Colombia - Public sector expenditure review

Report No. 7891 -CO Colombia Public Sector Expenditure Review January 15, 1991 Latin An-erica and the Caribbean Region Country Operations Department: Country Department IlIl FOR OFFICIAL UJS' ONLY 'Document of the World Bank 1his document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disckl.sed without World Bank authorization.  FISCAL YEAR January 1 to December 31 CURRENCY EQUIVALENTS Currency Unit: Peso (Col$) Exchange Rate as of: December 1, 1989 = US$1.0C = Col$424 December 1, 1990 = US$1.00 = Col$557 ACRONYMS BDP Banco de Proyectos CAIP Centro de Atenci6n Integral al Prescolar CAJANAL Caja Nacional de Previsi6n CFT Corporaci6n Financiera de Transporte CONPES Consejo Nacional de Politica Econ6mica y Social CORELCA Corporaci6n Eldctrica del Atlantico CSPF Consejo Superior de Politica Fiscal CVC Corporaci6n Regional del Valle del Cauca DNP Departamento Nacional de Planeaci6n ECOPETROL Empresa Colonbiana de Petr6leos EEEB Empresa de Energia EMPOS Empresas Municipales EPM Empresa Plablica de Medellin FEDESARROLLO Fundaci6n para la Educaci6n Superior y Desarrollo FEN Financiera Electrica Nacional FER Fondos Educativos Regionales FNC Fordo Nacional de Cafe FNCV Fondo Nacional de Caminos Vecinales FVN Fondo Vial Nacional HISP Health Integration Services Project ICBF Instituto Colombiano de Bienestar Familiar ICEL Instituto Colombiano de Energia Electrica IFI Instituto de Fomento Industrial ISA Interconexi6n Electrica, S.A. ISS Instituto de Seguridad Social IVA Impuesto al Valor Agregado JNT Junta Nacional de Tarifas MME Ministerio de Minas y Erergia MS Ministerio de Salud PROEXPO Fondo de Promoci6n de Exportaci6n PSAL Power Sector Acjustment Loan PSAP Power Sector Adjustment Program PSSC Power Sector Superior Council RDES Rentas de Destinaci6n Especifica RSNS Reorganizaci6n del Sistema Nacional de Salud SENA Servicio Nacional de Aprendizaje SSS Servicios de Salud Seccionales  FOR OFFICIAL USE ONLY Table of Contents Page No. EXECUTIVE SUMMARY ................. i CHAPTER I INTRODUCTION. 1 CHAPTER II MANAGEABILITY OF PUBLIC EXPENDITURE .5 A. Administration of Public Expenditures and Efficiency.... 5 B. Inefficient Planning and Budget Process. 5 Contradictory Regulations. 8 Limited Budget Coverage. 8 C. Reforms Undertaken l.1 Organic Law of the Budget l.1 Decentralization .14 Coffee Fund .15 Railroads .16 Procurement Statute .17 Other Initiatives .................................... 17 D. The Need for Further Rrf-iXrms: Administrative Reorganization u. the Public Sector .................. 17 CHAPTER III EARMARKING OF GOVERNMENT REVENUES ........................ 19 A. General Definition ................... 19 B. Trends in the Size and Structure of Earmarking (1970-87) .19 Size ati Structure of Earmarking .19 The Effects of Law 55 (1985) .21 Consideration of Social Security and Public Enterprise Profits .23 C. The Experience of Earmarking .24 The Origins of Earmarking .24 Recommendations of Previous Government Commissions .25 D. An Evaluation of Earmarking .27 Revenue Sharing ............... 27 National Government Level Earmarking ................. 32 Departmental Earmarking .............................. 38 Municipal Level Earmarking .39 Payroll Taxes .41 E. Summary and Recommendations .44 This report is based on the findings of a mission that visited Colombia in October 1988. The missicn was led by Mr. Luis Ramirez and comprised Messrs. L.J. Vasquez, William McCleary, and consultants D. Adkins and Salles. Messrs. A. Covarrubias and A. Raizen (consultant) also contributed to this report, but did not participate in the mission. Ms. Laura Santalla provided secretarial assistance irn the preparation of the entire report; Ms. K. Scalzulli asisted in the final stages. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - 2 - CHAPTER IV POWER SECTOR .......................................... 49 A. Overview .......................................... 49 13. Organization and Institutions .......................... 50 C. Power Demand .......................................... 54 D. Demand Forecast ........................................ 55 E. Power Supply .......................................... 56 F. Electricity Losses ..................................... 57 G. Sector Planning ........................................ 58 H. Investment Program ..................................... 60 Period 1989-1992....................................60 Generation .......................................... 60 Transmission ......................................... 61 Subtransmission and Distribution ..................... 62 Studies and Training ................................. 62 Investment Program 1993-1996 ......................... 63 Financing of the Investment Program .................. 64 I. Recommendations ........................................ 68 CHAPTER V HEALTH ..........................................' 0 A. Health Conditionq ........................ 70 The Health System ..... .............................. 71 B. Structure and Strategy of the Health System ............ 72 Coverage ............. ............................... 72 The Government System ................................ 72 The Social Security Subsector .74 Integration of the Public .ealth Sector .76 C. The Government's Program Goals Compared to Implementation ........................................ 77 The "Basic Health for All' Program ................... 77 Actual Performance vs. Planned Expenditure Targets ... 78 Actual Performance vs. Planned Financing Targets ..... 79 Planning for Basic Health Expansion .................. 80 D. The Financing of Expanded Health Care .................. 81 Earmarked Taxes and Revenues in Health ............... 81 The Financing of Expanded Health Care ................ 82 E. Impact of Past Deficiencies in the Budgetary Process on Health Programs ................... 86 F. Conclusions ................... 88  - 3 - CHAPTER VI EDUCATION ................................................ 91 A. The Educational System ................................. 91 Resource Mobilization for Education in Colombia ...... 