Report No. 7891 -CO Public SectorExpenditure Review

January15, 1991

Public Disclosure Authorized LatinAn-erica and the CaribbeanRegion CountryOperations Department: CountryDepartment IlIl

FOR OFFICIAL UJS' ONLY Public Disclosure Authorized Public Disclosure Authorized

'Documentof the WorldBank

Public Disclosure Authorized 1hisdocument has a restricteddistribution and maybe usedby recipients only in the performanceof their official duties.Its contentsmay not otherwise be disckl.sedwithout World Bankauthorization. FISCAL YEAR

January 1 to December 31

CURRENCYEQUIVALENTS

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 OFFICIALUSE 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 EARMARKINGOF GOVERNMENTREVENUES ...... 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 ...... 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 ...... 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 ...... 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 ExpenditureReview

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 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.rSector.

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 inflexibilityof public expendituresin Colombia has been due to weakneF es in the planning and budgeting process, and to a relativelyhigh degree of expendituresfinanced with earmarked taxes. There is no planning mechanism to ensure coordinationbetween current and capital expenditures,leaving investmentas a volatile residual after covering recurrent expenditures. Investmentprojects have been approved largely on the basis of short term financing availability,without due concern for long term financing requirementsand for the long run productivityof such investments. These projects sometimeshave originated through ad hoc procedures, committinigresources of future budgets, and limiting flexibilityin the use of available funds. The National Planning Department does not have a complete record of individualproject profiles,2 and disconnectedinvestment 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 legislativeprocess. Finally, the Central Governmenthas limited power over expenditure decisions of financialand non-financialnational and departmental decentralizedpublic enterprises,of local governments,of the Coffee Fund, and of the quasi-fiscalexpenditures of the Banco de la Repiblica. These characteristicsof Colombia'splanning and budgeting process have had, in many cases, a negative impact on the productivityof public expenditures.

4. The productivityof public expendituresis also affected by an interrelatedsystem of laws that contradict themselveson many aspects (planning,budgeting, procurement, personnel, organization, credit, and earmarking). Historically,the legal blueprintson each of these areas evolved independently,giving rise to a legal system with contradictions and inopholes that has adversely influencedthe efficiencyof public expenditures. For example, budgeting laws do not allow for estimates of contractually-agreedprice increaseswhile procurementlaws allow them. Budgeting rules limit payments to cash availabilities,while contracts, according to procurement laws, include agreed payment flows. Budgeting laws explicitlyforbid new earmarking,while new earmarking laws have continuouslybeen enacted. Budgeting laws do not allow for credits not yet legally confirmed to be included in the budget, while credit laws may have already authorizedthem. And, finally, according to organizationallaws, sector ministrieshave major planning and budgeting functions,but budgeting laws give to individualpublic entities the major responsibility foz preparing the budget.

5. These contradictionsand loopholes in the legal system governing public expendituresgenerate inefficienciesin most of the institutions controlledby the Organic Law of the Budget (centralgovernment, decentralizedentities and public enterprises). There are institutions that have escaped these contradictionsand loopholesbecause they are not

/ A 'Banco de Proyectos"has been created as part of the mandate of the new Organic law of the Budget. Investmentof public enterprises is in the process of being classifiedproject by project to generate project profiles. - iii - covered by '1hebudget 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-.ons089,090, 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 authoritieshave long been aware of the need to strengthen the Government'scapacity to manage public expenditures. Most notably, the reports of the Bird-WiesnerCommission of 1981 and the Gasto Pablico Commission of 1986 proposed importantmeasures 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 expenditureshas developed, and several important reforms finally materializedduring the Barco Administration:progress in decentralization,a new coffee contract, a restructuringand 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 contractingprocess and to achieve synchro- nization between planning, budgeting, procurement,personnel, organization, credit, and earmarking laws and regulations. Earmarking arrangements should be substantiallyreduced, especially if the country wishes to allocate a higher fraction of its resources towards social services for the poorest groups in the population. Implementationof all these reforms will require very significantchanges in the present administrativestructure to further transfer activities to local governments,to eliminate or reorient central agencies to avoid duplicationsin the decentralizationprocess, to select public entities for possible privatization,and to strengthen the planning system, especiallyDNP (DepartamentoNacional de Planeaci6n), sector ministries,public enterprises,Ministry of Finance, "Contraloria", and "ProcuraduriaGeneral de la Naci6n". A comprehensivereview of the administrationand organizationof the public sector is now the main issue for Colombia to increase savings and to improve the efficiency of expendituresin 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 Assessrnantof Government Initiatives(No. 7271-CO). CHAPTER T

Introduction

1. The Public Sector ExpenditureReview (PSER) is part of a progcam of economic reports that the World Bank prepares periodicallyto proviae a basis for the dialogue between Colombia and the Bank on macro and sector policies, investmentpriorities and Bank assistanceto the country. This PSER should be read in conjunctionwith 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- DecentralizingRevenues 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 IrrigationSubsector Review (Report No. 7380-CO, July 14, 1988), an AgriculturalSector Review of Selected Issues (ReportNo. 7130-CO, October 18, 1988), and a Food Marketing Review (ReportNo. 8048-CO, November 17, 1989). On Industry, a CommercialPolicy Survey (ReportNo. 7510-CO, December 15, 1989), and an Industrial Competitionand PerformanceReport (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/VocationalEducation & Training Subsector Study (Draft Report No. 7742-CO,April 27, 1989).

2. The Colombia Country Economic Memorandum (ReportNo.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 productivityin the private sector, the recommendationsfollow the lines suggested in the Commercial Policy Survey (ReportNo. 7510-CO) that liberalizingexternal trade should be considered an effective instrument. To enhance productivityin the public sector, the recommendationsare to strengthenpublic resource management, stresmline public investmentprojects, and improve the flexibilityof the budget allocation mechanism. These last issues are discussed in this PSER.

3. Colombia's capacity to manage public expenditureshas long been known to be weak. The Bird-Wiesnerreportl and the Public Expenditure Commission'sReport of 19862 made recommendationson several aspects of the administrationof public expendituresthat needed urgent remedial actions. Many of these aspects were consolidatedin 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 Administrationin Colombia and will clearly affect all future administrativereforms.

3/ EduardoWiosner and RichardBird, Finanzas Intergubernamentales in Colombia, Departam.nto Nacionalde Planeacidn,1981.

2/ Comisi6ndel CastoPGblico, Informe Final 1988. Ministerio do Hacienday Cr6ditoPGblico. - 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 Kenmnerermission 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 . 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 Intergubernamentalesin 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 evaluationof the plan is containedin Bank'sreport No 7271-CO,op. cit., Colombia, Social Problemsand Poverty allaviation:An assessmentof 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

Manageabilityof Public Expenditure

A. Administration of Public Expenditures and Efficiency

1. Public expenditureefficiencies (or inefficiencies)can be intro- duced at two broad stages of public sector management: resource allocation and expenditureexecution. Economistsare, in general, preoccupiedwith the elaborationof policies to improve the efficiencyof resource allocation, and have been less concernedwith how those policies can be carried out effectively. Efficiency,however, very often fails at the execution stage as much as it fails at the allocationstage. This chapter focuses mainly on inefficienciesgenerated at the execution stage.1

2. In Colombia, most public expenditure inefficiencieshave been created because the allocationand the execution of public expenditureshas been governed by: i) inefficientprocedures and time-consumingcontrols establishedby the budgeting process; ii) proceduresestablished by autonomousentities that escape the regulationsof the budget process; and iii) a network of interrelatedlaws and regulationsthat 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. InefficientPlannina and Budget Process

3. Planning has been extremelyweak in Colombia. There has been no planning mechanism to ensure coordinationbetween current and capital expenditures,often leaving investment--whichis supposed to be carefully programmed according to planning laws--as a volatile residual that fluctuateswith revenues after covering recurrent expenditures. DNP, the planning agency of the central government,is responsiblefor the investment budget only, while the Ministry of Finance prepares the current expenditure budget. But DNP is informed of investmentsof the decentralizedpublic sector only when investment is financed with government-guaranteedexternal credit. Many investments,therefore, are in practice outside of central governmentcontrol includingmanagement of debt, which is often contracted by public enterprises independently,with minimal supervisionor control of the central government.

4. The overall organizationof the public sector is partly responsible for the wc.aknessof plann-..gin Colombia. A comprehensivereform of the organizationof the State done in 1968 strengthenedthe need for planning, but left weaknesses that developed over time into an institutional incapacityto prepare sectoral plans that could be coordinatedinto a national developmentplan. Sector m-inistrieslike Education and Health have been overcrowdedmost of the time with day-to-dayoperational activities, with little capacity for developingpolicies and for elaboratingmedium or

1/ Allocativeefficiency is an absolutelynecessary condition, however. Sometimes certain taxes are advocatedjust becausethey can be effectivelyexecuted, risking the worseof the cases:an economicallyinefficient 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-esupplementary 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 firnancialexecution 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 regulatiorsthat govern public expenditures (henceforthcalled legal blueprints) includes laws regulatingi) budgeting, ii) planning, iii) procurement,iv) salaries and fringe benefits, v) organizationof 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 contradictionsand loopholes that has adversely influencedthe efficiencyof public expenditures. Also, each area developed its own elaborateprocedures designed mostly to prevent dishonestyand diversion of funds, but generatinginstead cumbersome and time-consumingcontrols that have greatly decreasedproductivity.

12. Many contradictionsand loopholes can be identifiedin the legal blueprints that regulate public expendituresin Colombia. Table 2.1 has been elaborated to single out some of the contradictionsstemming from the interrelationof the various legal processes. Each row shows aspects of a given legal blueprint (row) that are associatedto the interrelatedareas of all other legal blueprints (columns),except, of course, itself. Contradictionsand loopholes--andsometimes consistentcomplementarities-- can be found by comparing the off-diagonalsymmetrical 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 regulatingthat contracts should be limited to budgetary availability,and that payments should be limited to cash availabilities;while procurement laws authorizepayments according to actual costs, including all agreed upon payments like inflation adjustments and contractuallyagreed price increases. The result of this contradictionis a systematiclack of transparencyand underestimationof expendituresin the budget that is not due to unfor.-een factors.

13. Other contradictionsillustrated by Table 2.1 show that planning rules are apparentlyconsistent and complementaryto budgetary rules, i.e., planning rules define sector plans that translate into investmentquotas in the annual budget (cells 2.1 and 1.2). However, budgetary rules give priority to the payment of recurrentexpenditures (cell 1.2). As a consequence,investment becomes an unplanned residual. Also, while laws regulatingcredit may have already authorizedcredit contracts,the budget, by law, cannot include those credits unless they had been fully confirmed (cells 6.1 and 1.6). And, although laws regulatingthe organizationof the public sector give to sectoralministries a major planning and budgeting function,budgeting laws give to individualpublic entities the major responsibilityfor preparing the budget (cells 5.1-5.2 and 1.5-2.5). All these contradictionsand loopholes create confusion,delays, and opportunitiesfor corruptionwhich ultimatelytranslate into cost overruns.

Limited Budget Coverage

14. While by law the budget is intended to be comprehensive,it controls only a fragment of all governmentoperations. Several institutions which manage public expendituresare excluded from the budget process: the quasi-fiscaloperations of the Central Bank, the National Coffee Fund, some relativelyminor operationsof the Cajas de Compensaci6nFamiliar, and local governments. Also, although the Organic Law of the Budget is supposed to ------p ......

o"....wWI,,,, 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*. ¶ip~~~JO0,,,,.flU 01 ettilgin, lop.,, g.u,n.e ip.. *p.s.j l).ds.~~~~~O.e,

*i-50a....ecna a .,*i#ppvi .ipUd. SiIU "Un.S...h,00. -.. op

-11V ~ n,aiL..Oqap IV~~~~~~~~~1

!I~ ... P- ~I..*j ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~ggu., .stu

*I.op.q~ ~ ~ ~ ~ ~ ~ 1 en.ir. *l~0 n*.ee.~~~s p.%sns.e 41** "!- *Aijflflt *i-pfl fplV *5!t'1.0I p.tti,et*.1.60 ~5.4.; ~ ~ ~ ~ ~ ~ "- I, q n -- 40 P" sap,,. s*.ns 01 n011V .~~~~~~~~~~.-mn!peIed

Am J. 'j.'p~~~~~~~~~~~~~~~~ttO,pes. *1

to. qasin.n t.in. .?0"t

nagp.puigs.e tin a, 0...eesn,ns4 po *eOEA...... s.nas- ,o.9 seiglil

Bgg..95s,en eneg.. *nisd,fl0,A.- 0 'M #*05 ne.1..,~~.na *.P.,. t.n.t noes,.., *lo,IV t d4l .. ae. ne wpn,en.s

L3MVO]M Wrl 31.PW 314 tA RNnea OI34111130 SA MIOK 1 - 10 - include decentralized agencies, Social Security and public enterprises, the central goverrnment has in practice little control over the operations of these agencies and enterprises. While the contradictions and loopholes found in the budget process generate inefficiencies in most of the institu- tions controlled by the Organic Law of the Budget, most inefficiencies created by the limited coverage of the budget derive from lack of control. Of total revenues estimated at 31.6 Z of GDP in 1988 and current expendi- tures estimat-ed at 25.8Z of GDP, about one fifth (revenues of 5.6Z of GDP, and current expenditures of 6% of GDP) could be said to be directly under central goverrnment control. And, of total public investment estimated at about 8Z of GDP, only about 1.32 of GDP is in direct control of the central government--16Z of public investment--(Table 2.2 below).

Table 2.2: COLOMBIA - STRUCTURE OF PUBLIC SECTOR REVENUES AND EXPENDITURES (as percent of GDP)

-- 0ga-c a. o0 t'o Bcaget -- NAt oAa I Ca,** de ns~*C*Q~acTota. 0-ca. T*ota. CeAt'li 0.ceA'tr Soc'a. P.0o c Tots ,drala 06qt

C,,F.nt 1"-. 5 6 2.5 :. 2 a 22.4 3.1 0.6 26 : . 31.6 C.-snt *.edt, 6 0 1 9 9.6 :-8 2 9 0.5 2 8 a: 2588

C.'-Ont S~p.,- 0e N"C t) -0.4 :5 0: 2 8 ' a 0.2 0 : A 3 .' 5 8

:.3 .8 0.: 3.4 6.9 0 3 003 73 . 8 2

0,.,uiIS.'01.,s o (Dof C t) -:2 -o -o: -:0 -2.9 -o: Co -3.0 0.5 -2.5

F;nA*"c;g :_7 0 : . :0 2 9 0 : 0 0 3 0 -0.5 2.5

E.ta'". ~~~~~0.90.2 0.0 '.2 2.3 0.0 0.0 2 3 0 0 2.3 :Aflo.4 2.8 0.3 0 0 3.4 8.5 0.0 0 0 8.5 0 6 .6 A..o.t;z*t.*n 1.8 o.:- 0.0 2.2 '.2 0.0 0.0 4 2 0.1. 4 3 InternAs 0.2 0.0 0 : -0.2 0.1- 0.: 0 0 0.:- 0 0 0 0.5 0.0 o.:. 0.0 0.6 0.1- 0.0 0.6 0 0 0 6 Ar.c,~tat.oA'5 0.3 0.0 0.0 0 2 0.5 0.0 0.0 0.5 0.0 0.5 BaAk~ of Republic &Ad oth.. 0.5 0 0 0.0 0.0 0.5 0.0 0.0 0 S -0.5 0 0 - 11 -

C. Reforms Undertaken

15. The Government of Colombia has made significant efforts at improving short-term macroeconomic manag3ment. These efforts have been quite effective at maintaining tight fiscal and monetary controls, keeping the real exchange rate at competitive levels, and turning the economy around from the poor performance of the early eighties (average GDP growth of 1.6Z in real terms during 1980-83 became about 5X a year during 1985-88). To maintain growth in the long term, however, higher dcmestic savings and better capital product:vity are needed. The Barcc administration, aware of these needs, has initiated efforts at improving efficiency in the private sector by proposing trade reforms that would eventually increase competition, and it has also made some n.oves to reform the financial sector aiming at increasing private savings. The government has also substantially improved its overall capacity to manage public expenditures through the enactment of several initiatives. In what follows, we report on the status of these initiatives to improve the management of the public sector.

Organic Law of the Budget

16. The weaknesses of the budgetary process have been reported widely since the publication of the final report of the Misi6n de Finanzas Intergubernamentales.4 This report, together with several unpublished background papers, helped at creating political consciousness over the last eighth years, and contributed generating a process of change that finally materialized in a New Organic Law of the Budget, approved by Congress at the end of 1988. The importance of this law can not be overemphssized. Due to its content and far reaching objectives it will most certainly affect most future public sector administrative reforms.

17. The process of change, as it is usual in Colombia, has not been an easy one. Several attempts at modifying the old legal instruments that presently govern the budget --mainly the Constitutional Amendment of 1965, the Constitutional Reform of 1968, and Decree No. 206 of 1973-- failed. Onie of the last proposals to modify the budget process was submitted in 1984, but due to substantial modifications in its passage through the political discussions in Congress, it was allowed to expire. Then, in 1985, the Government appointed a Public Expenditures Commission (Comisi6n de Gasto Piblico, CGP), with a mandate to diagnose the main problems with the budget process and to suggest changes.

18. The report of the Expenditures Commission was ready in 1986 and helped again to bring discussions about budget reforms to the open. The Government's original plans were to submit constitutional changes to the most restrictive Constitutional norms dealing with the budget process. Parallel to the proposals to modify the constitution, the Government was planning to submit changes to the Organic law of the Budget in 1987; but due to unexpected political difficulties on having other aspects of the Government's plan approved, budget issues were postponed. Finally, at the end of 1988, a first round of Constitutional discussions ended (a constitutional change requires two rounds of legislative discussions), and the new Normative Law of the Budget was approved.

4/ Bird-Wiesner, op. cit., DNP, 1981. - 12 -

19. The most important elements of the new Organic Law of the Budget are: i) creation of the Consejo Superior de Politica Fiscal (CSPF); ii) increase in budget coverage; iii) eliminationof "additional"budgets; iv) formalizationof macroeconomicprogramming; v) establishmentof a "Banco de Proyectosn; and vi) extension of the Treasury's responsibilities.

20. The Consejo Superior de Politica Fiscal (CSPF)will coordinate the overall budgeting system and will report directly to the President and to the Consejo Nacional de Politica Econ6mica (CONPES). The CSPF is expected to eliminate the lack of coordinationbetween the diffetent institutions that presently prepare the budget, and to generate a desirable link between government'slong term plans and budgeting,between long term planning and short term macroeconomicmanagement, and between capital and current expenditures. The CSPF will be composed of the Minister of Finance, two ministers from the social sectors designatedby the President,the Economic Secretary of the Presidency,the Chief of Planning and two full time advisors. The executive secretariatof the CSPF will be the office of the budget (Direcci6nGeneral del Presupuesto),which will require substantial reinforcementto perform its new expanded technical functions.

21. The budget coverage has been considerablyexpanded under the new budget law. The profits of the decentralizedpublic sector will be part of the budgetary revenues of the Central Governmentwhich could freely allocate them back to the decentralizedagencies or to any other cempeting end. There will be only one budget delivered to Congress before July 20 each year, including the National Governmentand the public agencies together--presently,the budget of the National Governmentand of the public agencies is delivered separatelv.before July 20 and before September 10, respectively. Also, as part of *he government'scapital revenues, the new law adds the financialyields obtained from government'sdeposits, the higher peso value of foreign exchange disbursements,and the profits of the Banco de la Repciblica'sCuenta Especial de Cambios determinedby the Monetary Board. All external credit, includingCentral Government and decentralizedagencies, will be estimatedand allocated in advance. Also. although the Coffee Fund continuesto be handled outside of the budgetary process, most of the principles that regulate the budget have been included in the recently concluded new ten year agreementwith the National Federationof Coffee Growers. The only portion of public funds that has been excluded of budgetary regulationsis that of the quasi-fiscal operationsof the Banco de la Rep6blica.

22. Supplementarybudgets are expected to be fully eliminated. The constitutionalrequirement that the budget be balanced in a formal sense is being reviewed as part of the constitutionalreform presently under discussion. This requirementwas the basic reason for the existence of several supplementarybudgets during the budget period, since new supplementswere needed as new resourceswere becoming "legally"available. Meanwhile, as a transitorymeasure, the new budget law authorizesonly one supplementarybudget during the budget period. Credit financingcould now be incorporatedin the initial budget insofar as it has been authorized, (and is expected to be available during the proposed budget period) avoiding the requirementthat credit needed to be signed and formally finalized before it could be incorporatedin the budget.

