. 2. Dept of Statistics, SS that on average they had been closed for 1.75 days in the previous week. 4.19 Both rural and urban health centers suffer Table 4.11: Average Number of Months of Closure from the low quality of medical staff (e.g., auxiliary per Year in Rural Health Posts nurses with not much more than primary education Health region 1993 1996 1999 and inexperienced doctors who are completing their 0- Metro 0 0.1 0.3 compulsory social services), inadequate supply of I- Central 0.1 0.4 0.5 medicines and equipment due to budget constraints 2- Comayagua 0.4 1.1 nd 3- San Pedro Sula 0.8 1.3 nd and the inefficiency in SS's pharmaceutical 4- Sur- Choluteca 0.6 0.7 0.3 management, and short opening hours due to - Occidente 0.2 0.9 1.3 shortage of staff and lack of effective control over 6- AtlAntida 0.7 2.7 0.9 their working hours. On average, the CESARs 7- Olanc~ho 0.2 0.9 0.2 8- Gracias a Dios 0.3 n.d 2.6 Per 3 iTotal 0.4 1.1 0.7 percent of the list of 38 basic medicines and 55 Source: Dept. of Statistics, SS percent of the list of 17 basic items of equipment they were supposed to have available (ESA Consultores, 1999). Most health centers are open only in the mornings, and in order to be seen, it is often necessary to arrive before 7:00 AM and then wait for several hours before the nurse or doctor is available. According to the NBES, the average waiting time in SS clinics is two hours. A recent survey of 484 users of reproductive health facilities in the Metropolitan Region found that 84 percent of patients were in the clinic over two hours, but that 77 percent had consultations of less than 15 minutes. In response to this sort of problem, the SS together with the IDB have developed a program for the institutional reorganization and expansion of primary health services (Programa de Reorganizaci6n Institucional y Extensi6n de Servicios de Salud, PRIESS). This program provides for the use of NGO and private agencies in order to provide a basic package of health services under contract to the S. 4.20 Hspitals: In the hospital sector, there are also causes for concern on grounds of inefficiency and underutilization of existing capacity. During the 1 990s, the number of discharges rose by 3.3 percent a year, compared with a rise of two percent a year in the number of beds, and the bed turnover rate improved (Table 4.12), but the occupancy rate deteriorated. Like the health centers, most SS hospitals are fully staffed only in the mornings, as specialist physicians normally work in their private clinics in the afternoon, leaving junior doctors (internistas) under study in charge of hospitals. This leads to severe waste. In the absence of qualified specialist physicians, operating surgeries for elective surgery and the non-emergency outpatient clinics are closed after 1:00 PM. Underutilization of surgery facilities in turn lowers bed occupation rates in the corresponding wards. As discussed in more detail in Chapter 2, improving control of staff deployment and performance should be among the sector's highest priorities. Honduras Public Expenditure Review 53 Table 4.12: SS Hospital Productivity, 1991-1998 1991 1992 1993 1994 1995 1996 1997 1998 Beds 3,703 3,871 3,952 3,819 4,119 4,165 4,132 4,247 Discharges 172,775 172,451 167,489 186,749 194,582 194,985 205,375 218,199 Patient days 1,042,062 961,031 951,906 976,334 1,023,308 1,024,888 1,023,277 1,028,823 Bed days available 1,318,213 1,318,257 1,300,846 1,373,274 1,469,737 1,473,877 1,496,392 1,449,108 Average stay 6 5.6 5.7 5.23 5.26 5.26 4.98 4.73 Occupation rate 79.1 72.9 73.2 71.1 69.63 69.54 68.4 71.0 Source: SS, 1999 4.21 Co-payments and efficiency: Co-payments by users can be an important source of additional funding for public health services. When budgetary systems are rigid, making procurement a cumbersome process, co-payments can provide invaluable source of flexibility to local administrators. They can also act as a barrier to prevent diversion of public resources to cover high-cost interventions in favor of private individuals who could pay for them. And they can send valuable signals to users about which facility they should go to in the first instance. However, although the SS has been operating co-payments for over 20 years there is still confusion about the goals of the system and the official information about co- payment revenue is unreliable. 4.22 There are two systems authorized within the SS: the institutional system and the community system. The community system allows free use of the funds at local level, relying on community-based scrutiny to assure probity. This is the prevalent arrangement in CESARs. The institutional system requires the income to be reported to the corresponding health region. Hospitals are supposed to pass 10 percent of their co-payment income to the region, but few do so. CESAMOs operating on the institutional system are allowed free use of a proportion of their co-payment income, but most of it must be passed on to the Region. 4.23 According to preliminary results from a new study of the co-payment system underway during 2000, a quarter of UPS operate the institutional system, 42 percent operate the community system (including 55 percent of CESARs), 10 percent have some sort of hybrid between the two, and 23 percent of UPS have no co-payment system in operation. The 1998 National Health Accounts (based on the Health Module of the 1998-99 Household Income and Expenditure Survey) estimate that household spending on co-payments in SS facilities is L. 103 million, equal to six percent of the SS budget in that year. 4.24 The available evidence suggests that co-payments are not at present doing much to improve the rationality of user behavior or to protect the SS against the incursion of wealthy users to take advantage of extensive tertiary treatments. Data on unit prices gathered in the 1995 NHES showed that on average, Area Hospitals charged four times more for the same service as that charged by National Hospitals; and Regional hospitals charged two and a half time as much. These findings suggest that the existing prices might have a perverse effect on the pattern of demand, giving users an incentive to go to the National Hospitals instead of the lower level facilities. The same study reported that the user charge in the San Felipe Hospital for a course of radiotherapy was 10 percent of that in private clinics and endoscopic tests cost less than two percent of the private price. Paradoxically, the fees charged are high enough to keep out most poor users but are not nearly high enough to recover the cost of the treatment from the non- poor.32 32 World Bank, Honduras: Improving Access, Efficiency, and Quality of Care in the Health Sector. 1997 Honduras Public Expenditure Review 54 4.25 There is a clear need to establish means testing (including for coverage by the IHSS or private insurers) before allowing such huge subsidies for sophisticated tertiary treatments for chronic illness. Otherwise, as the sophistication of treatments available in SS hospitals grows, subsidies to the non-poor will progressively erode the availability of financial resources for primary programs. The following principles are suggested as a point of departure for a rationalization of the SS' co-payments system. 1. At all levels of the SS, the co-payment system should be equitable and transparent. The "community system" should be governed by clear, nationally agreed rules about what should be paid by whom for which services, and allowable uses of the resulting income, and clear records should be kept about all payments and the use of the funds. 2. In primary health, the main purpose of the co-payment system should continue to be the mobilization of additional funding to complement nationally assigned resources and permit greater flexibility in the operation of clinics. A secondary function should be to send signals to users about which facility to use, so that higher payments should be requested if users seek attention at the wrong level of the system. 3. For mother and child health, where the goal is 100 percent coverage, care should be taken to avoid imposing charges that might reduce the take-up of services. Charges should therefore be low and voluntary. No one who says they are unable to pay should be turned away. 4. For other types of primary clinical attention, the SS should decide its co-payment strategy in the light of its medium term goal for the development of the service. At present, the range of services offered is limited and their quality is generally very low. This has had the effect of restricting use of such services to those who have no alternative (the poor). As the quality of SS services improves, there is a medium-term risk that use by the non-poor will increase, leading to the crowding out of private alternatives and generating an unsustainable increase in the fiscal cost of these programs. There are two possible ways that this could be dealt with. One option is to establish cost-based (unsubsidized) charges for the non-poor and use means testing to limit access to subsidized services. However, generalized means-testing is likely to impose significant burden on the SS's limited administrative capacity, and may thus be difficult to administer at a local level. An alternative is to limit the scope (but not the quality) of primary services available through SS clinics, so that the non-poor will continue to seek privately financed services in the private sector." 4.26 Whatever strategy is adopted for co-payments in primary clinical services, immediate action should be taken to establish rigorous means testing for subsidized access to high cost hospital treatments. Without it, the cost of such treatments will quickly begin to squeeze the available funding for primary programs. At present, for example, the charges levied by SS hospitals for cancer treatments are around five percent of the private sector price. This "crowds out" private provision, but, perversely, still results in charges that are high enough to prevent most poor people from getting access to such services within the SS. The normal price for such treatments should be set at their full cost to the SS, and subsidized access should be allowed only following strict means testing, designed to prevent those with an ability to pay (including either private or social security insurance) from getting unneeded subsidy. The funds for such subsidies should be explicitly budgeted. " Ibid., pp. 31-34. Honduras Public Expenditure Review 55 Box 4.1: Means-testing in Chile: the ficha CAS In designing publicly subsidized social service programs for the poor, there is a danger that the non-poor may also use the service and crowd out the neediest citizens for whom the program %%as intended. Targeting, or means-testing, can be used to verify users' incomes and thus exclude those w%ho in fact are able to pay for the service. But means-testing can pose a considerble ch.illenge to a government with weak administrative capacity. The Chilean go%ernment's experience with the ficha CAS (now called CAS II) provides an example of a generalized means-testing instrument that reduces the administrative burden on public agencies, while still allowing the goernment to target subsidized programs to the poorest citizens. The ficha CAS, introduced in 1973, is a two-page form designed to determine households' eligibility for a range of subsidized social programs, including low-income housing, utility subsidies, childeare, and monetary transfers. The form is completed every three years, and points are awarded to each household based on the answers to questions regarding household size, income, housing conditions and certain assets held by the household. The subsidized programs target the poorest 20 percent of the population. Around 30 percent of Chilean households are interviewed by CAS social workers. There are two main advantages to theficha CAS. First, by using the same means test to determine eligibility for a variety of programs, administrative costs are reduced. The cost of a CAS interview is about US$8.65 per household. In 1996, administrative costs accounted for only 1.2 percent of the benefits distributed using the CAS system. Second, a quantitative evaluation of theficha CAS programs finds that all of them have a high redistributive impact, thanks to their effective targeting of the poor and very poor. Source: "The Targeting of Government Programs in Chile: A Quantitative and Qualitative Assessment" (draft) by Carinne Clert and Quentin Wodon, November, 2000. The Crisis in the IHSS 4.27 Honduras' social security system is one of the least developed in Latin America, even when compared with the poorest countries. The IHSS' coverage of is concentrated in the main cities, with 52 percent in Tegucigalpa, 35 percent in San Pedro Sula and 13 percent in other areas (Table 4.13). Despite the ISS' claim that it covers almost 25 percent of the national population, the 1995 NHES found that in reality only some 10 percent of the population is effectively covered by the IHSS. According to the NHES the lion's share of IHSS resources (68 percent) is received by the top two quintiles and only 13 percent goes to the poorest 40 percent of the population (Table 4.14). The IHSS is even more concentrated on the top two quintiles than the private sector. 4.28 The IHSS health programs are mainly financed by beneficiaries' contributions to the health and maternity insurance program. The annual contribution per worker to the IHSS is US$51 a year. Of the total, 33 percent is nominally for long term (pension) insurance and 67 percent is for health insurance. Employers pay 67 percent and workers, 33 percent. The contribution is based on 10.5 percent of earnings up to a L.600 a month, a ceiling that has been frozen for 40 years. When it was set it was worth US$300 but today it is worth only US$41 a month and is about a third of the minimum wage. As a result, everyone now pays the same, regardless of what they earn. The only significant source of potential subsidy to the IHSS health program has been transfers from the pension fund. These are not easy to quantify exactly, but Honduras Public Expenditure Review 56 are not apparently large. In 1995 they were estimated at 10 percent of the total cost.34 For more recent years, reliable data are not available.s 4.29 The social security system is Table 4.13: Insured Population of IHSS facing a profound financial and political crisis. Political leaders use 1995 1996 1997 1998 1999 % the IHSS for blatant patronage Type of beneficiary appointments. Employees' Contributors 518,389 535,496 553,167 566,996 581,171 41 conditions under the collective labor Spouses 187,562 193,752 200,146 205,149 210,278 15 contract are superior to what is Children <5 567,964 586,707 606,068 621,221 636,751 45 stipulated by the Estatuto del Area Mjdico Empleado, and IHSS Tegucigalpa 661,824 683,664 706,225 723,880 741,977 52 hospitals are generally over staffed San Pedro Sula 446,817 461,563 47,795 488,715 500.933 35 and inefficiently run. One recent Otras Areas 165,274 170,728 176,361 180,771 185,290 13 study concluded that IHSS costs per Total 1,273,915 1,315,955 1,359,381 1,393,366 1,428,200 100 patient day are almost four times % of pop. 24.0 24.1 24.2 24.2 24.2 greater than the SST36 Source: Dept of Statistics, IHSS 4.30 The inefficiency of the IHSS Table 4.14: Health Expenditure by Supplier ad Quintile has in turn led to strong resistance to% increases in contributions. The IHSS 6ile: 1 2 3 4 5 Average TO 4 Board, controlled by private sector Structure by quintile employers and unions, has refused to SS 72 58 47 29 22 41 raise the contributions ceiling.l7 Faced 4 1 3 4 5 8 5 with limited growth in contributions Private 27 39 49 66 70 54 income and steady erosion by Total 100 100 100 100 100 100 inflation of the real contribution per Structure by supplier beneficiary, the IHSS health system SS 22 24 24 17 14 100 has adjusted to rising costs through IHSS 2 11 20 23 45 100 deteriorating service quality, leading private 6 12 19 29 34 100 to growing displacement of insured Total 12 17 21 23 26 100 workers towards both SS and private 1/Excludes self-medication. sector facilities. Source: World Bank (1998), based on NHES 1995 4.31 The key step that needs to be taken to rescue the social security system is the complete separation of the three functions of retirement insurance, health insurance, and health service provision. An an's- length relationship based on contractual commitments should be established between the health insurer and health service suppliers. This would allow the IHSS to procure services from the SS and from private i World Bank, Honduras: Improving Access, Efficiency and Quality of Care in the Health Sector. 1997 e The accounts of the IHSS are not separated in a manner that allows precise analysis of the balance between the health fund and the pension fund. However, in general, the direct costs of the health program are covered by the 67 percent of contribution income which is slated for that purpose, while the contribution income for pensions plus the interest on the invested fund cover IHSS administration costs and pension payments. Although detailed actuarial studies are not available, most commentators agree that the pension fund is probably under-funded, as the result of some of its income being creamed off to pay for central administration in the teSS (an indirect subsidy to the health fund). However the main problem is not so much actuarial imbalance as the simple fact that the low contributions ceiling translates directly into low pensions rights, relative to earnings, as a result of which IHSS pensions have become almost worthless. 36 World Bank, Honduras: Improving A ccess, Efficiency, and Quality of Care in the Health Sector. 1997 37 The Colegio Midico is represented on the Board and generally supports the interests of its members who are employed in the IHSS, but does not have enough votes to override the private sector and union representatives who regard the payments to the IHSS as a tax and are reluctant to see it rise. Honduras Public Expenditure Review 57 sector suppliers, as well as from the existing IHSS hospitals, which would be placed under a separate administration from the health insurance fund. The latter would then be able to choose its suppliers on a competitive basis, in order to get the best value-for money for the insured population. This should eventually lead to a rationalization of the hospital sector, where at present there is wasteful overlap between the IHSS and SS especially in the major cities. 4.32 Mandatory contributions for health insurance should be increased to a level sufficient to cover the real cost of efficiently providing the basic insurance package. To this end, the health insurance system should be regulated by the Comisi6n Nacional de Banca y Seguros. The regulatory arrangements should include the determination of a fair contribution rate based on efficient insurance management and good value for money in the procurement of services. This would replace the existing system, where the decision is in the hands of a Board constituted of interested parties whose interests are in conflict with one another, leading to paralysis. Finally, the health insurance rules should allow the insured population to "top up" their basic social security insurance package with voluntary additional contributions that increase benefits. Conclusion: Institutional Constraints to Improved Sector Performance and the Agenda for Reform 4.33 In recent years the main emphasis in health has been on expanding coverage in primary services. Although there are still important differences in access to clinical services (especially among the rural poor), which would require more refined targeting of public services, the vast majority of the population is now covered by basic services and the results in the improving health status of the population are clear. The sector now needs to turn its attention to the institutional reforms needed to embed quality and value- for-money into the public health system and to promote the mobilization of private resources for non- primary treatments. This chapter has identified a series of institutional constraints that need to be tackled in that context. The following paragraphs summarize the most important of these constraints and indicate what measures are being taken. 4.34 Sector planning, budgeting and regulation: Although the Honduran Health Code (Cddigo de Salud) assigns it wide-ranging functions in the areas of sector planning, budgeting and regulation, the SS has in practice developed primarily as a service provider, neglecting its functions as planner and regulator. The SS is working to strengthen its activities in each of these areas. To provide a basis for more effective planning, it is working to improve its information about how resources for health are distributed in Honduras, through the development of National Health Accounts. This will help to underpin a more rational assignation of public resources. To translate plans into concrete actions, an effective budgeting system is needed. Budgeting in the past has been highly centralized and inefficient, while information systems have been fragmented and separated from the decision making process. The process of reform involves the development of modem management information systems, the linking of financial and physical programming and the decentralization of decision-making and responsibility, linked to performance targets. The process has begun with the implementation of the Integrated Financial Management System (SIAFI), but there is along way to go. A medium-term goal is to develop the capacity to budget on the basis of previous year's program performance within a medium-term budgetary envelope and realistic cost projection. An immediate step should be to formulate the budget on the basis of the previous year's budget execution. Currently, the sector's budget appears to be formulated on the basis of the previous year's approved budget, which does not accurately reflect actual spending pattern and needs (Chapter 3). 4.35 There is at present no effective regulatory framework to assure either value-for-money in public programs or acceptable quality in the provision of services in general. Measures are needed to develop the regulatory capacity of the SS to supervise the provision of health services and medicines. Honduras Public Expenditure Review 58 4.36 Improving quality and efficiency in the SS: SS services show the classic symptoms of public services that are managed by politicians and workers' representatives, without being subject to adequate regulation or accountability to users. As a result, its ambulatory and hospital services are generally low in quality and low in productivity. New mechanisms are needed which will establish clear obligations on service production and quality. 4.37 In primary health: * The government has developed an innovative project to contract with NGOs and private providers for the provision to 250,000 users of the "basic package" of mother and child health services and the establishment of community health funds. Learning from the successful community-based PROHECO schools, the government might consider expanding systematic use of community participation in running certain health posts. * The Integral Childhood Attention (AIN) program has been developed to shift the focus of the fight against malnutrition toward improved management including early identification of problems and support to families to correct dietary deficiencies. * The impact of the family allowance program, PRAF, is being optimized through a rigorous study of the respective impacts of supply side incentives designed to improve service quality and of demand subsidies to stimulate the take-up of primary services. 4.38 In the hospital sector: * The SS plans to implement a quasi-contractual relationship (Convenios) between hospitals and the ministry, which will specify performance targets. This will be coupled with the decentralization of budgets and the implementation of a modernized management model, including where appropriate the outsourcing of services to private suppliers. International experiences show that inter-institutional performance contracts are difficult to implement for a variety of reasons including the government's inability to effectively monitor performance and politically-motivated connivance of poor performance. Therefore, such measures should be introduced with care on a pilot basis with gradual devolution of budgetary and managerial flexibility only against demonstrated performance improvement. One of the essential pre- conditions should be timely and reliable reporting by hospitals on their budgetary execution. 4.39 A clear challenge for both primary and hospital services is that of negotiating changes to the labor statutes and collective agreements governing employee remuneration and working conditions that greatly reduce efficiency and raise the cost of services. New arrangements are needed which ensure that the public sector gets good quality services in return for fair pay to its staff. The SS needs to improve productivity through the modernization of labor contracts and linking remuneration to performance. The SEFIN has developed a new salary policy for the public sector that aims to improve the rationality of the system. However, real advance in the health sector is likely to require a shift away from the traditional pattern of "dual employment" in the public and private sectors, towards a single-employment norm, under which SS staff are expected to dedicate their full effort to their SS job. 4.40 Private sector development: Although there is an ample private market for ambulatory consultations in the urban area, only a small proportion of hospital attentions are privately delivered and financed. It is likely that the SS's highly subsidized hospital services are crowding out private hospital provision, in spite of the low quality of the SS's services. In the future, as SS service quality improves and the scope widens, this problem will become more severe. The SS therefore needs to define carefully the services that will be subsidized and establish full cost recovery for tertiary services (including from beneficiaries of the IHISS and private insurance) and introduce strict means testing for subsidized access Honduras Public Expenditure Review 59 to such services. A study is currently underway of the working of the co-payment system and the new World Bank health sector project will support the development of clear policies in this area. There is also a need to develop better regulation of the private insurance market in health in order to increase coverage in this area, which is presently infinitesimal. 