93 B. The Government's Primary Education Goals Compared to Implenentation .93 The Gap in Primary Education .93 The Escuela Nueva Program .94 The "Basic Education for All" Program .95 Actual Performance vs. Planned Expenditure Targets .96 Actual Performance vs. Planned Financing Targets . 97 C. The Government's New Investment Program for Primary Education ..................................... 98 D. Domestic Sources of Finance in Education ............... 98 Ordinary Budget Appropriations and the Situado Fiscal ....................................... 99 Earmarked Taxes and Other Earmarked Revenues ......... 100 Reform of the Education System ....................... 101 Cost Recovery and Privatization in Higher Education .......................................... 101 E. Conclusions .......... . 104 ANNEX I ...................................................... .106 STATISTICAL APPENDIX  - i - Public Sector Expenditure Review Executive Summary 1. Colombia's low level of productivity growth is limiting its capacity for long term economic development. To use domestic savings and net foreign inflows productively the country will have to improve substantially the overall productivity of investment. As suggested in the Country Economic 'Memorandum on Colombia, private sector savings and capital productivity can be enhan ed throuigh a variety of measures.1 In contrast, this report focuses on the capacity to increase savings and to improve the efficiency of expenditures in the public sector. Improving the efficiency of resources in the public se-tor will require significant reforms ir public administration. The Government of Colombia is aware of the neeci for these reforms and has expressed serious intentions to undertake them. Important measures have been taken in the area of taxation and of deficit control. However, management of public expenditures continues to be weak. Conscious of this weakness, the Barco administration, after several years of negotiations, was successful in securing Congressi-onal approval of a new Organic Law of the Budget. This important piece of legislation has the potential to improve significantly public sector management and to reduce the inefficiency of expenditures. The Executive begun reorganizing the administration of the public sector to implement the new law. Administrative reorganization and implementation of the Budget Law are now the key issues. 2. The primary goal of public sector expenditure reviews (PSER's) undertaken by the Bank is to provide an objective analysis of and hence recommendations to governments on how to increase efficiency by altering the level and composition of public expenditures. In Colombia, the size and composition of public expenditures has been difficult to control. The Central Government has traditionally had little power and few legal or institutional mechanisms to reallocate expenditures as needed. The primary goal of this PSER, therefore, is to analyze: (i) the degree of this lack of control over public expenditures, (ii) the progress achieved by the Colombian authorities towards increasing control and improving management of public expenditures, and (iii) the main reforms required to improve further the management of public expenditures. It is argued that the lack of control over the size and composition of public expenditures in Colombia is mainly due to weaknesses in planning and budgeting, and to the high proportion of expenditures financed with earmarked taxes. Thus, expenditures committed in past budget periods, without due concern for long term planning, have left little room to increase expenditures in priority areas like Health and Education and have overburdened the budget with expenditures in areas like the Pot.e.r Sector. 1/ See Conclusions of the Country Economic Memorandum on Colombia, Report No. 7629-CO. World Bank, September 15, 1989.  - ii - Planning and budgeting 3. Most of the inflexibility of public expenditures in Colombia has been due to weakneF es in the planning and budgeting process, and to a relatively high degree of expenditures financed with earmarked taxes. There is no planning mechanism to ensure coordination between current and capital expenditures, leaving investment as a volatile residual after covering recurrent expenditures. Investment projects have been approved largely on the basis of short term financing availability, without due concern for long term financing requirements and for the long run productivity of such investments. These projects sometimes have originated through ad hoc procedures, committinig resources of future budgets, and limiting flexibility in the use of available funds. The National Planning Department does not have a complete record of individual project profiles,2 and disconnected investment decisions are often made. Earmarked taxation is quite prevalent to finance both current and capital expenditures, and can only be changed by a lengthy and difficult legislative process. Finally, the Central Government has limited power over expenditure decisions of financial and non-financial national and departmental decentralized public enterprises, of local governments, of the Coffee Fund, and of the quasi-fiscal expenditures of the Banco de la Repiblica. These characteristics of Colombia's planning and budgeting process have had, in many cases, a negative impact on the productivity of public expenditures. 4. The productivity of public expenditures is also affected by an interrelated system of laws that contradict themselves on many aspects (planning, budgeting, procurement, personnel, organization, credit, and earmarking). Historically, the legal blueprints on each of these areas evolved independently, giving rise to a legal system with contradictions and inopholes that has adversely influenced the efficiency of public expenditures. For example, budgeting laws do not allow for estimates of contractually-agreed price increases while procurement laws allow them. Budgeting rules limit payments to cash availabilities, while contracts, according to procurement laws, include agreed payment flows. Budgeting laws explicitly forbid new earmarking, while new earmarking laws have continuously been enacted. Budgeting laws do not allow for credits not yet legally confirmed to be included in the budget, while credit laws may have already authorized them. And, finally, according to organizational laws, sector ministries have major planning and budgeting functions, but budgeting laws give to individual public entities the major responsibility foz preparing the budget. 