23. Macroeconomicprogramming has been quite successfulsince the present economic team has been in charge, but it has not been an - 13 - institutionalfunction of the budget process. A change in the team might well have put in danger the continuationof macroeconomicprogramming. The new budget law has reduced this risk considerablyby institution?lizing macroeconomicprogramming as part of the required steps of preparing the budget. The new budget process requires the preparationof a two (or more) years' financial plan, an annual investmentplan, and a yearly national budget. Each public agency will have to prepare a two-year financialplan that will be liuiitedby the requirementsof monetary, fiscal, or exchange rate constraints imposed by macroeconomicprogramming. The financialplan will be prepared by the Ministry of Fin&nce in coordinationwith DNP (NationalPlanning Department),and it will be submittedto CONPES after being approved by the CSPF (Consejo Superior de Politica Fiscal).

24. All investmentprojects will have to be incorporatedin the Banco de Provectos (BDP) before they could be consideredfor inclusion in the Annual Investment Plan. In addition, the budget will co-finance investment projects of public agencios only if i) the projects are duly registeredin the BDP, and ii) the public agency is currentwith its debt service payments. The Planning Department will be in charge of preparing the Annual Investment Plan, and will keep the Banco de Proyectosup to date. The BDP is defined in the budget law as a Pet of activities that have been selected as viable after a social, technical and economic evaluationhas been performed. DNP, together with the National Fund for Development Projects, will have a period of one year to prepare regulationsand to recommend changes in the organizationof the public sector that they may consider necessary to make BDP functional.

25. Substantialadministrative reforms will be needed to implement the operation of BDP. First, there will be a need to create a permanent coordinatingmechanism between DNP, the Ministry of Finance, and the planning offices of sector ministries. Second, the institutionalstructure affected by BDP will have to be defined, i.e., to determine the national, regional, departmental,and municipal agencies that will be affected by BDP at the level of decision making, planning, and execution. Third, a reporting system has to be created,with clearly defined obligationsto report to DNP by each agency affected by investmentdecisions. Fourth, the institutionalstructure to identify investments,to produce feasibility studies, to generate technical,financial, and economic evaluationof projects at different levels, national, regional,departmental, or municipal, has to be determined. Fifth, methodologiesof analysis will need to be homogenized and announced at all levels. And sixth, the needs for technical assistanceand for technologydevelopments to manage information will require assessment,and clear developmentplans.

26. The powers of the Treasury ha--ebeen expanded to include the pcssibility of short term borrowing to compensatefor transitory fluctuationsin liquidity during the fiscal year. The Treasury could sell or buy short term financing instrumentsin the capital market. It could also use a special line of credit with the Banco de la Rep(iblicaup to 8Z of current revenues of the previous year, but only to smooth revenue fluctuations. This special line of credit can not be used as additional financing of governmentexpenses, and it should be repaid within the fiscal year. The Treasury will need administrativereorganization in order to accomplishthese new expanded functions. - 14 -

Decentralization

27. The decentralization process now being undertaken in Colombia constitutes a r,ajor effort at strengthening the democratic process and at increasing local partlcipation in development planning and management. Together with the new Organic Law of the Budget, it will substantially affect the shape of the administrative reforms that will be required to enhance the capacity of the government for nlanning and budgeting.

28. The efforts of the government at decentralization are expected to affect seriously the management, the size, and the composition of public expenditures. ThLeBank has undertaken a separate study on decentralization which deals with the main aspects of taxation and administration of public resources. Three main aspects could be mentioned as of concern from the point of view of fiscal management: first, the probable effect of decentralization on the fiscal deficit due to transitory and permanent clairms for additional expenditures; second, the possible duplication of activities at the central level and at the decentralized level, which may unnecessarily increase public expenditures; and third, the potential need for occasional bailouts to local governments due to the possibility of financial mismanagement.

29. Decentralization will probably tend to increase the fiscal deficit. During the earlier stages, transitory expenditures will emerge as agencies of a national level disappear, and a need to cover expenditures associated with the layoff of national level employees arises--unless the same employees were to be transferred to the municipalities, wnich is highly improbable. Later, a more permanent effect will tend to generate deficits due to expected shifts in the demand and supply of public services. Due to decentralization, resources from the value added tax will be increasingly transferred--from about 35% in 1987 to 502 in 1992 (including 5Z to regional educational funds)--to the regions. If the demand for public services at the central level does not decrease in the same magnitude, a deficit will be created at the central level. Also, as resources are transferred to the local goverrments, thp demand for public services at the local government level may increase faster than the availability of resources, due to increased expectations. All these effects should be carefully measured to prevent future negative effects on macroeconomic management which, if not clearly addressed, may act against the decentralization process itself.

30. There is a danger of duplication of functions due to the process of decentralization, which will translate into unnecessary expenditures. Local governments have become responsible for the supply of w,ter and sewerage, public slaughterhouses, market places, some public works, all urban roads, park maintenance, first level of health services, technical assistance to small agriculturc, construction and maintenance of primary and secondary schools, construction of recreational and sport facilities, and the construction of cultural houses. Before decentralization, part of these services were supposed to be provided by the national level of government. However, although functions at the central level are being eliminated,the agencies that were providing the services do not seem to disappear. Such is the case in examples like Colturismo,Coldeportes, Colcultura, and Urban Development Enterprises. Further analysis is needed to determine all possible cases of potential duplication. These duplications will create - 15 -

further pressure on the fiscal deficit that may attempt against the process of decentralization.

31. Municipios or Departmentsmay occasionallyrequire L bailout. As part of the decentralizationeffort, a drastic solution was needed to remedy the financial insolvencyof the DepartmentalWater and Sewerage Corporations (EMPOS). The central governmentassumed their most important liabilities like external debt and accumulatedpayments due to workers. It was judged that a fresh start for the EMPOS would allow them to efficientlyexecute their new decentralizedfunctions. Although needed and welcome, this solution entails evident potential risks. In the absence of monitoring and adequate controls, the new companiesmay again develop inefficientpractices which will require further bailouts in the future. The precedent is a dangerous one. Excessive labor costs, politicalmanagement of public sector prices, and lack of control may potentiallyadd to fiscal deficits and attempt against the decentralizationprocess.

Coffee Fund

32. The Coffe- Fund has in prac':iceescaped many of the controls and regulationsof the budgeting process. Conceptually,it is equivalentto an earmarked tax managed by the National Federationof Coffee Growers, a private non-profitorganization representing coffee producers.5 The execution of expendituresis governed by private sector practices,with little or no accountabilityto the public sector. This lack of budoetary control over the operationsof the Fund has allowed its management to avoid the inefficienciesof the public sector budgetary process. But it has also allowed uses of the Fund in activitiesbeyond its original mandate of stabilizationof growers' income at levels consistentwith long term trends in world prices, and of provision of marketing, technical assistance, quality control, storage, and promotion of economic and social welfare in coffee growing areas. The Fund has developed interests in a maritime fleet, sugar, steel, meat processing,banks, insurance,savings and mortgages corporations,and it provides subsidizedinputs to coffee growers and subsidies to domestic consumptionof coffee.

33. During the Barco administration,in December 1988, a new ten-year coffee contract was signed between the Governmentof Colombia and the National Federationof Coffee Growers. In this contract, and in line with the reforms introducedby the new budget law to enhance public expenditure efficiency,the government introducedseveral modifications to improve fiscal control over the the Coffee Fund management: i) The budget of the National Coffee Fund will be prepared by the Federationof Coffee Growers to be presented to the National Coffee Committee,approved by the Minister of Finance, and ratified by a GovernmentDecree. The National Coffee Committee will have representativesof coffee growers with voting power not greater than that of the government representativesincluding the ministers of Foreign Relations,Finance, Agriculture,and Development,the Chief of DNP, and the manager of Caja Agraria. ii) The Coffee Federationwill contract the services of two technical advisors, appointedby the Minister of Finance, who will provide full time technical assistanceto the government representativeson all coffee matters. iii) The Fund will prepare separate accounts according to objectives,will prepare cost analysis for each separate activity,will develop indicatorsto measure the efficiencyof

6/ See Chapter III, section B. 4. - 16 -

expenditures,and will present a financial report to the National Coffee Committee every six months. iv) The Fund will prepare additionalmonthly reports to the government,through the Minister of Finance, covering monthly cash flows and monthly projections;it will provide a consolidatedbalance to December 31 of each year audited by the Office of the ComptrollerGeneral before the end of March of the followingyear, and will report timely on any informationrequested by the Office of the ComptrollerGeneral. v) Investmei.s of the Fund will be restrictedto those that benefit coffee areas only. The National Committee,with the approval of the Minister of Finance, should determinebefore June 30, 1989, the list of enterprisesand investmentsthat are consideredindispensable for the developmentand support of coffee policies. The National Coffee Conrmitteewill approve a program to sell all other enterprises,liquidating or selling them before December 31, 1993. vi) Every program modifying subsidies,promotion or diversificationof coffee, investments,etc, should be justified by a detailed economic analysis indicatingcoverage, profitability,relevance to the coffee industry, and financial,real, and social effects of each expenditure or program. And finally, vii) the Miniscer of Finance will have veto power in determiningcommercial policies, the selling price of coffee exports, and the methodology to determine the exporters selling price that fixes the retention scheme.

Railroads

34. The railroadshave representedone of the clearest case if poorly managed public enterprisesin Colombia. Aware of the problem, tild governmentachieved approval of legislationin February 1988 allowing a full restructuringof the sector. Through the proposed reforms to reorganizeit, the Barco administrationundertook an important initial step towards privatization,a potentiallyvery rich area for further improving the productivityof public expenditures. If the privatizationof the railrc ds is successful,it may open the door to privatizationof other activities presently in the hands of the public sector that could be made more efficient by being privatized.

35. The railroadshave become very inefficient. It has been estimated that no more than two thousand workers are needed to efficientlyoperate the system, as opposed to about seven thousand presently employed. Payroll expenses representan abnormal 802 of the total operating costs, with workers allocated to railroad stations and railroad tracks where the service does not exist, with excess personnel in administrativeduties, and not enough people working on maintenance. Only about 202 of the engines are operating,and less that 60Z of the wagons and about 802 of the railroad lines are usable, most of them in bad shape.

36. The governmenthas proposed to liquidatethe Colombian National Railways, to create a public enterprisein charge of the railway infrastructureand of the maintenanceand rehabilitationof the tracks, and to let the private sector take charge of the operation and of the maintenance of the motive power and rolling stock. The proposal has not yet become operative since the governmentis still seeking financing for about US$200 million to cover severance payments, early retirementprograms, to conclude pending contracts,and to repair and renovate capital equipment. - 17 -

Procurement Statute

37. Procurement procedures are presently under review. Congress, however, did not grant extraordinary powers to the Executive to review contracting laws; instead, it is expected that Coni :ess will enact a new law reviewing the main principles that define contracts, procurement processes, and the protection of national industry, and will later grant limited powers to the Executive to complete the procurement procedures. Apparently, no concern has so far been expressed about the lack of consistency of the present procurement statute and the full network of laws and regulations that govern public expenditures. Reviewing contractual procedures in isolation from laws regulating planning, budgeting, credit, and the overall organization of the public sector will not eliminate existing contradictions and loopholes in the present system. It would be desirable that the new procurement statute be a consistent element of a more comprehensive reform of the administration of the public sector.

Other Initiatives

38. Not all the initiatives of the Executive to increase the efficiency of public expenditure management have been successful. A recent proposal to eliminate earmarking was rejected by Congress. The proposal was included in the constitutional reform just approved by Congress in its first round (a second round is needed for final ratification) and intended to eliminate all earmarking with the exception of the Situado Fiscal and the VAT revenue sharing.6 Instead, a new earmarked revenue was created: five percent of the revenues of decentralized agencies will be earmarked to finance public health. The initiative for constitutional reform also included a proposal to create a Tribunal de Cuentas which would substitute for most of the present responsibilities of the Comptrollers General Office. This initiative was approved in the first round of Congress discussions, but was not ratified afterwards.

D. The Need for Further Reforms: Administrative Reorganization of the Public Sector

39. Most ot the potential benefits of the new Organic Budget Law could be lost without a comprehensive review of the administrative structure of the government. Implementation of the law could become impossible within the present ac.ministrativestructure. Action on several fronts is suggested.

40. First, the overall legal and administrative system should be reviewed to avoid inconsistencies, duplications and contradictory regulations, and to ensure complementarity among the many different laws and statutes that affect the management of public expenditures. The coordination mechanism among central units, sector ministries, and decentralized agencies to prepare and to implement the public sector budget needs strengthening at all levels.

41. Second, the Office of the Budget (Direcci6n General de Presupuesto) is not presently structured to serve as the technical supporting body of the

S/e Chapter III. 18 -

Consejo Superior de Politica Fiscal, as required by tne new budget law. The office will need to strengthen its capability to advise the government on fiscal policies and on how to improve the link between long term planning and budgeting, between long term planning and short term macroeconomic programming, and betwueen capital and current expenditures.

42. Third, the Consejo Superior de Politica Fiscal will need to suggest modifications to the exist 4 ng mechanisms of control to serve the purposes of the new budget law. There will be a need to monitor profits of decentralized agencies without interfering with their incentives to perform; to monitor government accounts yielding interest income, and accounts yielding financial gains due to exchange rate variations; to monitor the Banco de la RepCiblica's Cuenta Especial de Canbios and other quasi-fiscal operations; and, conceptually, to monitor Ell incomes generated within the public system independertly of uhether they are presently included in the budget or not.

43. Fourth, the implementation of the Banco de Proyectos (BDP) will require the creation of a permanent mechanism of coordination at a technical level between DNP, the Ministry of Finance, and the planning offices of the sector ministries. Also, the mechanism of implementation of the BDP at national, regional, departmental and municipal level has to be developed, including a reporting system,, and an institutional structure to identify and evaluate investment projects.

44. Fifth, sector planning needs reinforcement. The present capacity of the sector ministries to actively participate in generating the sector components of the national plan is weak. Also, sector budgets can not be just the sum of individual agencies' budgets, without vigorous involvement of the ministries in the budget process.

45. Sixth, the Treasury needs reorganization to accommodate its expanded responsibilities under the new budget law: especially to forecast short term financial needs due to fluctuating seasonal revenues, and to be prepared to efficiently buy and sell short term financial assets in the capital market.

46. Seventh, efforts at possible privatization of non-performing public enterprises should be expanded. For that purpose, a full evaluation of the efficiency of public enterprises is needed, including an analysis of alternatives to improve their management in the public sector, and to restructure or privatize them. The potential for action in this area is almost unlimited in Colombia, and it could become one of the most important sources of future improvements in the productivity of public expenditures.

47. Eighth, procurement legislation should be reviewed in the context of the overall administrative reform, including coordination with the rest of the legal system that govern the management of public resources.

48. Ninth, the decentralization process is responding to a profound need to strengthen the democratic process in Colombia by increasing local participation in fundamental decision making. The process itself may be in danger, h-wever, if serious consideration is not given to potential adverse effects on t; fiscal deficit due to expected shifts in the demand for public services, .Asible duplication of functions, and the eventual need to bailout municipal ana Aepartmental agencies if not properly managed. - 19 -

CHAPTER III

Earmarking of GovernrieiiiReuvmnues

A. General Definition

1. Earmarking is the term given to the p,ractice of assigning revenues from specific taxes or groups of taxes to certain government activities which may be broadly or narrowly defined. As suclh, it contrasts with general fund financing where revenues froi,,vazious sources are pooled together to be used for general government p;tposes; under general fund financing the connection between any particular government activity and any specific revenue source is remote or non-existent.

2. Earmarking can pertain to a wide gamut of public sector activities. At one end of the spectrum would be a fairly specific set of taxes set aside to be used for certain quite specific purposes. Examples would be gasoline taxes or tolls used for highway maintenance and construction, payroll taxes used to finance pensions and medical payments, and the after-tax profits of public commercial and industrial enterprises, the latter being included as earmarking to the extent that profits are retained within the enterprise rather than being used for general government purposes. At the opposite end of the spectrum would be earmarking in the broadest sense--the unconditional sharing of revenues between levels of government. Between these two ends of the spectrum lies a variety of earmarking varying irn the specificity of revenue sources, the specificity of government expenditures, the degree to which the benefit principle--or the overlap between tax/price payers and expenditure beneficiaries--applies, and the mix of financing between earmarked and non- earmarked sources.

B. Trends in the Size and Stru.ture of Earmarking (1970-1987)

Size and Structure of Earmarking

3 Earmarking (rentas de destinaci6n especifica or RDEs) has long been a prominent feature of Colombian public finances.1 During the past two decades, the proportion of earmarked (at all levels) government revenues has increased by about one half and the sources of earmarking have become more varied. In 1970, RDEs constituted just 1lZ of national government revenues and were equally large in size at the departmental level where they formed a much larger proportion of revenues (see Table 3.1 below). The most prominent sources of earmarking were beer, alcohol, and lottery rc.enues at the departmental level followed by the coffee export tax and gasoline tax at the national level, and the valorization tax for local improvements. Together, these four sorcrs accounted for about 89Z

1/ At first, our discussionwill be confinedto traditionalsources of earmarking--i.e.those funds and revenuesharing arrangementsthat aro usually listed as earmarking in Colombia. Later, non- traditionalsources of earmarking--payrolltaxes and the operatingprofits of non-financial enterprises--willalso be considered. There are two rear-ns for splittingthe discussion in this fashion. First, the traditionalsources are what is more commonlythought of as earmarking. Secondly, longertime series were availablefor tho traditionaloarmarking sources. - 20 -

of all earmarking (see Table 3.2 ahead). In total, RDEs constituted212 of revenues at all levels of government.

4. The upward trend in earmarkingsince 1970 has been the product mainly of greater earmarkingat the central governmentlevel; over the period, the proportion of national governmentrevenues earmarked tripled from 11% to a peak of 342 in 1984 before falling off slightly in the three years through 1987 (Table 3.1). The shares of earaarkingin departmental and municipal incomes have been generally trending downward over the period. Between 1970 and 1986 the share of the central governmentsRDEs in the total rose from 44Z to 75Z, while the shares of the departmentaland municipal governments fell from 471 to 232 and 9Z to 2% respectively.

5. By far the biggest contributingfactor to the increasewas the rise in revenue sharing between levels of government. "Situado fiscal" and "cesi6na las ventas" contributedabout 39Z of the total increase in earmarking over the period. Other major contributorswere: departmental alcohol/tobacco/gamblingtaxes (23Z), the gasoline tax (14Z), import duties to suport PROEXPO, IFI and Caja Agraria (11Z) and the coffee export tax (52).Z

Table 3.1: SHARES OF EARMARKEDREVENJES IN GOVERNMENTREVENUES a (percentages)

1970 1975 1980 1981 1982 1983 1984 1985 1986 1987

Share in National GovernmentRevenues 11.3 17.S 27.8 27.0 28.3 27.9 34.4 31.3 27.7 27.7 Share in Departmental GovernmentRevenues 93.9 82.5 88.6 85.8 86.6 85.5 78.3 768. 78.5 Share in Mur;cipsl GovernmentRevenues 30.1 30.2 31.3 30.6 29.3 33.9 27.0 22.9 16.3b Share in Government RevenuesAll Levels 21.1 23.5 33.4 32.3 31.1 33.5 39.1 36.5 32.0 a. Earmarked revenuesare listed in StatisticalAppendix Table 1, excludingpayroll taxes, the operating profits of non-financialpublic enterprisesand, at the national level, gother' sources of earmarking. Revenuesare total tax and non-tax revenuesat each level of government;at the national level, revenuesfrom the 4% ad valorem tax on coffee exports are added 1970-1983,because these monies did not pass through the budget in those years. b. Likely to be an underestimatedue to missing data.

Source: Estimates based on Tobles 1 and 2 in StatisticalAppendix.