4.41 Social security: The IHSS faces a grave institutional crisis that places in jeopardy the future of social health insurance in Honduras, threatening to channel more and more demand toward the SS and cause a large increase in the burden on the taxpayer. IHSS at present functions as an insurer for both health and retirement and as a (highly inefficient) provider of health services. These three functions should be completely separated. To promote cost efficiency, the new social security health insurer should not be obliged to use the existing ISS hospitals. They should be encouraged to procure services from third parties such as private hospitals and the SS. The old ISS hospitals and clinics most probably should be fused with the SS, but without a corresponding increase in the latter's budget. The SS would finance their operation by selling services to the new social security health insurer. 4.42 Contribution levels should be set that are sufficient to cover the efficient cost of providing a basic package of health insurance, including primary outreach programs for mother and child health, which the IHSS has traditionally neglected. Insured should be encouraged to "top up" IHSS insurance with optional private policies to cover risks not included by the IHSS. The establishment of adequate contribution levels would allow the coverage of the IHSS insurance system to expand without diluting quality. Special attention should be given to expanding social security coverage in centers of fast formal sector employment growth such as the Sula Valley. Honduras Public Expenditure Review CHAPTER 5 EDUCATION Overview of the Sector 5.1 Educational outcomes: Honduras has registered steady improvement in many key educational indicators over the last three decades (Table 5.1). Between 1974 and 1997 the average number of years' study doubled from 2.3 to 4.6; the illiteracy rate fell by over 50 percent; preschool attendance doubled and primary attendance of nine year olds rose by 33 percent. In each case, the rural population registered the most marked improvements. Still the urban-rural gap remained large in 1997. Table 5.1: Trends in Educational Outcomes 1974 1988 1990 1997 Urban Rural Total Urban Rural Total Urban Rural Total Urban Rural Total Average years of study, popln. aged 10+ 4.1 1.4 2.3 3.9 2.8 3.4 5.6 2.6 3.9 6.1 3.3 4.6 Illiteracyrate,% 19 51 40 17 42 32 11 30 22 10 23 17/1 Preschool attendance, children aged 5 and 6, % 33 18 22 46 13 33 38 19 26 60 34 44 Primary attendance, children aged 9, % 80 61 69 84 62 70 93 81 85 96 90 92 Source: UNAT (1999: 74-75), based on Census data for 1974 and 1988 and EPHPM for 1990 and 1997. Note: 1/ This figure, based on self- evaluation in the bi-annual household survey, is much lower than that of 30% reported in the WDR for 1996. 5.2 With public education Table 5.2: Education Outcomes Compared with International spending running at 4.6 percent of Benchmarks GDP in 1996, Honduras compared AvL.&M. relatively well with other Central income, American and Latin American HN CR ES GT NI LAC countries (Table 5.2). It spent Outcomes comfortably above the regional Adult illiteracy rate, 1997, /0/1 30 5 23 34 37 13 average of 3.7 percent and more than % of cohort reaching 5h grade, 1996/2 71 88 77 50 n.d n.d its Central American neighbors except Causes Costa Rica, which has contributed to Net primary enrollment, 1996, % 90 91 78 n.d. 78 91 the relatively high primary enrollment Net secondary enrollment, 1996, % 22 41 21 n.d 27 33 rate. But in spite of Honduras' Public edu. Spending % of GDP, 1996 4.6 5.3 2.2 1.7 3.7 3.7 relatively strong funding effort and Sources: World Bank, WDR 2000; SEP; Finance Ministry and UNAT, 1999. Note: I/ high coverage, its outcomes in The illiteracy data given here for Honduras are from the WDR and are nor consistent primary education are not notably with those cited in the previous table, based on the EPHPM, which show a much lower level of illiteracy in 1997. The goverment estimates illiteracy at 20% in the population superior to those of its neighbors, and aged 15 to 40 in 1997. The figure given here for spending as a % of GDP in Honduras its coverage of secondary education is taken from Finance Ministry data (see table x on page x). The comparative figures for compares unfavorably with other other countries are fr the WDR. Note: 2/The data for FN is for the 6th grade. countries. 5.3 Coverage: I the 1990s Honduras made progress in Table 5.3: Educational Coverage, 1990-99 expanding education coverage (Table 5.3). Preschool Preschool Primary Secondary coverage rose sharply from 17 percent in 1990 to almost 40 Gross Gross Net Gross percent by the end of the decade. Gross primary education 1990 17.1 94.5 81.5 29.6 coverage (defined as all children enrolled in primary 1993 26.4 98.9 87.2 28.4 divided by the estimated population of primary age) rose 1996 32.6 98.3 88.0 34.4 slightly in the same period and is now close to 100 percent, 1999/1 38.9 97.2 n.d. n.d. while net coverage (excluding repeaters) is 89 percent. Ref. Pop. 4 - 6 yrs 7 - 13 yrs 14-19 yrs Notes: 1/ Preliminary. Source: SEP- Computer Dept. Honduras Public Expenditure Review 61 5.4 Secondary coverage has also grown, but more Table 5.4: Age-specific Secondary School slowly, reaching 35 percent in 1999, compared with 30 Attendance by Cycle, 1999, % of Cohort percent at the start of the decade, according to official Age 13 14 15 16 17 18 19 20 figures. Household survey data show that the attendance at Ist cycle 16 25 30 22 16 9 6 5 secondary school during the correct age range is still 2nd cvcle 0 1 2 8 11 8 6 4 disturbingly low. In 1999, only 32 percent of children aged Total 16 26 32 29 27 17 11 9 15 were attending secondary school. The coverage of the Source: Calculated from EPHPM, March 1999 second cycle of secondary education during the correct age is especially low: for no age group is the attendance rate at this level above 1 percent (Table 5.4). In higher education, coverage is estimated in 1997 at 9.8 percent of the population in the relevant age group (19-25), up from 6.3 percent in 1980.3r 5.5 Principal services and types ofprovider: Table 5.5 summarizes the supply of pre-school, primary and secondary education in Honduras. The data show the dominance of the public sector provision at the primary level, and the relative importance of the private sector in the pre-school and secondary levels. The relatively small size of the secondary school population - 290,000 students in 1999 in contrast to 1.1 million at the primary level - is quite noteworthy and indicative of the supply gap in the education system. The fastest-growing sector is the informal private pre-schools, where enrollments grew on average by 20.1 percent in 1990-99, followed by private secondary schools with an annualized growth rate of nine percent, indicating the latent unmet demands in these two areas. Table 5.5: Supply of Preschool, Primary and Secondary Education in Honduras Preschool Primary Secondary Total Formal Informal* Total PublicPrivate Total PublicPrivate Establishments NumberStructure Number Structure NumberStructure 1990 910 n.d n.d 7,685 96% 4% 519 39% 61% 1993 1,929 63% 37% 8,076 95% 5% 660 30% 70% 1996 2,658 61% 39% 8,401 95% 5% 818 44% 56% 1999 ** 3,481 50% 50% n.d. n.d. n.d. n.d. n.d. n.d. Annualized growth, 1990-99 16.1% 7.4% 14.9% 1.6% 1.5% 4.1% 7.4% 11.2% 4.1% Students enrolled 1990 76,558 73% 27% 889,346 94% 6% 192,588 60% 40% 1993 123,475 55% 45% 991,216 95% 5% 203,192 59% 41% 1996 163,771 49% 51% 1,032,595 94% 6% 266,538 64% 36% 1999** 202,804 46% 54% 1,079,869 94% 6% 290,446 n.d. n.d. Annualized growth, 1990-99 11.4% 5.9% 20.1% 2.2% 2.1% 2.8% 4.7% 2.1% 9.0% Memo item: Pop. growth of 1.6% 2.2% 3.8% relevant age group, 1990-99 Notes: * Includes CENPENF and CCIES. ** Preliminary data. Source: Enrollment data: based on SEP data and our calculations. Population growth estimates: derived from the official projections implicit in EPHPM database for 1990 and 1999. 5.6 In higher education, there are two important state universities: the Universidad Nacional Aut6noma de Honduras (UNAH), with some 71 percent of all university students in 1999 and the Universidad Pedag6gica Nacional (UPN), with 13 percent. Total enrollment in these state universities in 1998 was estimated at 57,000, or 84 percent of the total." There are several private universities, including the Escuela Agricola Panamericana (EAP) El Zamoran,; the Universidad Catdlica, UNITEC, and the Universidad Josi Cecilio del Valle. Although the supply of private university places has grown in 38 "Estudio Sobre el Gasto en Servicios Sociales Bdsicos. Iniciativa 20/20." Secretarfa de Estado en el Despacho Presidencial, Unidad de Apoyo Tdcnico. Sept 1999. 3 There is also a small state agricultural university (Escuela Nacional Agricola, ENA, at Catacamas) and a state forestry school, ESNACIFOR, at Siguatepeque, which are not included in these data. Honduras Public Expenditure Review 62 absolute and relative terms during the last decade, they still account for only some 10,000 students - about 16 percent of all students enrolled in higher education. Education Expenditure in Honduras 5.7 National education accounts and the Table 5.6: National Education Accounts by Funding Source role of private spending: According to the and Financial Agency, 198, L million. national education accounts for 1998,40 an estimated 7.1 percent of GDP was spent on Funing sor educational activities. Of this, 80 percent (5.7 Public st Pie seto percent of GDP) was channeled through public agencies: National External housF agencies, such as the Secretariat of Public Public agencies Education (SEP), UNAH, the UPN, FHIS, PRAF, SEP 2,257 32 688 2,977 59.5 4.2 the professional training institute, INFOP and the UPN-FM 94 7 7 108 2.2 0.2 grant agency, EDUCREDITO. A further 1.4 percent of GDP was channeled through private and NGO-run schools, preschools and PRAF 34 15 50 1.0 0.1 universities, representing only 20 percent of the FHIS 35 77 112 2.2 0.2 total. INFOP 107 107 2.1 0.2 5.8 In terms of funding sources, funds for EDUCREDITO 10 3 14 0.3 0.0 education from public sources totaled 62 percent Sub total 2,943 131 828 110 4,011 80.2 5.7 of the total (4.4 percent of GDP) and private % 73.4 3.3 20.6 2.7 100 sources generated 38 percent (2.7 percent of Private agencies GDP) .4' However, households provide an Preschools 69 69 1.4 0.1 important complement to state funding of public 7 139 139 2.8 0.2 education programs. Although fees for public Secondary programs are uniformly very low, parents are schools 551 551 11.0 0.8 required to supply teaching materials, uniforms Universities 230 230 4.6 0.3 and other supporting resources and often are Sub total 0 0 990 0 990 19.8 1.4 required to contribute to the upkeep of the school% 0 10 0 to building. Once the cost of such inputs is taken into account, a fifth of the funding for the public TOTAL 2,943 131 1,818 110 5,001 100 7.1 education programs came from private % 58.8 2.6 36.3 2.2 100.0 households. In most cases public funds cover only %of GDP 4.2 0.2 2.6 0.1 7.1 the cost of staffing and the school building. Source: ESA Consultores, based on Finance Ministry data and estimates of Textbooks are also supplied free with the support private spending from diverse sources. of donor-funded programs, but their availability is patchy. The private agencies' programs are funded almost entirely with private resources in the form of user fees for enrollment and tuition.d e 40 The available information about private resources for education in Honduras is patchy. The estimates presented here are based on a survey of 39 private schools and colleges conducted by ESA Consultores. In addition, the SEP's data on the number of students enrolled at private schools are not very reliable. 41 Estimates made as part of the preparation of the Plan Decenal suggest that private spending on education in 1996 was two percent of GDP (UNAT, 1999:98). However, this did not include an estimation of private spending in support of public programs. lA nationwide sample of 806 households in the coverage area of 24 primary schools collected for the Ex-Post Evaluation of FHIS 2 undertaken by ESA Consultores for the World Bank in 1998 found that the average household spent US$25 per year for each student on books, materials, uniforms and exams and US$8.5 in contributions to the P arents committee. (These data are reported in the data annexes of the study. SA small number of private secondary schools are co-financed by the state and the educational grant agency, EDUCREDITO provides some funding for students attending private colleges. In Table 5.5 these resources are Honduras Public Expenditure Review 63 5.9 An obvious area in which Honduras is not doing enough to mobilize private resources for education is the university sector. The 1982 Constitution "earmarks" 6 percent of current government income for the UNAH. Although the budget normally assigns somewhat less than this (it was 4 percent in 1999), this has generated a culture of fiscal "rights" in the university and has promulgated an attitude that higher education should be state funded and virtually free. This policy is questionable from a poverty- reduction point of view. Students from the top two income quintiles consume an estimated 80 percent of university education in Honduras. These students benefit from a level of per-capita spending estimated at over seven times that registered in primary education, and nearly four times that in secondary education." Nevertheless, co-payments by UNAH students in the form of tuition fees and the purchase of books and materials amounted in 1998 to only 19 percent of the cost of their university training (L. 122 million of the total of L.645 million). Of this, only L.29 million was paid in fees to the university; the annual matriculation fee is just L.250 (about US$16.7). The total annual cost of attending the UNAH in 1988 was estimated at L.2,250, compared with average private university costs of L.23,000 (US$1,500). With UNAH costing just 11 percent of the average private university, it is hardly surprising that the private sector has not developed a significant market share. This is a clear case of crowding out, analogous to the impact on private hospital development of the availability of relatively sophisticated, virtually free tertiary treatments in the public hospital sector. 5.10 Trends in public spending: Of the Table 5.7: Public Education Spending by Agency 1990-99 pubic agencies involved in the sector, the SEP is by far the most important, accounting for 69 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 percent of the public spending, followed by the Structure (%) UNAH (15 percent), FHIS (seven percent), SEp 73 64 66 65 72 68 69 67 71 68 INFOP and the UPN (three percent each), FHIS 1 10 11 11 2 9 7 10 3 8 PRAF (two percent) and EDUCREDITO (one PRAF 1 3 3 4 3 3 3 3 2 2 percent). FHIS' spending fluctuated as new INFOP 4 4 4 4 4 3 3 2 3 4 tranches of funding were negotiated and EDUCREDITO 2 1 1 1 1 0 0 0 0 0 disbursed. The other components of spending UNAH 18 16 14 14 16 14 14 15 17 15 have remained relatively stable, although the2 2 2 2 3 2 3 2 3 3 haE' sreandelaiely sgtale, aelethough the TOTAL 100 100 100 100 100 100 100 100 100 100 SEP's share declined slightly, reflecting the(1990=100) transfer of its capital programs to FHIS. SEP 100 88 101 106 106 111 107 112 119 142 FHIS n.a 100 115 127 19 107 82 120 42 116 5.11 Through the decade, education spending PRAF n.a 100 100 144 90 112 124 122 62 89 fluctuated without a clear trend, and then shot INFOP 100 94 116 113 97 91 72 72 98 142 up in 1999, primarily due to the award of EDUCREDITO 100 65 53 64 45 30 32 31 32 39 teacher salary increase. By the end of the NAH 100 86 83 93 95 89 89 99 115 122 decade, the agencies that grew most strongly UPN-FM 100 110 121 129 161 153 166 163 218 256 were the SEP, FEHS, INFOP (whose funding is TOTAL 100 100 111 120 107 118 113 122 122 151 based on a mandatory one percent payroll tax and is therefore highly elastic) and the UPN (which achieved independent university status in the first part of the decade and received strong support from the SEP thereafter). EDUCREDITO, UNAH and PRAF registered declining real per capita spending. reflected in the top section and are not ascribed to private programs, due to the lack of budgetary details to allow this breakdown to be made. 44 UNAT (1999, 89). Honduras Public Expenditure Review 64 5.12 Structure of SEP spending: In 1999, Table 5.8: Structure of SEP Spending by Program, primary education programs accounted for 43 1990-99,% percent of SEP spending, secondary for 18 1990 19911992 1993 1994 1995 1996 1997 1998 1999 percent, transfers to the two state universities Primary 50 48 47 47 47 47 45 45 45 43 for 25 percent, and central administration for Secondary 16 16 14 18 18 19 19 18 18 18 10 percent. The shares of primary education Adult education 1 1 1 1 1 1 1 0 0 0 and that of transfers to the universities Sector direction! declined markedly in 1999 (due to the impact central admn. 2 2 6 3 3 2 3 7 4 10 of salary rises in the non-university sector) and unverst administration grew significantly, apparently Capital due to the opening of departmental offices expenditures 0 0 1 0 0 0 0 0 0 0 (Table 5.8). When analyzed by expenditure Building constn. group, the personnel line item dominates the and maintenance 1 2 3 2 2 2 1 1 0 0 ministry's spending: there is no other Projects/I 0 0 0 0 0 0 0 0 0 4 Total 100 100 100 100 100 100 100 100 100 100 significant item apart from the transfers to the I/From 1999, budgets for projects such as PROMEB and EDUCATODOS universities (Table 5.9). The virtual absence of are shown separately. Source: Based on Finance Ministry data capital spending is a reflection of the transfer of SEP's school building programs to FHIS. 5.13 Cost of inefficiency: The Table 5.9: Structure of SEP Budget by Expenditure Group, % rising personnel cost and the stagnant educational outcomes at the primary 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 level (discussed below) raise Personal services 67 65 65 65 66 67 66 65 65 70 important questions about efficiency Other services 0 0 2 1 1 1 1 1 1 1 of the resources assigned to the Materials and supplies 1 1 1 1 1 1 1 2 1 1 education sector. A recent study by Machinery and equipment 0.3 0.4 1 0.4 0.5 0.4 0.3 1 0 1 Construction 1 2 3 1 1 1 1 1 0 0 UNAT reported that the average Real estate 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0 0 number of years' study to graduate Transfers 30 31 29 30 29 29 31 30 32 26 from primary education in Honduras Global 0 0 0 0 0 0 1 0 0 0 is 9.4 (compared with six years if no- Total 100 100 100 100 100 100 100 100 100 100 one repeated or dropped out); for each * Excludes 13th and 14th month salaries, paid directly by the Finance Ministry of the three-year secondary cycles, the average is 4.2 years; and for the university (most of whose courses are programmed at five years' full time) the average is I11 years. These figures suggest, implicitly, that 36 percent of primary spending, 29 percent of secondary spending, and 55 percent of university spending is not used efficiently, in the sense that it does not produce a student qualified at the level being studied. In total, about 36 percent of SEP spending (including transfers to the university) would be spent inefficiently. This would amount in 1999 to 31 percent of all public education spending, and was the equivalent of 1.4 percent of GDP, or L. 1,050 million. This would imply that improved efficiency is the number one potential source for funding improved educational coverage and quality. 4 UNAT (1999, 76). The same study (p.100) estimates the saving from eliminating inefficiency at 16 percent of public education spending, but this is an underestimate. The estimate of 36 percent presented in the text is made as follows. If the average student takes 9.4 years to graduate from primary, instead of six, then 3.4 years' of tuition are wasted, representing 38 percent of total spending (3.4 / 9.4 = 36 percent). in secondary the wastage rate is 29 percent and in the university, 55 percent. Primary education accounted in 1999 for 45 percent of SEP spending; secondary for 18 percent and transfers to the universities for 25 percent, so the overall wastage rate is given by the sum: (36% 0.45 + 29% * 0.18 + 55% * 0.25) 35.2 percent of SEP spending. Since SEP spending in 1999 amounted to four percent of GDP, the implicit wastage was equal to 1.4 percent of GDP (4.0 * 0.352 1.41). Honduras Public Expenditure Review 65 Equity of Access to Education and of Educational Outcomes 5.14 Access to primary education: As Table 5.10: Attendance at a Primary School, March 1999 noted in Section 1, Honduras has done very Av,7 well expanding access to primary education. Age: 7 8 9 10 11 12 -12 13 14 15 16 Data from the government's household survey National 88 93 94 94 90 61 87 39 20 9 4 show that nationwide, 93 percent of eight year Tegucigalpa 96 98 96 99 93 56 90 27 13 4 5 olds, 94 percent of nine year olds, 94 percent S. Pedro Sula 97 96 98 98 91 62 90 34 14 11 8 of 10 year olds and 90 percent of 11 year olds Other urban 87 96 95 98 92 59 88 37 19 5 1 go to school (Table 5.12). There is a sharp fall Rural 85 91 92 91 88 63 85 42 22 11 4 in attendance among 12 year olds, to 61 Quintile 1 84 89 90 92 90 66 85 48 22 16 7 percent. Average primary attendance for seven uintile 2 87 9:3 94 94 93 58 86 46 23 10 3 to 12 year olds is 87 percent, compared with an ideal target of 100 percent. Quintile 4 91 99 97 98 92 58 89 27 13 5 1 5.15 Coverage is very similar across Quintile 5 96 97 97 98 88 54 88 25 11 1 3 geographical areas. The highest, in San Pedro Source: EHPHM, March 199. Sula and Tegucigalpa (90 percent) is not far above the 85 percent registered in rural areas. This impression of geographical equity is Ed.,.tin.,..1,percapita Income at Departmental Level, 1999 further reinforced by the absence of any apparent correlation between per-capita incomes and educational budget assignments at departmental level, indicating the absence of a biased allocation in favor of better-off 3 l =0 055 areas (Figure 5.1). Coverage is also remarkably even across income groups, varying only from 85 or 86 percent in the bottom three quintiles to 88 or 89 percent in the top two quintiles (Table 5.P10). Pco Mar 199 5.16 Average educational attainment has been improving through time. Of the cohort aged 15 in 1989 only 62 percent graduated primary and average years of primary schooling were 4.7 (Table 5.11). Five years later these figures had risen to 71 percent and 5.0, respectively. The expansion of primary coverage has mainly favored the poor. Among the bottom quintile, the average primary years completed and the share of the population who completed primary education improved respectively from 3.7 in 1989 to 4.6 in 1999, as compared to the change from 5.5 to 5.6 among the top quintile. Despite these encouraging signs, the improving trend of the 1980s petered out in the 1990s. In the cohort aged 15 in 1999, 70 percent had graduated from sixth grade and the average number of primary years completed was 5. 1, very similar to the figures the 1994 cohort. 5.17 Table 5.11 also shows a persistent, Table 5.11: Trends in Primary Attainment, 1987-97 marked difference in attainment between household income quintiles. Although only 52 Av. primary years completed % with primary complete tof bottom quintile children in the 1999 Cohort Quintile Quintile perentaged 15- cohort graduated from primary education, in: 1 2 3 4 5 Total 1 2 3 4 5 Total compared with 87 percent of top quintile 1999 4.6 4.8 5.2 9. 8 5.6 5.1 52 63 71 81 87 70 children. And average years of primary 1994 4.4 4.5 4.8 5.3 5.7 5.0 49 65 66 83 88 71 attainment are only 4.6 for the bottom quintile 1989g 3.7 4.2 4.6 5.3 5.5 4.7 40 50 63 77 86 62 compared with 5.6 for the top quintile. The Source: EPHPM, March 1999 persistence of considerable gaps is evidence that the main victims of primary education failure are still children from poor households. Although 85 Honduras Public Expenditure Review 66 percent of children aged 7-12 in the poorest quintile attend school, 48 percent of children in this quintile eventually leave school without completing the primary program, condemned to re-live their parents' poverty in the next generation.46 The divergence of outcomes between income groups is much greater than the divergence of access, and indicates the need for concerted efforts to retain poorer children in school once they enroll. 