5. These contradictions and loopholes in the legal system governing public expenditures generate inefficiencies in most of the institutions controlled by the Organic Law of the Budget (central government, decentralized entities and public enterprises). There are institutions that have escaped these contradictions and loopholes because they are not / A 'Banco de Proyectos" has been created as part of the mandate of the new Organic law of the Budget. Investment of public enterprises is in the process of being classified project by project to generate project profiles.  - iii - covered by '1he budget even when they do manage public expenditures (the quasi-fiscal operations of the Central Bank, the National Coffee Fund, and a few other entities like the "Cajas de Compensaci6n'). In these institutions, however, inefficiencies of expenditures stem more from not being subject :o the scrutiny of the budget process. Earmarking 6. The pervasive earmarking practices in Colombia are mainly the product of the last two decades. Whereas 26Z of income at all levels of government was earmarked in 1970, this figure reached a peak of 44Z in 1984. Since then, the Government has reduced it to about 38Z. These figures, however, understate the degree of inflexibility of government expenditures because they do not include after-tax operating incomes of non-financial public enterprises--which are like earmarking of funids since they are controlled by the enterprises. The same may be said of payroll taxes for social security, for which the gcvernment has no discretion over their use. The extent of earmarking in this expanded sense reaches about 55Z of total public income. The major items are, in descending order of importance: public enterprise profits (27Z), payroll taxes (24Z)--to cover expenditures for Social Security, for the Instituto Colombiano de Seguridad Social (ICBF), and for the Servicio Nacional de Aprendizaje (SENA)--, revenue sharing (20%), departmental alcohol/tobacco/gambling taxes (11Z), gasoline tax (7Z), import duties for PROEXPO (4%), the municipal valorization tax (2Z), and the ad valorem tax on coffee exports (2%). Together these eight sources account for 97Z of total earmarking. 7. Earmarking in Colombia is too complex to allow blanket recommen- dations for reform. Instead, we must limit ourselves to piecemeal suggestions based on considerations of whether earmarking is generating the appropriate amount of resources for the activity in question, whether these resources appear to be utilized effectively, and whether benefits are received by those who pay--benefit principle of taxation. With this view in mind, it is recommended that: First, the existing revenue sharing arrangements be strengthened since they are an important part of the decentralization effort. This can be done by a) eliminating certain anomalies in the sharing formulas;3 b) introducing more incentives for efforts to generate resources at the local level, and c) improving the capacity of the municipios, especially the small ones, to utilize their resources effectively. Second, for those taxes that in general follow the benefit principle, like the coffee export tax, the gasoline tax, and the municipal valorization tax, some modifications might be appropriate to improve efficiency. The coffee fund, which has strayed into activities beyond its original mandate of income stabilization and assistance to the coffee growing areas, should have better government surveillance. This is now possible after the recently signed new coffee contract, but very determined government action will be needed to implement reforms to focus 3/ See Chapter III. A comprehensive discussion of fiscal aspects of decentralization is presented in Decentralizing Revenues and Provision of Services: A Review of Recent ExPerience, World Bank, Report No 7870-CO, October, 1989  - iv - activities back to the original mandate. The gasoline tax earmarked for highways, insofar as it is considered as a tax in lieu of road user charges, has a strong benefit rationale. In Colombia, however, domestic gasoline prices at the consumer level have been mostly equal or lower than CIF import prices, and therefore, road taxes have been paid by ECOPETROL, not by road users. Earmarking is not justified, unless these taxes were modified to be added to the CIF import price of gasoline. In addition, real tax revenues have not kept up with inflation, and with road construction and maintenance needs. Falling real revenues and rising debt service paymo.nts are becoming a serious problem in the transport and in the energy sectors. Third, there are earmarked taxes where it is very questionable whether resource allocations are appropriate and which have little or no benefit rationale at all, like PROEXPO, IFI, Caja Agraria, SENA, ICBF, Cajas de Compensaci6n, and departmental alcohol/tobacco/ gambling taxes which finance the health sector. Earmarking of these taxes should be eliminated, forcing the above entities to finance their operations by competing for funds out of the general budget. An additional factor in the case of payroll taxes is their adverse employment effect and their regressiveness (the payroll tax is equivalent to an income tax where the tax rate decreases as income increases). Total payroll tax rate varies from 24X to 29Z of the wage bill, one of the highest in Latin America (including ICBF, SENA, Cajas de Compensaci6n and Social Security). And fourth, there is a large number of very minor funds which appear questionable and should be examined with a view towards possible reductions, like the Fondos