2/ While the taxes listedmade significantcontributions to increasesin earmarkingover the period, not all grew as fast as total governmentrevenues. We have already seen how departmentaland municipalearmarking failed to keep pace with own sources of revenueat those levels of government. The rising share of earmarkingin nationalgovernment revenues is explainedby rising shares for revenue sharing,PROEXPO (plusIFI and Caja Agraria), and (to a small extent) a variety of minor funds making up for falling shares for the codfee export tax and gasolin, tax. Together, the former three rose from 2.9% to 21.7X of nationalgovernment revenuesbetween 1970 and 1987 while the lattertwo fell from 8.4X to 5.9W. - 21 -

In addition, there was a proliferationof relativelyminor RDEs--e.g.hotel taxes to support tourism, taxes on movie tickets to support movie making and culture, taxes on coal production to support exploration and development,and a variety of others.

6. By the mid-1980s, the following six were the major sources of earmarking in descending order of importance: revenue sharing (situado fiscal and cesi6n a las ventas), departmentalalcohol/tobacco/gambling taxes, the gasoline tax, the import tax (PROEXPO,etc.), the coffee export tax, and the valorizationtax. Together they constitutedabout 94Z of total earmarking (see Table 3.2). The large number of remaining funds contribute very little to the total.

The Effects of Law 55 (1985)

7. Since the mid-1980s there has been a deliberate governmenteffort to curtail the extent of earmarking,most particularlyas the result of recommendationsby the Commissionon Public Expenditures. However, resistance to reductionsin earmarking has been strong and the impact of measures taken to date has been limited. SpecificallyLaw 55 of 1985 stipulated that the National Council of Economic and Social Policy (CONPES) would decide on the reassignmentof funds from recipientsof earmarking to complementaryor related activities. The RDEs affected by Law 55 and the activities to which funds could be reassigned are spelled out in Table 4 in StatisticalAppendix. The upper bounds or reassignmentswere set at 10% of earmarked revenues in 1985, 20f in 1986, and rising in 10X incrementsto a maximum of 502 in 1989. In implementingLaw 55, proposalswould be made by the relevant minister and the Chief of the National Planning Department based on programs submittedby the agencies receiving earmarkingfunds and related agencies and the final allocationwould be made by CONPES on a case by case basis. The reassignmentcould be effected by the earmarking recipient agency itself carrying out the activity or, if it was unable, by contractual agreementswith the governmentor with its various public organizationor enterprises.

8. Law 55 resultad in the reallocationof Col.$30 billion and 52 billion in 1986 and 1987 (Table 5 in the StatisticalAppendix), the equivalent of 11.9Z and 15.2Z of total earmarkingat the national level in the two years respectively. If these amounts were removed the remaining earmarking at the national level would have constituted24% of national revenues in the two years (in contrast to the 28% figures reported in Table 3.1). Particularlysignificant amounts--both absolutely and relative to total earmarking--wereinvolved in the cases of PROEXPO, the gasoline tax, ICBF and SENA.

9. There are a number of reasonswhy the foregoingoverstates the reduction in earmarking actually achieved by Law 55. First, Law 55 did not--and was not intended to--free up earmarkedmoney to be used entirely at the central government'sdiscretion. The alternativesfor which monies could be used were carefully spelled out and narrowly limited to activities in the same sector. Secondly,while intending to free up budgetary resources,Law 55 actually created two new (relativelysmall) earmarked

3/ The imit for PROEXPOwas set at 20% for 1986 afterwhich it wouldfollow the sameschedule as other funds. - 22 -

Table 3.2: SHAR£OF VARIOtS TRADMONAL EARMARKEDFUNDS IN TOTALEARMARKED FUNDS, 1970-86 a (percentage)

1970 1975 1980 1981 1982 1983 1984 1986 1988

NationalLeval 44.7 e6.2 72.2 72.4 74.7 70.8 71.9 73.5 74.7

A. Tax on Imports C.I.F. 1.3 4.1 8.1 7.5 8.4 6.8 8.6 8.6 10.6 Caja Agraria 0.0 0.0 0.0 0.0 0.0 0.0 1.1 1.1 1.2 IFI 0.0 0.0 0.0 0.0 0.0 0.0 1.1 1.1 1.2 PROEXPO 1.3 4.0 7.3 6.7 7.5 6.0 8.3 6.3 8.1

B. Gasoline Tax 14.2 8.9 20.8 21.1 18.1 16.3 13.6 14.8 13.8 FondoVial Nacional (FVN) 75% 10.7 8.7 15.6 16.8 13.6 11.5 10.5 9.9 9.7 Fondod- Ceminos Vecinales (FCN) 10X 1.4 0.9 2.1 2.1 1.8 1.5 1.2 2.0 1.6 Ferrocarrileg Nacional i0X 1.4 0.9 2.1 2.1 1.8 1.6 1.2 2.0 1.6 Corp.Fin de Transports6% 0.7 0.4 1.0 1.1 0.9 0.8 0.6 0.9 0.8

C. Tax on International Travel 0.7 n.a. 1.1 1.6 1.9 1.7 1.6 1.4 1.0

D. 4X Tax Coffee Exports 19.1 12.9 4.4 3.2 2.7 2.8 2.8 3.3 6.3 Fondo Nacional del Cafe n.a. n.a. 3.7 2.7 2.3 2.2 2.3 2 8 4.4 FEDECAFE,zonas cafeterias n.a. n.s. 0.7 0.6 0.4 0.4 0.6 0.6 0.9

E. 5% Tax on Tourism 0.0 0.7 0.4 0.5 0.5 0.5 0.1 0.1 0.1

F. 16% Tax on Movie TheaterTickets 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.1

G. Contributionsand Participations 2.8 2.2 1.5 1.0 1.6 1.9 1.9 3.0 2.7

H. Special Funds 0.1 0.6 0.1 0.1 0.1 0.4 0.8 1.2 2.4 b/

I. Revenue Sharing 6.4 36.9 36.7 37.7 41.4 41.7 43.1 41.2 38.9 Tax Transfer 0.0 23.4 23.8 27.4 28.3 29.9 28.7 26.9 22.6 Sales Tax Allowance 6.4 12.6 11.9 10.3 13.1 11.8 14.3 16.3 18.5

DepartmentLevel 48.7 29.6 23.4 23.1 21.0 23.8 23.5 22.7 23.2

Municipal Level 8.6 5.3 4.4 4.5 4.3 5.4 4.6 3.8 2.1 C/ a. See footnoteto Table2 in StatisticalAppendix for definitions. b. Includessome miscellaneousreallocations under Law 55. c. Understated becauseof missingdata.

Source: Estimates based on Table2 in StatisticalAppendix. - 23 -

funds--the tax on cinema tickets and the fees on registration of public documents. Moreover, in subsequent years, still further RDEs have been established e.g., the allocation of the VAT on cement for housing and water/sewerage and the national valorization tax on public works to be allocated for municipalities. Lastly, in numerous cases, the earmarking recipient agencies have ended up carrying out the activities mandated by Law 55 themselves. As examples, the Law 55 stipulation that a portion of gasoline tax revenues be used for road maintenance and repair effectively keeps the funds within the highway fund (Fondo Vial Nacional). Many of Law 55 projects in integrated rural development or in training for the Ministry of Education are carried out by SENA itself on a contract basis since other agencies are not prepared to carry out the work. While some of ICBF's monies were reallocated to Ministry of Health programs (e.g. , children's , control) 4a 1985-87, thereafter ICBFs Law 55 money will be dedicated to its own 'hogares program' of nutrition and daycare support (see paragraph 53). In sum, the reallocation of resources under Law 55 is in reality much smalier than appears or the surface; only in the case of PROEXPO has there beern a significant shift of resourzes (in this case, out of subsidized export credit toward a separate program of export subsidies).

Considerationof Social Security and Public Enterprise Profits

10. Considerationof the traditionalsources of earmarking--asin Tables 3.1 and 3.2 above--understatesthe true extent of the earmarking of public funds in Colombia.4 Public enterprise profits and the payroll tax for social security and SENA, ICBF, etc. are the major reasons for the understatemenc. Net profits of public enterprises (i.e. gross receipts net of operating expenses and taxes paid to government)may be considered earmarkingbecause they are left to the control of the enterprisesdespite the fact th,.tthey are owned in whole or in part by the government;that is, the government cannot or in practice does not attempt to allocate such resources at its discretion. Similarly,payroll taxes for social security imply an obligation to pay pensions and medical expenses now and in the future and the governmenthas no discretion about how payroll tax receipts are to be used. Depending on the rates charged relative to the mortality and morbidity problems faced by the population,present revenues may be greater or less than the present payment obligations. That part of the payroll tax that is allocated to SENA and ICBF commits the public monies to funding a variety of activitiesin training,nutrition, daycare, and social welfare. In reality, therefore,such profits and payroll taxes are little different than the traditionalcases of earmarking.

11. The result of adding these non-traditionalsources of earmarking to the traditionalsources is shown in Tables 3.3 in the text and 6 in the StatisticalAppendi.:. The extent of earmarki.±gin public income5 is, of course larger at all levels of governmentbut, most particularly,at the

4/ Quite apart from the fact that some relativelyminor funds are not included in our data. S/ Public income is derived by adding to the tax and non-tax revenuesof Table 2 the net operating profits of non-financialenterprises and payroll taxes. - 24 -

national and municipal levels. Earmarkingin total public income was 46Z in 1980 and rose to 55Z or more in 1984 and 1985. Over the period, the share of earmarkingcontributed by payroll taxes and public enterprise operating profits has been increasingwhile the share contributedby traditionalRDEs has been falling; by the end of the period the split between traditional and non-traditionalsources of earmarkingwas roughly fifty-fifty. Increasingpublic enterpriseprofits at the national and municipal level have been the major cause of this trend. In 1985, the major contributorsto traditionaland non-traditionalearmarking were in descending order of importance;public enterpriseprofits (27Z), payroll taxes for social security and SENA/ICBF (24Z), revenue sharing (20Z), departmentalalcohol/tobacco/gambling taxes (11Z), gasoline tax (7Z), import duties for PROEXPO, etc (42), the municipal valorizationtax (2Z)6 and the ad valorem tax on coffee exports (2%). Together these eight sources account for about 972 of total earmarking. Out of the large number of other cases of earmarking,no single source contributesas much as one percent to the total.

C. The Experience of Earmarking

The Origins of Earmarking

12. The extent of earmarkina found in Colombia is the product of several important features of its polit'-^--decentralization,political instability,and an elaborate budgetary process. Since Independence, Colombian politics have been dominated by powerful and opposing political forces favoring centralizationand decentralizationof power. Over the

Table 3.3: SHARES OF TRADITIONALAND NON-TRADITIONALEARMARKING IN PUBLIC INCOME (percentages)

1980 1981 1982 1983 1984 1986 1986 1987

Share in national public income 39.3 38.2 43.3 43.0 64.2 49.9 44.9 46.0

Share in departmental public income 90.9 88.6 89.3 88.4 82.0 80.8 -- --

Share in municipal public income 61.4 65.0 e6.9 67.9 63.4 67.2 -- --

Share in Public Income All levels 4656 44.6 48.7 49.1 57.8 54.6 -- --

a Public income is the sum of tax and non-tax revenues,operating profits of public non-financialenterprises, and payroll taxes for social securityand for SENA, ICBF, etc. The ad valorem coffee export tax is added to nationalpublic income 1970-83 since these revenuesdid not pass through the budget in those years. The earmarkingconsidered here is traditionalearmarking pl.a non-traditional earmarkingwhich includes non-financial public enterpriseprofits and payroll taxes for social securityand SENA, ICBF, etc.

Source: Estimates based on Tables 1 and 2 in StatisticalAppendix.

| Also includes Bogota'staxes on beer and tobacco. - 25 -

long sweep of one hundred or more years, centralizationhas been the dominant tendency, producing a central government that carries out the majority of public activities,collects an even larger share of public incomes, and dominates the decision-makingprocess on economic and social questions. In the last two decades, however, there have been greater efforts at decentralization,the hiving off of government activities to special agencies and to lower levels of government. This has necessitated a shift in public revenues--partlyon a discretionarybasis and partly on a more automatic basis through the sharing of government revenues or the setting aside of particular revenue sources for particular activities.

13. Colombia's politicalhistory has also been dominated by strong opposing forces, producing periods of extreme tension and often violence. In a climate of uncertainty--whenthe poll'ical situationcan shift abruptly and sharply, earmarking provides a means for providing somewhat greater assurance that certain importantactivities will be carried out. The wishes of one administrationor congress will be carried on to the next unless explicit measures are taken to stop the activity.

14. Last, but very important, is the budgetary process. As explained in Chapter 2, the process, before the new organic law of the Budget, was extremely cumbersome. For agencies faced with this process, earmarking has a number of advantages.7 It gives them greater certainty about the amount and timing of resources they will receive during the year. It is easier for them to forecast the amounts of the earmarked tax (even acknowledging possible errors in those forecasts)than to forecast the outcome of the budgetary process during the year. Moreover, procedures are easier for earmarked funds. Earmarked taxes may be collected by the Division of National Taxes and distributedto the relevant agency or, in some cases collected directly by the agency itself. In either case, earmarkedmonies are subject to far less detailed controls L.i3ngeneral budgetary resources. In sum, earmarking gives recipientsmore certainty about available resources, allowing them to plan their spending programs more effectively and to enter into longer-termcontractual arrangements.

Recommendationsof Previous Government Commissions

15. The Commission on IntergovernmentalFinances of 1981 (the Bird- Wiesner Report)8 took a generally skepticalview of the earmarking of taxes, but recognized that such earmarkinghad its origins in the weakness of Colombian budgetary processes and could be justified--insome cases--by appeal to public finance principles. Budgetary processes are designed to curb deviations and dishonesty and the result is bothersome,time-consuming and uncertain. Earmarking provides simpler and clearer procedures and more certain outcomes. The Bird-WiesnerReport took the pragmatic view that a more perfect world of rational budget processesmoving resourcesto their highest-priorityuses is not readily obtainable and earmarkingmay be justified because it works and because it protects priorities. Thus Bird-

/ These points were cited in numerous interviewswith officialsfrom agenciesthat are recipients of eormarkedfunds. I Eduardo Wiesn-r and Richard Bird. 'Fnanzas Intergubernamentaleson Colombia",Departamento Nacional do Planeaci6n,1981. - 26 -

Wiesner approached earmarking on a case-by-case basis asking whether the following criteria were fulfilled: (a) is the benefit principle involved with a clear identity of tax/price payers and beneficiaries? (i.e. it is not appropriate to use earmarking where benefits are not easily assigned or where income redistribution objectives are involved); (b) does earmarking provide the sole source of financing (or, at least the preponderance of funds) for the entity involved? where large amounts of funds from the general budget are involved, earmarking may not be doing any harm but it is also not clear that it is having any impact on the allocation of resources; and (c) is there a match between the revenues generated and requirements of the recipient agency or is there a systematic tendency for the tax/pricing procedures to under- or over-fund activities?

16. With these criteria in mind, the Bird-Wiesner Report made a number of recommendations:9

o Earmarking which takes the form of allocating a fixed percentage of total government expenditures (or total government revenues) to a specified activity (e.g. justice or education in Colombia's national budget) serves no justifiable purpose and should be eliminated.

o Revenue sharing arrangemenss between levels of government may be justified but the sharing formulas need to be modified to assure that no jurisdiction obtains amounts in excess of its needs.

o Some uses of the payroll tax can be clearly justified on benefit principle grounds (e.g., SENA) but others (e.g. ICBF and the Cajas de Compensaci6n) clearly cannot and in no case was it clear that the payroll taxes were generating the appropriate amount of resources for the sector.

o Earmarking of taxes with no clear benefit rationale cught to be abolished; this would eliminate, for example, the use of departmental taxes on beer and alcohol for the health sector or the use of tobacco and amusement taxes to support sports.

o There is no justification for leaving public enterprise sector profits at the disposition of the enterprise and public enterprises should be subject to income taxes.

17. Since the Bird-Wiesr.erReport, there appears to be growing pressure to eliminate or at least substantially reduce the role of earmarking in Colombia. Starting from the same set of criteria as Bird- Wiesner, the Comisi6n del Gasto Publico (1986) comes to a much harsher conclusion. Having examined most of the prominent cases of earmarking, it finds that cases of identity between payers and beneficiaries are in fact the exception (only the funds for highways and aviation seem to fit this criterion very closely); cases of extremely tenuous or non-existent connections predominate (even SENA and PROEXPO fail to produce a significant overlap). Moreover, the Commission is very skeptical of the

9/ Eduardo Wiesner and Richard Bird. FinanzasIntergubernamentales en Colombia",Departam.nto Nacional de Planeaci6n,1981. - 27 -

argument justifying earmarking by reference to Colombia's inefficient budget processes. Given these ir.efficienciesit may be desirable, from any agency's point of view to obtain earmarked resources, but there is no evidence that recipients of earmarked resources are more efficient than other agencies or that--g ven the lack of identity between payers and beneficiaries--the appropriate amount of resources is being allocated to these sectors. Hence, it is not clear that earmarking leads to an efficiency gain from the point of view of the country.

18. On the basis of the above arguments, the Commission conc]udes: (a) earmarking should be eliminatedl ; (b) the budgets of all agencies ought to be subject to the same budgetary process; (c) improvements in the administration of the revenue sharing a.rangements appear to be desirable. Wlhile not pursuing this matter further, the Commission also expressed concern over the effects of payroll taxes on the cost of labor and hence upon employment.

D. An Evaluation of Earmarking

Revenue Sharing

19. One of the prominent characteristics of Colombian public finance is the significant transfer of revenues between levels of government and from the various levels to a number of decentralized agencies. In the main, the flow is from the central government proper to its own decentralized agencies and to lower levels of government and their decentralized agencies. Part of the transfers are discretionary, the amounts being decided each year as part of the budgetary process, but a growing proportion of transfers are automatic, the result of turning over revenues from specified taxes or fees to decentralized agencies and the sharing of revenues between levels of government according to predetermined rules. The dominant factors in the latter have been the tax allowance (situado fiscal) and the sales tax transfer (cesi6n de ventas or IVA). Together, their share in central government RDEs and in central government current revenues grow from 14Z to 56Z and 2Z to 15X between 1970 and 1987. The tax allowance is roughly a third larger than the sales tax transfer, but the latter has been growing more rapidly in recent years.

20. The Tax Allowance (Situado Fiscal). Law 46 of 1971 provided that a portion of annual ordinary revenuesll be allocated among departments, national territories, and the Special District of Bogota with 302 to be shared equally among 33 jurisdictions and 70% to be divided on the basis of population shares as computed from 1964 census. Law 46 called for the shares (of ordinary revenue to be transferred) to be 13%, 142, and 15% in 1973, 1974 and 1975, thereafter increasing by 2% per annum to a maximum of 252 provided that ordinary revenues were rising by at least 15X per year. These 2% increases were never formally implemented and the proportion shared has varied around 15Z since 1975.

10/ Reference here seems to be to a subset of earmarkingthat excludes revenuesharing, payroll taxes (except for SENA, ICBF, and the Cajas de Compensacl6nFamiliar), and the profits of public enterprises. 11/ Ordinary revenuesare defined as current revenuesnot earmarked by logienftionfor any specific purpose or use. - 28 -

21. Law 46 also stipulatesthat 74Z of the amount transferredeach year shall be used to fund the operating costs of primary education and 26X to finance those of the health sector. The national government retains the right to alter the split between education and health, but has not done so specifically. The departmentaland territorialgovernment have a legal obligation to provide, in addition to the tax allowance,the same share of their ordinary income to finance primary education and health operating costs as they provided in 1972. According to the law, the funds provided through the tax allowance are to be administeredthrough the Regional Education Funds (Fondos EducativosRegionales or FERs and Section Health Services (ServiciosSeccionales de Salud or SSSs). Despite the seeming tightness of the legal provisionsfor implementingthe tax allowance, there has been considerableroom for flexibilityin how they have been interpreted. There are a number of reasons for this: first the requirementsof individual FEPs and SSSs are consideredand the amounts are totalled to make sure that thwy do not exceed the amount available to be distributed. Second, where the sum of the tax allowance and local resources in any jurisdictionis not sufficientto meet requirements, supplementscan be added. Similarlywhere there are excesses, the excess may be diverted to the finance of investmentwith the approval of the Ministry of Finance. Third, the process of additions to the initial budget distorts the process still further since the additions do not follow the sharing formula precisely.