5.18 There are three basic causes of the failure to graduate more poor children from primary education. First, there is a problem of late entry: only 88 percent of seven year olds had started school in 1999. In the bottom quintile, the figure dropped to 84 percent. Second, there is a high level of failure, especially in the lower grades. The failure rate for first grade students is 18.5 percent. The cumulative effect of these two factors (late entry and first grade failure) affects the poor drastically. Of children aged eight in March 1999, 39 percent of those in the bottom quintile had not completed the first year of primary education, compared with only nine percent in the top quintile. Since the difference in the proportion in school at age seven is not great, it must be the failure rate that accounts for most of the difference in age for grade. Third, there is a strong tendency to drop out of school when children reach the age of 13, regardless of whether they have completed primary education.47 Therefore, once a child gets behind, the probability that it will not complete primary education is great. Together, these three factors add up to the failure to graduate 30 percent of all children and 48 percent of the poorest children from primary school. 5.19 Access to secondary and university education: The education system's failure to give the poorest children a fair deal at primary level is reproduced, in a magnified form, at higher levels of the system. Naturally, unless a child graduates from the primary level, they cannot continue at secondary and university levels. The lack of public provision in secondary education compounds the problem: with half the available places provided by the private sector, the poor have little effective access to secondary schooling, especially outside Tegucigalpa and San Pedro Sula. 5.20 The educational coverage gap Table 5.12: Age-specific School Attendance by Household between the quintiles is much larger for Income Quintile, 1999 pre-grade (5-6) and secondary ages (13- 18) than it is for primary ages (7-12) Age: 5 6 7 8 9 10 11 12 13 14 15 16 17 18 (Table 5.12). At the age of five, only 37 1stquintile 37 66 84 89 90 91 91 74 62 42 37 29 17 14 percent of children from the 1st quintile 2nd quintile 46 75 89 93 94 94 93 75 68 53 39 29 25 16 are in school, compared with 76 percent 3rd quintile 53 79 86 94 94 93 88 83 60 57 41 37 29 17 for the fifth quintile, a gap of 39 percent. 4th quintile 60 81 92 99 97 98 94 88 79 59 65 46 39 21 During the primary years the gap falls as 5th quintile 76 86 98 97 97 98 98 94 85 70 62 47 48 34 low as seven percent. But in the Gap 5th- secondary range (13-18) it climbs back 1st quint. 39 20 13 9 7 6 6 21 23 28 25 19 31 20 up to the range 20 to 30 percent. The Source: Calculated from EPHPM, March 1999 bottleneck at secondary level, in turn, serves to filter out most poor students before they can get to university level, explaining the very low participation of poor students at that level, in spite of the very high level of subsidy available there. 46 No comparable data are available on the enrollment levels or educational outcomes of Honduras' ethnic minority groups, such as the Garifuna, Chorti, Pesch, Tawaka and Misquitos, which between them constitute around 10 percent of the population. Ethnicity-specific data are hard to come by because there is no ethnic identifier in any of the major surveys. However, given the absence of educational materials in minority languages, coupled with the generally known correlation of ethnic minority status and poverty, it is highly likely that these groups experience lower than average enrollment and below average outcomes. With World Bank support, the government is currently developing minority language materials for primary education, which should help to improve both coverage and outcomes for these groups. 47 J. Edwards, B Fuller and S. Parandenkar. "Primary Education Efficiency in Honduras. What Remains to be Done?" 1997. Honduras Public Expenditure Review 67 5.21 Table 5.13 summarizes the results in Table 5.13: Educational Attainment of 25 year olds terms of educational attainment. Among 25 year by Income Quintile % olds in 1999, in the bottom quintile 60 percent had not completed primary education; only three percent had completed secondary Highest level attained Total Q1 Q2 Q3 Q4 Q5 education and none had reached the university. Less than primary complete 38.3 60.449.637.623.3 14.2 In contrast, in the top quintile, only 14 percent Primary complete 31.928.931.638.338.721.6 had failed to complete primary; 18.5 percent Secondary incomplete 13.5 7.9 12.5 11.7 17.7 19.4 had completed secondary school and a further Secondary complete 9.8 2.9 5.7 9.914.118.5 26.2 percent had gone on to university studies. Higher education 6.4 0 0.6 2.8 6.126.2 The predominance of the non-poor in university Total 100 100 100 100 100 100 education is particularly striking. EPHPM data Avee years of schooling 6.0 4.4 5.2 5.6 6.7 8.2 show that 80 percent of those with 51y Source: EPHPM, March 1999 university level education are in the top two household income quintiles. The foregoing analysis leads to the following conclusions: L Tackling primary education failure of the poor remains the number one challenge. Emphasis should be on getting the child into school at the right age and providing sufficient educational quality to reduce repetition and failure rates, rather than expanding physical supply of primary schools as such. Measures to improve teaching quality and management efficiency are critical. * The second major challenge is expanding secondary coverage overall, and in particular for the poor. There is a need for a considerable expansion of state provision in this area, while encouraging private provision (both formal and informal) and avoiding crowding out when state provision expands. s The third major challenge is improving equity, efficiency and effectiveness in the University sector in order to reduce inequitable subsidies to the non-poor. Improving Equity, Efficiency, Quality and Coverage in Public Education 5.22 Devising strategies to the remaining problems of equity, efficiency and quality would require (i) more detailed look at the causes of inefficiency and low quality in public primary education programs and seek appropriate measures for improvements, (ii) examining strategic options for improving secondary coverage, and (iii) exploring means to make progress in university reform. 5.23 Primary and pre-school education - causes offailure: Table 5.14: Impact of Single Teacher The causes of educational failure are complex, but some of the Schools on Incomplete Grade Coverage key problems with the Honduran primary system are well %ta oe known. In the first place, a considerable proportion of rural No of teachers % of schools 6 grades schools do not cover six grades, making it virtually impossible n 44 66 for the children to graduate from primary level. Nationally, 16 2 27 96 percent of primary schools do not provide fUl six-grade 3 0 6 17 99 coverage, and almost all of these are single-teacher schools. In 7 to 10 3 100 total, 46 percent of all primary schools in Honduras are still p a to 20 5 100 single-teacher institutions and a further 21 percent have two 21oton25 2 100 teachers. SEP data on staffing and grade cover in a sample of >25 2 100 2,589 schools for 1998 shows that 34 percent of the single Total 100 84 teacher schools fail to cover all six grades, while virtually all Source: Based on SEP data for 2,589 primary schools with two or more teachers provide full coverage (Table schools and 9,594 teachers in AtlAntida, Copdn, Choluteca, El Paraiso, Fco. Morazn, Gracias 5.14). The long-established goal of eliminating single-teacher Dios and Valle, 1998 (No data were available for schools clearly makes sense in this context. the rest of the countrd). Honduras Public Expenditure Review 68 5.24 However, increasing the number of teachers to eliminate single-teacher schools may not be a feasible option in many cases either due to the lack of fiscal resources to finance such an increase, or due to the difficulty in attracting teachers to remote areas, where single-teacher schools tend to be concentrated. When these problems hamper employment of additional teachers, or where enrollment is not enough to justify more than one teacher, special care should be taken to train the teacher to use the available multi-grade texts for teaching third-sixth grades. 5.25 There is also compelling evidence of inequitable Table 5.15: Dispersion of Resource Assignment and resource distributions within the primary sector. This Outcomes in Nacaome phenomenon occurs as much within local areas as between "rich" and "poor" departments, so it is not No of Pupil Repe- Deser- Total necessarily picked up by aggregated analysis at schools in Cost per teacher tition tion failure departmental level. Often, it reflects an urban-rural Group group student ratio rate rate rate division. For example, data from a 1997 sample of 79 2 2 25 27 14 1 17 schools in Nacaome show that while 30 percent of the 3 2 17 31 16 3 19 schools had pupil-teacher ratios (PTR) below 30 and 4 20 1275 36 14 3 17 resources averaging over L.2,000 per student, at the 5 4 792 44 22 0 22 other end of the scale there was a small group of schools Av. 1766 31.57 15.2 2.8 17.9 with a PTR of 44 and a quarter the level of resources, S,,,, UNAT, 1999:93. Based on SEP data compared with the best funded group (Table 5.15). While it is not straightforward to associate resource levels to educational outcomes, especially on the basis of such limited data, it does appear that there is room for improving equity in resource allocation at the school level. 5.26 Whatever the role of resource assignments, it is Table 5.16: Incidence and Cause of Closures in clearly not sufficient to assign teachers to schools. In a Sample of 24 Primary Schools, 1998 order to get outcomes, the school must be open and the Days open last week (average) 3.75 teacher must teach. As reported above, the level of Reasons for closure productivity of Honduran teachers is not high and there Staff rinin 8 r 3tf 32r133n3i16g 19 is considerable evidence of school closure leading to Staff vacations or permits 25 low numbers of class days per year. For example, the Holiday 22 ex-post evaluation of FHIS 2 looked at a sample of 24 Parents strike 25 primary schools nationwide, and found that they had on Source: Survey of 24 schools slated for MIS support, with a average been closed for 1.25 days (25 percent of tle total of 7600 students and 249 staff, ESA Consultores, 1998 time) in the previous week. Over half the closures were attributable to staff absences for one motive or another (Table 5.16). The study also found that 44 percent of the children had missed one or more days. About half the days missed were for demand-related reasons, such as children's illness, travel or working. The other half was missed for supply-side reasons, such as school closures or teacher absences (Table 5.17). Table 5.17: Missed Days in Primary Schools: Household Survey Evidence No. of days missed last week Reasons given for missing school % of children % of children % of children 0 56 Demand-related 48 Supply-related 46 1 17 Illness of child 17 School holiday 11 2 10 Child away from home 5 Absence of teacher 13 3 6 Child working 9 School closed for other cause 22 4 3 Child helping at home 2 Bullying 1 5 9 Child refused to go 9 Other causes 6 Child withdrawn from school 6 Source: Sample of 806 children in primary schools slated to receive FHIS investments. ESA Consultores, 1998. Honduras Public Expenditure Review 69 Box 5.1: Diagnosing the Causes of Educational Failure The causes of educational failure for poor children are complex. They have roots in the disadvantageous family situation (lower education of the parents, lack of resources for study materials, absence of a perception that education matters) and are reinforced by the relatively deficient supply of educational services in poorer communities. Various recent studies have used multivariate analysis to cast light on the relative importance of different factors in causing educational outcomes in Honduras. * The World Bank-supported Unidad de Medicidn de Calidad Educativa (UMCE) attached to the UPN is studying educational outcomes in 568 schools in third grade and sixth grade (ongoing). It has found that socioeconomic considerations, coupled with students' attitudes, are decisive determinants of results in third grade performance tests. There is also evidence that pedagogical procedures are an important factor. Other supply side variables are not found to be important. * A study of "Repeating and academic performance in grades 1-3" funded by USAID in 1997 sought to establish whether there is any objective basis for failing so many students at these levels. It was based on a sample of 17 schools, stratified by the level of attainment of the school (high, medium and low). It found a weak relationship between scores in attainment tests and the probability of being failed. Almost half the students who had been failed got test scores of over 60 percent in Spanish and math papers used by the researchers. The study found that supply side factors such as teaching quality and pedagogical strategies, classroom hours and attendance at preschool all contribute to determining outcomes. Pupil teacher ratios were not found to have a significant impact. The "Escuelas con Exito" project sponsored by USAID during 1998 in 39 schools in Francisco Morazan and Intibuca used mentor teachers to improve teaching strategies (relevant materials, student centered learning) and failure rates were reduced. However, there is also a need to improve the production of simple descriptive statistics on pupil teacher ratios, teacher and pupil attendance rates, textbook availability and similar supply-side indicators. 5.27 Strategies to improve quality and efficiency at the primary level: To push up the primary completion rate from the present average of 70 percent to the 85-90 percent range already being achieved by the top quintiles, a variety of initiatives are needed. Among them are the following: * Monitoring and reporting on educational quality and outcomes (being addressed through the work of UMCE). * Curriculum reform coupled with improved teacher training (to be supported through a new World Bank project). * Improved availability of texts (with support from Germany), * Get more primary schools to cover all six grades by adding additional teachers to single teacher schools and to those with high PTRs. * Provide a preschool year for more students. * Establish mechanisms to improve teacher performance (including the regulation of attendance rates through parental supervision, and better staff motivation through performance related bonuses). Honduras Public Expenditure Review 70 * Establish effective mechanisms to encourage families to enroll their children at the required age. The "Save the first grade" campaign which begun in 2000 with support from the World Bank's PROMEB project aims to address this problem. * Consider the automatic promotion of all students with an acceptable attendance rate in first and second grades. 5.28 Necessary elements of these strategies are already included in various reform agendas both by the government and by the civil-society-led Foro Nacional. However, there is a need to develop priorities and cost alternatives in order to turn this agenda into reality. This process would be helped by more evidence about the potential impact of alternative strategies. Cost-effectiveness analysis should be used to help identify the best strategies. 5.29 PROHECO as a new strategy for improving coverage, quality and efficiency: One of the most important causes of failure to complete primary education is late entry, due in part to access difficulties in remote areas. Even where schools exist, teacher and student absenteeism are rampant. To remedy this problem in remote areas, the PROHECO model aims to improve attendance by both teachers and students by involving the parents in their supervision. Local community associations called Asociaciones Educativas Comunitarias (AECOs) run the PROHECO schools. This gives the parents a decisive role in the administration of the school, including teacher hiring. The SEP assigns the AECO with legal status to carry out these functions, using delegated authority from the President of the Republic. The World Bank- supported PROMEB project pays the salaries of the PROHECO staff for the first year and the SEP picks up the bill thereafter. Parents donate their time and effort for the administration of the school, so that formally these are "semi-official" schools - the first time this has existed at primary level. 5.30 PROHECO teachers are paid according to the Estatuto del Docente, and since they are mostly new graduates at the bottom of the scale, their monthly pay is just over L.3,000 (about US$200). However, unlike mainstream SEP staff, they do not go onto the SEP payroll or acquire a "plaza" through an acuerdo ejecutivo, which thereafter is theirs by right. The PROHECO post belongs to the community and the teachers are contracted by the AECO, which receives a transfer of public funds in order to pay them. The teachers' unions are campaigning for these posts to be transferred to the standard arrangement (their so-called "officialization"), while parents are opposed to such an idea. 5.31 During 1999, PROHECO created 506 schools and 646 community based AECOs, surpassing program targets. To date (October 2000) 809 PROHECO schools have been created, covering over 20,000 children. The demand had been covered in Ocotepeque and virtually covered in Lempira, Copan and Valle. FHIS has been contracted to build 515 schools for PROHECOs. As with past FHIS school building activities, there have been problems of coordination with the educational authorities. In the case of PROHECO, this is compounded by ambiguity about whether the SEP or the community is to own the building. 5.32 The emphasis of PROHECO to date has been increased coverage in remote areas coupled with the creation of a new institutional relationship between the community and the school, which should result in improved efficiency due to better teacher and student attendance. However, the project has worked within the limitations of traditional pedagogy, which is not well geared to community-based education or to multi-grade teaching, and PROHECO schools report similar levels of grade-failure to traditional schools. 5.33 A new SEP-World Bank Community Based Education project due to begin in 2001 will build on the achievements of PROHECO. As well as building new PROHECO schools in communities where demand is sufficient to justify it, the project will support the transfer of failing traditional schools to the PROHECO model and finance the creation of new pre-grade classes using a cost-effective method. At the same time it will attack the problem of educational quality by promoting locally developed quality Honduras Public Expenditure Review 71 improvement strategies for preschool and basic education, working with AECOs (in the PROHECO model) and with Asociaciones de Desarrollo Educativo Local (ADELs), to be created in "official" schools. The AECOs and ADELs will be involved in the development of School Improvement Plans and will administer resources to provide additional support for teachers. Without doubt, the development of effective parental and community involvement in the management of schools in the SEP's mainstream program would mark a major step forward in Honduran primary education. The project will also strengthen the support from educational departments and districts to community-based and bicultural schools (in indigenous communities) and will promote improvements in teacher training and the adjustment of curriculum design to the needs of such schools. 5.34 Expanding post-primary coverage: The expansion of coverage in the first cycle of secondary education is the next major challenge for Honduran education policy. There is a need for an overall approach that specifies the role of the different initiatives and assures that the necessary resources are available to meet the agreed targets. The strategy should include both the expansion of the formal system and the development of vocational and distance learning alternatives. From the standpoint of public expenditure management, the dilemma facing the government is how to increase public funding for secondary education while minimizing the "crowding out" of existing private funding and avoiding the undermining of good private secondary schools that are already functioning. 5.35 Modalities for the supply of basic secondary education - the role of the private sector: A flexible approach should be adopted toward the expansion of post-primary coverage. The eventual goal should be universal coverage of basic education for grades zero through nine in the formal sector, standardizing the notion that children should stay in school through to 16 years of age. In many areas of the country, the most cost-effective strategy will be to widen the scope of existing state primary schools, adding three "middle school" years, and where possible, using the primary classrooms in the afternoon. This is much more economical than setting up traditional "Institutos Polivalientes" covering grades 7-12, which require heavy building investments and whose costs-per-student are nearly double those of primary schools. In some areas of the country, the most cost-effective strategy for increasing access at the basic secondary level might be giving grants to enable students to attend existing private institutes that have established programs of satisfactory quality. 5.36 However, many communities will remain outside the scope of secondary education for the next few years, and many students who have already left the system can only be reached by distance learning. Therefore, in parallel with the expansion of formal post-primary education, Honduras should continue to develop vocational programs using private agencies to deliver services, and should continue to develop distance learning for teenagers who are working or who for some other reason cannot access formal programs. This is particularly important for poor teenagers, many of whom need to work in order to increase their family incomes. The USAID project Educatodos is working in distance learning, for both primary and secondary education. Through CADERH, USAID is also supporting the development of vocational training, establishing 34 NGO-operated training centers, which will be co-financed by INFOP. 5.37 Financing the expansion ofsecondary education - minimizing "crowding out": There is currently little secondary education in Honduras and much of it is privately financed. However, if the definition of "basic education" is expanded to cover grades seven-nine, the role of public funding will inevitably increase. Then, sector policy should seek to minimize the "crowding out" of private finance when publicly funded secondary schooling expands. It is normally considered that basic education should be provided essentially free by the state. The existing co-payment mechanisms (provision of materials and other inputs plus parental contributions to school administration and upkeep) should be maintained to increase users'/parents' "ownership" of educational services, but no student should be excluded from basic education due to the lack of parents' willingness to pay. Honduras Public Expenditure Review 72 5.38 If the principle of free provision of basic education is followed, it is probably impossible to avoid all "crowding out" of private funding. If grades seven-nine become universally available on a free basis, some parents, who are at present willing to pay because there is no free alternative, will prefer to switch to the free services. However, the extent of this crowding out is likely to be fairly low, because private provision in grades seven-nine covers only around 15 percent of the cohort. There will also continue to be a market for private institutes and colleges based on quality differentials. Nevertheless, this market is also likely to be eroded as the quality of state programs improves. Therefore, once the policy decision is taken to extend basic education to cover the first cycle of secondary education, the main future growth in demand for private sector financing is likely to arise in the upper secondary cycle (grades 10-12). In this context, sector policy should seek to channel the private resources freed in grades seven-nine toward other educational applications, so they are not lost to the sector. 5.39 Financing and quality of higher education: The UNAH is a classic example of an institution trapped in a vicious circle of mediocrity and decline, and where institutional constraints have prevented the crystallization of a workable reform strategy. It provides a parallel from the education sector to the unresolved crisis of the IHSS in the health sector. The key obstacles to change are the UNAH's statutory autonomy, coupled with the mandate from the 1982 constitution that six percent of state revenues should be assigned to the University, with no effective accountability for its use. For years, there has been a growing consensus in civil society on the need for a transformation of the UNAH but no headway has been made, because a two-thirds majority in Congress is required to change the status quo, so that successive administrations have regarded confrontation with UNAH as quixotic. 5.40 The diagnosis of the problem is well known. Most students with UNAH places are from non-poor households, leading to highly inequitable distribution of the very high subsidy available to this sector. Students from the top two income quintiles consume an estimated 80 percent of university education in Honduras. These students benefit from a level of per-capita spending estimated at over seven times that registered in primary education, and nearly four times that in secondary education." Nevertheless, co- payments by UNAH students in the form of tuition fees, etc., amounted in 1996 to only seven percent of the public expenditure on the University. At the same time, the quality of UNAH's education is in a state of deterioration. 