Especiales, Contribuciones y Participaciones, tourism tax and cinema tax. 8. In summary, che number of earmarked funds, and the extent of government monies earmarked, as the planning and budgeting process improves, should be substantially reduced, thus increasing the flexibility of the government budget. Earmarking should be limited to revenue sharing, payroll taxes for social security, and those few cases where benefit connections are clear4 and pricing arrangements appear to generate appropriate levels of resources. Even in these cases, the commitment to earmarking ought not to be open-ended. In each case, a date should be set to review the earmarked financial arrangement, cr to terminate it unless expressly renewed. Power Sector 9. Resources already committed to the Power Sector add to the inflexibility of public expenditures. On average, the power sector's investment program has represented about 24Z of the total public investment program over the last twelve years, and about 25% of the total public debt. The sector needs to continue investing to meet increases in demand estimated at 5.7Z a year:5 US$1.6 billion are needed just to finish ongoing projects during the 1990-92 per.od, and US$408 million to initiate a new least-cost expansion program of about US$2.5 billion for the overall 1990-96 period. The sector, however, has already reached a level of 4/ In the sense of having substantial overlap between beneficiaries and tax payers or buyers of the services. 5/ Present estimates are under review, however, to accommodate a lower than expected growth in GDP in the near future.  indebtedness of about US$5.3 billion, of which US$4.3 billion are foreign debt (about 28% of the total public external debt of the country at the end of 1988). Tariffs, which are collected in local currency, are weIl belc.w long-run marginal costs, and have not been adjusted fast enough to kEep pace either with actual inflation or with the rate of devaluation. The Government of Colombia, however, has recently implemented significant changes to the level and structure of electric tariffs.6 The average rate at national level increased 10.6% in real terms in October 1990, and will continue to increase in real terms until it reaches about 92X of the Average Long Term Incremental cost of producing electricity by 1995 (an accumulated increase of 22.6Z in real terms from June 1990 to June 1995). However, institutional deficiencies, such as the poorly define! role, objectives, functions, and authority of each of the most relevant institutions in the power sector, make it difficult to properly plan activities and to improve financial planning and monitoring. Inefficient management and political control of electricity pricing (in spite of a National Tariff Policy) continue to be a major issue. A National Energy Policy is called for, including clear guidelines on energy pricing, on alternative means to meet demand, on indebtedness and finances of the sector, and on institutional aspects. Health and Education 10. The inflexibility of public expenditures affected the implementation of many economic and social objectives of the Colombian government. The strategy of the Plan de Economia Social's 1987--1990 (PES)--the four-year development plan of the Barco administration--to realign public expenditures in favor of the social sectors could not be fully achieved in spite of being the leading objective of the plan. This was due to (i) general fiscal restraint irposed by sound macroeconomic policies; (ii) inability to transfer resources from earmarked taxes that have been allocated according to old priorities, and that have excessive financing in relation to present needs or in relation to present relative priorities; and (iii) limited inplementation capacity of the social sector. Total financing foL health was supposed to rise by about 65% in real terms during the 1987-90 plan period, and financing for basic health was expected to increase almost four times in real terms over the same period. Instead, total real finance for the public health sector increased by only 1.2Z in 1987 and declined by an estimated 1.1% in 1988. Similarly, a shortfall in educational expenditures under the Plan de Economia Social can be traced to a shortfall in the amount provided by the national budget. 61 Decree 1555 of June 16, 1990; and Resolut-.ons 089,090, and 091 of December 9, 1990.  - vi - In spite of the high priority given by the government to its "Basic Education for All" program as part of its "war on poverty program", the planned allocation of resources did not materialize.7 Conclusions 11. Colombian authorities have long been aware of the need to strengthen the Government's capacity to manage public expenditures. Most notably, the reports of the Bird-Wiesner Commission of 1981 and the Gasto Pablico Commission of 1986 proposed important measures to improve the efficiency of public expenditures. But reforms in Colombia are carried out slowly, preventing sudden shocks to the system, and allowing time to develop consensus. Over the years consensus on many issues in the area of management of public expenditures has developed, and several important reforms finally materialized during the Barco Administration: progress in decentralization, a new coffee contract, a restructuring and privatization of the rai roads, and a new normative law of the budget. While these reforms represent important improvements, further reforms are needed to simplify the complexity of the contracting process and to achieve synchro- nization between planning, budgeting, procurement, personnel, organization, credit, and earmarking laws and regulations. Earmarking arrangements should be substantially reduced, especially if the country wishes to allocate a higher fraction of its resources towards social services for the poorest groups in the population. Implementation of all these reforms will require very significant changes in the present administrative structure to further transfer activities to local governments, to eliminate or reorient central agencies to avoid duplications in the decentralization process, to select public entities for possible privatization, and to strengthen the planning system, especially DNP (Departamento Nacional de Planeaci6n), sector ministries, public enterprises, Ministry of Finance, "Contraloria", and "Procuraduria General de la Naci6n". A comprehensive review of the administration and organization of the public sector is now the main issue for Colombia to increase savings and to improve the efficiency of expenditures in the public sector. 