22. As a result, there is considerablevariation in the proportion shared, the share going to each jurisdiction,and the split between education and health by jurisdictionor in total. (These results show up clearly in StatisticalAppendix Tables 7 through 9). Moreover, there appear to be a number of patterns in the data:

o Departments consistentlyreceiving more than implied by the sharing formula: Antioquia, Bogota, Choc6, C6rdoba, Cundinamarca, Norte Santander, Santander.

o Departmentsconsistently receiving less than implied by the sharing formula: Atlantico,Bolivar, Boyaca, Caldas, Cauca, Magdalena, Nariflo,Quindio, Risaralda, Sucre, Valle de Cauca.

o The territories'share in the tax allowance is always greater than their share in populationbut they consistently (approximately85% of cases) obtain less than implied by the sharing formula.

The remainingdepartments have had a more mixed experience of positive and negative differences from amounts implied by the sharing formula. Lastly, Table 9 (S.A.) shows that the shares devoted to education and health in the tax allowance have deviated--quiteoften substantially--fromthe formula. For the total, education shares have varied from a low of 71% (1986) to a high of 80% (1980) with health shares, of course, varying conversely. For individualjurisdictions, the variations have been even wider.

23. While the flexibilitywith which the situado fiscal sharing formula has been applied is welcome, the formula contains a number of flaws--regardingits treatmentof small versus large population areas and its failure to provide incentivesfor departmentalresource mobilization-- which should be corrected. First, the sharing formula for the tax - 29 -

allowance favors less populated jurisdictions at the expense of more populated areas as compared with a formula based on population alone. The full dimensions of the bias toward areas with smaller populations may not be appreciated, however. Given the shares to be allocated on population and non-population bases, the formula says that all jurisdictions whose population is greater than 3.037 will share in the tax allowance in a lower proportion than their share in total population while all those with populations less than 3.03Z will obtain shares in the tax allowance that are higher than their share in total population.1 2 Thus Antioquia with 14.2Z of the population gets 10.857 of the tax allowance while San Andr4s with 0.1Z gets 0.98Z. Put another way, according to the sharing formula, the per capita transfer that a jurisdiction receives varies inversely with its share in the total population. As compared with Antioquia (14.27 of the 1964 population) residents of Norte de Santander (3.1Z of the population) would receive 307 more; Sucre (1.8Z) 587 more; Caquet& (0.62) 190X more; and San Andr6s 1180Z more. However desirable it might be to make some additional resources available for smaller areas (say to create and maintain a critical minimum level of public education and health services or to compensate for the dispersion of population), it would be better to do so on the basis of estimated costs of compensating for smallness. The differential gains and losses for various jurisdictions cited above is not related to wealth or poverty in any systematicway and cannot be justified on any rational basis.13

24. A second flaw in the application of the formiula is the continued use of the 1964 census for the population-based part of the formula. This would favor those jurisdictions whose populations have grown relatively slowly at the expense of rap-ily growing areas. It would be desirable to use the 1985 census when applying the formula.

12/ Given that 30% of the tax allowanceformula is shared equally by jurisdictionsand 70% is allocatedon the basis of populationand that thereare 33 jurisdictions,the differencebetween a formulrbased on populationand the present tax allowanceformula is __7 | 7_ Pop 33 Pop .s Pi .3 =_ - __ , when Pi is the populationof the tth Pop 33 jurisdictionand Pop is total population. This expression is zero when the populationshare of the tth district is 3.03%, positivefor greater populationshares, and negativefor smaller. Not, that the point at which the expression is zero depends solely on the number of jurisdictionsand is Independentof the shares allocatedon a populationand non-population basis. 13/ Some additionalnotion of the arbitrarinessof tho formula may be obtainedfrom recognitionof the fact that the cutoff populationshare--separating jurisdictions receiving above- and below- average per capita transfers--dependssolely on the number of jurisdictions. The number presentlystands at 3.03% because there are 33 jurisdictions,but if--forwhatever reason- -Colombiashould decide to consolidateto (say) 16 jurisdictions,the cutoff populationwould become 6.2SX and areas would begin receivinghigher or lower per capita amounts for no reason other than a change in the number of departments/territories. - 30 -

25. The third flaw in the sharing formula is that it fails to encourage greater resource mobilization efforts at the local level. A factor needs to be added to the sharing formula to provide incentive for localities to add resources to those provided from the national level.14 The present formula does not encourage greater tax effort on the part of departments and available evidence suggests that local tax effort has been reduced. For exam,p.e, a 1978 study by the Foundation for Higher Education and Development (FEDESARROLLO) found that the non-matching nature of the formula would in general tend to discourage local effoft and that the evidence shows that the tax allowance has had no effect on the level of department expenditures in education and health.15 A study done by Slack and Bird, as part of their background work for the 1981 Mission on Intergovernmental Finance, shows that a one peso increase in the tax allowance leads to an increase in expenditures less than one third of 1Z as large, suggesting that the tax allowance substitutes for local resources.16

26. Sales Tax Transfer (Cesi6n a las ventas). The sales tax transfer was created by Law 33 of 1968, and through a series of laws during the 1970s. The uses to which the monies could be put and the proportions devoted to each were altered. Basically these resources were devoted to social security pensions (departmental), secondary education, and most importantly, the municipalities.17 Because of the size and elasticity of the sales tax, the transfer became an important source of municipal finance.

27. Law 12 of 1986 is designed to bring about improvements in the finances of smaller cities and brought major changes in the sales tax transfer.18 Law 12 has three important objectives:

o To strengthen the smaller municipalities, both by giving them more resources and by encouraging their own fiscal efforts.

o To transfer to municipalities certain national responsibilities equal in amount to the amount of additional revenues transferred.

14/ Sea R.M. Bird, IntergovernmentalFinance in Colombia--FinalReport of the Mission on IntergovernmentFinance (InternationalTax Program,Law School of Harvard University,Cambridge, Massachusetts,1984), pp. 148-163 and N.E. Slack and R.M. Bird 'Local Response to IntergovernmentalFiscal Transfers--ThoCase of Colombia"Public Finance (No. 3, 1983), pp. 429-437. 16/ Fundaci6nPare la Educaci6nSuperior y el DesarrollovEl Efecto del Situado Fiscal y la Cesion del ImpoventesSobre los Ingresosy Ia Estructuradel Gasto Departmentaly Municipal' (unpublished,Bogotg, October 1978). i,/ N.E. Slack and R.M. Bird, op. cit. 17/ In 1983, the sales tax was expanded to a value-addedtax (IVA). Vta Decree 232 in tlat some year, 30% of IVA revenueswere allocatedas follows:25% to departmentalmunicipalities via the 30-70X sharing formula,1.6% for the nationalterritories to be investedin local government, and 3.6% for tF^ social securityfunds of departmentalgovernments. In addition part of the revenuesthat were to have gone to municipalitieswere divertedto departmentalFERs in the followingproportions: 30% for municipalitiesof 100,000-S00,000population and 60X for those over 600,000. 18/ The discsalionof Lsw 12 (1986)--itscontent and >hects and the issues it raises--closely follove IBRD Colombia- A Review of Recent DecentralizationMeasures (Washington,D.C.; April 8, 1987), pp. 33-46. - 31 -

o To increase the efficiencyof the local sector; by reducing the number of agencies working at the level and by increasingthe capacity of local institutions.

28. The major changes introducedby Law 12 were: (a) the increase in the proportion shared of the value-added tax from 30% to 50% by 1992 and especially the share going to small cities (see Table 9 in the Statistical Appendix for details);19 (b) the introductionof "a tax effort" formula for small municipalities; (c) the use of populationas reflected in the 1985 census as the sole basis for sharing (modifiedby "effort' in small municipalities)--i.e.the eliminationof the 30X share that formerly was divided equally among departments; (d) the additionalfunds for municipalitiesin excess of 25.8% (1986 share) to be spent on investments (includingmaintenance, debt service and investment in FFDU bonds); (e) new shares for the Public AdministrationInstitute (ESAP) and National Cadastre Agency (IGAC),agencies which provide training for local level officials and revised estimates of property values respectively;(f) and a timetable for the national government to reorganizeand transfer functions toward the 1992 deadline.

29. Law 12 representsa welcome attempt to transfer greater resources to municipalities,give them incentivesto raise their own resources (at least for smaller municipalities),and initiate some measures to increase municipal absorptivecapacity. The full implementationof Law 12 will have a major impact on the finances of municipalities. By 1992, municipalitiesin the aggregate will receive 85% more resources than they would have under the old sharing formula.20 As such, however, the Law raises a number of issues about:

(a) the form of the sharing formula for small municipalities;

(b) the restrictionof tax effort just to those municipalities;and

(c) the speed with which municipal absorptivecapacity can be increased.

30. First, the eliminationof the 30X portion of IVA shared equally by departmentsand the introductionof a 'fiscal effort" factor for small municipalitiesare both improvements. However the Government'soriginal sharing formula for small municipalitiesis flawed in that it provides an incentive for raising property tax rates but not for increasingcadastral valuations.

31. Secondly, the "fiscaleffort" formula is too narrowly focussed both as to the taxes covered and the types of municipalities. By focusing on property taxes, it not only fails to encourage the developmentof other tax bases but may actually encourage the substitutionof property taxes

19/ Note that Law 12 maintainsthe retentionfeature of the previousallocation system in that for cities with populationof 100,o0-600e,000,30% of their potentialtransfers will be redirected to departmentFERs while for cities over 600,000,the amount would be 50%. 20/ See IBRD (1987),op cit, p. 39 - 32 -

for other types of taxes with little overall gain in local resource mobilization. To forestall this problem, considerationshould be given to including all municipal revenue sources in the formula, even while continuing to use cadastralvalues as the (actualor potential) tax base. Moreover the applicationof "fiscal effort" to small municipalitiesalone is undoubtedlyundesirable. Large municipalities(above 100,000 inhabitants)bulk large in total local government finance and the lack of an incentive for effort invites them to substitutenational resources for local resources.

32. Thirdly, Law 12 and its follow up Decree 77 (1987)which identified those public services to be transferredto local government,21 reassignedresponsibilities among various levels of government and presented timetables for the alteration/liquidationof a number of government agencies,have mandated a major shift in responsibilitiesand in available resources among levels of government. At this stage it is not yet clear whether local governmentsare well prepared to handle the additional responsibilities,especially within the short time-frame available.

National GovernmentLevel Earmarking

33. Coffee Fund. The Fondo Nacional de Cafe (FNC) was created in 1940 in response to the new "QuotasAgreement" of the InternationalCoffee Organizationwhich called upon signatoriesto establishmechanisms for withholding surpluses above quota limits. Management of the Fund is by the National Federation of Coffee Growers, a private non-profit associationrepresenting almost all coffee producers. The Federation is responsiblefor the domestic and external marketing of Colombia's coffee. The nature of its activities and its relationshipwith government are spelled out in ten-year contractswith government,the latest of which co'Tering1989-99, was signed in December 1988.

34. The objective of coffee policy is the stabilizationof grower's real incomes at levels consistentwith long-run equilibriuminternational prices. The major stabilizationinstruments--the ad valorem export tax and the retention quota--alsogenerate important sources of income for the Fund. Of the 6.5% tax on the value of coffee exports, 2.52 goes to the central government general budget and 4Z to the Coffee Fund, the latter thus representingan earmarked tax. In line with government policy, the ad valorem tax share going to the central governmenthas been falling-- from 22Z at the inception of the tax in 1967 to 82 in 1980 to 2.52 in 1983. The 4Z tax rate, 0.82 (or 202) of which is reallocatedto Department Growers Committees to be used for economic and social projects in the growing areas, has been constant since 1967. The retention quota is a

21/ Among the public servicesto be transferredto the local level are: provisionof drinkingwater, sanitation,garbage disposal;management of morkets and slaughterhouses; construction, operationand maintenanceof health posts and centers and regionalhospitals; construction and maintenanceof schools and athleticfacilities; provision of agriculturalextension and technicalassistance; constructionand maintenanceof feeder roads, bridges,and river ports; extensionof electricityservices; constructionand up-gradingof low income shelters; and cofinancing(with central agencies)of rural developmentprojects. - 33 -

fixed proportionalamount of coffee--involume or equivalentvalue--which exporters must give to the Coffee Fund for every 100 bags of coffee exported. The retention tax ects like a tax on exports and because its proceeds are directed to the coffee fund it should be viewed as an earmarked tax.

35. Table 3.4 shows the amount of revenues the Coffee Fund obtained from its 42 share of the ad valorem export tax and the retention quota.22 The former is usually listed in any accountingof central government earmarked taxes, but the latter is not. Because of discounts and bonuses the Fund gives on private exports and because a significantshare of exports in any year comes from the Coffee Fund itself, the retention figures in the table are likely to exaggerate the net income accruing to the Fund. Bearing in mind, therefore,that these are upper bound figures,23 they would represent an amount equivalent on average to over 45% of total national government earmarked revenues 1980-87.

Table3.4: EARMARKEDTAXES FOR THE COFFEE3ECTOR, 1111-19-7

1980 1981 1982 198l 1984 198s 1986 1987

Ad ValoremExport Tax 3005.0 2823.0 3112.6 3377.0 4698.6 7298.7 14905.2 11,671.1 RetentionQuotas 31116.2 9596.9 23631.2 29262.1 61,742.0 81,607.4 174,143.8194,199.4 In Kind 31098.2 9688.8 23510.6 29146.6 61,639.9 80,821.1 172,667.8 n.a. In Voluo 18.0 8.1 20.7 11S.6 202.1 786.3 1676.0 n.a.

Total 34121.2 12419.9 26843.8 82639.1 68,440.5 88,9068.1189,049.0 205,870.6

Source: StatisticalAppendix Table 12.

36. The Fund engages in a number of activitiesbeyond its stabili- zation function (See Chapter 2). Through the DepartmentalGrowers Committees, it channels resources to health, education, rural electri- ficatimn, and infrastructurein the coffee growing areas. Investmentsare decided upon in consultationwith the national government. Whatever assets are created in this process are owned by the government and operated and maintained, using the government'srescurces, as part of the normal public system.

22/For a more completepicture of the CoffeeFunds' revenues and expenditures,see Statistical Appendix,Table 12. 2/ Theyare also an exaggerationof retention'scontribution to earmarkingbecause at timesthe CoffeeFund has loanedmonoy to the centralgovernment. - 34 -

37. It is clear that the Fund's activities have taken it far beyond the stabilization objectives envisaged some 50 years ago. Movement back toward stabilization and other coffee-sector related activities would be desirable. The recently signed contract with government (covering 1989-1999) is an inportant attempt to refocus the Fund's activities and to increase the government's influence on its operations. The outstanding features of that agreement are: (a) the Fund's budget is to be approved by the National Coffee Committee (which has as many government represen- tatives as industry representatives) and the Minister of Finance; (b) the Fund's different activities will be accounted for more explicitly by the objectives they ser-ve; (c) its investments will be restricted to those which directly benefit the coffee growing areas and a list of non-related holdings and a schedule for their sale by a target date will be drawn up; and (d) the Minister of Finance will exercise veto power over commercial policies related to coffee. These are important steps to returning the use of public money toward legitimate public sector activities such as. stabilization and economic and social welfare projects together with trade association type functions (e.g. quality control, export promotion). One might still wish to question the use of export taxes to finance public investment in coffee areas. While there is a general feeling in Colombia that the monies for such investments have been utilized effectively, there is no reason to suppose that coffee taxes will generate the right amount of economic and social infrastructure in these areas. It will be important to monitor such investments to see that they do not result in overbuilding or a maldistribution of infrastructure between regions. Lastly, subsidies for the domestic consumption of coffee remain a questionable use of public money and the provision of subsidized fertili7c- has been shonm in many other countries to lead to wasteful use. Continued progress toward the elimination of both these subsidies would be desirable,

38. NationalHighwayus Fund (Gasoline Tax). The Fondo Vial Nacional (FVN) was created by Law 64 of 1967 in response to a sharp decline in road infrastructure investment due to a fragmented public works budget and lack of continuity in budget resources. Its mandate was to provide funding for the planning, constructior, and maintenance of the national highways and inland wraterways networks.2 4 Under Law 64 of 1967, as modified by Law 30 of 1982, FVN receives 75Z of the earmarked tax on petroleum automotive fuel while 10% goes to the Fondo Nacional de Caminos Vecinales (or FNCV), 10% for the Ferrocarriles Nacionales de Colombia (or FNC), and 5Z for the Corporsci6n Financiers de Transporte (CFT) to fund urban transport operations.

39. During the 1980s, expenditures on the construction and maintenance of national highways have not expanded in real terms and hence have not kept pace with the expansion of general economic activity (see Table 3.5 below. The failure of gasoline taxes to raise sufficient revenues and the diversion of increasing amounts of FVN resources to debt service appear to be the major reasons. Article 2 of Law 30 (1982) modified the gasoline tax from a 5X rate based on the retail price to a specific amount equivalent to Colombian $13.50 per gallon of regular motor gasoline and

24/ For its onponcst-es and sources of revenue see Tables 13 and 14 in the Statistical Appendix. - 35-

diesel fuel (gasoline for aircraft and marinc dicse) were exempted). The amount of the tax is to be adjusted each time the gasoline price changes by an amount which takes into account the percentage change in the prices of gasoline and diesel and the percentage change in the heavy construction price index. Prices and taxes have in fact been adjusted annually but not enough to keep up with inflation generally and revenues from the gasoline tax have been falling in real terms (see Table 3.5 below). The rise of debt service (mainly on external debt) has beeni a recent phenomenon. From merely 0.7Z of FVN's expenditures in 1980 it rose to 10.4Z in 1986 and 17.402 in 1987 (see Table 13 in Statistical Appendix). While transfers from the central government budget have been rising, they have not been sufficient to offset the squeeze from fallirng gasoline revenues and rising debt service and FVN expenditures on construction and maintenance were lower than in 1980 in all years but one (1981); even a major expansion in 1987--financed mainly by external borrowing--left expenditures some 12% lower than in 1980.2

40. Gasoline taxes are in general a legitimate application of the benefit principle of taxation. There normally is a strong link between the tax Jase and the service received from the highway system because the road user has to buy gasoline. Gasoline taxes are, therefore, taxes in lieu of road user charges. While it is difficult to calibrate taxes paid to the exact amount of damage done by different types of uses (trucks causing much greater damage than private passenger vehicles), reasonable approximations can be established and a significant proportion of the cost of constructing and operating the system recovered from the users. In Colombia, however, the link between gasoline taxes and road maintenance has beer: broken: First, road users do not pay gasoline taxes since the price paid by gasoline consumers is quite similar to the CIF import price. Second, the price received by the producer of gasoline, ECOPETROL, is below the CIF price by an amount similar to the gasoline tax, i.e., gasoline taxes are paid by ECOPETROL, not by road users. (Table 3.5).

41. The general impression is that FVN's resources are well managed. Projects are subject to feasibility studies, and, while there is room for improvement, there is no reason to suggest that a significant proportion of projects would fail to meet economic criterion. Moreover, highway officials are convinced that the existence of earmarked funds protects them from the complexities of the budget process and allows themi to engage in long-term contracts with the private sector that would not otherwise be possible. However FVN (and FNCV) have fallen victim to the same problems highway funds in many LDCs have faced--the failure of the automatic tax arrangement to generate the appropriate level of revenues to carry out the long-term construction And maintenance program. In the late 1970s, it appeared that the gasoline tax was generating too much resources;2 6 in the 1980s, falling real gas taxes--and rising debt service--appear to be

26/ Similar results would be obtained if the deflatorfor heavy construction activities wero usod instead of the GDP deflator. 26/ See Eduardo Wiesner and Richard Bird. OFinanzasIntergubernamen'fles en Colombia",Departtmento Nacional de Planeaci6n,1981. - 36 -

Table 3.5: COLOMBIA - AVERAGE GASOLINE PRICES (US$ per bbl.)

1988 1989 Jan.1990

Prices

Retail price 21.18 21.44 24.38 CIF price 19.68 23.17 24.98 ECOPETROL price 13.50 13.70 15.29 (Retail-ECOPETROL) 7.68 7.74 0.09

Taxes and distributioncosts

Road Tax 4.85 4.91 5.87 Other taxes 1.08 1.08 1.27 Distributioncost 1.75 1.75 1.95 Total 7.68 7.74 9.09

leading to a shortage of resources. If highway funds are to be retained in the future, it would be desirable to set gasoline taxes (and other road user taxes such as tolls, licenses, and registrationfees), such that the highway program can be carried out.