5.41 The main elements of a reform package for the UNAH have been identified in many previous studies. They include an end to the principle of "paridad estudiantil" which gives the student organizations equal power to the authorities in the administration of the university. There is also a need to establish serious entrance exams to prevent the entry of students who are not equipped to deal with the rigors of university training. Similarly, the indefinite repetition of failed courses should be ended. To restore UNAH's academic prestige, it should organize staff exchange programs with internationally credible academic bodies and establish regulatory services to guarantee quality and regulate exams. 5.42 There is also a need for user fees that reflect program costs, which would make students much more demanding in relation to educational quality. It would also significantly reduce enrollment, especially by students whose possibilities of graduating are limited. It is well established that most of UNAH's students could pay cost-reflective fees and means-tested grants could be established to avoid the exclusion of the poor. Another option would be to allow the repayment of fees through a levy on future earnings, using a "graduate tax" of the sort recently established in the UK. Graduates working in the public welfare services (health and education) could be given a reduced (or zero) rate. 5.43 UNAH's statutory autonomy makes it nearly impossible to achieve reform without consensus with UNAH authorities. And so long as they see proposals for university reform as being essentially motivated by the desire to reduce state funding to UNAH, they will continue to oppose it. However, the 48 UNAT (1999, 89). Honduras Public Expenditure Review 73 principal purpose of all the measures proposed in the previous paragraph is to increase the effectiveness of UNAH education, and not to reduce public spending. There is therefore space to develop a reform proposal that UNAH authorities could support. Regulatory Failure and the Agenda for Reform 5.44 Regulating educational quality: Besides the problems of equity, efficiency and quality in public education programs at all levels discussed so far, the sector also suffers from serious regulatory failure, which exacerbates the other problems. The most important goal of educational regulation is to determine the quality of education, establishing norms for what is taught (the curriculum), and how teaching takes place, in order to ensure the relevance of schooling, and ensuring the proper training of teaching professionals. 5.45 Curriculum and textbook development: The curriculum is the basis of any education system. Curriculum development needs to crystallize in textbooks which effectively institutionalize the implicit "scopes and sequences" of teaching. Various donor-funded projects currently underway are designed to address curriculum reform and development. 5.46 Teacher training: The capacity and commitment of the teachers in the classroom is probably the single most important factor in educational quality and outcomes. The organization and regulation of teacher training is therefore one of the most important functions in the sector. The challenge is to transform a teaching force of over 43,000 people, many of whom are at present poorly equipped to do their jobs. This is not simply a matter of inadequate pedagogical techniques. Tests have found that the average achievement of a primary school teacher in math is at third grade level. Teachers who do not understand the material they are teaching are unlikely to be very successful. There is a need to transform the teacher-training curriculum, placing more emphasis on establishing competence in the material that is to be taught and on the teacher's general cultural development. However, there is also a need to promote modem approaches such as student-centered learning, in place of outdated pedagogy. 5.47 It is also necessary to rationalize the supply of teachers in accordance with the effective demand and the public sector's fiscal capacity to pay for them (and the demand from private schools). Honduras at present trains most of its teachers in eleven Normal Schools, which provide three years secondary education (grades 10-12). There is a National Teaching University (the UPN) which produces Licenciados and also Profesorados. However, only 20 percent of Normal School teachers have Licenciaturas; a further 40 percent have Profesorados; the remaining 40 percent are themselves simply Normal School (i.e. high school) graduates. The Normal Schools graduate annually about 2,500 new teachers, which is far in excess of the system's needs. Between 800 and 1,000 new teachers are absorbed annually by the SEP. Many people study teacher training simply as a form of secondary education. There is a need to rationalize the Normal Schools and to link them to the UPN. The government has established the norm that all primary level teachers should have at least two years' university level education following graduation from the Normal level. 5.48 Regulation of public provision: Mechanisms are needed in public schools to ensure that staff work the allotted time and are assigned in the right places. There is also a need to regulate the behavior of students and their parents to ensure attendance at schools and prevent children being put to work for money or in domestic labor in prejudice of their schooling. These goals should be pursued with carrots, as well as with sticks. Poor performance such as teacher absenteeism should be punished effectively, for example, with salary deductions, while good performance should result in positive incentives. 5.49 The development of parental committees with a regulatory function at a local level is a promising approach, as the PROHECO project is showing. Building on the PROHECO experience, Honduras is now planning to involve parental committees in the regulation of all primary schools, through the ADEL Honduras Public Expenditure Review 74 model. Parental committees will work with the school staff to develop and implement school improvement plans, and be charged with administering a budget for materials and supplies. Separately, PRAF is undertaking a pilot exercise with non-salary incentive bonuses to staff with clear rules for payment linked to attendance and educational outcomes, which began disbursing funds in late 2000. 5.50 A fundamental pre-requisite for effective regulation in the public sector is the separation of the SEP's regulatory function from that of service provision. While the SEP remains principally a service provider which is managed by people who have spent their professional careers as teachers and whose promotion has been the product of negotiations between union organizations and the ministry, it is difficult to imagine that it will become a more effective regulator of the state schools. This problem is, if anything, even more marked in the education sector than in the health sector, due to the lower degree of professional stratification in education. For this reason, legislative modernization to establish a separate regulatory authority in state education may be considered. 5.51 Regulating private schools and universities: Great care should be taken to ensure that the regulation of private educational institutions is not used to limit private sector development and to shore up the vested interests of gremios in the public sector. The two obvious dangers here are price controls in private education and spurious regulation of quality in the private university sector under the tutelage of UNAH, making the creation of new institutions of higher education very difficult. The economic regulation of private school fees should be approached cautiously. If private provision is made unprofitable, schools will close and the state's problem will be increased. The large number of providers in private education precludes the problem of monopoly. It would therefore be inappropriate for the state to rule on the level of fees. Economic regulation of the private education sector should not go beyond assuring transparency and fairness (e.g., preventing large real increases in annual fees after parents have paid large entry premia to enroll their children in a given school). The regulatory authorities should, however, ensure that the quality of provision in the private sector complies with the norms required in state schools. 75 CHAPTER 6 INFRASTRUCTURE SECTORS49 Introduction 6.1 The last Public Expenditure Review for Honduras which covered the infrastructure sectors was carried out jointly by the World Bank and Inter-American Development Bank in 1995. The review recommended reforms in several areas to support the development of modern, efficient and well functioning infrastructure services. These included: decreasing the overextended role of the state in many infrastructure sectors and stimulating greater competition in the provision of infrastructure services; reforming a rigid and obsolete structure of regulation combined with the application of a clear, de- politicized, policy for setting tariffs; eliminating excessive micro-management by government of public enterprises; and increasing attention to maintenance requirements in many of the sectors. 6.2 Since then a number of important developments and changes have occurred. Most significantly Hurricane Mitch, which hit Honduras in October 1998, wiped out a large amount of the country's productive infrastructure with total direct and indirect damages estimated at US$5 billion. While considerable progress has been made in restoring the stock of infrastructure to pre-Mitch levels, this event set back Honduras in its development objectives of providing infrastructure access to its population. Over the last few years, Honduras has also put in place a number of initiatives to invite private sector participation in service provision. Progress on this front, however, has been uneven and key policy issues related to improving access and efficiency in infrastructure sectors have yet to be effectively addressed. 6.3 This chapter concludes that despite a number of important initiatives, and some advances on improving access in infrastructure, Honduras has not yet overcome the fundamental institutional and policy weaknesses identified in the 1995 Review. Infrastructure performance remains an impediment to competitiveness and growth and the World Micro-Economic Competitiveness Index ranks Honduras 58 out of 59 countries, with only Ukraine ranked lower."o Although public expenditures in the infrastructure sectors appear adequate, virtually all indicators of access, efficiency and service quality continue to be low in absolute terms and in comparison with other countries in the region. In large part this is due to the fact that sector structures and the related regulatory environment are not grounded in clear objectives of improving operational efficiency and quality of service to consumers. This applies equally to all infrastructure services whether publicly or privately managed. Public spending has been channeled predominantly to address the question of affordability rather than addressing the issue of greater access, especially to the poor. The consequence of this approach has been to support policies which favor consumers connected to existing service networks. On the issue of maintenance, Honduras has made some progress in the roads sub-sector with the establishment of the road maintenance fund and the use of community-based micro enterprises. Limited progress has been made in other sectors. 6.4 Improving infrastructure services remains an important priority for Honduras to stimulate growth and reduce poverty. Levels of infrastructure and infrastructure services are crucial to growths" and recent surveys of the poor52 show that their highest priorities revolve around access to key infrastructure services (clean water and sanitation, electricity and transportation). To improve its infrastructure services, it is 49 This chapter draws on analytical work underway for the Honduras Country Framework Report on Infrastructure Policies, being funded by the Public Private Infrastructure Advisory Facility (PPIAF). More detailed analyses of the various aspects of this chapter are available in this forthcoming Report. 50 Honduras' ranking was particularly poor in electricity (58), ports (59), and internal logistics network (58). Roads (54), airports (50) and telecoms did somewhat better (44), but remained below other Central American countries. 51 See Leipziger, Danny, "The Unfinished Poverty Agenda - Why Latin America and the Caribbean Lag Behind" appeared in the Financial & Development, March 2001 issue, Volume 38, Number 1 52 Voices of the Poor, World Bank, 2000. 76 crucial that Honduras define a clear and coherent strategy for private sector participation and public spending for infrastructure services. Major policy recommendations of this Report include:53 * Given large infrastructure gaps, private participation in the infrastructure sectors will be essential for the future; * To maximize development impact, private participation arrangements should be structured to give greater weight to improving access, quality and efficiency rather than maximizing short term fiscal gains; * The greatest impact on improving the quality and efficiency of infrastructure services is likely to come from better sector regulation, promoting competition wherever possible, and regulating with clear objectives of improving efficiency and quality of service; * Regulatory autonomy and effectiveness can be improved by merging the various regulatory entities into a single multi-sector regulatory body; * Government subsidies in infrastructure sectors need to be restructured, toward improving access to infrastructure services which is more likely to have an impact on reducing poverty, and away from improving affordability to those with existing network access; and * Expenditure patterns should place priority on maintenance activities. Sector Structure and the Role of the Public Sector 6.5 The public sector has traditionally played a dominant role in the infrastructure sectors in Honduras. The last decade and a half has seen several initiatives to invite private participation in infrastructure. In the electricity sector, private thermal providers now account for 40 percent of the power generated, and a new privately financed generation plant is being planned in Puerto Cortes to supply electricity additionally to neighboring El Salvador and Guatemala. Another successful initiative for private participation has been for airports, with a 20-year concession of the four major airports granted in 2000 to Inter Airports, a consortium led by San Francisco International Airport, including US$19 million of emergency investments. In roads, most domestic passenger and freight services are privately run. 6.6 Other infrastructure services remain in public hands, with some on-going private participation initiatives. In water and sanitation, almost all utilities are run either by municipalities, or by the SANAA, the national public utility which operates the water and sewerage systems for Tegucigalpa and the water systems alone for some 20 secondary cities. Private participation in the water sector has thus far been limited to partnership with communities for rural and marginal urban water supply systems and one formal private sector provider. In October 2000, the second largest city, San Pedro Sula, awarded a 25-year concession to Italian-owned Aguas de San Pedro. In Puerto Cortes, the water and sanitation system has been "corporatized" through the formation of a municipally owned company (Sociedad An6nima), which is operating the system on a five-year lease. 6.7 Honduran ports also remain in public hands and there is as yet no official policy stance on the privatization of the Empresa Nacional Portuaria (ENP), though studies for restructuring the sector are planned both for ports and the publicly run railway. Most roads construction has thus far been managed by the public sector, though in the early 1990s, maintenance started being undertaken through private contracts. The newly established Road Maintenance Fund will now finance maintenance from user taxes and with maintenance contracts with private enterprises. " More detailed sector-specific recommendations will be presented in the forthcoming Country Framework Report on Infrastructure being prepared with support from the Public-Private Infrastructure Advisory Facility (PPIAF). 77 Table 6.1: Principal Agencies Involved in Infrastructure Services in Honduras Private Agencies Public Sector Agencies Transport Road construction Fondo Cafetalero, IHCAFE SOPTRAVI (Direcci6n General Carreteras), municipalities, FHIS, SAG, COHDEFOR, Road maintenance Financing of Fondo Vial through Road Maintenance Fund (Fondo Vial), municipalities user charges Transport regulation SOPTRAVI (Direcci6n General de Transporte) Traffic management Secretaria de Seguridad, municipalities Ports Empresa Nacional Portuaria Airports Inter-Airport SOPTRAVI, Direcci6n General de Aerondutica (private concessionaire) Civil, Superintendencia de Concesiones,. Railways Ferrocarril Nacional de Honduras Water Urban water NGOs, private communal water SANAA, municipalities, FHIS, committees, San Pedro Sula water concession, Rural water NGOs, private communal water SANAA, Health Secretariat, FHIS, committees Regulation _CNSSP, Health Secretariat, Environment Secretariat Planning and policy Health Secretariat, SANAA, municipalities Electricity Production Private generators (40%) ENEE, Transmission ENEE Distribution ENEE, privatization of distribution under consideration Regulation Comisi6n Nacional de Energia Policy Gabinete Energdtico Telecoms Operation CELTEL HONDUTEL / COHONDETEL (fixed line, public) (wireless) J Regulation CONATEL Main Characteristics of Infrastructure Performance in Honduras 6.8 Infrastructure access indicators: Starting from a historically low base, infrastructure access indicators have improved in many sectors, but except for water and sanitation, they remain among the lowest in the continent. In water and sanitation, Honduras has advanced greatly on access between 1985- 95, increasing water coverage by 15 percentage points and sanitation coverage by 25 percent. As a result, with water coverage of just under 80 percent and sanitation coverage of 82 percent, Honduras is second only to Costa Rica in the region. These substantial improvements in the water sector have been achieved by publicly financed investments in rural and urban-marginal water systems working with community- based models for project identification and system administration. Further progress in improving access indicators is likely to encounter some challenges as the remaining areas of access are concentrated in small and dispersed rural settlements with special technical problems and high costs of constructing and maintaining water systems. Household data indicate that this pattern of rapid improvement in coverage was interrupted in the late 1990s due to the impact of Mitch. Water coverage declined in rural areas, while the impact in sanitation was felt in urban and rural areas. 78 6.9 In electricity, coverage has also increased rapidly over the last decade from 33 percent in Table 6.2: Household Survey Data on Trends in Access 1989 to 52 percent in 1999, but still ranks among to Water and Sanitation the lowest on the continent. However, official 1990 1993 1997 1999 statistics mask high levels of illegal connections ational % ofhouseholds without and household data suggest that actual coverage ater 27 15 8 9 may be closer to 69 percent, which though higher, is still low in comparison to regional counterparts. Per capita electricity consumption, at 350 kwh per rban capita, is the lowest in the region with the ater 18 16 7 6 exception of Nicaragua. One of the interesting anitation 13 7 6 8 features of Honduras is that access indicators have ural improved the most in the water sector, though they ater 33 15 9 11 have remained low in electricity. This is somewhat anitation 50 26 26 27 different from the experience in most Latin American countries, where electricity access rates are higher than water access rates. Table 6.3: Honduran Infrastructure Performance Compared with Central America Year WN CR ES GT NI Water coverage (% of households) 1995 77 99 80 54 55 Sanitation coverage (% of households) 1995 82 98 65 49 n.d. Electricity consumption per capita, Kwh /1 1996 350 1,349 516 364 256 Electricity distribution losses, % /1 1996 27 12 13 13 28 Phone lines (fixed) per 100 population 1999 4.42 20.41 7.61 5.46 2.98 Road maintenance (% of network in poor condition) 1999 24 6.10 In telecommunications, Honduras has made limited progress in recent years. The number of fixed lines per 100 population has risen to 4.42, compared with 2.7 in 1995. This is ahead of Nicaragua and close to Guatemala, but well behind El Salvador and Costa Rica. Access to public phones is also low. There are only 1,300 public phones, or 0.22 per 1,000 population. This compares poorly with the Central American regional average of 0.69, and the figures for El Salvador and Costa Rica, of 0.9 and 2.06, respectively. 6.11 Even though access indicators have begun to improve, access remains highly determined by income levels and geographic location. This trend is consistent with the regional trend in Latin America - the lower the access indicators, the higher the inequity in access by income and geographic region. In the water and sanitation sector, currently only 24 percent of households in the bottom quintile of the income distribution have traditional public services (SANAA or municipal), while 52 percent rely on collective or private systems (operated by community water committees). In contrast, 80 percent of households in the top quintile get traditional public services. This lack of equity in piped water accentuates income distribution issues as empirical studies indicate that the net value of a water connection is estimated to be seven percent of the per capita income of households in the bottom quintile of the income distribution.54 6.12 Similarly, in electricity, coverage is only 30 percent for households in the bottom quintile of the income distribution, compared with 93 percent in the top quintile. Access is also highly correlated with geographic location. Ninety-two percent of the urban households are connected compared with only 49 percent of the rural areas. But unlike in the case of water, few poor households have established a private supply of electricity, either individually or collectively. Nationwide, less than 0.5 percent of households e Honduras Poverty Diagnostic, World Bank 2000. 79 are covered by such systems, rising to three percent in rural areas, leading to a much higher differential in rural-urban access than in water. Unlike water and sanitation, expanding access of electricity has not been a priority during the last ten years, and only US$10 million has been spent on rural electrification. Currently ENEE is working on a US$380 million rural electrification plan to increase rural coverage to 75 percent by 2010. In telecoms as well, as elsewhere in Central America, telephone services are concentrated in urban areas, especially in Tegucigalpa. More than 100 of Honduras' 298 municipalities do not have any telecommunications service and 50 have only telex and radio. 6.13 Quality and efficiency of infrastructure services: Operational efficiency and quality of service remains low in most sectors. In water and sanitation, even though coverage indicators have improved, the quality of the public services is very low. SANAA, the national utility based in Tegucigalpa, has serious problems of over employment and inefficiency, while most municipal systems are understaffed and lack adequate professional skills. Prices for water services tend to be low, undermining financial sustainability, particularly in municipal systems. A 1997 study comparing municipal and SANAA systems in small towns found that monthly tariff income per connection was L. 11.6 in the municipal systems, less than half of the L.23.7 reported in the SANAA systems. 6.14 In electricity, the estimated 20 percent distribution losses are far above the 12-13 percent reported in Costa Rica, El Salvador and Guatemala. Illegal connections in barrios marginales account for a good part of ENEE's distribution losses. Underestimated bills to commercial clients remain another long-standing problem. On the cost side, ENEE has 476 employees for every 100,000 customers. In contrast, the least efficient private power distribution utility in Latin America has about 100 employees per 100,000 customers with most in the range of 3 to 12. 6.15 In telecoms, the unregulated HONDUTEL monopoly has also resulted in poor service quality. Some 36 percent of lines report a failure each year, compared to the international benchmark of one percent. Only 60 percent of failures are repaired in a day (versus benchmark of 75 percent). Although the expansion of lines in the second half of the 1980s (implemented by Siemens and ATT) increased the digitalization of the network from 83 to 95 percent, the call completion rate remains very low, at 67 percent for local calls, 51 percent for domestic long distance and 52 percent for international calls. In each case the benchmark is 95 percent. Staff was reduced from 4,700 to 4,200 between 1995 and 2000 and further reduced afterwards by an additional 700 positions. This increased the number of lines-per- employee between 1995 and 2000 from 34 to 66, and to about 80 in 2002 but Honduras is still well behind Costa Rica (179) and Guatemala (147). 