7/ A full evaluation of the "Wer on Poverty Program" is contained in the World Bank's Report on Colombia: Programs and Poverty Alleviation: An Assessrnant of Government Initiatives (No. 7271-CO).  CHAPTER T Introduction 1. The Public Sector Expenditure Review (PSER) is part of a progcam of economic reports that the World Bank prepares periodically to proviae a basis for the dialogue between Colombia and the Bank on macro and sector policies, investment priorities and Bank assistance to the country. This PSER should be read in conjunction with recent related general economic reports that address topics on poverty (Colombia - Social Programs and Poverty Alleviation: An Assessment of Government Initiatives, Report No. 7271-CO, December 2, 1988), macroeconomics (Colombia - Country Economic Memorandum, Report No. 7629-CO, March 31, 1989), and decentralization (Colombia - Decentralizing Revenues and the Provision of Services: A Review of Recent Experience, Report No. 7870-CO, July 6, 1989). Other related sector reports include: On Agriculture, an Irrigation Subsector Review (Report No. 7380-CO, July 14, 1988), an Agricultural Sector Review of Selected Issues (Report No. 7130-CO, October 18, 1988), and a Food Marketing Review (Report No. 8048-CO, November 17, 1989). On Industry, a Commercial Policy Survey (Report No. 7510-CO, December 15, 1989), and an Industrial Competition and Performance Report (Report No. 7921-CO, June 29, 1990). On Energy, an Energy Pricing Study (Green Cover Report No. 8321-CO, December 29, 1989. And on Education, a TechnicaL/Vocational Education & Training Subsector Study (Draft Report No. 7742-CO, April 27, 1989). 2. The Colombia Country Economic Memorandum (Report No.7629-CO) concluded that total factor productivity (TFP) growth has provided little to Colombia's economic growth since the middle of the 1970's. The report suggested that progress on several fronts is needed to improve TFP growth. In particular, to enhance productivity in the private sector, the recommendations follow the lines suggested in the Commercial Policy Survey (Report No. 7510-CO) that liberalizing external trade should be considered an effective instrument. To enhance productivity in the public sector, the recommendations are to strengthen public resource management, stresmline public investment projects, and improve the flexibility of the budget allocation mechanism. These last issues are discussed in this PSER. 3. Colombia's capacity to manage public expenditures has long been known to be weak. The Bird-Wiesner reportl and the Public Expenditure Commission's Report of 19862 made recommendations on several aspects of the administration of public expenditures that needed urgent remedial actions. Many of these aspects were consolidated in the reforms enacted through the new Organic Budget Law approved at the end of 1988. This can be considered a landmark in the history of Public Administration in Colombia and will clearly affect all future administrative reforms. 3/ Eduardo Wiosner and Richard Bird, Finanzas Intergubernamentales in Colombia, Departam.nto Nacional de Planeacidn, 1981. 2/ Comisi6n del Casto PGblico, Informe Final 1988. Ministerio do Hacienda y Cr6dito PGblico.  - 2 - 4. Chirter II of this PSER analyzes the weaknesses of the Colombi,'s budgeting process. the improvements achieved with the new law and the need for further administrative reforms. The budget process modified by the new budget law was old and has long been known for its cumbersome procedures at managing public expenditures. According to the Bird-Wiesner report3 the budgeting system was established after the Kenmnerer mission of 1923 and, in spite of several modifications over the years, the system remained in 1980 ,substantially the same as originally designed". Most modifications had not changed the fundamentals of the budget process; they rather arose from the need to protect the Treasury from fraud and from politically oriented use of public resources. The system plus changes resulted in an unmanageable system of controls. To avoid the comnlexities of these controls, public officials have been induced to develop mechanisms to mrnage expenditures outside the budget, generating a disproportionate growth of expenditures financed with earmarked taxation4 and of expenditures administered by the decentralized sector. In Chapter II we also argue that, in addition to a cumbersome budgeting process, public expendiftures are also regulated by an interrelated system of laws that sometimes contradict themiselves (planning, budgeting, procurement, personnel, organization, and earmarkirng), adding legal confusion and further opportunities for inefficiencies to an already cumbersome system. Some of the problems of budgeting were finally solved by the enactment of the New Organic Law of the Budget at the end of 1988, but many of the benefits of this new legislation could be lost if related contradictory legislation that affects public expenditures is not reviewed and modified. 5. The inefficiencies of the budget process in Colombia have induced many public officials to seek ways to avoid the uncertainties and complexities of the budget to finance their programs. This is one of the teasons, e powerful one, that explains the widespread use of earmarked taxation in Colombia. This earmarked taxation has limited the flexibility of central government managers to such a degree that important present priorities can not be funded because the resources have al:eady been committed via earmarked taxes to old priorities. Since the budgetary process is expected to improve substantially after implementation of the new law, the justification to finance expenditures with earmarked taxes has decreased. 