42. PROEXPO. PROEXPO is one of a three part set of export incentives which has been in existence for more than twenty years and which includes Plan Vallejo, which provides registeredexporters with access to duty free imports; the CERT system which provides rebates for domestic indirect taxes (and tariffs as well for unregisteredexporters); and PROEXPO itself which provides subsidizedcredit for exports. It offers 33 distinct lines of credit for: pre- and post-shipmentfinancing; long-term investment, working capital, promotion of exports, and technical assistancefor quality control. A relativelyminor share of PROEXPO'sportfolio is in equity investments (about 20Z) and for the past two years policy has been to discourage such investments. PROEXPO does not raise money in capital markets to any extent but instead relies on rollover of existing loans (60-70Z),interest earnings (7-102) and earmarked taxes (15-20Z) to provide funding for its loans (see Table 15 in StatisticalAppendix). The earmarked tax is its share of a general 18Z tax on c.i.f. imports which is divided in the followingway: National Treasury - 10.4%; PROEXPO - 6Z; IFI - 0.8Z; and Caja Agraria - 0.8Z. Overtime,PROEXPO has reduced the extent of the subsidy in its credit lines, but the subsidies still remain sizable. Most of its credit carried a 19% rate 1980-83 and 22Z 1984-86, compared to commercialbank rates which averaged 442.27

27/ Soo I.B.R.D. Colombia- CommercialPolicy Survey, World BankReport No. 7510-CO,1988, pp. 42-43. - 37 -

Table3.6: REALOUTLAYS IN THEHIGHWAY SECTOR - FONDOVIAL (millionsof ColombiaI in 1976 prices)

1980 1981 1982 1983 1984 1986 1986 1987

TotalExpenditure 5267.4 8093.9 6888.8 5317.0 6214.6 6286.3 4862.6 6336.3

Construction, Rehabilitation and Maintenance Expenditures 3824.4 3975.1 2853.4 241656 2338.3 2667.8 2403.1 3179.2

MemorandumItem

RealGasoline Tax Revenue 4239.1 4386.5 4044.4 3260.7 3197.7 3156.1 3021.7 3243.3

GasolineTax in currentprices) (Col.S/gal.) 8.71 12.60 13.50 16.46 19.64 22.06 26.49 82.68

GasolineTax Adjusted for Inflation(Col. $/gal.0 1976 price.) 2.90 3.29 2.93 2.97 2.94 2.61 2.44 2.41

Source: Data from StatisticalAppendix Tables 13 and 14 deflatedby the GDP deflator. - 38 -

43. PROEXPO is one of the earmarked funds covered by Law 55 (1985), but part of the intended reallocationappears to have been offset through increases in resources that it has been able to obtain. According to Law 55, 20% of PROEXPO's earmarked tax proceeds are to be reallocatedto the financing of the CERT (or other financial incentives for exports) in 1985 and 1986 rising by 10% per year to a ceiling of 502 in 1989. And transfershave indeed taken place in line with this schedule. At the same time, however, PROEXPO has received a major increase in resources through a rise in its share of the tax on imports c.i.f., an expansion in the proportion of imports covered by tariffs, and a surge in impoBts. As a consequence,the amount of earmarked resources--evenafter subtractingout transfers under Law 55 was no lower in real terms in 1986 than in 1985 and was only 15% lower in 1987 than in 1985.

44 The logic of financing PROEXPO with an earmarked tax on imports is highly questionable. There is no connection between those who must bear the cost (consumersand producers using imported products) and those who are obtaining the benefits (domesticproducers obtaininghigher profits and/or foreign consumers receivingcheaper goods). Hence, shielded from the budget process where these allocationswould be evaluated against other priorities and from capital markets where the ability of PROEXPO to cover its cost of raising funds and its administrationcosts would be considered,there is no reason to suppose that the amount of loans being subsidized is any sense efficient or desirable from Colombia's point of view. The realizationthat resourcesdepend on total imports and infrequent changes on the duty on imports c.i.f. rate allocated to PROEXPO (1.5% 1968-75, 5Z 1975-86; 6Z, 1986) only enhances the suspicionthat its unlikely that the appropriate level of resources is being provided for this activity. Moreover, while at first glance PROEXPO appears to be an attempt to offset the bias against exports inherent in the protection given to import substitutingactivities, it is not evident that this is so. The subsidies to exports in PROEXPO loans are made possible by taxes on importswhich in themselveswould discourageexports. At best then, PROEXPO loans and their financingwould encourage the production of traded goods at the expense of non-traded goods. But, since not all of PROEXPO's loans are directly related to exports and since the subsidies to export vary sharply across sectors,28 the offset to greater protection is incomplete and introduces distortionsinto economic activities. The financing of PROEXPO with earmarked taxes should, therefore,be ended and the question of subsidizingcredit for exports--regardlessof source of finance--shouldbe re-examined.

DepartmentalEarmarkinR

45. The principal sources of departmentalincome are from earmarked revenues. Three of the major revenue earners are earmarked for health: beer for the SectionalHealth Services, SSS (with a minor part for the National Fund for NeighborhoodRoads); liquor for the SSS; and profits and taxes from lotteries (and race course betting) which go mainly to the SSS with a minor portion going to schools for the blind and the National Federation for the Blind. A fourth major source of departmentalrevenue-- cigarette taxes--is also earmarked for Sectional Sports Commission, of

28/ See IBRD, Colombia- CommercialPolicy Survey, op. cit., p. 44. - 39 -

which 30% must be passed on to the Colombian Institute of Youth and Sports (COLDESPORTES) which in turn is required to pass on 101 to Colombia Institute of Culture (COLCULTURA).

46. The type of tax (and lottery earnings) earmarking found at the departmental level has no rational basis whatever. There is virtually no link between taxpayers and the beneficiaries of government services provided. Hence, there is no reason to believe that tax revenues are generating an appropriate amount of resources for the activities in question. Moreover, there is no reason to believe that the tying of revenues to expenditures enables the government to raise more revenues because taxpayers see the proceeds going for a worthy cause. Taxes on alcohol and tobacco raise considerable sums in most countries regardless of how the proceeds are used. Even if the recipient agencies are worthy, the appropriate questions are the level of services required and the division of financing responsibilities between the departmental and national governments. The correct solution is not in the arbitrary earmarking of such taxes which apparently has not been sufficient to prevent the virtually complete substitution of national for departmental resources, but in the reform of the sharing formula for the tax allowance to take account of departmental needs and to induce greater local fiscal efforts.2 9

Municipal Level Earmarking

47. The majcr sources of revenue for Colombia's municipalities are the property tax, the industry and commerce tax, and transfers from higher levels of government. Other taxes (e.g. on billboards, vehicles, gambling, and public spectacles) generate relatively minor amounts of revenues. Of these sources, the major contribution to municipal level earmarking is the valorization tax, a betterment tax levied on real property that has benefitted from the construction of local public w-rks. 30

48. Valorization represents one of the purest forms of earmarking. This purity results from several characteristics:

o the fact that taxpayers and beneficiaries overlap to a substantial degree;

o the use of benefit-cost type considerations in making investment decisions ("appraisal" or "analysis' might be too strong a word) and the conscious and conscientious attempt to allocate tax assessments proportionally according to how people benefit from the project; and

29/ See our discussionof the 'tax allowancewin pars. 21. 30/ For an evaluationof the practiceof valorizationin Colombiasee W.G. Rhoads and R.M. Bird. 'FinancingUrbanization by Benefit Taxation'Land Economics43 (Ncvember1987), pp. 403-12; IBRD ValorizationCharges as a Method for Financing Urban Public Works: The Example of Bogota, ColombiaStaff Working Paper No. 264 (March 1977); and InternationalTax Program, IntergovernmentalFinance in Colombia:Final Report of the Mission on IntergovernmentalFinance (The Law School of Harvard University;Cambridge, 1984), pp. 280-282. - 40 -

o the strong link between valorization revenues and expenditures (i.e. no supplements from or diversions to general fund financing, at least in principle).

In addition, valorization appears to reverse the usual sequence of earmarking whereby available revenues drive expenditure levels: instead, the identification of desirable projects, often wthin the scope of prepared investment programs, appears to call forth efforts to raise financing from prospective beneficiaries.

49. Upon a project's selection, a "zone of influence" is demarcated, the area over which benefits are expected to be felt. Benefits are estimated--as the total resulting rise of site values--and allocated across the properties within the zone according to formulas which take into account a number of characteristics such as size, shape, topography, frontage, distance from project, and a number of economic factors. Over time the scope of valorization has been extended to include street construction, local paving, parking facilities, urban development, and what are called 'green spaces" and "zones of historical interest". In principle, there is nothing to stop authorities from attempting to recapture some or all of the project (net) benefits. but in practice only full recovery of the costs listed above' is sought. In addition, consideration is given to problems that the poor or persons with few liquid assets might have in paying the tax.

50. The valorization tax as practiced obviously has a number of desirable features--the benefit connection; the efforts to select economically desirable projects, and the flexibility with which it has been adopted in differing circumstances. As a tax on the unimproved value of land, the tax would have no adverse incentive effects. On the other hand, valorization does have some drawbacks, primarily centered around collections problems.3 1 First, quite aside from the fact that authorities attempt to recover only project costs, collections fall far short of 100Z.32 The reasons appear to be cost overruns, generous exemptions (e.g. the Catholic Church, charitable institutions, and public enterprises), and generous exceptions/payment schedules given to the poor and those with liquidity problems. Problems have particularly arisen when projects have been designed to improve conditions in low income areas. Second, the concept of the "zone of influence' has its limitations: in a number of cases, a sizable fraction of beneficiaries fall outside the zone, and this has led to collection problems because persons within the zone were reluctant to pick up the added burden. It is also hard to implement with a number of projects simultaneously affecting a number of overlapping zones. Lastly, valorization revenues have proven to be quite unstable in part because of collection problems and in part to administrative deficiencies in planning and implementing projects; thus revenue growth seems to be characterized by periods of stagnation followed by sizable increases.

51. Despite its imperfections, valorization deserves a favorable rating. Its popularity and its ability to raise revenues would justify its

31/ Sea WorldBank, World Development Report 1988 (OxfordUniversity Press; June 1988),p. l60. 32/ Apparentlythis has been the experiencewith bettermenttaxes around the world. - 41 -

continuationand extension into cities which have not taken full advantage of it. Its weaknesses could be reduced by cutting back on exemptions and the introductionof interest payments and penalties on arrears (the former was introduced in 1981). Further, where subsidizationof low income groups is a key project objective, the need for transfers should be made explicit and financed from general budget revenues.

Payroll Taxes

52. Payroll taxes in Colombia are earmarked to provide financing for two broad sets of activities--socialwelfare type activities and traditional social security. Included in the former are programs of technical skills and labor training (ServicioNacional Aprendizaje or SENA), programs of preschool education and nutrition for children and family protection (InstitutoColombiano Bienestar Familiar or ICBF) and family allowances and various private sector activities (Cajas de Compensaci6nFamiliar). Included in the latter are a variety of programs for pensions, health-maternitycare, and occupationalrisk covering employees in the private and public sectors. These are financed through payroll taxes and transfers from the government budget (the latter are especiallyneeded to cover the aeficits of the various public sector social security programs). Table 17 in the StatisticalAppendix shows the contributionof employees and the contributionof employers (as Z of wages). Several features of the contributionsystem stand out. The range of contributionis large, being quite high in the private sector and lower in the public sector, especially in the programs for the military and police. The combined employer/employeecontribution of 15-20X for social security is already quite high--seventhhighest in Latin America in 1986--andwhen the 9% tax is added to cover Cajas de Compensaci6n (4Z), SENA (22) and ICBF (3Z), total payroll taxes are adding 24-29Z to the cost of Colombian labor.

53. Funds for Social Welfare Objectives. ICBF is a public agency responsible for matters pertaining to natrition and family welfare. Its program covers such activities as food programs for children, assistance to families and children with judicial problems, aid for orphans and children with behavioral problems, and special projects for Indian minorities.33 In particular, its traditionalprogram (CAIP) of daycare and supplemental feeding for children under 6 years involved supervisedprograms in well- equipped day centers with trained personnel. The program suffered from high administrativecosts, lack of targeting for the poorest, and lack of access. To increase replicability,IBCF launched a new "programade hogares de bienestar infantii" (HBI) which involves nutritional support, daycare facilities run in the homes of volunteer mothers who are provided training and loans for home upgrading, and the targeting of the poor through the location of the facilities.34

13/ Se. StatisticalAppendix Table 20. 34/ ICBF was one of the agencies called upon to reallocateearmarked services to other activitiesin the health sector under Law 66 (1986) and actually reallocatedabout 8% of its 1986 and 1987 budgets to children'shospitals, vaccinations, and malaria control. From 1988, the reallocationsthat would have been requiredunder Law 66 will be directedto the 'hogares' program. - 42 -

The "hogares" program is supposed to expand from 45Z to 85X of ICBF's budget by 1990 and its coverage from 400,000 to 1.5 million children by 1992. The increased resources to pay for this expansion is to come from an increase in the payroll tax for ICBF from 2% and 3% and from foreign borrowing. Law 89/1989, approving that increase, was recently passed. Quite apart from unflavorableemployment effect (see below), this is a questionable use oa earmarked resources. There 4.sno connection between the beneficiaries and the payers of the payroll tax. There is therefore no assurance that the amount of taxes generated are appropriate for the activities involved. If the 'hogares program" is justified--and it appears to have many attractive features (including redistribution)-- it should be financed from general revenues where the matching of needs and resources can be more continually assessed.

54. SENA is a public sector agency responsible for the training of various labor skills. It has a decentralized structure to better enable it to tailor its programs to local labor market conditions. Financed from a payroll tax (2% per annum on the payroll of public, private, and mixed enterprises and 0.5% per annum on all government departments at the national, department and municipa; Levels) which has provided about 80X of its resources in the 1980s, it at first confined the bulk of its training activities to the needs of the entities t.aying the payroll tax.3 5 However, it has gradually expanded into construction, agriculture, and self- employment areas, areas which are only thinly covered by the payroll tax. This process has been accelerated by Law 55 which has moved SENA into broader areas of education--agricultural research and extension, artisans and handicrafts, rural devElopment and rehabilitation projects in guerilla infested areas. SENA would appear to have a benefit rationale in that employers obtain trained workers, but at best the link is not strong since employers have no options about whether or not to pay the payroll tax or how much to pay and hence there is no reason to suppose that the payroll tax generates the rigiv amount of resources for training. The likelihood of such matching grows less with SENA expanding into activities that are more remote from those that pay payroll taxes. There are a number of indications that the amount of resources going to SENA has been excessive: its training facilities are operating substantially below capacity; its agriculture centers have been overbuilt; and it has been unable to utilize its budget in most years in the 1980s. Thus, it would be desirable to cut SENA off from automatic sources of financing and force it to compete for budgetary resources with other governmental departments and agencies.

55. The Family Compensation Funds (Cajas de Compensaci6n Familiar) are private non-profit institutions, relying heavily on a 4Z payroll tax for financing. The bulk of the monies are supposed to be paid out as "family subsidies' to workers in participating firms and the remainder is to be

36/ The use of payroll taxes to finance training is popular among Latin Americanand Caribbean countries,particularly in the lower-middleper capita income bracket (US8401-1836GNP per capita). These countriescharge payroll tax rates of 0.05-3.0% per year; among them only Jamaica and Venezuela have rates as high or higher than Colombia. See J. Whalley and A. Ziderman 'PayrollTaxes for FinancingTraining in DevelopingCountries' (World Bank, Policy, Planning,and Research Working Paper No. 141; January 1989). - 43 -

used for various social programs directed at assisting the same group of workers. There are several questionable aspects of Caja activities. Their coverage is very narrow, reaching only about one-fifth of the employed population. While cash and in-kind allowances have been paid out, a large proportion of rcsources have gone into investment--e.g. supermarkets, theaters, and recreational facilities--which are luxurious and beyond the reach of ordinary workers. The use of public monies for luxury inrvestments is quite simply undesirable and the use of an earmarked tax for this purpose is again questionable. The Cajas should be cut off from earmarked taxes and forced to rely on their own resources.

56. Social Security. The social security system in Colombia consists of the Institute of Social Insurance (ISS) covering the private sector and almost two hundred funds covering civ>l servants at all levels and employees of public decentralized agencies. Although not commonly thought of in this fashion, the payroll taxes that fund these various private and public funds represent earmarked taxes in that they are the result of a contractual obligation to pay pension, death and medical benefits and there is no aiternative use to which these monies may be put (except financial investments toward the future payment of these obligations). To the extent that the funds are being run on an actuarially sound basis--i.e. that premiums charged (payroll taxes) plus interest accumulations are sufficient to cover the expected payments to beneficiaries based on age distribution, mortality, and morbidity characteristics of population, payroll taxes would be a good indication of the trend in obligations of these funds. Differentials between payments and premiums (payroll taxes) in any year could be regarded as temporary deviations. However, there are a number of reasons that this is not the case in Colombia. Firstly, a number of funds rely explicitly on the government budget for funding (the tax on the employee is small and there is no tax on the employer forcing continued reliance on beneral budget financing). Second a number of public funds--most notably CAJANAL which covers the majority of civil servants, at least at the national level--have been running continuous deficits. Thirdly, even in ISS, which has run surpluses in all years but one in the 1980s, premiums have not been increased in line with actuarial needs and pension funds have been depleted by transfers to cover the costs of medical programs. The public programs--most importantly CAJANAL--have no actuarial base and no financing program other than an ad hoc and open ended reliance on the budget. For these reasons, payroll taxes understate the extent of potential future obligations the Government of Colombia may face. Nonetheless payroll tax revenues for social security are the appropriate indicator of their contribution to total earmarking; they represented 18Z of total earmarking at all levels of government in 1980 and 19Z in 1985 (see Table 6 in Statistical Appendix).

57. Emploment Effects of Payroll Taxes. Formal sector wage determination is closely related to questions of employment and . Employment in Colombia has been shown to be sensitive to movements in real wages.3 6

36/ See El Problema Laboral Colombiano:Piagnostico, Perspectivas y Po:iticas-Informe Final de la Misi6n de Empleo (EconomiaColombiano - Revista de la ContraloriaGeneral de la Republica: agosto - septiembrede 1986) and IBRD Colombia- Country EconomicMemorandum, World Bank Report No. 6828-CO. (October15, 1987), Ch. III. - 44 -

Rising formal sector wages reduce employment by inducing a direct substitution of capital for labor and by discouraging investment through reduced profitability. Since the mid-1970s, total labor compensation in Colombian has risen by more than 502 in real terms--a large increase both absolutely and relative to the increase in labor productivity or the cost of capital which has scarcely changed. Non-wage compensation (e.g., leave arrangements, dismissal payments, and payroll taxes) have been rising faster than wage components. As a consequence, mainly because of higher real wages and slower real growth in the 1980s, the open unemployment rate has risen to 12-14Z (as compared with the 1958-79 average rate of 9.5Z) and the gap between formal and informal sector wages has widened. Urban poverty, which has been shown to be largely a question of employment in low productivity jobs in the informal sector, is likely to have increased in the process.