6.16 In transport, airports show the best operational indicators. Honduras' four international airports (Tegucigalpa, San Pedro Sula, La Ceiba and Roatin) have shown rapid passenger growth over the last few years, at 8-9 percent a year. The new San Pedro Sula airport is the largest in capacity and volume, and is experiencing particularly rapid cargo growth of over ten percent a year, linked to the maquila sector. Honduran airports have relatively high labor productivity, at 4,134 passengers per employee, toward the top end of the internationally acceptable range. Passenger charges are high at US$25 per departure (the top of the international range of US$5 to US$25), but aircraft charges are mid-range (US$660 for an Airbus 310, compared with a range of US$200 to US$1000). The main problems of the air travel sector are: unreliable and poor quality service at Tegucigalpa, lack of electronic aids for landing and the low capacity of the passenger terminal; high ticket prices due to low volumes, the lack of competition (the market is dominated by one carrier), and the lack of effective consumer protection mechanisms; inadequate cargo handling capacity at San Pedro Sula; and weak air traffic control for lower airspace. 6.17 In ports, Puerto Cortds accounts for 89 percent of the country's port traffic, shipping 4.98 million metric tons in 1999. Though unit costs for cargo handling and tonnage per meter of quay are within 80 international ranges, the port has very high ship waiting times and the rate at which cargo is handled is far below international norms. These time delays imply that the total costs for users of Puerto Cort6s are much higher than international standards. 6.18 For roads, a recent road survey found that paved roads are in relatively good condition,5 perhaps in large part due to the high budget allocations to SOPTRAVI. Only eight percent of the asphalt concrete roads and 11 percent of the hydraulic concrete roads rated as poor. In contrast, 53 percent of the surface treated roads and 63 percent of the dirt roads are in poor condition, reflecting perhaps both the lower sensitivity of paved roads to inadequate maintenance activity as well as the possible lack of focus on the secondary network. The major issues in the roads sector are the relatively poor quality of regulation and traffic management, which have led to congestion and poor quality of public transport services. Broad Patterns of Public Expenditures 6.19 Through the last decade, the annual spending of the major entities in the infrastructure sectors56 has averaged 14 percent of GDP. These significant allocations imply that the main constraints to infrastructure performance do not come from the allocation of resources but from other issues related to sector management and regulation. The total share of infrastructure spending, however, declined substantially after 1996 as Honduras tightened fiscal policy as part of its stabilization efforts, averaging just about 11.4 percent in the following three years. As a result, at the end of 1999, real spending on infrastructure sectors was 12 percent lower than in 1990. 6.20 The biggest spender, by far, has been ENEE (averaging 41 percent of the total) followed by HONDUTEL (29 percent), and SOPTRAVI (20 percent). ENP and SANAA, in contrast, have spent, between six percent and seven percent of the total. As total spending in the infrastructure sectors declined over the decade, telecoms and the roads agencies continued to show positive growth in spending, while SANAA, the water agency, and ENEE, the electricity agency, showed declines in real spending. Table 6.4: Public Investment in Infrastructure, 1990-99 Average 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 1990-99 Total public spending in infrastructureas%ofGDP 16.0 13.7 16.3 15.7 15.6 15.9 15.7 11.4 11.7 11.2 14 Total public investment, % ofGDP 6.6 7.2 10.4 12.4 11.1 10.0 8.0 5.9 5.2 6.1 8.3 6.21 Capital expenditures for the infrastructure agencies averaged 7 percent of GDP, more than 80 percent of the total public investment of 8.3 percent of GDP estimated by the Central Bank over the last decade.7 In real terms, investment expenditures in infrastructure sectors have also declined by 12 percent over the decade 1990-99, though there was a large increase in public investment between 1998 and 1999 due to the post-Mitch reconstruction efforts. SOPTRAVI was the most important investor in the last decade, accounting for 36 percent of the total," followed by HONDUTEL (33 percent), ENEE (20 ' Honduras has an official network of 13,603 km of primary, secondary and tertiary (rural) roads managed by SOPTRAVI and an unknown extension (between 7,000 and 15,000 km) of other roads developed by other state agencies. Primary roads constitute 24 percent of the official network, secondary roads 19 percent and the remaining 57 percent are rural roads. 20 percent of the official network is paved, 68 percent is gravel and 12 percent is dirt. 56 This analysis includes SOPTRAVI (the transport secretariat), the ENP (the port authority), HONDUTEL (the phone corporation), ENEE (the electricity corporation) and SANAA (Tegucigalpa water corporation). ' There may be some definitional differences between the Finance Secretariat data and the BCH's estimates of rublic investment for the GDP data, but they would not likely affect this conclusion significantly. 8 In this context it should be remembered that SOPTRAVI classifies road maintenance spending as investment. 81 percent), SANAA (10 percent) and ENP (4 percent). Over time, however, capital expenditures by ENP showed the largest increase in real terms, with SOPTRAVI and HONDUTEL also showing positive increases in investments. Real investment in ENEE and SANAA declined. The decline in public investment in electricity was offset by an increase in private investment in electricity generation, and likely by municipal investments in the water sector, though no data are available at this point. Table 6.5: Distribution of Public Expenditures and Investments by Sector Public spending by infrastructure agency Public investment by infrastructure agency Distribution, 1999 real spending as Distribution, 1999 real spending as 1990-1999 average % of 1990 spending 1990-1999 average % of 1990 spending SOPTRAVI 20 134* 36 182* ENP 6 187 4 331 HONDUTEL 29 157 33 157 ENEE 41 67 20 70 SANAA 7 82 10 74 otal 100 88 100 88 * Figure is for 1998 spending as 1999 numbers were not available. 6.22 Sources offinance for public expenditures: The two key issues in infrastructure finance are: (i) the extent to which user charges cover the costs of the service (including operation and maintenance, depreciation and financing charges); and (ii) the financing of capital programs. Table 6.6 breaks down total expenditure (current plus capital) for each of the sub-sectors according to the source of finance: operating income (mainly from user fees); other internally generated finance (including depreciation allowances and movements in short term asset and liability positions in the balance sheet); national funds (transfers or loans provided by the government or borrowed directly); and external funds (net transfers and borrowing from abroad). 82 6.23 The three public Table 6.6: Infrastructure Spending by Financial Source % of Total corporations, ENP, HONDUTEL Spending - Current plus Capital - of Each Sub-sector and ENEE depend on the 1990 payment of user charges for the 19901991 1992 1993 1994 1995 1996 1997 1998 1999 -99 majority of their expenditures - Operating income 94 percent for ENP, 84 percent SOPTRAVI 0 0 0 0 0 0 0 0 0 n.d. 0 for HONDUTEL and 73 percent ENP 95 92 86 79 144 63 105 97 95 82 94 for ENEE. Depreciation HONDUTEL 76 73 93 87 67 84 74 109 94 79 84 allowances (which are current ENEE 35 70 55 77 69 64 75 109 101 76 73 expenditure but do not give rise SANAA 17 22 22 21 11 29 38 44 48 24 28 to a cash outflow) and Total 42 55 48 50 50 58 63 83 80 73 60 adjustments in net short-term Depreciation and balance sheet adjustments liabilities provide between 11 and SOPTRAVI 0 0 0 0 0 0 0 0 0 n.d 0 24 percent of the financing for ENP 21 21 31 4 -44 32 -41 42 20 22 11 total expenditure of the public HONDUTEL 6 19 19 23 46 27 -26 11 -15 23 13 corporations. As a result, the ENEE 84 30 31 33 -1 12 26 -9 -1 30 24 public corporations do not SANAA 7 17 9 23 13 10 17 7 12 14 13 receive any significant current Total 46 20 19 17 13 16 2 3 -3 25 16 budgetary transfers from the Net transfers and loans - national funds state. Rather, HONDUTEL and SOPTRAVI 63 53 36 32 26 38 54 49 76 n.d. 47 ENP make significant transfers ENP -13 -2 6 0 0 0 0 5 0 0 -2 to the national and local HONDUTEL -1 0 3 -2 1 0 0 0 0 0 1 governments. In 1999, transfers ENEE -16 0 4 0 39 15 6 0 0 -1 4 from HONDUTEL to the central SANAA 0 0 0 0 0 0 0 0 0 0 0 government amounted to slightly Total -2 10 7 11 19 12 11 9 12 0 9 government0amloned 30 Net transfers and loans - external funds over US$20 million (L.300 million) or the equivalent of 0.4 SOPTRAVI 37 47 64 68 74 62 46 51 24 n.d. 53 percentENP -3 -11 -10 16 0 5 36 -35 -15 -5 -2 peren HofNDPTL' and venen HONDUTEL 19 8 -9 -9 -13 -11 52 -20 21 -1 4 of HONDUTEL's revenue.ENE3 0 1 -0 -7 0 -7 0 0 50 Statutory transfers from the ENP ENE 76 61 69 5 7 1 45 4 0 62 5 to public agencies are smaller, TA 14 1 26 22 1 4 4 4 62 5 (L.84 million in 1999), though Ttl1 5 2 2 1 4 2 1 21 (L.8 milionin 999) thughSource: Our analysis of data from various offices in the Finance Secretariat. they represent a larger proportion Note: National funds and external funds include both budgetary transfers from central of its revenue (averaging 13 government and loans made directly by other public or private financial agencies to percent during the 1990s). The the agency concerned. All such loans bear sovereign guarantees. law also mandates that local authorities in port cities receive four percent of the ENP's revenue, and Puerto Cort6s, a major beneficiary of this clause, recently used this income stream as an implicit guarantee for a US$15 million loan from the TDB for the construction of its sewerage system. 6.24 SOPTRAVI and SANAA are in a different category. SCIPTRAVI does not have operating income, while SANAA funds only 28 percent of its total expenditures from tariff income. Both of these receive significant transfers from the government. SOPTRAVI transfers are financed almost equally from national and internationally mobilized funds during the decade 1990-99. SANAA on the other hand, does not receive transfers from national funds, but financed 60 percent of its total spending in the last decade from nationally guaranteed but internationally mobilized funds. Policy Issues and Priorities 6.25 Infrastructure performance in Honduras has been held back only in part by fiscal constraints. More significant issues, however, have been a persistent weakness of clear objectives in sector 83 management, weak institutional capacity and the lack of an effective regulatory framework to elicit efficient performance from existing infrastructure sectors. Poor quality of service has further eroded consumers' willingness to pay and the weak institutional capacity and regulatory framework has limited external financing. The key priorities facing Honduran policy makers in this regard are to: (i) mobilize private investment and management; (ii) structure private sector transactions to improve access, efficiency and quality of services rather than maximizing fiscal gains; (iii) complement this with a clear regulatory framework designed to promote competition improve quality of services; (iv) develop regulatory capacity and autonomy; (v) refocus public financing of infrastructure to enhance access to infrastructure services for poorer groups rather than on affordability subsidies; and (vi) ensure adequate expenditures and incentives for maintenance activities in the sectors. 6.26 Roles of the public and private sectors: There is an increasing acceptance among most parties concerned that some form of private participation will be essential for Honduras to close the performance and financing gap in infrastructure sectors. Though private financing is not a panacea, there remains considerable scope for deepening private participation in the infrastructure sectors in Honduras. In electricity, for example, there is consensus within the sector on the need for greater private financing, more efficient private investment and the need for a modem regulatory framework. Over the next five years, the sector's investment requirements will average some US$170 million per year, of which roughly 70 percent will be for generating capacity, 20 percent for expansion of the transmission grid and 10 percent for the distribution system. ENEE expects all future investments in generating capacity to be made by the private sector. In the telecommunications sector, notwithstanding the failure of the HONDUTEL auction in late 2000, the private sector can finance a significant proportion of the necessary investment, given an adequate legal and regulatory framework and a well designed bidding process. 6.27 Similarly, in the transport sector, the port sector urgently needs an infusion of capital and management expertise in order to expand capacity and reduce end user costs, and the government is now ready to ensure alternative models to accomplish this. In the roads sector, within the next five years, most sections of the main north-south highway will have average daily traffic (ADT) volumes above 10,000, which is normally regarded as the threshold level for economically viable private concessions for construction and management. In water, the success of the San Pedro Sula negotiations opens the possibility a similar operation under the next municipal administration in Tegucigalpa. 6.28 Inviting private participation, however, offers only part of the solution, and many areas will continue to require public support, either through direct provision of services or public financing of some part of the sector. In roads, for example, except for a few concessionable roads, such as the second ring road around San Pedro Sula and some sections of the north-south highway, much of the roads sector is also likely to remain in the public sector. Priority for public expenditures should be given to meeting the full costs of the maintenance requirements of the system (see below), and only then to low cost projects to expand access in remote areas by improving gravel roads. Some investments in widening sections of the north-south highway to provide crawler lanes on dangerous slopes are also urgently needed for safety purposes. 6.29 Similarly, most capital finance for water and sanitation for small towns as well as rural and marginal urban systems is likely to continue in public hands, as will the management and construction of the road network. In these sectors, public investments should be conditioned on full cost recovery (including operation and maintenance, capital consumption and interest), and on compliance with performance indicators that push the operator towards financial viability as a borrower in national and international capital markets. The government and 1DB have adopted this strategy for a new municipal water and sanitation loan, which will support 24 cities, beginning in 2001. In the rural sector, investments to increase water and sanitation coverage in dispersed communities using sustainable, community-based models should have a high priority in the PRSP and in the national planning framework as a whole. FHIS 84 is working to develop a model that aims to attack the remaining coverage deficits. Appropriate technologies and promotion of community-based operation and maintenance can also significantly enhance the scope and quality of these services. 6.30 Maximizing efficiency and access in private participation design: As a relative latecomer to private participation, Honduras can learn from the experience of previous reformers to best utilize private participation to advance development goals. A survey of 600 concession contracts around the world indicates that in most cases contracts have been tendered for the highest transfer or annual fees, suggesting that governments are most concerned with relieving fiscal constraints than tariff reductions or access improvements.s However, defining the objectives of privatization as improving efficiency or maximizing access can lead to different transaction designs which are more effective in achieving policy objectives, albeit with a smaller fiscal windfall. In fact, private participation does not have to be anti-poor and many poor will benefit from the service expansion that may be possible through privatization. 6.31 Honduras' large performance gaps on access and operational efficiency make it important that private participation transactions be designed to improve access and operational performance rather than maximizing fiscal impact. The San Pedro Sula concession offers an interesting design that gives priority to efficiency gains and avoids large tariff rises linked to over-dimensioned investment programs. This concession gives priority to improvements in utility performance from improved metering and collections. The long-term goals for the concessionaire are defined in terms of service improvement goals rather than investment goals. It is expected that once the concessioned company has gained credibility with improved performance, higher tariffs and increased investments would be more politically palatable. 5 A Review of Concessions in Latin America, Guasch, Jose L. et al, World Bank (forthcoming 2001). 85 6.32 The importance of a regulatory framework to promote competition and consumer protection: A strong regulatory structure is important both to reduce costs of financing for the private sector and to protect consumers. Argentina's experience shows the relative importance of regulation. Utilities privatization yielded operational gains in the infrastructure sectors equivalent to 0.90 percent of GDP, but effective regulation can add gains worth 0.35 percent of GDP (16 percent of the average expenditures on utility services). Moreover, the benefits of effective regulation as a proportion of existing expenditures on utility services were highest for the lowest income quintiles because regulation acts as a mechanism for transferring rents from the owners of capital to the consumers of the services produced with that capital.o Box 6.1: The San Pedro Sula Water Concession The San Pedro Sula water concession, approved in 2000, illustrates what can be done if the design of a concession is made to reflect local preferences. Unlike many other water sector concession schemes in Central America - such as the one that failed to get off the ground in Panama in 1997 - the San Pedro Sula negotiation gave priority in the short run to efficiency gains, rather than large new investments. This happened because the politicians supervising the process were not prepared to go to the public with a plan that involved a rate rise. As a result, even though the existing tariff is low by international standards, the concession will lower it further for many consumers. A household that consumes 35 m' a month will now pay US$0.09 per meter, compared with US$0.11 beforehand. Only consumers of more than 50 m3 a month will pay more. The concessionaire will depend for increased income on the improvement of metering and billing. At present many households with estimated bills pay for much less water than they really consume. Increased coverage (of water and sewerage) will also bring increased revenues. Another important feature of the concession is that the required investments, estimated at US$200 million over 30 years, are indicative rather than contractual. It is the service improvement goals that are binding. In the first five years the specified goals are for 100 percent water, water metering and sewerage coverage. The target for beginning wastewater treatment does not come into force for eight years and there is a provision for a tariff revision before this. After fihe %ears. the concession should have established its credibility through turning efficiency gains into improved services and it should then be easier to promote a public discussion of the need for higher rates to finance wastewater treatment. 6.33 Honduras continues to have inadequate regulatory frameworks for infrastructure services. For one, the regulators lack clear mandates in the areas of universal service access, operational efficiency and consumer protection. Further, in most sectors, policy making and regulatory roles remain confused. As a result, tariff setting remains an intensely political exercise with limited reference to cost recovery. In water and sanitation, for example, the regulation of SANAA's tariffs is charged to the Comisidn Nacional Supervisora de Servicios Pfiblicos (CNSSP). CNSSP does not seek to guarantee either efficient operation or a tariff sufficient to recover both operating and capital costs. Nor are municipal or SANAA systems regulated to require the operator to be efficient or to guarantee the users the right to a connection and to decent service. In the absence of clear objectives, existing regulatory institutions have supported the status quo, which in many cases has been the employee or special interest capture of the sector policy and implementation. Low service quality of service and poor customer responsiveness are pervasive complaints in Honduras. 6.34 The problem of the lack of autonomy and institutional capacity of the regulatory bodies is not limited to the water and sanitation sector. Regulatory agencies, one for every sector, each with limited 60 As quoted in Estache, Foster and Wodon, Infrastructure Reform and the Poor, World Bank forthcoming. 86 institutional capacity, are attached to the executive branch of the government. Most regulatory authorities have the legal status of organismos desconcentradas, meaning that their budgets are at the complete discretion of the central government. This arrangement obscures the distinction between policy and regulation. In recent years, Honduras has experimented with new regulatory institutions under new sectoral laws in electricity and telecoms. However it has yet to make sectoral regulation truly autonomous of the policy and service provision functions. In the electricity sector, the 1994 law created the Comisi6n Nacional de Energia (CNE), but the law was amended in 1997 to strengthen the control of the Executive branch over CNE. The draft electricity law being discussed in Congress does not correct this anomaly. In telecoms, the 1995 law created a sector regulator, the Comisi6n Nacional de Telecomunicaciones (CONATEL). However, CONATEL has also been involved in the formulation of sector policy, confusing its role in the sector. In water and sanitation, tariff regulation of SANAA is presently assigned to CNSSP which is composed of political and interest group representatives and does not operate as an independent technical body. Proposed new sectoral legislation currently before Congress provides for gradual municipalization of water services with the creation of a national regulatory office attached to the Secretariat of the government in charge of municipal affairs. 6.35 To address the issues of limited capacity and lack of autonomy of regulatory institutions, Honduran policy makers should consider setting up a multi-sectoral utility regulation structure, as has been established in Bolivia. Such a multi-sector structure would provide economies of scope by pooling several functions such as consumer protection, conflict resolution, and economic analysis for several sectors. It would also go a long way to providing regulators with sufficient critical mass, credibility, and perhaps, the means to better withstand normal political pressures arising from short term gains on difficult pricing decisions and the long term requirements for greater access and financial sustainability. 6.36 Infrastructure pricing and subsidies: Tariff policies in most infrastructure sectors are not optimal, and in many cases include subsidies that are poorly targeted and regressive. In the water and sanitation sector, the prices charged by utilities are highly subsidized: SANAA, for example, funds only 28 percent of its total expenditures from user charges. Additionally, most of its capital spending is funded by transfers from the central government, implying a significant subsidy from general tax revenues to those that are connected to piped water. This was corroborated by a 1995 survey conducted in Tegucigalpa's marginal barrios which showed that at that time, households having piped water saved US$7.70 per month compared to those that did not have piped water, and were able to increase their water consumption by 26 cubic meters." Since access, especially to piped water, is correlated with income levels, this subsidy has a regressive incidence. Government resources are also used for significant subsidies for the construction costs of community managed projects in rural and marginal urban areas. 