6. Chapter III reviews earmarked taxation in Colombia and proposes reforms that would considerabjy decrease, but not eliminate, earmarked taxes. Among free-market LDCs, Colombia--together with some other South American countries--has long been known for extensive earmarking. Table 1 of the Statistical Appendix--which cites the major instances of earmarking in existence today with the agencies involved, the tax bases and rates,and the purposes to which revenues are put--gives some notion of the breadth of earmarking as currently practiced. It should be clear from the outset that earmarking in Colombia runs the full spectrum of experience. It has earmarking which follows the benefit principle (e.g., the highway funds, 3/ Op. Cit, Finanzas Intergubernamentales in Colombia, Departamento Nacional de Planeaci6n, 1981. 4/ See Chapter III.  -3- the airport funds, the valorization tax). It also has taxes with no clear benefit rationale .e.g., taxes on alcohol, gambling and tobacco used for health, sports, and neighborhood roads). It has payroll taxes used for traditional social security purposes (pensions and medical payments) and also for other purposes-vocational training, family welfare and nutrition. It also has revenue sharing between levels of government and an extensive set of decentralized agencies and enterprises at all the levels of government. 7. Chapter IV of this PSER is devoted to an analysis of public investment in the power sector, which is the dominant part of the Energy Sector. Historically, the Energy sector has committed a large share of total public investment and indebtedness. Energy represents about 20% of total investment in the Plan de Economia Social 1986-89, of which, the power sector is about 12% of the total (Table 1.1). Table 1.1: COLOMBIA - PUBLIC INVESTMENT PLAN DE ECONOMIA SOCIAL 1986-1989 (percent) 1986 1987 1988 1989 1986-89 Health 24.1 24.9 24.3 23.5 24.1 Education 19.5 20.1 19.7 18.4 19.3 Energy 16.2 22.1 20.3 19.8 19.8 (of which, electric) 13.7 (11.9) (11.4) (11.3) (11.8) Sub-total 59.8 67.1 64.3 61.7 63.2 Other Sectors 40.2 32.9 35.7 38.3 36.8 Total Investment 100 100 100 100 100 Total Investment (billion Col$) 842.1 1,035.3 1,426.8 1,984.6 5,288.8 Source: Departamento Nacional de Planeaci6n, Programa Macroecon6mico para 1989, Anexo E., enero de 1989. 8. The inflexibilities of the budget process have had a major impact in seriously limiting the capacity of the Government of Colombia to allocate as many resources as needed to implement the Barco's Plan de Economia Social.5 Since the main emphasis of the plan was to focus on poverty issues, Health and Education represent key sectors from the point of view of public expenditure analysis. About 86% of all social investment of the Plan de Economia Social 1986-1989 was supposed to be spent on Health and Educatior (Table 1.2 below). S/ An evaluation of the plan is contained in Bank's report No 7271-CO, op. cit., Colombia, Social Problems and Poverty allaviation: An assessment of Government initiatives.  - 4 - Table 1.2: COLOMBIA - SOCIAL INVESTMENT PLAN DE ECONOMIA SOCIAL 1986-1989 (Z share) 1986 1987 1988 1989 Social Investment 100.0 100.0 100.0 100.0 Agriculture 3.0 2.8 2.9 2.8 Health and Welfare 49.5 48.0 47.8 48.1 Education 41.3 40.2 38.8 38.0 Justice 5.8 8.8 10.2 10.7 Other Social Expenditures 0.3 0.2 0.3 0.4 Source: Programa Macroecon6mico para 1989, Departamento Nacional de Planeaci6n, enero 24, 1989. 9. Chapters V and VI of this PSER are devot2d to an analysis of public expenditures on Health and Education, respectively. Although it is estimated that about 25Z of the Colombian population does not have access to adequate basic health services, and in spite of Government efforts to increase its financing of health, total public expenditures on health declined somewhat as a percentage of GDP in the period 1985-88, entirely due to a decline in the social se-urity subsector, which is undergoing a serious financial crisis. Public expenditures on education have also suffered and have fallen well behind planned levels in spite of a serious gap in the accessibility of educational opportunity, especially for rural children. 10. LI summary, the three sectors reviewed in detail in this PSER account for about 63Z of total planned investment in the Plan de Economia Social 1986-89. (See Table 1.1).  - 5 - CHAPTER II Manageability of Public Expenditure A. Administration of Public Expenditures and Efficiency 1. Public expenditure efficiencies (or inefficiencies) can be intro- duced at two broad stages of public sector management: resource allocation and expenditure execution. Economists are, in general, preoccupied with the elaboration of policies to improve the efficiency of resource allocation, and have been less concerned with how those policies can be carried out effectively. Efficiency, however, very often fails at the execution stage as much as it fails at the allocation stage. This chapter focuses mainly on inefficiencies generated at the execution stage.1 2. In Colombia, most public expenditure inefficiencies have been created because the allocation and the execution of public expenditures has been governed by: i) inefficient procedures and time-consuming controls established by the budgeting process; ii) procedures established by autonomous entities that escape the regulations of the budget process; and iii) a network of interrelated laws and regulations that sometimes contradict themselves. A new Organic Law of the Budget approved at the end of 1988, if properly implemented, has the potential to solve most of these problems. B. Inefficient Plannina and Budget Process 3. Planning has been extremely weak in Colombia. There has been no planning mechanism to ensure coordination between current and capital expenditures, often leaving investment--which is supposed to be carefully programmed according to planning laws--as a volatile residual that fluctuates with revenues after covering recurrent expenditures. DNP, the planning agency of the central government, is responsible for the investment budget only, while the Ministry of Finance prepares the current expenditure budget. But DNP is informed of investments of the decentralized public sector only when investment is financed with government-guaranteed external credit. Many investments, therefore, are in practice outside of central government control including management of debt, which is often contracted by public enterprises independently, with minimal supervision or control of the central government. 