58. The effects of payroll taxes on employment must, therefore, be considered, both because of the relatively high unemployment rate and because future increases in payroll taxes may become necessary to put the social security system on a sound financia' and/or actuarial basis. Payroll taxes have unfavorable effects on employment by raising the cost of labor, penalizing labor intensive sectors and making it economic to substitute machinery for labor. The effects on employment will be greater the higher is the tax, the greater is labor's share in the cost of factors of production, and the greater is the elasticity of substitution. The recent studies enable us to get some idea of the amount of employment that would be generated by reducing payroll taxes. In a study prepared for the Chenery employment mission, Gomez Buendia and Cifuentes projected that eliminating the payroll taxes supporting SENA, ICBF and the Cajas de Compensaci6n would add labor force to total urban employment. Using estimates from the World Bank's later study of the demand and supply and labor in the Colombia's formal and informal sectors suggests that the addition could be in the order of 72,000-75,000 jobs. The reduction of 0.5-1.1 percentage points on a 132 unemployment later implied by these studies is not trivial and suggests that finding different ways of financing these activities merits serious consideration.3 7

E. Summary and Recommendations

59. Over the past two decades, earmarking in Colombia has trended upward and the sources of earmarking have become more varied. The rise in overall earmarking (in the 'traditional"sense) from 21 to about 35? of revenues at all levels of government is the product of a substantialrise in earmarked national government revenues earmarked (from 112 to 282) and falling proportionsof revenues earmarked at the departmentaland municipal levels (from 942 to 77Z and 302 to about 20?, respectively). During this period the share of the national governmer.tin earmarking at all levels rose from under one half to over three quarters. In 1970, departmental taxes on alcohol/tobacco/gamblingwere the largest source of earmarking

37/ H. Gom6z Buendfa and A. CifuentesNoyes *SeguridadSocial y Empleo' in El Problema Laboral Colombisno,Tomo 2, J. Antonio Ocampo y M. Ramirez (ods.), (ServicioNacional de Aprendizaj., DepartamentoNacional de Planeaci6n,Contralorta General de Is Repablica;Bogoti, 1987), pp. 116-139;and IBRD, op. cit., Annex 3. - 45 - followed by the coffee export tax and gasoline tax at the national level and the valorization tax at the municipal level. During the past two decades, revenue sharing (the tax allowance and sales tax transfer), import duties for PROEXPO (et. al.), and, to a lesser extent, a number of new minor funds have increased the variety of earmarking. Concentration on traditional sources of earmarking, however, understates the extent to which government monies are tied up. "Non-traditional" sources--public enterprises net operating profits and payroll taxes for social security and SENA/ICBF also represent earmarking in that they are public incomes which cannot be used flexibly for whatever purposes the government might deem necessary. Including non-traditional sources raises the share of earmarking in total public incomes to about 552 today. About 97Z of earmarking comes from just eight sources: public enterprise profits (27%); payroll taxes for social security and SENA/ICBF (24Z); revenue sharing (20Z); departmental alcohol/tobacco/ gambling taxes (11X); the gasoline tax (72); import duties for PROEXPO et. al. (4%); and the ad valorem coffee export tax and municipal valorization tax (2X each). The remainder comes from about 20 relatively minor sources (e.g. tourism, cinema, coal), no one of which contributes as much as one percent to the total.

60. The key measures taken by the government in the 1980s to reduce the extent of earmarking are Law 55 of 1986 and the recently passed Organic Law of the Budget of 1988. The former is an attempt to reallocate specified proportions of earmarked money to different activities in the same field. Because of this limitation, a significant increase in flexibility in the use of earmarked monies was not to be expected and in practice the Law 55 activity was carried out by the original earmarked recipient agency itself because it was best qualified to carry out the activity or because changing government priorities meant that reallocations should not take place (as for example with ICBF's "hogares program"). Thus, Law 55 has increased the variety of activities that can be financed with earmarked i..ney, but has not significantly increased the flexibility of the budget. The new Organic Law is likely to have an impact on earmarking both directly and indirectly. It prohibits the creation of new earmarked funds from existing sources of revenue and it also states that the operating profits of public enterprises belnng to the central budget until (or unless) the enterprise justifies their use with an appropriate investment program. The latter should help to foster better medium-term enterprise planning and introduce some flexibility into the major source of earmarking. In addition, by greatly restricting the number of mini budgets within one year, the new Organic Law would sharply reduce variance between original and final budgets, greatly reducing uncertainty about allocations each year. If cumbersome budget procedures were also reduced, these reforms of the budgetary process could substantially reduce the pressure for earmarking. Thus tha new Organic Law introduces some important changes, but further measures are likely to be necessary to reduce the scope of earmarking and improve budgetary performance.

61. Earmarking in Colombia is too complex to allow blanket recommen- dations for reform. Instead we content ourselves with piecemeal sug- gestions based on assessments of whether the automatic funding arrangements involved in each case are likely in principle to bring about desirable results and whether they are working in practice. Our assessment in each case involves considerations of whether there is a substantial overlap - 46 -

between beneficiariec and tax/price payers, whether the tax/price mechanism appears to generate an appropriate level of resources, and whether these resources are being utilized effectively for the purposes intended. With these considerations in mind, there are five broad recommendations some of which wi.l be elaborated upon in subsequent paragraphs. If recommendations (a)-(c) below were to be implemented in full, approximately one quarter of earmarking would be eliminated and greater flexibility in use of resources imparted to an additional one quarter. The recommendations are:

a. There are a number of earmarking arrangements which have no strong benefit rationale and which are prime candidates for elimination; i.e. those for PROEXPO (plus IF! and Caja Agraria); SENA/ICBF/Cajas de Compensaci6n; and the departmental taxes on alcohol/tobacco/gambling for health, welfare, and sports.

b. Further potential candidates for elimination should be sought in the large number of very small funds which make up about 3X of total earmarking (see Statistical Appendix Table A2, esp., the categories 'Special Funds", 'Contributions and Participations', and "Other' at the national level). While this Report has not had the time to review these cases, only a small number (e.g. the taxes on tourism, cinema tickets, coal) appear to be benefit-related and, in all cases, questions should be raised about whether appropriate levels of funds are being raised and whether the activities being financed are indeed a desirable use of public funds.

c. Those parts of the new Organic Law of the Budget pertaining to earmarking--namely, the limitations on mini-budgets, the prohibition of new earmarking from existing revenue sources, and the claim of the central budget on public enterprises' operating surpluses--should be implemented expeditiously.

d. Existing revenue sharing arrangements between the central government and the department or municipalities need to be strengthened since they are an important part of the decentralization effort.

e. Gasoline taxes need restructuring. Earmarking for road maintenance is not justified insofar as the tax is paid by ECOPETROL. Earmarking could be continued if consumer taxes were added on top of the CIF import price, i.e., if the tax was paid by the consumer.3 8

38/ This chapter only deals with the issue of earmarking,but the analysis indicatesthat the whole structureof gasolinetaxes and of financingfor road constructionand maintenanceshould be reviewed. This includesreviewing the system of toll roads, registrationfees, and gasoline taxes, to improvo the link betweGn the users of the road system and the burden of those who pay the tax. - 47 -

f. There is a number of earmarkingcases which follow the benefit principle quite closely (e.g. coffee fund, and the municipal valorizationtax) and which--with some adjustments--couldsuitably he continued.

62. Earmarking arrangementsfor PROEXPO are a prime candidate for elimination. There is little connection between those who pay the import duties and those who benefit (exportersor foreign consumers);the arbitrary level of the duty makes it problematicwhether an appropriate amount of export financing is being provided; and the differentialcredit subsidies across sectors has introducedan additional distortion into the allocation of resources. Payroll taxes for SENA/ICBF/Cajasde Compensicifn Familiar are also questionable. The benefit connection is virtually absent in the case of ICBF and is not strong in the cases of the Cajas or SENA which have ventured into activitieswhich have little impact on the payers of payroll taxes. The result is that any rate of payroll tax would end up being arbitrary and there is no assurance that the automatic funding arrangementsthat earmarking entails are generatingan appropriatelevel of resources. In addition, Colombia'spayroll taxes (includingthose for social security) are ac the upper end of the spectrum among Latin American countries and add 24-29X to the cost of labor; eliminatingthe payroll taxes for non-social security purposes would lower labor costs and likely take 0.5-1.0 percentage points off the unemploymentratio. Lastly, on the same grounds as above, the earmarking of departmentaltaxes on beer/alcohol/tobacco/gamblingfor welfare, health and sports has no rational basis; such arrangementsshould be terminated and these activities made to rely on general budget resources.

63. The tax allowance and sales tax transfer--neitherof which exipted 20 years ago--have played important roles in the government'seffort to decentralize,the former for the departmentsand the latter (mainly) for municipalities. The government is committed to continue in this direction and our recommendationshere are intended to assist in this process. First, certain anomalies in the sharing formulas ought to be corrected,in particular the undue favoritism to small jurisdictionsoffered by the non-populationbased part of the tax allowance and by the disincentiveto increase cadastral values contained in the present sales tax transfer. Second, all revenue sharing formulas ought to contain incentives for greater local resource mobilization (as the formula for small 'lunicipa- lities presently does) to prevent national resources from substitutingfor local resources and encourage greater efforts in priority areas. And lastly, it is essential that efforts be intensifiedto ensure that municipalities-- especially small ones--have the capacity to utilize additional resources effectively.

64. Even for those earmarkingcases where the benefit principle seems to fit and where price/tax arrangementsoffer the possibility of generating an appropriatelevel of resources for the sector, a number of improvements are necessary. The coffee fund, which has strayed into a number of activitiesbeyond its mandate of income stabilizationand assistanceto coffee growing areas, needs increasedgovernment surveillance. The new coffee contract 1989-1999makes this possible, but very determined governmentefforts will be needed to ensure that activitiesare refocussed - 48 - back toward the original mandate. As for earmarking in highways, gasoline prices and taxes have been adjusted annually but still have not kept pace with inflation and the amount of resourcesdevoted to maintenance and constructionhas been falling in real terms. If earmarkingfor highways is to continue, gasoline prices/taxes (and tolls, license, and registration fees) should be set so that they can cover an appropriate expenditure program and debt service for the sector. Lastly, municipal level valorizationtaxes represent a quite pure form of earmarkingwith considerablereliance on benefit-costconsiderations in investment decisions. Valorizationappears to be popular and--in some instances-- offers a way around resistanceto property tax increases. Its extension to cities which have yet to make substantialuse of it (i.e. beyond the major cities) merits consideration.

65. The recommendationsmade above would substantiallyreduce the number o.,fearmarked funds and the extent of governmentmonies earmarked, thus imparting greater flexibilityto the governmentbudget at all levels. Earmarkingwould be limited to revenue sharing, payroll taxes for social security,and those few cases where benefit connectionsare clear and pricing arrangementsappear to generate appropriatelevels of resources. Even in these cases (especiallythe first and third), the commitmentto earmarking ought not to be open-ended. In each case, a date should be set when the automatic financing arrangementwould be reviewed, or better, when the arrangementwould be terminatedunless expressly renewed. This would permit adjustments to changing circumstancesand force a review of the pricing arrangements,the quality of investmentsfinanced, and the past and future growth of infrastructure in the sector relative to needs. - 49 -

CHAPTER IV

Power Sector

A. Overview

1. Historically, the power subsector has been responsible for a large fraction of total public investment and indebtedness. This trend is expected to continue in the medium and long term. In the last 12 years the average participation of the power subsector investment in total public investment has been 23.7Z, and its average share in the balance of the public debt was 25.5Z.1 The reason for this is mainly to be found in the chronic low value of the average electricity price charged to the consumers,2 and in the ambitious power expansion plans which were undertaken in the early 1980's based on fast growing demand scenarios (about 10Z per year) which did not fully materialize. Although future least-cost power expansion plans have been recently developed based on demand scenarios growing at a more moderate pace (about 5.7Z per year) the resulting investment program will continue to require, on average for the period 1989-1996, about 23Z of the total public investment, and about 28Z of the balance of the public indebtedness if tariffs are adjusted at a slow pace.

2. In the period 1990-1992, the power subsector is expected to invest about US$2.1 billion. Of this investment, the completion of the power system expansion plan initiated in the early 1980's accounts for about US$1.6 billion, and has to be carried out (the Guavio and Rio Grande Scheme power generating plants have to be in service by 1992/93 to meet the demand through 1996, and the associated transmission and distribution systems completed to deliver the power to final consumers). The recently approved generation and transmission least cost expansion plan through year 2000 including the hydroelectric plants Urra-I, Porce-II and Miel-II, plus the necessary power distribution development, accounts for projected invest- ments of about US$408 million in the period 1990-1992 and US$2.1 billion in the period 1993-1996.

3. The power expansion program for the short and medium term contain the most economic generating and transmission projects. Financing the program, however, will require substantial increase in the sector's internal cash generation, even after allowing for reasonable net transfers from the government, and for plausible contributions from domestic and foreign borrowing. In particular, the perennial policy to subsidize some electricity tariffs (particularly residential) should be reviewed and

/ Source: DepartamentoNacional do Planeaci6n.Situaci6n Financiera del Sector El6ctricoa Mediano Plazo. DocumentoDNP: 2408: UNIF, Bogot4, Enero 11, 1989.

!/ Averags electricitytariff n 1988 was US$4.0/kwh,which is 60% of the estimatedUS$6.44/kwh averag, long run incrementalcost (soe World Bank's Energy Pricing Study, November 1989). - 50 -

replaced by the sound policy of having the consumer pay for the economic cost of the service (except, perhaps, for temporary subsidies to the poorest consumers).3

B. OrRanization and Institutions

4. Six main util'lies currently dominate the power subsector: Empresa de Energla Elsctrica de BogotA (EEEB), Empresas Pdblicas de Medellin (EPM), and Corp. ra,_ . Aat6nc... Regional del Cauca (CVC); these are strong and independent uti;lties serving the major markets of Bogota, Medellin, and Cali respectively. The remaining three utilities are: Instituto Colombiano de Energia E16ctrica (ICEL), which is the main shareholder of thirteen smaller utilities, Corporaci6n Electrica de Costa AtlAntica (CORELCA)--the main shareholder of nine utilities; and Interconexi6n Eldctrica, S.A. (ISA), a generation and transmission utility. Five other smaller utilities play a minor role. The power sector services about 4.1 million customers; it has about 20,000 employees (Statistical Appendix, Table 41).

5. EPM4is fully ox-nedby the municipality of Medellin, Colombia's second largest city. It operates power, water, sewerage and telephone services. It is responsible for serving about 10% of the customers and represents 202 of total national electricity sales in the country. Although EPM is generally well run its power losses are relatively high (21Z).

6. CORELCA and TCEL are fully owned by the Government and together account for aL'out 50Z of the customers and 35% of electricity sales. The twenty twqo subsidiaries face serious problems, stemming from pocr management, excpssive ,oliticization and generally poorer, isolated market areas. Control by the two parent companies is poor, particularly in the case of ICEL. A.fter in-depth analysis, whiich was carried under the World Bank's Power Sector Adiustment Loan (PSAL),4 these institutions are now implementing satisfactory management improvement programs.

7. CVC is a generally well run regional development corporation whose main function in the sector is to sell bulk power to the integrated system and to the distribution company serving Cali (EMCALI). Neither EMCALI nor CVC have major problems.

8. EEEB serves 22Z of customers and has 31% of electricity sales. It supplies the city of Bogota. Over the initial half period of its 20-year relationship with the Bank, EEEB had a history of good management and satisfactory financing policies. In the last ten years, however, performance of EEEB has deteriorated. The simultaneous construction of the Mesitas and Guavio hydroelectric projects, which the Bank helped to finalce through Loans 1628-CO (signed April 9, 1979) and 2008-CO (signed March 8, 1982), strained both EEEB's financial capability and its management. Relatively frequent changes in senior management added to the problem and

3/ Althiough n strong case could be mndo to handlo poverty alleviation by means other than elo4ctrlcic- subsldies.

4/ Ponor Soctor Adjustn.- Loan, Roport No. P-4876-CO,dated November 10, 1987. - 51 - the zeal of the main shareholder (the municipality of Bogota) in preserving the autonomy of the company from interference by the Central Government added a political dimension to efforts to redress EEEB's problems. Changes in management of EEEB made in the last two years are effectively redressing this situation.

9. ISA is a key institution, with broad responsibilities for sector planning, interconnection of regional systems, and operation of the integrated system. It was established in 196? as a corporation whose shares were owned by the main regional and municipal utilities. The Bank played an important role in ISA's establishment, which reflected a concern dating from the early 1960s with the fragrrentationof decision-making power among increasingly autonomous utilities. The Government is represented on ISA's Board since 1985, when ISA's by-laws were amended to allow Government non-voting representation on the Board. The lack of direct Government representation between 1967 ancL 5 was an important impediment which prevented the Government from g effective action to ensure that power sector policies were fully in line with macro-economic programs.

10. ISA has evolved into an important and effective institution. Its role in the critical area of overall sector planning has, however, been seriously constrained by the strong pressures from individual region-based utilities to develop their own gen,eration facilities and their general resistance to intervention. In 1979, at the insistence of the regional utilities which were facing the threat of electricity rationing, the Government agreed to permit the regional utilities to initiate their own generation projects. However, it was agreed by the Government and ISA's shareholders that ISA would operate the trunk transmission lines, act as national dispatch agency and, in the long run, own at least one third of the total generation capacity in the sector.

11. With the benefit of hindsight, Lhe decisiolnto allow the regional utilities to own and to construct large generation projects was not sound, as it led to competition among them for projects in their respective areas, without due consideration of national interests. This weakened ISA's ability to ensure that the investment program constituted the national least-cost solution and to adjust programs to changing circumstances. Once projects were included in the investment programs of the municipal or regional utilities, ISA lacked the authority to adjust the pace of execution to take into account changes in demand. In order to redress this situation, it was agreed in the PSAL that ISA will prepare a least-cost expansion of the power system for the period 1990-2000. In 1988 and early 1989, in close consultation with the Bank, ISA prepared the least- cost expansion of the generation and transmission system for the power sector under terms of reference satisfactory to the Bank. As a result of this consultation, several changes were introduced during the preparation of this study. The initially relatively high load forecast proposed by ISA was scaled down to a moderate average growth rate of 5.7Z per year in the period 1988-2000 (para. 20); the methodology for selecting the least-cost expansion solution was modified to account for boundary conditions at the horizon of the planning period; the costs and benefits of multipurpose hydroelectric projects were adequately considered; the methodology for hydroelectric-plant cost estimates was reviewed; market prices were - 52 -

converted into efficiencycosts for the economy, etc. The fruitful exchange of views between ISA and the Bank led to a draft least-cost expansion plan in November 1988--almostfully satisfactoryto the Bank-- which was approved by ISA's Board of Directors in December 1988. The Bank provided minor additionalcomments on it in January 1989. The draft of the expansion plan was agreed with the Bank in March 1989, and approved, by the ISA's General Assembly in April 1989. The plan will be reviewed,however, to adjust to a lower than expected growth in demand in 1988 and 1989, and to a possible interconnectionwith Venezuela in the near future.

12. Despite the progress achieved in strengtheningISA, a recent study of the World Bank's OperationsEvaluation Department,5 found that persistent drawbacks zan only be addressed through an in-depth re- examination of the institution,followed by adjustmentsof ISA's structure and role. One of the main questions is whether ISA should remain one company or whether it should be split up in order to pursue with less constraint objectives such as: (i) a transparentdecision making process shieldingplanning from political input at an early stage; (ii) a role for Government more commensuratewith its financial obligations;and (iii) a realistic definition of the role of the central generating and transmitting utility.

13. The Ministry of Mines and Energy (MME) plays an important policy and regulatory function in the power sector and specificallyis responsible for issuing regulationson the provision of electricityand on technical standards. The tariff board, JNT (Junta Nacional de Tarifas) which is under the National Planning Department (DNP), is empowered under Decree 201 of 1974 and Decree 149 of 1976 to ensure sound electricitytariffs, but does not have full regulatory functions to ensure compliance and to assess the efficiency of individual utilities. Although the head of JNT is a highly competent professional,the group is not properly staffed to be in a position to analyze tariff proposals for a large number of utilities. Therefore, unless this constraintis released, JNT could not play a more meaningful role than at present. Until recently, some utilities interpretedJNT's tariff resolutionsas setting only an upper limit for the tariffs they could charge to consumers. This contributedto some of the existing tariff distortions,as some utilities charged lower tariffs than those approved by JNT, especially to low-incomeconsumers. Since 1986, when the Government establishedclearly, through policy directives and by law, that JNT sets the actual tariffs to be charged, the utilities have been complyingwith the JNT's tariff resolutions.