61 Households without piped water paid on average US$2.90 per cubic meter for water from tankers or similar sources and consumed on average only 3.7 cubic meters a month, for a total expenditure of US$10.7 per month, which represented 7 percent of these households' average income. In contrast, those with piped supplies were paying $0.10 per cubic meter and consumed on average just over 30 meters a month, paying a bill of US$3.0 (1.9 percent of their income). (Walker et al, 1997: 19) 87 6.37 In electricity, the average price of Table 6.7: Targeting Performance of the electricity for domestic consumption is Electricity Subsidy in Honduras L.1.09/Kwh (US$0.073). By comparison, the average residential rate of private utilities in El Consumption Share of Share of Share of Salvador and Guatemala is approximately level in kwh clients by clients in subsidy spent US$0.10 /Kwh, while private utilities in category poverty by on non-poor Argentina, Bolivia, Chile, Peru and Colombia category households have residential rates (before taxes) in the rage 0-20 20.31 44.93 1.38 of US$0.07 to US$0.14 /Kwh. As well as 20-100 22.69 35.66 7.34 having a low average tariff, Honduras has one 100-150 12.63 16.82 13.14 of the most convoluted pricing systems for 150-200 11.16 10.98 19.24 residential electricity service in the world with a 200-250 9.25 15.64 20.35 complex rate schedule with multiple blocks and 250-300 7.43 17.09 20.36 an electricity subsidy for users with 300+ 16.53 10.15 - consumption of up to 300 Kwh per month.62 Total 100 24.96 81.81 Households in the consumption range 100-300 Kwh per month (41 percent of residential customers) get 86 percent of the subsidy, while those that consume less than 100 Kwh (44 percent of consumers) receive only 14 percent of the total. Since there is a strong correlation between electricity consumption and income, this distribution is clearly regressive. The government has currently allocated a maximum annual budget of L.240 million or about US$16 million to fund the subsidy, and it is estimated that 82 percent of the subsidy goes to the non-poor." 6.38 In telecoms, HONDUTEL charges for local calls are among the lowest in the continent and far less than the economic cost of the service.' Local calls are cross-subsidized by international long distance, and by business customers who subsidize residential services. For example, a call from Honduras to the USA costs US$1.14 per minute (reduced from US$1.40 a minute in February 2001), compared with US$0.60 for the same call in the opposite direction. As a result, 90 percent of the traffic originates in the USA, as many Hondurans use the call back feature, and HONDUTEL receives US$60 million a year (1998) from US carriers in settlement of these calls. This distorted tariff structure lies at the root of HONDUTEL's low domestic market penetration, because it enables the company to make comfortable revenues from a small proportion of high-usage customers: one percent of the users generate 24 percent of all revenues. At the same time, it gives the company little incentive to expand coverage among low-income users, as the bread-and-butter business of local calls is a loss maker. However, in the medium term this strategy is unviable, due to competition from callback and internet and mobile providers. It is also damaging to the economy as it discriminates against business customers, reducing their competitiveness in the international market place. Recognizing this, Honduras has begun a tariff rebalancing process that will reduce international call charges by 62 percent and increase local charges by 62 percent between 2000-05, which would end cross-subsidization to residential consumers. To the extent that the poor make relatively more local calls than international calls than the non-poor, this change would have a regressive distributional effect. However, since few poor have access to a phone service at present, this effect is not likely to be great in the case of Honduras. 6.39 Government subsidies also exist for urban public transport. Subsidies in bus fares potentially might have a greater immediate poverty alleviation impact in Honduras because these form a large part of 62 At 100 Kwh per month the subsidy is L.40; at 200 Kwh it is L.95 and at 300 Kwh it is L.152. When consumption reaches 301 Kwh the subsidy drops to zero, so that the marginal cost of the 301st Kwh is L.153. 63 Honduras Poverty Diagnostic 2000, World Bank, 2000 and Estache, Foster and Wodon, Infrastructure Reform and the Poor, World Bank forthcoming, 2001. 6 The client receives 200 free minutes in return for the monthly rental of US$2.05 and the marginal rate thereafter is US$0.021 per minute. 88 the expenditures of poor households. In Tegucigalpa, the government provides a bus subsidy to operators on a lump sum basis based on the monthly value of the passengers carried. Bus fares currently stand at L.1.5 per journey (about US$0.10), less than half the level in other parts of Latin America. The cost of a one way journey is about 3.3 percent of the daily minimum wage. Collective taxis, in contrast, charge about L.6.5 per trip, or 13 percent of the minimum wage. Fare revenues are supplemented by the subsidy to bus owners, which generates about 25 percent of total revenue. Compared to other subsidies, this subsidy is somewhat self-targeting, as reliability of buses is low and buses are dirty,. overcrowded, and dangerous with petty crime, which would mean that the use of buses is not a first choice for those who can afford alternative transportation modes. In fact, the poor quality of bus services has led to a modal shift to other forms of public transport, with the share of buses falling from 80 to 65 percent between 1990-7 in Tegucigalpa and San Pedro Sula, leading to over-crowding and congestion. However, there are various problems with the urban transport subsidy in its present form. First, this subsidy does not benefit the most marginal urban areas where bus routes do not operate due to lack of road access. Second, the allocation mechanism of this subsidy to bus operators inhibits competition on these routes, and exacerbates the poor quality of service by lowering the operators' dependence on the sale of services for revenue. It is noteworthy that there was no major public outcry when a similar subsidy was removed in San Pedro Sula, two years ago indicating that the need for a bus subsidy may be exaggerated. An alternative option of least cost subsidy may be a more effective way of implementing the same system. 6.40 Overall, it is clear that the pricing and subsidy policies in the infrastructure sectors are poorly targeted. The poorly targeted electricity and urban transport subsidies alone are double the size of the PRAF's relatively well targeted transfers to low-income rural households (L.185 million in 2000). There is a strong bias towards affordability subsidies that benefit existing users of traditional services, a group biased towards the non-poor. Some politicians support such consumption subsidies on the grounds that they help the poor. However, empirical evidence from Honduras and elsewhere suggests that the beneficiaries are primarily the non-poor. The benefits can only occur to those that have access to services, typically the non-poor. The subsidies do not even target those that already have access. In the case of electricity, World Bank estimates suggest that 81.8 percent of the subsidy is spent on the non- poor. Where willingness to pay is higher than service cost, access rather than affordability, is the binding constraint for poor households. There is considerable evidence that this is often the case in Honduras, since many of the poor households pay more for off-grid infrastructure services than they would pay if they had access to a grid connection. It is estimated that a connection to the public electricity network results in a 19 percent decrease in energy expenditures. Similar data from Guatemala indicate that the price of energy is reduced by 25 to 31 percent when the household has access to electricity. 6.41 The poor are more likely to benefit from public support toward increased access to infrastructure services. Additionally, access subsidies are one-time subsidies that do not imply a continuing commitment on public resources. They are easier to target as they are likely to go to those who do not yet have a connection, more likely to be the poor, and such subsidies have lower administrative costs. Access subsidies also are more likely to mobilize external financing than subsidies on consumption. From a development point of view, access deficits are particularly high in many infrastructure sectors in Honduras, especially in electricity and telecoms, implying a need for a greater push on improving access in these areas. Of course, any programs for improving access must be grounded on financially and institutionally sustainable mechanisms. 6.42 Thus, it is recommended that Honduras refocus its subsidy programs to enhance access rather than those that improve affordability. In the past, public support to improve infrastructure access has largely been limited to water sectors, allowing Honduras to achieve regionally comparable access indicators. The government has now indicated plans to establish similar initiatives for rural electrification and rural telephony funds financed from prospective privatization revenues. These access subsidies, however, should be conditioned on program designs for financial and institutional sustainability. This is 89 an important lesson from the experience with the large increase in improving access to water and sanitation sector where some projects have not been sustained. 6.43 To the extent some consumption subsidies continue, there is a need for much more analytical work to develop more efficient targeting mechanisms. In fact, World Bank research in the electricity sector indicates that electricity consumption per se is not the most efficient instrument for reaching the poor. A combination of socio-economic and housing characteristics were more accurate in identifying the poor and the extreme poor than electricity consumption. Thus, one of the priorities for the government should be to improve the analytical base for targeting subsidies. 6.44 Maintenance: As in most developing countries, maintenance has often been inadequately integrated into budgetary planning, and Honduras' poor record of maintenance was highlighted in the 1995 WB-IDB Infrastructure Report. Preventive maintenance activities are rarely performed and lead to high levels of system breakages and losses documented above in the section on operational efficiency. 6.45 The roads sector, however, has made important progress Table 6.8: Public Spending on on this front in recent years with the privatization of maintenance Road Investment and from force account to contracted maintenance in the early 1990s,'5 Maintenance followed in 2000 with the establishment of a Road Fund (Fondo Vial) funded with part of the taxation income generated from the 1999 real 2000 bud fuel surcharge. This will provide a predictable source of revenues Investment for maintenance activities. Over the past decade, new construction in roads was estimated to have absorbed two thirds of spending on roads, leaving around a third for maintenance. In Paving 11.6 9.4 2000, however, with the creation of the Fondo Vial, there has Reconstruction 0.7 5.4 been a major shift in this public spending policy. The share of Rehabilitation 8.9 8.8 maintenance shot up to 68 percent due to the creation of a Fondo Sub total 45.4 32.3 Vial. The Fondo Vial was to receive an estimated income of Maintenance US$40 million in 2000. Ensuring that these earmarked funds are used effectively will require a strong focus on institutional and accountability mechanisms. As part of this strategy, routine Periodic 38.8 0.0 maintenance activities under the Fondo Vial will be carried out by Routine 0.3 0.0 micro-enterprises selected from target communities that are Repairs 7.6 13.4 located along the road to be maintained, and their contracts will Sub total 53.3 67.7 include performance based monitoring indicators and penalties for Unclassifiable 1.2 non-compliance. This approach has been tried successfully in other countries such as Colombia and Peru, and the selection of Total 100 micro-enterprises from beneficiary communities has been shown Total US$ 52.9 76.1 to make a big difference in that micro-enterprise members have Source: SOPTRAVI budiets strong ownership of and interest in the road segment they maintain, and their contracts are performance rather than input based. 65As a result, the number of employees in the Maintenance Directorate decreased from 5,500 to 1,000 and all maintenance equipment was sold off. 90 6.46 Without a doubt the Fondo Vial Table 6.9: Estimated Maintenance Needs for the Roads marks a significant improvement of System financing for road maintenance. An Length Needs/km Needs important priority for the government km (US$/km) m. US$ should be to ensure that the transfers Official Paved 3000 10908 32.7 mandated by the current law actually take Unpaved 11000 2609 28.7 place on a continuing basis. Currently, the Total 21000 15731 76.9 amout bingtrasfered s lss hanthe Source: Calculated from UPEG Calculations on needs/km. and amount being transferred is less than th ESA Consutores estimates on road length. this expanded amount of resources is less than the total maintenance needs of the roads system which are estimated to be US$61 million a year (Table 6.9), and which will only be partially met from the US$40 million transfers expected from the Fondo Vial. These figures do not include additional US$15.5 required to maintain unofficial roads (7000km), which are not responsibility of the Fondo Vial but which would be important for improving access to roads by the poor in remote areas. This implies that Honduras would be best advised to seek financing strategies that do not add to the burden of public spending on maintenance. For example, wherever possible, concessions financed by tolls should be used for the construction and operation of high volume highways. The feasibility studies for any new publicly financed capital investments in the roads sector must be scrutinized very carefully with respect to their displacement effect on existing maintenance commitments. Another issue for the government is to enforce the clause of the law that specifies that no more than 10 percent of the transferred resources to the Fondo Vial be used for rehabilitation activities. Such rehabilitation expenditures should be the responsibility of the Roads Directorate and not dilute the focus of the Fondo Vial or divert funds from rightful maintenance activities. Conclusions 6.47 The main conclusions and recommendations of this Report are: (i) Improving infrastructure performance in Honduras is more an issue of improved operational efficiency and regulation rather than simply one of mobilizing additional capital resources. (ii) Any strategy to reduce the large gaps in infrastructure access and performance gaps will require an increased role of the private sector. (iii) Rather than focusing on the beneficial fiscal impact of share sales in public utilities, private participation transactions should be designed to improve operational efficiency, reduce unit costs, improve service quality improve access by giving higher priority to these indicators in contract design. (iv) Regulatory frameworks for all sectors, whether publicly or privately managed, need to be clarified to focus on sector efficiency, improved access and consumer protection. (v) Regulatory agencies need more institutional capacity as well as operational autonomy to achieve these objectives. One possible way of achieving this would be to consider the creation of a multi-sector regulatory agency. (vi) Public support in infrastructure sectors should focus public resources on increasing access through one-time capital subsidies rather than enhancing affordability. Affordability subsidies typically tend to be regressive and prone to leakages. Some affordability subsidies may be justified in special cases but need a very strong data base to minimize such leakages. (vii) More detailed sector-specific recommendations will be reviewed with the government during the discussion on the PPIAF funded Country Framework Report on Infrastructure. 91 ANNEX 1 DEBT SUSTAINABILITY ANALYSIS - A METHODOLOGICAL NOTE To implement the debt sustainability analysis we assume that the fiscal deficit is financed with external resources. This is reasonable for the case of Honduras where (i) domestic financing of the deficit has been traditionally small (in 1999 domestic debt represented less than 5 percent of GDP), and (ii) the medium term macroeconomic framework assumes that the deficit remain fully foreign financed. Formally, the change in domestic currency terms of the stock of debt between times t and t- 1 is expressed as: E(t)F(t)-E(t- 1)F(t- 1), where E is the nominal exchange rate in domestic currency per unit of foreign currency (i.e. Lempira per US$) and F is the stock of foreign debt denominated in foreign currency. The previous expression highlights that changes in the stock of debt denominated in domestic currency can be due to changes in the stock of external debt F and to changes in the exchange rate E. Next it is assumed that the fully foreign financed deficit (D) is converted in domestic currency at the end of the period, and that the nominal depreciation of the exchange rate between periods t and t- 1 is given by c(t), so that E(t)=(1 +s(t))E(t- 1). Hence, it is possible to write E(t)F(t)-E(t-1)F(t-1)=E(t)(F(t)-F(t-1))+(E(t)-E(t-1))F(t-1)=D(t)+ 6(t)E(t-1)F(t-1), or in terms of GDP, denoted by Y(t) E(t)F(t)/Y(t)-E(t-1I)F(t-1l)/Y(t)=D(t)/Y(t)+ s(t)E(t-1I)F(t- 1)/Y(t). With inflation running at a rate n, and real growth at rate a, Y(t)=(l +n(t))(l +o(t))Y(t- 1), and denoting with lower case letters ratios to GDP, it is obtained f(t)-f(t-1)=d(t)-k(t)f(t-1), (1) with ic(t)=(R(t)+ G(t)- s-(t))/(l+7E(t)+ CF(t)). 92 Box 1 An additional interesting feature of equation (1) is that by examining the control factor i, it is possible to infer whether the debt dynamics are in a stable path (K positive), or in an explosive path (x negative), and hence it can be used to analyze historical episodes. The table below reports the country's inflation, GDP growth rate, exchange rate depreciation, the estimated control factor K, and the deficit of the NFPS over the period 1961-99. It also reports the equilibrium debt for Honduras (as a percentage of GDP) implied by the dynamic debt model above and actual public sector debt. Inspection of this table indicates that over the period 1961-70, Honduras fiscal policy was consistent with debt to GDP ratios below 15 percent of GDP. The situation during 1970s was slightly different with larger deficits that resulted in an equilibrium stock of debt of almost 50 percent of GDP. As implied by the model, debt started increasing and duplicated over the decade. It is important to note that even if actual debt in 1980 represented 30 percent of GDP the continuation of the macroeconomic framework observed during the 1970s would have led the debt to GDP ration to its equilibrium value of 47 percent of GDP and then stabilized. 1961-70 1971-80 1981-90 1991-99 Inflation 2.9% 7.8% 6.9% 18.9% Growth Rate 5.0% 4.8% 2.5% 3.1% Depreciation 0.0% 0.0% 16.8% 12.2% K 0.07 0.11 -0.07 0.08 Deficit 0.9% 5.2% 7.3% 3.4% Equilibrium Debt (1) 12% 47% EXPLODING 42% Actual Debt (1) (2) 15% 30% 118% 85% (1) In percent of GDP. (2) Includes both foreign and domestic public and publicly guaranteed debt. As for the 1980s, there is a dramatic change with respect to the 1960-80 period. Lower GDP growth rates (half of those observed in the previous two decades) and the sharp devaluation of the Lempira in 1990 (which implies and average devaluation of almost 17 percent) resulted in an explosive path for the debt to GDP ration. This explosive path is evidenced by a negative K Actual debt increased during the 1980s to about 120 percent of GDP. Finally, the 1990s show a return to more sensible fiscal deficits, which together with a pick up in growth and the real exchange rate appreciation results in a lower equilibrium debt level (42 percent of GDP). In other words, the continuation of the macroeconomic scenario of the 1990s would have halved Honduras debt burden over the long term. Unfortunately, this economic scenario would not only project inflation running at about 20 percent per year, but also an unrealistic constant real exchange rate appreciation of about 4 percent per year over the medium term. If instead it is projected that over the long term, inflation stays at about 7 percent per year and the real exchange rate evolves according to the purchasing power parity (PPP) hypothesis, the equilibrium debt should be revised upward to close to 80% of GDP, or a similar level to that at end 1999 before HIPC. 93 ANNEX 2 SELECTED STATISTICAL RESULTS OF THE STAFF TRACKING SURVEY Table 2.2.1 - Proportion of Health Employees Really Employed at Site Given in Central Database Prof. Aux. Promoters & Specialist GP nurse Nurse educators Technical Others Total Column % Employed at the site 91.9 79.9 93.8 91.9 75.4 87.5 95.2 90.6 Not employed at site 8.0 20 6.1 8.1 24.6 12.6 4.8 9.3 Previously employed here 2.9 11.7 4.8 7.2 22.9 6.6 2.5 6.9 Pensioned 0 0 0 1.1 1.2 1.1 1.2 1 Permanently disabled 0 1.5 0 1.6 0 0.6 0 0.7 Transferred 2.9 10.2 4.8 4.5 21.7 4.9 1.3 5.2 Not known here 5.1 8.3 1.3 0.9 1.7 6 2.3 2.4 al number in universe 567 583 778 4,950 1,349 1,318 4,909 14,454 Average monthly salary, L. 21,500 17,000 7,000 3,600 3,000 7,000 2,615 5,013 Est. earnings of unknown staff, L./month 621,716 822,613 70,798 160,380 68,799 553,560 295,252 1,738,990 Table 2.2.2 - Proportion of Education Employees Really Employed at Site Given in Central Databases Non teaching (Ministry, Primary Secondary teachers Total Depts, teacher Basic Commerce & Poliva- Bach Acad./ Directors) (7-9) computing lientes Normales Percentages Employed at the site 89.4 95.2 95.6 99.1 91.8 100.0 94.1 Not employed at site 10.6 4.8 4.6 0.9 8.3 0.0 5.9 Previously employed here 8.5 0.2 2.9 0.9 8.3 0.0 3.0 Pensioned 0.0 0.0 0.0 0.0 0.4 0.0 0.0 Transferred 8.5 0.2 2.9 0.9 7.9 0.0 3.0 Not known here 2.1 4.6 1.7 0.0 0.0 0.0 2.9 Total number in universe 9,932 21,684 5,107 2,653 3,326 1,347 44,048 Average monthly salary, L. 5,215 4,612 6,062 5,912 5,123 6,397 5,021 Est. earnings of unknown staff, L/month 1,087,703 4,600,303 526,296 0 0 0 6,413,785 94 Table 2.2.3 - Attendance at Work Over the Last Week of H.S. Employees Really Employed at the Site Promoters & Specialist GP Prof. nurse Aux. nurse educators Technical Others Total % ofpossible days that were worked Central ministry 100 55 100 100 75 100 96 Regional office 88 59 24 0 23 33 33 29 National hospital 67 48 74 75 91 79 83 77 Regional hospital 95 82 81 81 99 79 97 87 Area hospital 83 55 69 72 100 70 72 72 CESAMO 22 72 63 75 78 57 53 68 CESAR 99 82 48 100 77 CLIPER 20 13 53 39 60 34 37 Total 74 61 70 74 69 69 76 73 Table 2.2.4 - Reasons for Absence from Work Over the Last Week, by Type of Institution Central Regional National Regional Area Percentage ofabsences ministry office hospital hospital hospital CESAMO CESAR CLIPER Total Training 0 4 5 0 2 11 4 0 5 Work meeting offsite 13 1 2 1 0 4 13 0 2 Vacations 46 0 38 11 18 17 30 34 21 Illness 0 0 22 30 18 4 0 11 12 Family member ill 0 0 0 2 0 1 0 0 0 Leave to compensate for extra time worked 0 0 24 19 22 6 0 50 16 Other paid leave 0 1 0 0 0 2 0 0 1 Unpaid leave 0 0 0 0 1 6 0 0 1 Strike 0 0 3 0 7 5 0 0 3 Unknown 42 95 7 37 33 45 53 6 39 Total 100 100 100 100 100 100 100 100 100 Table 2.2.5 - Reasons for Absence From Work Over the Last Week, by Type of Employee Prof. Aux. Promoters & Percentage of absences Specialist GP nurse Nurse educators Technical Others Total Training 12 15 9 5 9 4 0 5 Work meeting off site 6 6 0 2 10 0 0 2 Vacations 31 24 25 23 0 15 24 21 Illness 8 0 16 17 4 3 15 12 Family member ill 2 1 1 0 0 1 0 0 Leave to compensate for extra time worked 18 27 11 24 1 10 11 16 Other paid leave 0 0 0 0 2 0 1 1 Unpaid leave 0 3 6 0 5 1 0 1 Strike 0 0 0 0 0 25 1 3 Unknown 25 25 32 30 68 41 47 39 Total 100 100 100 100 100 100 100 100 95 Table 2.2.6: Attendance at Work of Education Employees that are Really Employed at the Site, and the Reasons Given for Absences Non teaching Primary Secondary teachers Directors), teacher Basic CommercH & Poliva- Bach Acad./ Total (7-9) computing lientes Normales Percentages % of possible days that were worked over last week 84 93 80 69 76 63 86 Reasons given for absence Training 19 0 0 0 0 2 5 Meeting out of workplace 10 14 0 1 1 0 6 Illness 28 24 7 4 20 15 18 Family member illness 13 2 10 0 0 0 5 Paid leave / vacations 23 45 3 12 7 48 23 Unpaid leave 3 2 45 12 35 5 16 Strike 1 0 10 38 34 8 12 Other 0 0 0 0 0 0 0 Unknown 4 13 26 33 3 22 15 Total 100 100 100 100 100 100 100 Source: Institutional survey Table 2.2.7 Health Secretariat Posts by Contract Type- STS data Type of contract Percentages Acuerdo Contract Nothing Total Senior management 87 11 3 100 Specialists 95 4 0 100 GPs 69 18 13 100 Professional nurses 89 7 4 100 Auxiliary nurses 94 4 2 100 Promoters & educators 100 0 0 100 Technical 96 2 2 100 Other medical professionals 89 9 1 100 Others 86 6 8 100 Total 90 5 5 100 Source: Institutional survey 96 Table 2.2.8 - Ratio of Real to Budgeted Posts in the Health Secretariat - STS Data/ 1 Total - Total - Health Nat. Reg Area CESAMO weighted simple Percentages Central regions Hospital Hosp Hosp /CMI CESAR CLIPER average average Specialists 96 100 100 88 80 120 n.a n.a 91 98 GPs 139 114 149 111 110 110 200 117 162 130 Profnurses 111 116 79 106 84 144 100 108 101 105 Aux. Nurses 100 129 100 437 96 125 94 100 119 155 Promoters & educators 94 101 100 100 0 126 89 0 79 74 Technical 100 104 118 116 79 155 n.a 100 106 112 Others 91 117 103 110 95 150 350 118 252 149 Total 95 119 101 142 94 134 116 109 116 117 Source: Institutional survey. Note: 1/Real posts includes staff with acuerdo, contract staff and other staff. Table 2.2.9 - Education Secretariat Posts by Table 2.2.10 - Ratio of Real to Budgeted Posts in the Education Contract Type Secretariat/ 1 Type of contract Total- Total - Percentages Acuerdo Contract Nothing Total Depart- Secon- weighted simple Directors / senior Percentages Central ments ry dary average average management 95 5 0 100 Senior Teachers 90 5 5 100 management 100 85 95 91 95 93 Primary 89 6 5 100 Teachers 102 100 102 101 Secondary 93 3 5 100 Other Other professionals 85 0 15 100 oers 108 87 12 9 107 Other non professional 76 1 23 100 Total 103 90 112 101 111 102 Total 89 4 7 100 Source: Institutional survey. Note: I/Real posts includes staff with acuerdo, Source: Institutional survey contract staff and other staff Table 2.2.11 - Correspondence Between the Real Function, Nominal Function and Education of SS Staff Prof. Aux. Promoters & Percentages Specialist GP nurse Nurse educators Technical Others Total Real function of employee Administrative 19.3 11.9 27.1 3.6 58.9 27.2 47.5 26.3 Medical services 80.7 85.0 68.8 91.4 33.0 70.0 25.4 61.9 Other 0.0 0.0 1.7 0.0 8.2 2.8 25.1 9.0 Unknown 0.0 3.1 2.4 5.0 0.0 0.0 2.0 2.8 % whose function corresponds to nominal post 92.0 84.9 65.6 88.0 84.8 82.7 79.1 83.0 % whose function corresponds to education 96.1 91.1 81.5 81.6 78.2 85.6 70.7 78.9 97 Table 2.2.12 - Correspondence Between the Real Function, Nominal Function and Education of SS Staff Non teaching Primary Secondary teachers (Ministry, Depts, teacher Basic Commerce & Poliva- Bach Acad./ Total Directors) (7-9) computing lientes Normales Percentages Real function of employee Administrative 91 23 24 9 21 2 38 Educational services 9 77 76 91 79 58 61 Other 1 0 0 0 0 0 0 Unknown 0 0 0 0 0 42 I % whose real function corresponds to their nominal post 80 96 90 92 89 58 90 % whose real function corresponds to their education 70 86 90 97 94 60 83 Table 2.2.13 - Proportion of Workforce that has Migrated Since their acuerdo was Issued (%) Where they work now Central Regional National Regional Area CESAMO/ CESAR CLIPER Total office hospital hospital hospital Clinica MI Type of employee (migrants as % of same type of employee in the place they work now) Specialists 45.5 100.0 46.6 18.2 0.0 100.0 - - 38.2 GPs - 100.0 31.3 0.0 57.1 100.0 - - 60.0 Prof nurses 100.0 100.0 44.2 32.2 29.3 100.0 100.0 0.0 45.0 Aux. Nurses - 100.0 49.0 6.9 22.9 37.5 51.9 - 36.7 Promoters & educators 0.0 81.6 100.0 0.0 71.4 36.6 0.0 - 48.7 Technical 34.3 24.3 66.7 0.0 57.6 100.0 - 100.0 56.3 Others 0.0 0.0 44.4 0.0 10.2 83.7 - - 35.3 Total 14.6 50.5 48.9 7.7 25.5 53.8 49.0 73.9 40.0 Place of original assignment (migrants as % of all employees who migrated to the present place of work) Central 0.0 12.6 0.0 0.0 0.0 0.0 0.0 0.0 0.6 Regional office 3.2 0.0 0.3 0.0 3.9 8.6 0.0 0.0 2.5 National hospital 11.4 0.0 24.8 0.5 1.3 0.0 0.0 0.0 9.3 Regional hospital 0.0 0.0 2.7 0.0 0.0 2.5 0.0 0.0 1.4 Area hospital 0.0 0.0 14.9 0.5 8.5 0.6 25.1 0.0 8.8 CESAMO/Clinica MI 0.0 21.4 2.9 6.0 9.4 8.3 6.9 0.0 6.5 CESAR 0.0 0.0 3.4 0.6 2.4 7.0 17.1 0.0 4.4 Unknown 0.0 16.5 0.0 0.0 0.0 26.8 0.0 73.9 6.4 Total 14.6 50.5 48.9 7.7 25.5 53.8 49.0 73.9 40.0 98 Table 2.2.14 - Other Paid Employments of SS Staff- Survey Findings Prof. Aux. Promoters & Specialist GP Nurse nurse educators Technical Others Total Percentage of HS employees % with another paid job 53.6 33.4 13.5 6.9 15.6 20.6 4.8 11.2 Compared with HS post, second job is Percentage ofallsecondjobs Similar in nature 75.4 89.0 74.7 61.2 23.2 71.6 18.0 59.0 Different 27.7 18.3 25.3 38.8 76.8 21.4 82.0 41.1 Unknown 0.0 0.0 0.0 0.0 0.0 7.0 0.0 1.1 Place of second job Other HS hospital 23.3 8.3 54.9 23.7 0.0 19.6 0.0 17.8 Other HS health post 0.0 0.0 0.0 10.3 0.0 7.0 0.0 3.5 Private clinic 45.6 49.7 14.5 37.5 7.8 44.8 18.0 33.8 Other 31.7 25.0 30.6 11.4 39.4 5.9 46.1 25.0 Unknown 2.5 24.2 0.0 17.1 52.8 22.7 35.8 21.2 Shift worked in second job Morning 3.1 7.2 0.0 0.0 0.0 0.0 0.0 1.3 Afternoon 12.5 3.7 5.2 0.0 28.0 3.8 0.0 6.6 Night 17.3 2.8 21.0 0.0 7.8 12.8 18.0 10.6 Unknown 70.2 93.6 73.8 100.0 64.2 83.4 82.0 82.8 Earnings from second job L. / month 17,563 10,153 9,106 2,203 3,578 5,801 1,586 6,512 % of earnings in sampled HS post 90.2 59.5 125.7 57.4 87.3 103.6 52.6 131.1 Table 2.2.15 - Other Paid Employments of Education Staff - Survey Findings Non teaching (Ministry, Primary Secondary teachers Total Depts, teacher Basic Commerce & Poliva- Bach Acad. T Directors) (7-9) computing lientes Normales Percentages % with another paid job 16.1 19.7 48.2 41 15.5 49 23.1 Compared with sampled ES post, second job is Similar in nature 98.3 89.9 98.4 82.6 80.3 96 92.6 Different 1.7 10.1 1.6 17.4 19.7 4 7.4 Place of second job University 0.2 0 0 4.4 0 26.9 1.8 Public technical or vocational institute 3.2 0 2.7 0 7.9 4.1 1.7 Public secondary school 7.6 39.4 25.2 64.9 27.6 24.3 31.4 Public primary/basic school 29 30.1 17.6 5.1 48.7 36.8 26.5 Other public school 0.8 9.7 1.6 13 15.8 4 6.6 Private secondary school 0 0 9.4 0 0 0 2 Private primary school 59.1 14.2 0 0 0 0 16.3 Unknown 0 6.6 43.5 12.6 0 4 13.5 Earnings from second job L. / month 3,709 5,605 3,365 4,297 5,890 5,378 4,652 % of earnings in sampled ES post 71 122 56 73 115 84 93 Note: No data could be gathered on hours of work in the second job. 99 ANNEX 3 STAFF TRACKING SURVEY METHODOLOGY As part of the Honduras Public Expenditure Review (PER) undertaken during 2000, the government and the World Bank decided to finance a staff tracking survey (STS). The STS in Honduras is a variant of public expenditure tracking surveys (PETS), which have been undertaken in various countries to determine the correspondence between budgeted and real expenditures and to establish their relationship between outcomes and the funds assigned in the central budgetary process. STS seek to improve understanding of the relationship between the real trend in budgetary assignments and observed outcomes at the local level, in order to orient the design of policy reforms to improve outcomes. In some cases, although budgetary assignments for primary programs rise (due to national shifts in policy priorities), there is no corresponding improvement in primary health or education outcomes, because the resources never reach the programs in whose names they have been consigned. STS have identified the cause of such incongruities in the leakage of funds through straightforward corruption and through misdirection of funds to unbudgeted purposes. In other cases, they have revealed that the programmed spending does take place, but does not result in any worthwhile outcome, due to inefficient service delivery arrangements. The design of the STS study for Honduras took account of some important differences between Honduras and other countries where STS have been undertaken. In Honduras, although there have been some advances in decentralization of the administrative and fiscal system, the payments system still highly centralized. The Finance Ministry executes almost all payments (both for staff and for supplies) and there is little transfer of cash resources to other secretariats, either at national level or to departmental or local managements. Therefore, there is relatively little opportunity for siphoning off cash to unbudgeted uses as it passes down the administrative chain. However, under the Honduran system, there is possible scope for misappropriation of funds through the payment of salaries to fictitious employees and for the misallocation of resources through the transfer of jobs to other workplaces or functions. The centralized payment systems deal with enormous numbers of employees, contracts and payrolls. It is difficult - if not impossible - for the users of the central payroll information systems to know whether a given employee really exists and is actually working where they are supposed to be working and doing what they are supposed to be doing, according to the budgetary assignment that pays for their contract. It is part of Honduran folklore that - as part of the political rewards system - governments often nominate supporters to fictitious posts where they simply draw a monthly check without working in any meaningful way. This practice is known as paracaidismo (parachuting). It may be reflected in payments to people simply unknown in their supposed workplace, or in the appointment of "protected" employees who are allowed to come and go as they please and are not expected to do any serious work. However, although the newspapers often carry stories alleging such practices (usually sourced to the political opponents of the government of the day), there has been no effort to quantify the problem at national level. The problem of the migration of posts between administrative units within ministries is also known to be important in Honduras. The SEFIN receives requests from the line ministries to create new posts. Once the SEFIN approves the funding, it is incorporated in the annual budget of the ministry, duly approved by Congress. It is then in the discretion of that ministry to nominate the individuals to fill the posts, subject to any specific legislation governing hiring, such as the Civil Service Law, the Estatuto del Midico Empleado (1985), or the Estatuto del Docente (1998). The acuerdo (agreement) to hire the individual 100 must be published in the government's official newspaper, the Gaceta.' The cancellation of a post also requires a formal ministerial resolution. When a post is cancelled (due to death, retirement or dismissal) the budgetary assignment becomes available for the minister in question to issue a new acuerdo in favor of someone else. These arrangements mean that a plaza or nombramiento, once assigned, is effectively the property of the nominee, rather than belonging to the administrative unit (school, health post, hospital etc.) whose needs justified its original creation. If the holder of the plaza negotiates with the minister a transfer to another workplace, the original workplace loses the funds. A post holder may lobby for transfer from a remote posting to more comfortable assignments in large towns and from primary to non-primary programs. However, the central budgetary information systems may still show the post as being assigned to the original place and function. Another cause of the migration of posts is the inflexibility of the budgetary process. Given the difficulty of creating new posts, system administrators often solve urgent staffing problems by negotiating with the director of the unit concerned to use a vacant post to cover the needs of a different institution from that to which it is assigned. Whatever the cause of the migration of the post, it leads to misinformation regarding the amount of resources being assigned to remote areas and primary programs. Previous reports and studies have concluded that the personnel information systems in the health and education secretariats are inadequate and have suggested that there is a problem of the "migration" of posts from the place or function for which they were originally assigned. A Health Secretariat staff census conducted in 1995 found that 38% of all staff was no longer working in the place to which they were originally nominated (quoted in the World Bank's 1998 study "Honduras - Towards Better Health Care for All", page 45). For these reasons, it was decided that the Honduras STS should concentrate on tracing the congruence between budgetary and real assignments for the employment of staff, and indicators of staff attendance at work, in order to try to quantify the importance of the problems of paracaidismo and migration of posts. Due to resource constraints, the scope of the survey was limited to staffing expenditures in health and education, which in 1999 constituted 47 percent of the total budget of the Health Secretariat and 70 percent of that of the Education Secretariat.' Study questions The principal study questions were the following: 1. What correspondence exists between real staff assignations and what the central records state? Two dimensions of this question were explored: a. Paracaidismo: What proportion of staff assigned in different types of entity and to different types of function, really work for the system? (i.e. identify "ghosts"). b. Migration ofposts: What proportion of staff work where they are supposed to be working according to the central or regional records, what proportion of posts of different types have "migrated" in the past, and what migratory patterns exist? 66 In the past, these resolutions required the signature of the President of the Republic but nowadays for most cases the Secretary of State in the relevant secretariat is the person who must sign the agreement. 67 Another important difference between Honduras and other countries where STS have been carried out is that, in Honduras, the efficiency and efficacy of service production units is relatively well studied, especially in education and health. For this reason, the survey concentrated on the tracing of personnel and did not seek to evaluate the service delivery dimension at local level. 101 2. Absences from work over the last week and the reasons for this. 3. How many people have multiple jobs, a) within the same ministry and b) combining ministry employment and outside employment? 4. What proportion of employees does what they are supposed to be doing (correspondence of real function with officially recorded function). 5. What proportion of employees does what they are best qualified to do (correspondence or real function with training). 6. How do people get employed in the system in the first place? 7. How could the central staffing information systems be improved? Study phases The study had two phases: 1. Analysis of central information related to staffing and remuneration. In this phase the staffing and payroll records of the two ministries were gathered together and analyzed to provide a sampling frame for the field study. At the same time the study team evaluated the personnel information systems, developing recommendations for their improvement, and analyzed the national level databases to reach conclusions about some of the study questions using the universal data sets. 2. Survey of a nationally representative sample of staff for verification of the official records and to answer the other study questions. In this phase sample data were gathered in the field on staff's existence, assignment to the correct place of work, attendance at work and personal employment histories. The results were analyzed to answer the study questions. Sample design for the primary investigation The sample for the primary investigation was based on national payroll database information in the health and education secretariats. The total sample size and its clustering in specific workplaces was constrained by budgetary limits which allowed for a total sample of 1,465 employees clustered in 64 different workplaces. Subject to this constraint, the sample design sought to optimize probable sampling errors taking account of the expected variances of the study variables in different sampling strata.68 Sampling universes Health From the total of 15,308 employees in August 2000, the following categories were excluded: a) doctors on call-out contracts (midicos de guardia), who would not be expected to be at work for a spot visit survey, so there was no point including them in the sample (331 in total) and b) individuals whose place of work could not be identified from the database, who totaled 422; c) individuals whose post was not identified (20); and individuals whose place of work was given as a private hospital or clinic (40)9. Initially, the agreed sample size was 1205 employees in 60 workplaces. This was to be divided into 805 in health, in 40 workplaces and 400 in education, in 20 workplaces. However, during, the design phase the study team decided to increase the education sample to 660 in 44 workplaces, in order to lower the expected errors of the education survey data. 69 The HS staff posted in private clinics were removed because the intended sampling procedure was a two step process based on the selection of workplaces in the first step, so that no-one who did not work in a HS establishment could possibly be selected. 102 The total removed from the database for sampling purposes was 813, representing 2.7 percent of the Ministry's staff. The final database for sampling purposes totaled 14,495, located in 873 different workplaces or health service production units (Table 3.1) Table 3.1 - Sample frame, Health National Regional Type of employee Central Regions hosp. hosp. Area hosp. CESAMO CESAR CLIPER Total N N N N N N N N N Specialist physician 32 15 352 78 71 17 0 2 567 (5.6) (2.6) (62.1) (13.8) (12.5) (3.0) (0.0) (0.4) (100) General physician 22 66 85 68 121 204 21 16 603 (3.6) (10.9) (14.1) (11.3) (20.1) (33.8) (3.5) (2.7) (100) Prof. Nurse 25 76 285 89 141 125 21 17 779 (3.2) (9.8) (36.6) (11.4) (18.1) (16.0) (2.7) (2.2) (100) Auxiliary nurse 2 176 1678 640 936 825 659 37 4,953 (0.0) (3.6) (33.9) (12.9) (18.9) (16.7) (13.3) (0.7) (100) Promoters 102 273 78 16 149 513 217 1 1,349 (7.6) (20.2) (5.8) (1.2) (11.0) (38.0) (16.1) (0.1) (100) Technicians 106 176 427 133 231 228 16 17 1,334 (7.9) (13.2) (32.0) (10.0) (17.3) (17.1) (1.2) (1.3) (100) Non medical 554 500 1365 551 1219 503 109 109 4,910 (11.3) (10.2 (27.8) (11.2) (24.8) (10.2) (2.2) (2.2) (100) Total employees 843 1,282 4,27( 1,575 2,868 2415 1,043 199 14,495 (5.8) (8.8 (29.5k (10.9) (19.8 (16.7 (7.21 (1.4) (100 Total centers 1 S7 7 2 23 3 8731 (0.1) (1.0), (0.8), (0.8) (2.3) (74 (67.2) (0.3) (100A Source: Health Secretariat database, June 2000. Excludes doctors on call contracts and 482 cases, which could not be linked to a Health Secretariat workplace. Eduation The Secretariat supplied the Escalaf6n databases for primary teachers (31,954) and secondary teachers (14,066). The Secretariat also supplied lists for employees assigned to the central ministry (814); non- teaching staff employed under Civil Service rules outside the central ministry (3,725); and for teachers with administrative duties (3,812). This gives a total of 54,371 employees or posts within the Education Secretariat. The Escalaf6n databases permitted the identification of the workplace of all the teaching staff. For sampling purposes, preschool teachers and teachers employed on distance learning programs were removed, as the study did not cover preschool and it would not be feasible to make a visit survey to distance learning staff. The Central Ministry list included 86 people who were shown as already retired, who were removed for sampling purposes. A further 84 whose workplace could not be identified (10 percent of the total in the Central Ministry) were also removed, leaving a total of 671 in that stratum of the sampling database. For reasons of survey economy, the sample frame excluded ancillary staff. The only non-teaching staff from outside the Central Ministry included in the sample frame were the Departmental Employees (those registered as such, not the staff on loan from local schools); and the directors of each school in the sample frame. The total number of employees or posts finally included in the final sample frame was 43,702 103 (about 80 percent of the total of the secretariat's employees or posts), who were located in 9,159 workplaces or educational centers (Table 3.2). Table 3.2 - Sample Frame, Education Cype of employment Central Ministry Departments Primary Secondary otal Non teaching 671 117 1,910 1,354 4,052 (16.6) (2.9) (47.1) (33.4) (100) Primary teacher - - 27,218 12,432 39,650 - - (68.6) (31.4) 100 Secondary teacher Ciclo Comdn/Bdsica - - - 5,107 - ComerciolBachiller Computacidn - 2,652 - Polivalente/Bach. Agric./Bach. Elict. - 3,326 - Bachillerato/Normal/Artisticas - 1,347 - Sub total 12,432 Total jobs 671 117 29,128 13,786 43,702 (1.5) (0.3) (66.7 (31.5 (100) Total centers I 14 8,444 70 9,159 (0.0) (0.2) (92.2) (7.6 100 Source - SEP databases, Sept 2000. Primary excludes preschool teachers and Secondary excludes staff assigned to distance learning. Sample selection A total sample was designed of 1,465 staff nationwide, 805 for health and 660 for education. The sample was drawn from a representative group of 35 workplaces in health and 44 in education (giving an average number of interviews per workplace of 23 in the health sector and 15 in the education sector). Fieldwork was carried out in September and October 2000. Health The first step was to select 35 workplaces, stratified as follows: the Ministry; 2 regional offices; 3 national hospitals, 2 regional hospitals, 3 area hospitals, 12 CESAMOS (urban health posts7o), 10 CESARS (rural health posts) and 2 CLIPERS (peripheral clinics in Tegucigalpa and San Pedro Sula). In the second step, a sample was selected of 115 of each sort of employee. The same number of each type was studied in order to facilitate comparisons between the groups and to generate similar confidence levels for the inferences about each group. The sample was distributed by workplace type according to the distribution observed in the universe. For example, 62 percent of specialist physicians in Honduras work in national hospitals, so 62 percent of the sample of specialist physicians was drawn from this stratum. 70 Mother and Child clinics, Clinicas Materno Infantiles (CMI), were grouped with CESAMOS for sampling purposes. 104 Similarly, 13 percent of auxiliary nurses work in CESARS, so 13% of the sample of auxiliary nurses was drawn from CESARS." Table 3.3 summarizes the designed and implemented samples. Education The first step was to select 44 workplaces, stratified as follows: the ministry, 3 departmental offices, 20 primary schools and 20 secondary schools. The workplaces were ordered geographically. In the second step the following numbers of staff were selected in each type of workplace: 140 non-teaching staff (including 80 in the central ministry, 20 in Departments, 20 head teachers in primary schools and 20 in secondary schools2); 200 primary teachers, and 320 secondary teachers. The latter were further sub- divided into four sub-groups with a sample of 80 teachers in each group (Ciclo Comtin, Comercio, Polivaliente y Bachillerato). These sample sizes were chosen bearing in mind the expected variance within the different sub-strata. For example, the larger total sample size for secondary school teachers, compared with primary teachers, reflects the greater complexity of types of institution in the secondary level. The sampling at workplace level was less complex than for the case of the health ministry, because all employees of a given type are concentrated in a particular stratum. For example, all the primary teachers are working in primary schools, and all the secondary teachers are working in secondary schools. All employees of the chosen type within the group of centers pre-selected in each stratum were listed and the sample was selected using systematic random procedures. Table 3.4 summarizes the designed sample and the implemented sample. 71 This rule was followed as closely as possible, subject to the limitations posed by the need to select whole numbers of interviewees and subject also to the availability of sufficient numbers of the type of employee in question in the sampled workplaces. 72Ancillary staff such as cleaners, security guards and secretaries were not included in the sample frame. 105 Table 3.3 - Health sample Nat. Reg. Central Regions hosp. hosp. Area hosp. CESAMO CESAR CLIPER Total 0/ 0/ 0/ %/ 0/ 0/ % %/ 0/ Designed sample Specialist physician 6 3 71 16 15 3 0 1 115 (5.2) (2.6) (61.7) (13.9) (13.0) (2.6) (0.0) (0.9) (100) General physician 4 14 17 14 22 41 0 3 115 (3.5) (12.2) (14.8) (12.2) (19.1) (35.7) (0.0) (2.6) (100) Prof Nurse 4 11 44 14 22 16 1 3 115 (3.5) (9.6) (38.3) (12.2) (19.1) (13.9) (0.9) (2.6) (100) Auxiliary nurse 0 4 40 15 22 19 14 1 115 (0.0) (3.5) (34.8) (13.0) (19.1) (16.5) (12.2) (0.9) (100) Promoters 10 26 8 1 15 49 6 0 115 (8.7) (22.6) (7.0) (0.9) (13.0) (42.6) (5.2) (0.0) (100) Technicians 9 15 38 12 20 20 0 1 115 (7.8) (13.0) (33.0) (10.4) (17.4) (17.4) (0.0) (0.9) (100) Non medical (ancillary) 13 12 33 12 29 12 1 3 115 (11.3) (10.4) (28.7) (10.4) (25.2) (10.4) (0.9) (2.6) (100) Total employees 46 85 251 84 145 160 22 12 805 (5.7) (10.6) (31.2) (10.4) (18.0) (19.9) (2.7) (1.5) (100) Total centers 1 2 3 2 3 12 10 2 35 (2.9) (5.7) (8.6) (5.7) (8.6) (34.3) (28.6) (5.7) (100) Implemented sample Specialist physician 6 2 70 16 14 3 0 1 112 (5.4) (1.8) (62.5) (14.3) (12.5) (2.7) (0.0) (0.9) (100) General physician 4 14 17 14 22 41 0 3 115 (3.5) (12.2) (14.8) (12.2) (19.1) (35.7) (0.0) (2.6) (100) Prof. Nurse 4 11 43 14 22 16 1 3 114 (3.5) (9.6) (37.7) (12.3) (19.3) (14.0) (0.9) (2.6) (100) Auxiliary nurse 0 4 40 15 22 19 14 1 115 (0.0) (3.5) (34.8) (13.0) (19.1) (16.5) (12.2) (0.9) (100) Promoters 10 26 8 1 15 49 6 0 115 (8.7) (22.6) (7.0) (0.9) (13.0) (42.6) (5.2) (0.0) (100) Technicians 9 15 38 12 20 20 0 1 115 (7.8) (13.0) (33.0) (10.4) (17.4) (17.4) (0.0) (0.9) (100) Non medical (ancill.) 13 12 33 12 29 12 1 3 115 (11.3) (10.4) (28.7) (10.4) (25.2) (10.4) (0.9) (2.6) (100) Total employees 46 84 249 84 144 160 22 12 801 (5.7) (10.5) (31.1) (10.5) (18.0) (20.0) (2.7) (1.5) (100) 106 Table 3.4 - Education Sample Central Ministry epartments Primary Secondary kotal X /X Designed sample Non teaching staff 80 20 20 20 140 (57.1) (14.3) (14.3) (14.3) (100) Teachers - - 200 320 520 - - (38.5) (61.5) (100) Ciclo Comzin/Bdsica - - - 80 - Comercio/Bachiller Computaci6n 80 - Polivalente/Bach. Agric./Bach. Eldct. - 80 - Bachillerato/Normal/Artisticas 80 - Total designed sample of staff 80 20 220 340 660 (12.1) (3.0) (33.3) (51.5) (100) Total designed sample of centers 1 3 20 20 44 (2.3) (6.8) (45.5) (45.5) (100) Implemented sample Non teaching staff 77 20 20 20 137 (56.2) (14.6) (14.6) (14.6) (100) Teachers 200 290 490 (40.8) (59.2) (100) Ciclo Comrin/Bdsica 79 Comerciol Bachiller Computaci6n 73 Polivalente/Bach. Agric./Bach. Eldct. 