4. The overall organization of the public sector is partly responsible for the wc.akness of plann-..g in Colombia. A comprehensive reform of the organization of the State done in 1968 strengthened the need for planning, but left weaknesses that developed over time into an institutional incapacity to prepare sectoral plans that could be coordinated into a national development plan. Sector m-inistries like Education and Health have been overcrowded most of the time with day-to-day operational activities, with little capacity for developing policies and for elaborating medium or 1/ Allocative efficiency is an absolutely necessary condition, however. Sometimes certain taxes are advocated just because they can be effectively executed, risking the worse of the cases: an economically inefficient allocation that is well executed (i.e., payroll taxes in Colombia).  - 6 - long term planning.2 Other ministries like Economic Development and Agriculture became heads of very large decentralized agencies. These agencies sometimes became stronger that the ministries therselves, eliminating any possibility fDr effective sector planning and control. The result was a very weak planning and policy making capability in most line agencies, including weak project monitoring systems, making it extremely difficult to plan or budget future investment outlays. Due to the weaknesses of the sector ministries, the sector units of DNP found themselves elaborating the sector components of this Government's Plan de Economia Social, but with no capacity or institutional structure for implementing it or for integrating the plan into annual budgets. 5. The budget has been purely an annual short term exercise, not integrated with medium-term planning, and it h,as in practice lost its capacity to allocate resources. After covering recurrent expenditures, which are in general inflexible in the short term, the residual left for investment has also become inflexible: it partly covers wages and payments for internal and external debt service, which have been classified as investment for budgeting purposes. Also, due to the priority given to counterpart resources to attract disbursement of external credit, the degrees of freedom to allocate budget resources left to central government authorities are in practice nil. As a consequence, the government has been unable to allocate as much cash resources as originally planned to programs that were included in its Plan de Economia Social with proclaimed first priority: At the beginning of 1988 it was already known that more than 50% of DRI, 75Z of PNR, and 90% of the Absolute4 Poverty Program planned for the year could not be executed. 6. Budget allocations of current expenditures had become a routinized exercise that simply repeats expenditures of the past with no economic efficiency criteria. Partly, this was the result of a constitutional requirement that the budget may not exceed by more than 10% the previous year's budget. In practice, compliance with the law meant to automatically increase previous expenditures by 10% without questioning. And, when more than 10% increases were needed, public officials had found a way to bypass the limit by creating new institutions, or by separating the responsibilities of one agency into two, creating "new" expenditures. Since inflation has in general been higher than 102, new institutions or public agencies may have been created just to keep real current expenditures the same as in the past. 7. The budget as an allocation device and as an instrument to provide information about future expenditure plans had become irrelevant. A constitutional requirement that the budget be balanced in a formal sense had mandated that expenditures not be higher than the total of legally confirmed financial sources. As credit and other revenue sour:es became legally confirmed through the budget period, additional budgets were required, where no attention was normally paid to the composition of expenditures o. to the initial intention of the legislators. More that fiv-e supplementary budgets were needed recently in a single budgetary period, with discussions about the appropriateness of expenditures carried out only in the initial budget review which had sometimes covered less than 50Z of total expenditures of the period. 2/ See Chapters 4 and S.  - 7 - 8. As a result, it has been almost impossible to evaluate the economic efficiency of budgetary allocations ir. Colombia. In addition to the multiplicity of additional budget allocations added during the budget period, many allocations are not associated with clearly determined projects or programs, but with "partidas", for which no specific output could be associated.3 9. The execution process is also inadequate. The firnancial execution of the budget is complicated, and the contractual procedures to execute it are extremely cumbersome. In order to actually receive the funids allocated by the budget, public agencies need to go through several ex-ante controls and lengthy authorization procedures, with such controls directed at determining the legality of expenditure actions, rather than their economic soundness. First, certification is needed to show that the expenditure item has been included in the 'Acuerdo de Obligaciones", which is a quarterly expenditure resolution that has to be approved by the Cabinet, and is used to distribute expenditures through the year according to seasonal patterns. The same exercise is repeated at the level of "Acuerdo de Gasto', which is a monthly allocation of