14. Another key sector institutionis FEN (FinancieraElectrica Nacional) a financial intermediaryset up by the Colombian Government in 1982, to help finance the investmentprograms of the power sector, to oversee the sector's finances, and to coordinate the sector's financing strategies. FEN's shareholdersare the National Government (95Z) and the sector utilities (5Z). FEN was to act as a financial agency for the sector, as well as an advisor to the Government on financial issues facing the sector and on funding schemes for large projects. These objectives

6/ The PowerSoctor and the World Bank, 1970-1987, OED,August 1989. - 53 - have not yet been fully attained. FEN's role over the past few years has been to channel to the power sector the funds obtained from the Bank (Loan 2401-CO) and from cofinancingand other foreign sources, and to raise funds in Colombia's financialmarkets which it has on-lent to the sector utilities. FEN is efficientlyrun and its organizationand staffing are adequate for the functions it is carrying out. Its profitabilityhas been consistentlysatisfactory, and the return on equity has been 21Z-22Z a year. Recently it has begun to act as advisor to the Government on financial issues facing the sector. Under a recent loan from IDB for electricityloss reduction,FEN has agreed to evaluate subprojects 3ubmitted by the beneficiaries,following criteria which meet the appraisal standards of IDB (which are similar to the Bank's). To do this, FEN has set up a Projects Evaluation Division and has staffed it with qualified personnel. This function will expand FEN's role to one which more closely resembles a developmentinstitution for the power sector. Recently, the Government has proposed to the Colombian Congress a bill to expand FEN's responsibilitiesto cover the financialneeds of the entire energy sector. Thus, FEN will become FinancieraEnergdtica Nacional.

15. The PSAL had as one of its goals the improvementof the power sector regulatory frameworkby creating the National Energy Board and establishingthe Power Sector Monitoring Committee. The Energy Board bill, approved by the Colombian Congress in mid-1989, will permit to steer and oversee energy pricing, and sector finances and management, and to ensure the country's energy needs are met at least-cost for the economy. Meanwhile, an advisory group (IntegratedEr-rgy Planning Group) has been functioningsince 1987 as an ad-hoc National Energy Board. The Power Sector Monitoring Committee functioned satisfactorilyfrom early 1988 to June 1989 assisting the power sector in its dialogue with the Bank and in coordinatingthe implementionof the various action plans of the PSAL. The Power Sector Superior Council has now replaced the Monitoring Committee with expanded functionsand responsibilities.

16. The creation of the National Energy Board is an important step towards improving coordinationin the energy sector. The Board is expected to organize all functions presently exercised in the area of energy sector regulation by the Ministry of Energy and Mines, DNP, JNT, FEN, and ISA. It will be able to translate Government objectives and policies into sector rules, and to monitor their implementation. The Board will face a challengingtask to exercise the following functionsthat have been found to be only loosely coordinatedat present:6

- define for the individual utilities areas of autonomy as large as possible in order to encourage the deployment of regional dynamism, initiative,and entrepreneurshipbut limited such as to avoid endangeringthe pursuit of clearly identified national objectives;

- strengthen sector planning and operationalcoordination, i.e., in the present setup, support ISA in carrying out its main tasks;

The PoworSector and tho WorldBank, 1970-1987, op. cit.,Vol. I, and Vol.II, Ch. II and Ch. III. - 54 -

- establish and oversee the decision-makingmechanism connecting technical and economic sector planning, financing,and project implementation;

- set, in accordancewith the general Governmentobjectives for the sector, the operationaland financial targets for the individual utilities;

- oversee creation and operation of joint ventures to build and own new plants, following rules defined in accordancewith the above first point above;

- monitor operations,identify shortcomings,and agree with the utilities on remedial action;

- set the tariffs in agreementwith the general objectivesand rules set forth by Governmentabove and to oversee their implementation;

- link up with the proper Governmentinstitutions, in particularthe Ministry of Finance, to make sure that Governmentcontributions establishedon the basis of the sector's investmentand operationalplanning in accordancewith the general Government objectives,come forth at the proper time, and in the proper amounts;

- define the rules of interchangebetw:een utilities and, in particular, of sharing surpluses and deficits within the sector.

C. Power Demand

17. Electricity use is extensive in Colombia and is increasing at rates higher than those for total energy consumption. About 58Z of Colombia's population of 28.4 million had access to electricity. Electrification is higher in urban areas (about 95Z in BogotA, Medellin and Cali), but large areas of the country still lack electricity service. Electric power provided 28Z of total commercial energy in 1987, which compares with 24Z in 1980.

18. In 1988, nearly 50Z of electricity consumption was by residential users. In the interior and in some urban areas, low-income households use cocinol, a type of kerosene which is unsafe and highly subsidized. LPG is also subsidized but available only in limited amounts. Industry and commercial sectors accounted for about 30Z and 102 respectively of total electricity consumption.

19. Electricity demand increased at a rate of about l0.lZ in the 1970s but growth has dropped substantially since 1981 (30.79 TWh in 1988 or 5.6Z higher than electricity demand in 1987). The high consumption growth during the seventies and delays in completing some generation projects caused electricity rationing during 1981 and 1982. This led to excessively conservative planning expansion for the power sector in the 1980s, on the assumption that demand would continue to grow at about lOZ per year. In the period 1980 to 1987 the actual yearly average growth rate was 6.1Z. - 55 -

D. Demand Forecast

20. The demand forecast for the period 1988-2007 was prepared by ISA in December 1988 7 using an econometric model developed by ISA based on an old model on which the Bank made comments. Although the Bank has not made a detailed review of the econometric model used by ISA, the resulting reference demand forecast -- which indicates an average annual growth rate of 5.7% for the period 1988-2000 -- has been found acceptable by the Bank for the purpose of studying the least-cost expansion of the power generation and transmission system. The econometric model forecasts the electricity sales to four sectors at a national level (industrial, commercial/services, official and public lighting) as well as the sales to the residential sectors served by EPM, EEEB, CORELCA/Electrificadoras, ICEL/Electriricadoras and CVC. The model has four sets of exogenous variables: (a) macro-economic (per capita income; industrial investment; value added of commercial/services; value added of official sector; and urban GNP defined as total GNP less agriculture, mining, and petroleum products GNPs); (b) demographic (total population; number of dwellings; urban population; number of customers for each sector): (c) prices (electricity prices; petroleum products prices; LPG and natural gas prices for the residential sector; coal prices; and residues prices for the industrial sector); and (d) policies of the MIE (electric service coverage targets; loss reduction programs; and tariff policies).

21. The growth of GDP has been found to be the dominant factor influencing power demand growth at the sector level.8 The lower than planned growth of demand in the early eighties (6Z instead of 1OX) was mainly due to the 1981-84 collapse of GDP growth. Therefore, there is still room to review dovnward the demand forecast to account for the actual GDP growth rates in 1989 and 1990, and to consider the evolution of the specific electricity consumption per user in the residential and commercial sectors. Also, the demand forecasting models can be improved further by a better knowledge of the market for electricity uses and a better assessment of the impact of tariff structure changes and substitution of gas for electricity.

22. The values of the exogenous variables used in developing the reference demand forecast are indicated in the statistical appendix, Tables 42 to 49, and the results depicted in Table 50. Additionally, Table 51 shows the results of 14 sensitivity studies to changes in the main exogenous variables. Based on these analyses, a high annual growth rate of 6.0 and a low annual growth rate of 5.3Z demand forecast were chosen with a mid-point of 5.7Z average annual growth rate for the period 1988-2000. These values were used for the sensitivity analysis of the least-cost expansion plan for generation and transmission.

7/ ISA: Proyeccionesde demanda 1988-2007con modelo econometrico,Medellin, Diciembre 1988.

8/ The Power Sector and the World Bank, 1970-87,op. cit., 'Demand Forecset',Vol. II, Ch. 4. - 56 -

23. Electricity demand in Colombia has two basic components: the electricity consumption, and electricity losses. The former include electricity sales and electricity theft (unbilled energy), and depend on the macroeconomic and demographic evolution, and on the price policies and service coverage. The electricity losses and electricity theft reflect the inefficiency of the power utilities, and may also be caused by inappro- priate price policies at commercial and industrial levels.

E. Power Supply

24. Since Colombia's power system9 is predominantly hydroelectric, the level of power reserve in the system is defined by the availability of firm energy and not by the effective installed capacity. This is so because the hydrological variations and the size of the reservoirs constrain the amount of electric energy that hydroelectric plants can supply with a desirable reliability level. In 1988, the energy expected to be available in the system (36,000 Gwh) was 17X higher than the actual demand (30,790 Gwh). This reserve margin will gradually decrease to 14% in 1990, then increase again to 18% in 1992/1993 when the Guavio and Rio Grande II hydroelectric plants enter into operation, and would decrease to almost zero by 1996, when new generation is likely to be necessary. These reserve levels are within acceptable margins for excess enrrgy, as a result of economies of scale in planning the expansion of hydroelectric systems with annually regulated reservoirs. In contrast, the effective installed capacity in 1988, 8,335 MW, of which 6,507 MW hydro and 1828 thermal power plants, exceeded by 54Z the 1988 peak load of 5,428 MW. However, it will decrease to 202 in 1992, increase again to 45Z in 1993 when the Guavio and Rio Grande hydroelectric plants enter into operation, and will decrease to 23Z in 1996, just before the next power generating plant would be added to the system, to remain at about 25% until year 2000.

25. The overcapacity described above results primarily from an overestimation of demand and concern with the consequences of rationing in 1981-82. Another contributing factor may have been ISA's lack of power to adjust the pace of execution of projects (para. 11). This led to a large expansion plan approved in 1981 which included the installatior. of 9,160 MW (of which 8,270 hydroelectric) during the period 1981-1991 and demanded yearly investments of US$1,150 million in 1982, US$1,357 million in 1983, US$1,252 million in 1984 and US$928 million in 1985. As the constraints on national resources for investments and external borrowings, and the likelihood of excess capacity became apparent in 1984-85, the Government reviewed the program and decided on drastic reduction in its scope. These decisions were made in consultation with the Bank which was assisting in the financing of some of the projects concerned. As a result, the construction of four large hydro projects was postponed (Miel-I, Urra, Calima III, and Cafiafisto,with an installed capacity of 3,290 MW), and the completion target dates of the Rio Grande and Guavio projects, totalling 1,323 MW, were postponed to 1992/93. This decision resulted in a

9/ This discussionis centeredin the powersupply systems of publicutilities, because those producingfor theirown use, mainlyindustrial, now accountfor no more than 6X of *'a power supply, a situationwhich is expectedto continue in the future. - 57 - reduction of 512 in the capacity expected to be installedduring the period 1981-1991 in relation to the previous plan. The recently approved least- cost expansion p an has dropped the Miel I, Calima III and Caftafisto projects and replaced them by Urra I, Porce II and Miel II, totalling 1112 MW only.

F. ElectricityLosses

26. Electricity losses (technicallosses--or energy that cannot be consumed--pluselectricity theft) representa serious problem for the sector. They have increased from 16.8Z in 1970 to 20.1Z in 1980 and to 24.5Z in 1988. About 55% of the total losses are technical losses due to some extent to insufficientinvestment in and inadequatemaintenance of distributionfacilities which has resulted in their deteriorationand overloading. The insufficiencyof investment in distributionis attributableto deficienciesin distributionplanning and management. About 451 of the total losses are non-technicallosses, mostly energy theft attributableto commercial and industrial theft and to social problems (proliferationof illegal land occupationsand correspondingelectricity connections). The non-technicalloss is unbilled energy consumptionand, consequently,represents a financial loss. Also, since the stolen energy is not paid for, its consumptiontends to increase demand for energy, resulting in an often not recognizedadditional economic cost to society.10

27. While the non-technicallosses can be reduced to almost zero, the technical losses have an economic lower limit (about 10?) determinedby equalizing the incrementalcost to reduce them, and the incrementalbenefit accrued. Lack of both an appropriatelegal frameworkto punish energy theft and of an effective system to detect it, compoundedwith a weak commitment of the power companiesto combat it, have aggravated the problem. Recently, the Government has issued a decree defining the various kinds of electricity theft and the sanctions the electric power companies should apply to consumerswho steal electricity.

28. As part of the Power Sector Adjustment Program (PSAP) initiatedby the Government in late 1987 and supportedby a World Bank loan, an ambitious program of loss reductionwas introduced. The main objective of this program was to reduce total electricitylosses by improvingthe technical characteristicsof the electric system. The implementationof the Loss Reduction Program establishedin the PSAP was delayed, mainly due to inadequate planning and inadequate financial resources. The latter was the consequenceof a delay of about 18 months in the effectivenessof Loan 237/IC-CO fru...the InteramericanDevelopment Bank (IDB), which supports the loss-reductionprograms of a number of power companies, and a similar delay in implementationof Loan 2634-COwhich finances the loss-reductionprogram of EEEB. As a result, losses,which in 1987 were within the targets of t.,e original plan (24.0 compared with 23.8? of total net energy generation), increased in 1988 (to 24.42 compared with a target of 23.0Z). Early in 1989, the Government committed itself to prepare an emergency plan for the power companies to reduce the losses at an acceleratedpace and in April 1989 the Emergency Loss-reductionPlan was completed. The plan, which was further refined in June 1989, defines detailed action programs for each of

The Power Sector ond the World Bank, 1970-87, op.cit., Ch. 4. - 58 - the power companies. The loss levels of the emergency plan, as compared with the levels set out in the PSAP, include a somewhat higher loss target for 1989 (22.7% against 21.9Z), maintain substantially the same target for 1990 (21.2Z), and establish lower loss targets for 1991 and 1992 (19.92 and 18.5Z against 20.12 and 19.2Z, respectively). The main features of the plan include: (i) installing 1,285,000 meters in the period 1989-92, of which some 232,000 (18Z) will be installed in 1989 which will reduce the number of customers without ,ieters from 11.4Z in 1989 to 3.12 in 1992; (ii) inspecting 762,000 meter/connection installations in 1989 and 3,895,000 by 1992 (75, of IlLeteredcustomers), concentrating on consumers where fraud is known to be high (40% of commercial and 10OZ of industrial customers will be inspected in 1989); and (iii) legalizing some 67,000 irregular customers in 1989, and some 276,000 customers by 1992. During the implementation of the plan, specialized crews are expected to effect some 644,000 service cuts to fraudulent customers, of which some 141,000 would take place in 1989.

29. The Emergency Plan is being implemented according to schedule, and actions taken show a clear intensification of measures carried out during the period April-May 1989. The following actions required to support the emergency plan have already been taken: (i) the Government issued on June 19, 1989, decree No. 1303 establishing sanctions against electricity theft (this is considered a real milestone, since it provides, for the first time, a clear definition of the various infractions that constitute electricity theft, and the corresponding fines and penalties that can be applied by the electric power companies); (ii) MME has instructed the power companies to take legal action against consumers who steal electricity; (iii) the Government issued on June 21, 1989 decree No. 1327 establishing the functions and organizational structure of the Power Sector Superior Council (PSSC), a new body which will replace the Power Sector Monitoring Committee with expanded functions and responsibilities; (iv) the PSSC, through the Loss Reduction Unit of its Technical Secretariat, has set in place a monitoring system designed to review monthly the performance of the 25 electric power companies vis-a-vis their respective emergency plans for loss reduction; (v) the MME has awarded the contract, starting in August 1989, for a public advertising campaign and a consumer-education program, both designed tc inform and orient consumers on how to avoid and stop electricity theft; (vi) MME has prepared a detailed plan of action for the acquisition of one million meters and other essential equipment urgently required for the loss reduction program starting in 1990; and (vii) ISA, will purchase 900,000 imported meters to be financed under IDB Loan 237/IC- CO. Bids for these purchases are scheduled to be issued in the last quarter of 1989. Due to the complexity of the program, however, it is not yet possible to judge if the program's target of reducing losses to a programmed 22.7% level can be met by December 1989.

G. Sector Planning

30. Sector capabilities for p'anning generation and bulk transmission expansion are relatively strong from a technical standpoint. Present overcapacity problems can be attributed above all, to the difficulty to anticipate the slowdown in economic growth. Until the 1960s, power generation expansion plans were prepared independently by each regional utility. When ISA was created, one objective was to centralize in it - 59 - generation and transmission planning. ISA was initially assisted by Harvard University, and adopted Harvard's models for power expansion planning. ISA subsequently developed a comprehensive methodology for analyzing alternative expansion plans, consisting of several mathematical models, includi:±ga stochastic treatment for hydrology, as well as methodologies and models for analyses of the transmission system. In the early 1980s DN4Pand ISA, with technical assistance from the Federal Republic of Germany, developed an inventory of hydroelectric resources, and optimization and simulation models for long-term strategic planning. Though the planning methodologies used have been sound, some distortions occurred in the past in the planning and decision making process.

31. The distortions referred to above were addressed by ISA in the study of the least-cost investment program for 1991-2000 prepared in close consultation with the Bank. -his has implied, among other considerations, the introduction cf shadow-pricing analysis, inclusion of environmental costs and macroeconomic constraints, reliability analysis, and a wide spectrum of sensitivity studies to test the robustness of the proposed least-cost expansion plan to changes in the demand forecast, increases in the hydroelectric plants capit.' investment costs, variation in the annual discount rate, etc. The proposed least-cost expansion plan for the generation and transmission system--and the associated reference demand forecast of 5.7Z average growth rate in the period 1988-2000 on which it is based--were approved by ISA's Board in December 1988, found satisfactory by the Bank in March 1989, approved by ISA's General Assembly in April 1989 and approved by CONPES by end June 1989. The generation expansion plan recommends the addition of three hydroelectric plants by 1996/1997: Urra-I (340 MW), Porce-II (392 MW) and Miel-II (380 MW), and the completion of Rio Grande II scheme (323 MW) by 1992 and Guavio (1000 MW) by 1993. A lower demand growth (average 5.3Z) delays these plants by about one year, a higher one (average 6.0Z) advance them a few months. Consequently, the Bank recommended that the design of 150-300 MW gas-fired plants (single or combined cycle) be completed and the sites selected as a strategy to cope with unforeseen delays in the commissioning of hydroplants. Also, the investment plan will be reviewed again to take into account lower rates of GDP growth in the period 1988-90, and a possible interconnection with Venezuela, which may further delay the construction of the three selected hydroelectric plants.

32. In order to achieve a harmonious development of the power sector and to ensure the implementation of the least-cost plans, the Government has established under the Power Sector Adjustment Program that, as a general policy, all generating plants with an installed capacity of more than 100 MW will be owned by ISA (with exception of the Urra I project). In order to utilize the sector's technical and administrative capacity ISA may be allowed to sub-contract the construction and/or operation of new plants to its member companies while retaining control over major decisions. By the same token, it may be also allowed to sub-contract the future operation of those plants. Such delegation must be approved by the Government. Further, the Government established the policy that ownership of new transmission lines will be governed by the following criteria: i) all lines required for the connection of a generation project will be considered part of the project; ii) ISA will own all components of the main high-voltage network interconnecting the areas of its member companies; and (c) all lines between regions of the saL. market will be owned by the company responsible for that market. - 60 -

H. InvestmentProgram

33. The investment program for the period 1989-1996 will be examined in two 4-year periods: (i) the period 1989-1992 corresponding to the finalization of the investment program initiated in 1984 and since then modified several times to account for project implementation delays and lower than expected demand growth; and (ii) the period 1993-1996 corresponding to full implementation of the proposed investment program needed to meet the demand from 1997 until the year 2000.

Period 1989-1992

34. The power development policy through 1992 places emphasis on transmission, subtransmission and distribution works and completion of power plants under construction. This will not only increase the coverage and quality of power service but also restore the balance between the generation and the distribution infrastructure.

35. Accordingly, sector investments are aimed at:

- Strengthening of transmission system. - Expanding the distribution systems. - Impreving the quality of power service and reduction in power losses in urban centers. - Completing the power generation projects currently under construction (Rio Grande II in 1992 anidGuavio in 1993).

36. It is to be noted that on the basis of current demand growth projections, and lead time required by hydroelectric projects, there would be the need to start construction of new generation projects in 1990, since the existing firm energy generation capacity (plus the addition of Rio Grande II and Guavio) is expected to be fully absorbed by 1996.

37. The investment program contemplates a total equivalent to US$2,422 million plus US$650 million in interest during construction. Included in this amount is US$303 million corresponding to the investment to be incurred during 1990-91-92 in initiating the projects of the expansion plan to be completed in 1996. The size of the US$2,063 million investment program for the period 1987-1990 is in agreement with what the Bank found acceptable to include within the Power Sector Adjustment Loan (PSAL) No. 2889-CO and significantly lower than the US$6,000 million investment level between 1980-1984. The major projects to be completed during the period 1989-92 are the 1000 MW Guavio hydroelectric plant, the 322.5 MW Rio Grande Scheme (300 MW Tasajera and 22.5 MW Niquia hydroelectric plants), the second Power Distribution Project in nogot4, the Guavio and Rio Grande II transmission system, and the expansion of the power distribution systems of the CORELCA and ICEL groups of "electrificadoras" and EPM. All these projects are at a well advanced stage of construction.