71 Bachillerato/Normal/Artisticas 67 Total implemented sample of staff 77 20 220 310 627 (12.3) (3.2) (35.1) (49.4) (100) Note: The 20 non-teachers in primary and secondary schools are self-selected as the directors of the schools in the sample frame. Ancillary staff were not sampled outside the Central Ministry and Departments 107 Expansion factors The findings from the surveys that are presented in this report are estimated values for the relevant study universes and sub-universes, based on the field data together with the appropriate weightings or expansion factors which reflect the (inverse) probability of selection in each stratum of the sample. Questionnaire forms The questionnaire in each type of institution had two sections. The first is an institutional questionnaire, directed to the director or manager responsible for the center, and the second is an individual questionnaire, directed to the employee who was selected in the sample. Copies of the questionnaire forms are in Annex 3. In the institutional questionnaire, a register was taken of the staff of the center, detailing their contractual status and comparing the total for each sort of employee with (a) the number of posts authorized for the center, according to the national budget and (b) the number of posts that in the opinion of the interviewee was needed for the center to operate well. For each individual in the survey sample, the survey established: (a) if they are known to the manager; (b) if are working at the center at present; (c) if not, whether they ever did work here, when they left this center, and where they work now (d) when they started work here, (e) their post, (f) their contractual status, (g) their salary and (h) their (contracted) number of hours per day. The attendance register was then studied to determine, for each person in the sample who really does work at the center, what was their attendance at work over the last week and the motive for any absences". The individual questionnaire is directed to the sampled employees who turned out really to work in the sampled workplace. For each individual, the following data were recorded: (a) are they present today? (if not, where possible, the rest of the data is gathered from a colleague who knows them well), (b) their education (c) present functions (d) date of starting work in the secretariat (e) how they got their job (f) contractual status now (g) contractual status on initial entry to the ministry (h) time on contracts before getting an acuerdo (i) hours of work - both formally assigned and really worked at present (j) salary (j) other jobs and the earnings they generate (k) job satisfaction (1) a detailed register of their activities on the previous day. 73 By analyzing absences over the previous five days, the study team sought to reduce possible distortions from staff turning up especially on the day of the survey to "window dress" the institution's performance. In any case, although Directors were informed that the survey would visit their institution, they were not told the exact date, further reducing the probability of inducing atypical attendance patterns. Therefore, the incidence of absences reported by the survey is likely to be a good reflection of the real patterns. 108 ANNEX 4 HONDURAS: TRACKING POVERTY-REDUCING EXPENDITURE ASSESSMENT Assessment of Tracking Capacity' Involvement of Different Levels of Government Less than 10 percent of poverty-related spending is currently carried out by subnational levels of government. Although the PRSP (which will define the vehicles to channel HIPC assistance to the social sectors) is not completed yet, there are no plans for or expectations that HIPC funds would flow differently. The subsequent discussion applies to central government only. Budget Formulation Comprehensiveness Coverage of the budget There is a close fit between the government sector in Honduras and the GFS definition of general government. However, the budget does not provide a comprehensive picture of central government revenue, expenditure, and the deficit. The budget includes the transfers to the deconcentrated entities and the decentralized agencies. But the budgets of the deconcentrated agencies are not sent to Congress for approval and are not made public. Those of the decentralized agencies are usually approved late in the budget year. Retained fees and user charges by the Ministries of Health, Education, and Agriculture are not included. Inclusion of donor funds. Donor activities are included in the budget with the exception of donor support to local governments-substantial-and donor assistance in-kind where information is not always available. Both capital and current donor funds are included in the budget and are supported by outturn data provided by executing agencies to the Ministry of Finance (SEFIN) in a timely way and in harmony with the rest of budget reports. Donor outturn at the local government level is not known. Classification Standard classification. The central government has administrative, economic, functional, and program classifications, although the functional and program dimensions are flawed. Functional classification is too aggregated (22 categories in the 2001 budget), reflects principally the administrative classification, does not fit international standards, and does not have a relationship to any program classification. Unclear spending categories such as global allocations and centralized services-used for both budget formulation and budget execution-undermine the transparency of the budget classification. Poverty-reducing expenditure classificatio. The authorities are proposing the creation of a virtual poverty fund (PF), as an institutional mechanism for channeling-among others-HIPC-related funds to priority programs identified in the PRSP. The PF will be included in the annual central government budget and supervised by a national council chaired by the SEFIN and composed of members of the government, civil society, and bilateral donors. The institutional mechanism has already been incorporated in the 2001 Budget, but the draft of the PF law has not been presented to Congress yet. In the 2001 budget, the PF is defined as a new source of financing" along with other traditional domestic and external financing sources (Box 1). Some of the poverty-related programs financed by the PF will be carried out by the most important autonomous agencies. These agencies have differing expenditure management systems but, in general, have the capacity to properly track these projects. Among them are 74 Compiled by: Dorotinsky and Poverty-Reducing Expenditures Tracking mission team. 7 The PF is identified by the code 29. 109 decentralized entities like FHIS (Honduran Social Investment Fund) and PRAF (Family Assistance Program), and non-financial public enterprises like EENE (National Electricity Company), and SANAA (Water Supply and Sewerage Company). The budget registers only the transfer from the PF to these entities. Box 1. Honduras: The Poverty Fund in the 2001 Budget * The PF in 2001 budget includes 400 million lempiras. In the budget, this figure is allocated to four institutions: Ministry of Education (229.3), Ministry of Health (63.4), Ministry of Agriculture (4) and Centralized Services (103.3). * The 2001 budget presents a summary of each institution's expenditure by source of financing and economic category. The Ministry of Education is given as an example. Ministry of Education Summary of expenditure by source of financing and type of expenditure (million lempiras (ML)) Type of expenditure (code) Domestic Sources PF External Financing (including PF) Wages and salaries (100) 4070.8 119.7 150.1 Non-personal services (200) 59.8 15.0 38.4 Goods (300) 49.3 6.2 22.7 Capital expenditure (400) 74.6 1.8 9.5 Transfers (500) 1160.0 71.3 91.2 Debt service (700) 1.5 0.0 0.0 Global allocations (900) 8.6 15.0 15.0 Total 5425.0 229.3 327.1 * PF resources received by Ministries are either directly executed by them, transferred as grants to deconcentrated or decentralized agencies, or included as global allocations. When executed directly,, the budget is transparent because it identifies in detail the program, subprogram, region and type of expenditure that is being financed by the PF. When the resources are transferred by Minsitrres to anoter agency, however, the budget only reports on the receiving agency and the program but it does not provide any additional information. A couple of programs from the Ministry of Education are provided as illustration. Ministry of Education Expenditure by program, subprogram, activity and type of expenditure Program Subprogram Activity/ Region Type expenditure Benefiaciary Amount(ML) Basic education (12) Schools (02) Morazan region (08) Basic salaries(1 11) 5.4 Paper (331) 0.01 Teaching supplies(437) 0.15 Subtotal 5.6 Education assistance(16) PROHECO(04) Coordination (01) Capital transfer (542) FHIS (schools) 3.5 * The resources allocated to centralized services includes transfers to PRAF (3.3) and to a housing program (100). It's not possible to know these allocations in detail. 110 Multi-year projections The budget documents include the spending priorities and the main policy lines of the central administration. Medium-term projections are endorsed by Cabinet and provide the starting point for the annual budget process. However, these estimates are low-quality, and at a somewhat too aggregated level, focusing on the main fiscal variables: revenue, expenditure, and deficit. The multiyear forecast is not based on the ongoing costs of current policies. Budget Execution Internal control Currently, Honduras does not have either visa or warrant controls, although for some pilot ministries they are moving to warrant controls. Commitment controls are in place through quarterly allotment of funds by the SEFIN to ministries. There are very few or no expenditure arrears. Supplementary budgets are used to modify the pattern of government spending. A unique internal control system needs to be established to increase its effectiveness and to reflect a clear allocation of responsibilities. Legally, there are two different internal control systems. The first, named "pre-intervention," is exerted by officers of the General Budget Directorate within the ministries. The second is carried out by auditors appointed by the General Comptroller's Office (GCO). Both types of internal control focus on the legality of procedures and availability of financial resources. In-year tracking of expenditure The integrated financial management system (SIAFI) that is being adopted produces reports based on accounting data. In-year spending reports are available on a functional and program basis, but these are not consistent with chart of accounts. There is no routine reconciliation of government bank accounts with the general ledger. Budget Reporting Regularity of reports Spending unit expenditure tracking reports are received monthly between two-four weeks of the end of the reporting period. The SIAFI will reduce this to less than two weeks. However, the system is operative in only four Ministries. Tracking reports are routinely supplemented with information on commitments. There is progress with the preparation of follow-up and final reports. Last year, the SEFIN started reporting to Congress-on a quarterly basis-on the budget execution of the central government. Information on the budget execution of autonomous entities is not made public. Final audited accounts The General Accounting Office prepares final annual reports in about two months, following the closing of the fiscal year. These reports are quite comprehensive because they include decentralized entities and Ill decentralized agencies. However, they are not prepared from audited information and are not made public. As for external audit, there has never been a complete audited report on the financial situation of the public sector. The GCO has a dual role (internal and external control) that may be generating a conflict of interest because the GCO becomes the audited and the auditor at the same time. Besides, it seems the GCO lacks both financial and human resources to fulfill its functions. The GCO reports are not published and there is no mechanism by which to enforce its recommendations. Budget Evaluation The Bank is supporting the implementation of public expenditure tracking survey (PETS) in the health and education sectors that would support the monitoring of the quality and efficiency of resources. The UNDP has also been supporting some efforts to assess the impact of poverty spending at a program level through targeted evaluations. Capacity for Change The authorities have made a strong effort to improve budget management over the past three years assisted as follows: * The Bank's Economic and Financial Management TA operation has components that would support: (i) the implementation of a medium term expenditure framework; (ii) the expansion of the SIAFI beyond the four Ministries where it is already functioning; (iii) the instituional strengthening of the GCO; and (iv) the devlepopment of performance evaluation capacity. The Bank is also in the process of completing a Public Expenditure Review (PER). The PER has made use of PETS in health and education which as noted above will support the monitoring of the quality and efficiency of resources. * The IMF is assisting the government in developing a compliance action plan on fiscal transparency. * Other assistance comprises the IDB on procurement reform, ASDI on the creation of the National Institute of Statistics, USAID on strengthening the GCO, PAHO on budget management of the health sector, and the UNDP on poverty spending incidence tracking. Action Plan to Upgrade PEM Capacity to Track Poverty-Related Expenditure Strategy Based on the revised assessment, the mission suggested a three-year action plan to upgrade the public expenditure management capacity of Honduras to track PRPE. In particular, the mission defined four broad areas that require special attention and the extent to which these areas are being addressed by government efforts or by current or projected TAs from international organizations or donors. The mission suggested what additional TAs would be needed to comply with the basic PEM requirements over a three-year period. The four broad areas defined by the draft AP to be strengthened are: The medium-term planning capacity The main need in this area is the implementation of a Medium-Term Expenditure Framework (MTEF) that should contain initially at least estimates of the resources available over the medium-term and the costs of continuing government policies, with an appropriate dissaggregation of poverty reducing 112 programs. The authorities still need to identify other PRPE programs financed by sources other than the proposed Poverty Fund. This action is important to establish a baseline to track the level and composition of overall government spending on poverty-related programs. The Bank's ongoing Economic and Financial Management Technical Assistance operation will support capacity building for the implementation of the MTEF through the strengthening of the SEFIN's Planning and Evaluation Unit. The operation will also support sound planning, monitoring, and evaluation capacity in line ministries, which will provide a sound basis for the performance budgeting aspect of the MTEF. The classification and comprehensiveness of the budget There are no current government efforts in this direction. The mission advised the authorities on revisions to the budget classification to improve the functional classification and eliminate existing-unclear spending categories such as global allocations and centralized services. Priority should be given to classifying all poverty-reducing programs. The mission pointed to the importance of a clear regulation concerning the maximum size and possible use of this fund. A provider of a technical assistance in these areas needs to be defined. To improve the comprehensiveness of the budget, the mission suggested to the authorities the simultaneous presentation of the budgets of the central government and autonomous agencies (deconcentrated entities and decentralized agencies) to the Legislature. This action does not need specific TA. It requires a reprogramming of the budget preparation process in order to: inform the autonomous entities in a timely way of the transfers they will receive from the budget; and speed up the process of issuing the compulsory report by the SEFIN on the budget of these entities. The reports evaluate how these budgets adjust to the objectives and priorities of national fiscal policy. An appropriate tracking of the 2001 HIPC spending requires: first, inclusion in the budgets of autonomous agencies executing programs financed by the PF, in similar detail to that provided for programs directly executed by ministries (program, subprogram, activity, type of expenditure): and second, publication of the budget of these entities. The internal and external control systems There are two issues in this area: one is a redefinition of the functions and responsibilities of each type of control: the other is the creation of sufficient institutional capacity to perform them. On these issues, the mission recommended assigning internal control functions exclusively to the Executive and focusing the GCO on external control. The authorities do not have a plan for allocating the responsibilities on internal control, or what kind of controls should be implemented. In the short-run the authorities will need to generalize an internal control mechanism like the preintervention, in all ministries and agencies executing programs financed by the PF. There are, at present, no plans to provide TA on this issue. Given the relatively long time span required to fully implement an internal control system, the need for assistance in this area is urgent. However, this is a sensitive issue and the authorities first need to reach internal agreement on the appropriate type of reform. Closely related to the control issue is the lack of a real treasury system capable of monitoring receipts and payments, implementing cash management, to ensure the timely delivery of resources for payments, and providing the basis for reconciliation between bank accounts and accounting data. The project on the Financial Management Law provides the basis for a legal solution. The authorities are in the process of contracting private assistance to address this topic. The mission considers that TA may be required in this regard. 113 Concerning external control, the GCO is already in a process of strengthening this area, with support from the Bank (Economic and Financial Management program) and USAID (Strengthening of the CGO). The mission did not have the opportunity to investigate whether or not these two projects are, in fact, closely coordinated. Reporting In this regard, the mission advised the authorities that the main step needed to expedite a a reliable and timely reporting system is to proceed with the current plan to make the SIAFI operative in all ministries by the end of the year. Priority should be given to integrating into the system those Ministries responsible for the execution of PRPE programs. The mission also suggested to include in the budget execution quarterly reports sent to Congress, information on decentralized entities, as well as a breakdown following the administrative, economic, and functional classification. In particular, the quarterly report should include expenditures of autonomous agencies executing activities financed by the PF. Routine reconciliation of the fiscal and monetary data is also required: this area could be evaluated initially by the Fund desk economist. 114 ANNEX 5 AN ASSESSMENT OF BUDGETING SYSTEMS: HONDURAS76 PART A: QUESTIONS 1. INVOLVEMENT OF DIFFERENT LEVELS OF GOVERNMENT 1. What amount of total government poverty reducing spending is carried out by subnational levels of government? a. Is it 0-10 percent of total? b. Is it 10 percent to 50 percent? c. More than 50 percent? 2. What amount of expected additional spending (following receipt of HIPC debt relief) will be executed by subnational levels of government? a. Is it 0-20 percent? b. Is it 20 percent to 50 percent? c. More than 50 percent? IL BUDGET FORMULATION A. Comprehensiveness 3. Does the budget reporting entity match the GFS definition of the general government sector? a. Identical b. Very close fit c. Considerable differences 4. How far are general government activities funded through defined extra budgetary funds? a. Insubstantial (estimate level as a share of total spending) b. Substantial (estimate level as a share of total spending) c. Very substantial (estimate level as a share of total spending) 5. Are non-tax receipts or other non-budgeted funds available to spending units? a. None b. Likely to be some, but not very material (estimate level as a share of total receipts) c. Likely to be material (estimate level as a share of total receipts) 6. How would you describe the budget outturn relative to the originating budget? a. Very close b. Quite close c. Not close Please explain if the outturn is managed by allowing substantial expenditure outside of the budget recording system. Bold answers apply to Honduras. 115 7. Are donor funds included in the budget? a. All b. Incomplete c. None 8. Are outturn data provided on expenditures financed by donors? a. All b. Incomplete, but on some other basis (describe) c. None 9. Are outturn data on such expenditures timely? a. More timely than other budget reporting b. In harmony with the rest of budget reporting c. Later than the rest of budget reporting B. Classification 10. Indicate the levels of classification that apply to budget transactions? a. Administrative, economic, functional, and programmatic b. Adm./ec./fun. or Adm./ec/prog c. Other 11. To what level does the functional classification break down? a. Sub-subsector level (more than 80 functional categories) b. Subsector level (between 20 and 80 functional categories) c. Sector level (less than 20 functional categories) 12. Are there any (additional) classification or institutional systems to record HIPC related or some definition of poverty reducing spending? a. Use of a virtual poverty fund, or existing classification system (detailed) b. Use of an actual poverty fund c. Other - yet to be determined. C. Multiyear Projections 13. How would you describe the quality of the out year estimates (medium term) for spending? a. Integrated into the budget formulation cycle b. It's not integrated into the budget formulation cycle c. None 14. What level of detail is provided for such estimates? a. Same level of detail as general classification system b. A level of detail to allow coding to HIPC- related spending, per question 12 above or through other means c. A level too aggregated to consider the functional additivity of the HIPC funds 116 III. BUDGET EXECUTION A. Internal Control 15. What are the principal forms of control to ensure that spending reflects the authorized budget? a. Ex ante visa controls or ex ante warrant controls b. Other c. None 16. (i) Do these controls involve or are they supplemented with commitment controls? a. Yes b. Partial c. No (ii) If cash rationing is a feature of the control system, describe the basis of decisions made about which activities should receive cash and which should not? 17. What do you estimate as the level of the current stock of expenditure arrears? Please, describe the level of accumulation of new arrears over past year. a. Very few or none b. Some (up to 5 percent of total expenditure) c. Significant amount (more than 5 percent of total expenditure) 18. Are supplementary budgets used to add to or substantially change the pattern of government spending? a. Yes, but for small (less than 2 percent of total spending) amounts b. Yes, and for large (more than 2 percent of total spending) amounts c. No 19. How would you describe the internal audit function? a. Effective - run by ministry of finance or by spending unit b. Partial - run only by some spending units c. None 20. Is internal control supplemented by public expenditure tracking surveys that follow funds to the ultimate service provider or beneficiary? a. Yes, regularly b. Yes, sporadically c. No B. In year tracking of expenditures 21. What in-year reports are available for tracking budget performance? a. Functional classification is reflected in the chart of accounts b. There are inconsistencies between functional classification and the chart of accounts c. There is no tracking on a functional basis 22. Is there regular reconciliation of all government bank accounts (those held in the central bank and the commercial banks) with the government's ledger records? a. It occurs in a routine way 117 b. It occurs in an untimely way c. It does not occur 23. In (monthly or otherwise) reports of the consolidated operations of government, how large is the discrepancy between revenue minus expenditures (derived from MOF data) and financing items (derived from banking data)? a. Less than 0.1 percent of GDP b. Between 0.1 percent and 0.5 percent of GDP c. More than 0.5 percent of GDP IV. BUDGET REPORTING A. Regularity of Reports 24. When are budget tracking reports from line ministries and treasury received? a. They are received within two weeks of the end of the relevant period b. They are received between two weeks and four weeks of the relevant period c. They are received more than four weeks after the end of the relevant period 25. Are the tracking reports supplemented by information on commitments? a. Yes, routinely b. Yes, but not routinely c. No 26. How often does the public and/or the legislature receive budget tracking reports? a. More often than quarterly, within four weeks of the relevant period b. Quarterly, within four weeks of'the relevant period c. Other B. Final Audited Accounts 27. What is the period between the end of the fiscal year and the closure of the 'books'? a. Within two months b. Between two months and six months c. Other 28. How soon after the end of the relevant year are the accounts forwarded to external audit? a. Within six months b. Between six months and one year c. Other 29. How soon after the end of the relevant year are the audited final accounts presented to the public and/or the legislature? a. Within six months b. Between six months and one year c. Other 118 V. BUDGET EVALUATION 30. Are any efforts being mounted to assess the effects of spending on the incidence of poverty? a. Tracking studies of HIPC-related funds or targeted evaluation of sector or other poverty programs b. Broad social indicators of poverty c. None