the budget determined mostly by a delegate of the Ministry of Finance in each public agency. In some cases, the approval of the office of the Comptroller General ;. --,so required as an ex-ante control. Then, an "Acuerdo de Gir. finally authorizes the release of the money. Most of these steps were initially thought of as procedures tc program expenditures over the budget period, not as the legal controls that developed in practice. The bottom line is that having resources budgeted for a given purpose is no guarantee that the resources will be actually given, or if given, that they will be available on time. The money might be available in the Treasury, and budgeted for a given purpose, but still not released due to short term macroeconomic management needs, which vary month to month. Or, the Treasury might be temporarily out of money due to seasonal shortages, and not be able to dcsburse because it is not authorized to borrow in the short term. The final outcome is that government contractors protect themselves from this uncertainty by considerably overinvoicing government bills. 10. The execution of expenditures is also adversely affected by cumbersome contracting pr3cedures. Most public officials interviewed in Colombia tend to agree that these procedures may constitute the main obstacle to efficient execution of expenditures. Bids are required to comply with many requirements that have nothing to do with the traditional aim of bidding procedures, like clauses to protect the national industry, to protect national engineering, to ensure that bidders are in compliance with tax obligations, etc. To obtain a contract, clearances are required from I the Secretariat of the Presidency, the Council of Ministers, the State Council, the Comptroller General, and Incomex. To avoid these procedures, public officials have found ways to by-pass the system, like the presertly expanding practice of "contratos de fiducia&. Through these contracts funds are transferred to private agents who agree to manage public expenditures according to the instructions received by the public manager who entrusted the funds to them. By this system all legal requirements and controls imposed by the budget process are in practice avoided because these controls do not apply to the private sector. 3/ The new Organic budget Law Iimits additional budgets to only one revision after the initial budget has been approved.  -8- Contradictory Regulations 11. The network of laws and regulatiors that govern public expenditures (henceforth called legal blueprints) includes laws regulating i) budgeting, ii) planning, iii) procurement, iv) salaries and fringe benefits, v) organization of the public sector, vi) public credit, and vii) earmarking. Historically, the legal blueprints of each of these areas developed independently, resulting in a legal system with contradictions and loopholes that has adversely influenced the efficiency of public expenditures. Also, each area developed its own elaborate procedures designed mostly to prevent dishonesty and diversion of funds, but generating instead cumbersome and time-consuming controls that have greatly decreased productivity. 12. Many contradictions and loopholes can be identified in the legal blueprints that regulate public expenditures in Colombia. Table 2.1 has been elaborated to single out some of the contradictions stemming from the interrelation of the various legal processes. Each row shows aspects of a given legal blueprint (row) that are associated to the interrelated areas of all other legal blueprints (columns), except, of course, itself. Contradictions and loopholes--and sometimes consistent complementarities-- can be found by comparing the off-diagonal symmetrical squares. As an example, by comparing row 1-column 3 with row 3-column 1, we see that the Organic Law of the Budget affects procurement by regulating that contracts should be limited to budgetary availability, and that payments should be limited to cash availabilities; while procurement laws authorize payments according to actual costs, including all agreed upon payments like inflation adjustments and contractually agreed price increases. The result of this contradiction is a systematic lack of transparency and underestimation of expenditures in the budget that is not due to unfor.-een factors. 13. Other contradictions illustrated by Table 2.1 show that planning rules are apparently consistent and complementary to budgetary rules, i.e., planning rules define sector plans that translate into investment quotas in the annual budget (cells 2.1 and 1.2). However, budgetary rules give priority to the payment of recurrent expenditures (cell 1.2). As a consequence, investment becomes an unplanned residual. Also, while laws regulating credit may have already authorized credit contracts, the budget, by law, cannot include those credits unless they had been fully confirmed (cells 6.1 and 1.6). And, although laws regulating the organization of the public sector give to sectoral ministries a major planning and budgeting function, budgeting laws give to individual public entities the major responsibility for preparing the budget (cells 5.1-5.2 and 1.5-2.5). All these contradictions and loopholes create confusion, delays, and opportunities for corruption which ultimately translate into cost overruns. Limited Budget Coverage 14. While by law the budget is intended to be comprehensive, it controls only a fragment of all government operations. Several institutions which manage public expenditures are excluded from the budget process: the quasi-fiscal operations of the Central Bank, the National Coffee Fund, some relatively minor operations of the Cajas de Compensaci6n Familiar, and local governments. Also, although the Organic Law of the Budget is supposed to  ------------------- . ......p o"....w WI,,,, j....lpo *e,., . ,. .. ,, , , . e gao .oJP." go .. ... .*so .... pt2ttqe .,i,nj p.v..,.t ipi*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t10- *neili'1. 00 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~Jqd"P._s. *ip.sO g*05.1*. 0,,,,.