Generation

38. The generation component of the 1989-1992 expansion program (see Statistical Appendix, tables 53 and 54) consists of: - 61 -

(i) the Guavio project, a 1,000 MW hydropower plant that is being implemented by EEEB and financed partially by loans from the Bank and IDB, to be completed by 1992193.

(ii) the Rio Grande project, a multipurpose project with a power component of 323 MW, which is being implemented by EPM and financed partially by the Bank and IDB, to be completed by 1992; and

(iii) initiation of the construction of Urra-l, Porce II and Miel II hydroelectric plants.

Transmission

39. The transmission component consists of 560 km of 500 kV and 2,500 km of 230 kv transmission lines and corresponding substations. The 500 kV line is the second interconnection line between Medellin and the Atlantic Coast, which is required by 1991 to maintain service reliability in the coastal region and to transmit energy from the Central Region to the coast regularly from 1994 on. The 230 kV lines extend the interconnected system to the south, the west and the southeast and would reinforce the existing interconnected system.

40. The following transmission projects are being undertaken:

Company Project

EEEB-ISA Guavio-Villavicencio-Bogota ICEL (CHB) Betania-Popayan (commissioning) ICEL Popayan-Pasto ISA Cerromatoso-Uraba (tentative) ISA San Carlos-Sabanalarga, 2nd circuit ISA Ancon Sur-Esmeralda ISA San Carlos-Barranca ISA Betania-Ibague ISA Bucaramanga-Arauca ISA La Mesa-Ibagte ICEL Bucaramanga-Ocafla-Cucuta CORELCA Sabanalarga-Fundaci6n CORELCA Valledupar-Cuestecita CORELCA Chinu-Monteria CORELCA transmission CHEC Esmeralda-La Hermosa ESSA Barranca-Bucaramanga CVC asto-Tumaco

41. The transmission lines included in the program were selected on the basis of reliability analyses of the interconnected system completed with line by line cost-benefit analyses to compare with alternative options and to define the completion date for each line. The reliability analyses were made by simulating the operation of the system with and without each of the proposed lines. The comparisonwith alternativeoptions, such as generation plants instead of interconnectionlines, showed that the - 62 - proposed lines are the least-cost solution to supply electricity to the regions to be served by those lines. The cost-benefit analyses to define the completion date for each line considered the cost of power rationing and losses. The cost of power rationing was based on a detailed power rationing cost study by sector of consumption previously carried out by ISA.

Subtransmission and Distribution:

42. The subtransmission and distribution component consists of:

(i) improvements to the existing distribution facilities to reduce energy losses;

(ii) expansion of urban distribution systems to maintain current coverage level; and

(iii) extension of rural electrification lines through four programs; PERCAS (Rural Electrification) II by CORELCA, PPNER (National Rural Electrification-Program) II and North Caqueta by ICEL, and Cundinamarca by EEEB.

43. The distribution component has been defined to meet three basic objectives: (a) to reduce energy losses by improving the existing distri- bution facilities; (b) to maintain current service coverage level in the urban areas; and (c) to extend rural electrification at a similar pace as in the past. Although a figure higher than the proposed amount for urban distribution programs would be economical, distribution investments have been limited taking into account the sector's financial constraints and its execution capabilities. New rural electrification projects, included in the investment program, respond to Government-supported programs as part of integrated rehabilitation programs in priority regions affected by guerrilla activity, drugs or poverty. Decision on the execution of these projects will be reviewed once their studies are completed.

44. The rural electrification program includes the execution of the Second National Electrification Plan of ICEL, and the second stage of the CORELCA PERCAS program with a target of providing power to 212,000 dwellings. Other programs are executed by EEEB and other government agencies. Based on these plans, the country's rural power service coverage will rise from 26.4Z at the end of 1986 to 40Z by 1991.

45. Urban subtransmission and distribution programs are being carried out by EEEB (Second Power Distribution Project in Bogota), EPM, CVC and CORELCA. In addition, the National Program for Loss Reduction and Improved Urban Service (FEN-IDB) is being implemented.

Studies and Training

46. The studies and training component, which amounts to 2Z of the total investment program, consists of:

(i) prefeasibility and feasibility studies, to widen the spectrum of generation alternatives to be considered in planning the system expansion; - 63 -

(ii) a study, on rehabilitation of old power stations, which is expected to provide attractive options to postpone generation expansion planning;

(iii) the following special studies: (a) a management study of EEEB and management studies of the electrificadoras in the ICEL and CORELCA groups (these studies funded by FEN are almost completed); (b) improvement of planning methodologies, mainly: consideration of opportunity costs instead of market prices; consideration of socioeconomic and environmental factors in the comparison of expansion alternatives; a power tariff study that would also include improvement of demand models; development of a control system to follow-up the sector plan implementation, and improvement of data base for project selection; (c) studies, to be carried out by ISA for improving tariff policy, mainly: socioeconomic surveys for stratification purposes and demand studies by consumer blocks to assess the impact of tariff changes; and (d) management studies on ICEL and CORELCA groups for efficiency improvements and assessing of training needs; and

(iv) ongoing training programs under Bank and IDB financing.

(v) implementation of a Power Sector Information System (SINSE) and of a Uniform Accounting System to facilitate consolidation of the sector finances.

Investment Program 1993-1996

47. The investment program (see Statistical Appendix, table 52) is based on the least-cost expansion plan through year 2000 approved by ISA in December 1988. It contemplates a total investment equivalent to US$3,228 million plus US$661 million in interest during construction. Of the US$3,228 million the generation and transmission expansion plan accounts for US$2,060 million, and the distribution system expansion plan accounts for US$1,168 million. The main projects included in the power expansion plan for completion in or before 1996/97 are: Urra-I (340 MW), Porce-II (392 MW) and Miel-II (380 MW) hydroelectric plants and associated transmission lines; the reinforcement of the 500 kV and 230 kV transmission systems, and the implementation of the distribution plans of all power companies. Also construction of other projects to be completed beyond 1996 would be initiated during this period (Nechi A., Porce, and Porce-III hydroelectric plants and related transmission systems).

48. Traditionally, expansion plans have suffered time and cost over- runs caused by unforeseen contingencies. However, in the case of the Guavio and Mesitas projects, better planning and detailed engineering aimed at coping with adverse geological conditions could have helped to decrease time and cost overruns. Dalays in the construction of projects with long construction time involve potential rationing of electricity. Cost overruns in .rease the cost of the electric service and burden the finances of the power company. How to minimize the effects of time and cost overruns has been a difficult problem to solve. The strategy recommended by the Bank to the Government to cope with future delays in the implementation of hydroelectric plants is to have complete designs and specifications of thermal power plants in the 150-300 MW range (gas turbine or combined-cycle plants) which could be constructed in a relatively short - 64 -

period of 2-3 years. In turn, these type of projects require low invest- ment (but have high operating costs). It should be mentioned that the selected future hydroelectricprojects Urra-I, Porce-II, and Miel-II are located in areas of excellent geologicalcondition.

49. Overall, the sector has managed to generate a more balanced investmentprogram, with a greater share for subtransmissionand distribution,greater flexibility,and improved comprehensivenessof the alternativesconsidered. The investmentplan is under revision to include larger variations in the main project parameters, including a much bolder approach to sensitivityand risk analysis.11 Large hydroelectricplants will be postponed for a few years to take into account possible lower GDP growth, and a possible interconnectionwith Venezuela;and complete designs and specificationsof smaller thermal power plants in the 150-300 MW are being prepared to be constructedquickly if demand turns out to be greater than projected or if the constructionof hydroelectricplants faces major delays.

Financing of the InvestmentProgram

50. Financial analysis of the power sector is difficult due to a significant lack of reliable 'nformation. Financial statisticsare zharacterizedby incompleteconsolidation of partly inconsistentdata of individualcompanies. 1 Utilities use different accounting systems and apply principles that are not uniform. ISA, FEN, and other entities involved in the sector are applying great efforts to compile consistent information,but successhas so far been elusive.

51. For the analysis of the short term (1987-90),FEN prepared in June 1989 a comparisonof the financial results that were expected to happen at the time the Power Sector AdjustmentLoan was approved in 1987, and the actual results, including current expectationsup to the end of 1990. Annex 1 shows this analysis. The main conclusionsare: (a) The cost of the investmentprogram for the 1987-90 period, exclusive of interest during construction,increased by about US$200 million (10Z). (b) Working capital has become negative, and in the absence of quick action, is expected to rise to only US$44 million at the end of the 1987-90 period instead uf the US$23 million programmedin 1987. (c) Gross funds generated internally were US$236 million lower in 1987-88, and, unless tariffs are adjusted soon, are expected to be US$514 million lower over the four years' period in comparisonwith the expectationswhen the projectionswere being made in 1987. (d) Governmentcontributions are now estimated to be US$92 million less that projected in 1987. (e) Foreign borrowingwas US$436 million lower in 1987-88 compared to the projected figure in 1987. Local borrowing

11/The revisionof the Investmentplan presently being undertaken follows the recommendationsof the draftreport 'Colombia: The PowerSector and the WorldBank, 1970-1987^, OED, August 1989.

12/ The PowerSector and the WorldBank (1970-1987),op. cit.,vol. I, and Vol.II, chapt,r2. - 65 -

was US$70 million higher. f) Total requirementsfor amortizationof debt were US$23 million lower than expected for the 1987-1988period. g) the net increase in debt for the four year period is projected to be US$311 million more than the US$516 million projected in 1987.

52. For the analysis of the long term the Bank has three different sources of data: (a) Projectionsfor the 1987-2000period provided by ISA with detailed financial statements--thisinformation was prepared as part of the studies of various alternativesto determine the least-cost expansion plan. (b) DNP data prepared in January 1989 covering the period 1987-1996. (c) A more recent projectiondone by FEN in October 1989, which was prepared to serve as the basis for discussionsof future joint Bank-IDB lending to the sector. Annex 1 shows an analysis of ISA projections and makes reference to the views of DNP. In what follows, an analysis is made of the latest figures provided by FEN in October 1989.

53. Our main conclusion is that the least cost investmentprogram adopted by the Government seems to be economicallysound and will probably not change after including reasonableuncertainty in the main determining parameters. Marginal adjustmentsare being made to consider a lower expected rate of G.'r growth in the near future, to include a possible connectionwith Venezuela,and to add the design of gas-fueledplants of a 150-300 MW' size as standbys--torespond to possible delays in the completion of the proposed hydro-electricplants, and to cover the risk of demand growing above the current estimates.13 These adjustmentswill postpone the proposed new hydro-electricgenerating plants for a few years, but will probably not change the composition,size, or sequence of the least cost investmentprogram. Financing of this program, however, will require a very serious effort to raise the sector's internal cash generation even after allowing for reasonablenet transfers from the Government, and for plausible contributionsfrom domestic and foreign borrowing. In addition, an adequate mechanism to transfer surpluseswithin the sector needs to be devised. Otherwise a rise in internal cash generationmight be used for purposes different than financing the least- cost investmentprogram.

54. The financingplan prepared by FEN includes expendituresof US$4.4 billion for the 1990-92 period, of which US$1.5 billion would be financed by increasing the sector domestic and foreign debt, and US$2.9 billion would be financedwith resourcesgenerated by the sector (includingminor direct transfers from the governmentand others)--seeTable 4.1. This financing plan can be substantiallyimproved and it is presently under review. Basically,the plan's internal cash generationcould and should be increased,expenditures could be reduced and, therefore,debt financing could and should be reduced.

13/ The revisionof the investmentplan presentlybeing undertakenis very much in agreementwith the recommendationsof the draft report Colombia:The Power Sector and the World Bank, 1970-19870,OED, August 1989. Table 4.1: COLOMBIANELECTRIC SECTOR INVESTMENTFINANCING

FinancingPlan 1990-92 ShorttallAnalysis (USS1lTio-ns-) W/o Expan. W/Expansion 1990 1991 1992 90-92 Plan Plan Other Total

1. Investments 816.7 726.8 698.1 2040.6 1832.1 408.6 - 2040.6 In-excution 502.6 B11.37 1632.1 lT2.71 - - 1832.1 Now (expansionplan) 114.1 114.6 179.9 408.5 - 408.6 - 408.6

2. Workingcapital increase) 161.6 136.4 264.1 562.0 482.0 0 70.0 552.0 Working capital (sector) 140.5 114.4 227.1 4827. 482.0 - 482.0 Working capital (ISA) 15.0 15.0 20.0 50.0 50.0 50.0 Other (FEN) 6.0 7.0 7.0 20.0 _ _ 20.0 20.0

3. Shortfall1989 114.4 - - 114.4 - - 114.4 114.0

4. Interest 532.4 563.0 694.4 1689.8 1836.1 53.7 - 1689.0 IDC */ of investmentsin execution 65.4 79.9 80.5 225.8 225.8 - - 225.8 IDC of new investments(expans.plan) 5.2 16.8 31.9 63.7 - 53.7 - 63.7 Operationalinterests 410.3 430.6 466.6 1307.3 1307.3 - - 1307.3 FODEX interests 51.6 36.0 16.6 103.0 103.0 - - 103.0

5. Investment& Interests 1426.0 1426.2 1548.6 4396.8 3760.2 482.2 184.4 4396.8 (1)+(2) +(3) +( -i 0%' B. InternalResources 843.7 963.0 1101.9 2908.6 2798.3 110.3 - 2908.6 Gross internalgeneration (GIB sector) 790.7 912.2 1036.9 2738.8 2738. - - 2738.8 Private contributions 16.8 15.6 14.7 47.1 47.1 - - 47.1 Gov., ICEL & CORELCA contributions 6.4 6.3 4.4 16.1 16.1 - - 18.1 Other Resources (GIB not explained) 30.8 ,o.9 48.8 110.3 - 110.3 - 110.3 Minus G18 non-transferable -1.0 *1.0 -1.7 -3.7 -3.7 - - -3.7 7. Deficit -681.3 -462.2 -444.7 -1488.2 -961.9 -351.9 -184.4 -1488.2

8. Deficit Financing 200.6 -91.8 -109.0 0 0 0 0 0 a) Existingloans (Net) -109.0 -2 5 -218.0 -537.5 -637.5 o_ -S37.5 - Disbursements 486.8 376.5 305.3 1167.6 1187.6 - - 1167.8 - Amortizations -694.8 -587.0 -523.3 -1706.1 1705.1 - - 1706.1 b) New loans 309.5 118.9 109.0 537.4 637.4 0 0 637.5 - Defined disbursements 26T4. 74.0 64.0 402.8 402.6 - 402.8 - Expecteddebt relief 44.9 44.9 46.0 134.8 134.8 - - 134.8 9. Shortfall(7)*(8) -380.8 -653.8 -563.7 -1488.2 -961.9 -361.9 -184.4 -1488.2

*/ Interestduring construction. - 67 -

55. The Power Sector could and should increase its internal cash generation. It has been estimated that for each additional US$0.01 per Kw for the average tariff, total revenues would increase by about US$250 million. Therefore, financing of expenditures with the sector's own internal cash generation could be substantially increased. Actually, the single most important problem of the sector has been low prices. In mid- 1989 average revenue was around US$0.04 per Kw, while long run marginal cost (LRMC) was estimated to be about US$0.065 per Kw. 14 This situation of underpricing has continuously been a characteristic of the power sector for many years, at least since the early 1970s. In 1978, the average tariff was less than half (49%) the level of the marginal cost of the interconnected system, 75Z in 1983, and 62% in 1987. 15 A program to raise the average tariff to a level close to the LRMC in a reasonable period of time is badly needed.

56. The sector has continuously generated deficits that have been covered by borrowing and by transfers from the central government. As a consequence, total debt has accumulated to an abnormally high level of US$5.3 billion at the end of 1988, of which US$4.3 billion is external debt (about 25% of the total external debt of the country). If the 1990-92 financial plan were to be carried out as is, the total debt of the sector would increase by about US$1.5-US$1.6 billion, or about 302 in dollars in the three-year period.

57. There is little room for expenditure reductions in the investment plan. Expenditures for the 1990-92 period (US$4.4 billion) include only US$2.0 billion for actual investments (US$1.6 billion to finish projects under construction, and US$0.4 billion to initiate the new expansion plan). The rest (US$2.4 billion) is to increase working cap-.tal (US$0.6 billion), to pay interest (US$1.7 billion) and to cover a deficit left without financir.g from 1989 (US$0.1 billion). The biggest share of expenditures is quite inflexible: US$1.6 billion to finish projects under construction, and US$1.7 billion to cover interest payments, of which US$1.5 billion represent interest on past debt. However, US$0.3-0.6 billion may be saved for the period by reducing to a minimum the working capital requirements and by postponing the new expansion plan.

58. Interest payments on past debt are high (US$1.5 billion), about 552 of the sector's total gross generation of internal resources. This is due not only to the very low level of the average tariff compared to average production cost, but also to the accumulated effect of financing past investments with excessive debt. In addition, the terms of debt have worsened over time, with very unfavorable maturities and grace periods in relation to the structure of revenues. In the early 70's, loans were obtained with maturities of 25-30 years and grace periods of 5-7 years, a

14/ A careful revisionof the estimatesof Long Run MarginalCost (LRMC) is needed. Long Run IncrementalCost is calculatedon the basis of an arbitraryperiod of time for which an investmentplan is defined, and it may change substantiallydepending upon the period selected. LRMC in turn, normallychanges as a functionof the Kw's providedby the system.

15/ The Power Sector and the World Bank, 1970-1987,op. cit., vol. II, ch.6. - 68 -

time span relativelyclose to the actual economic life cf the hydro- electrics plants financedwith these loans (depreciationis normally computed on an average useful life of forty or more years for hydro- electric plants). In 1976, lending conditionsworsened, as is clearly pointed out in the Bank's report from OED already mentioned before... "even for such long lead projects as hydroplants to a maximum of 17 years, with grace periods of 4 years, while simultaneouslylevelling principal repayments. This meant for hydro projects like those in Colombia that: (i) principal repaymentsjumped to a level some three times that of depreciationcharges, and (ii) loan amortization,on average, started 15 months before the plant started operation and 23 months before the Bank made the last disbursementfrom the loans".16 The consequenceis that even projects with very high financialinternal rates of return could face serious cash flow problems during the first years. The situation is even worse in Colombia due to the very low levels of tariffs, which has produced financial rates of return of the order of 7Z-8Z. But, if tariffs were raised too much, to fully eliminate the early years' cash flow problem, big surpluseswould be generated in the later years unless they were allowed to erode with inflation. Marginal cost pricing, together with a more balanced correspondencebetween maturities and economic life of the projects, and between constructionperiods and grace periods is certainly desirable.

I. Recommendations

59. In conclusion, financingof the power sector investmentprogram requires a definitionof the overall sector strategy, designed to take into account the medium and long term macro-economicand energy policies, including the following elements in the key areas indicated below:

(a) electricitypricing: Ci) improvementsin the LRMC-based electricitytariffs establishedby Resolution (86, to reach levels closer to LRMC more rapidly; and (ii) eliminationof unnecessary subsidies,particularly to medium- and high-income residentialconsumers and to non-residentialconsumers;

(b) investmentprogram: evaluationof alternativeprograms which involve: (i) a lower reliabilityof electricity supply (what level of optimally-managedrationing would be acceptable?);and (ii) possible postponementof generationinvestments to adjust to a lower expected rate of GDP growth starting in 1989, and through the constructionof an interconnectionwith Venezuela whos_ technical feasibilityhas recently been demonstrated.

(c) sector finances: Ci) formulatingand implementingsector financial policies to limit the increase of the sector's debt to a level that is sustainable,which has reached very high levels, through increased generationof internal resources and transitory Governmenttransfers; and (ii) establish transfer mechanisms to ensure that funds internallygenerated by the sector be used for purposes that match governmentpriorities;

16/ The Power Sector and the World Bank, 1970-87, op. cit., vol I. - 69 -

(d) institutionalaspects: reforming the sector's institutional structure,management and coordinationto eliminate excessive politicizationof electricitypricing, and to improve financing planning and monitoring. The sector needs to: (i) define unequivocallythe role, objectives,funcJions and authority of each of the most relevant sector institutions,such as Financiera Electrica Nacional (FEN), particularlyin view of its much- enlarged proposed role as FinancieraEnerg6tica Nacional; Interconexi6nElectrica S.A. (ISA); the recently created Energy Board; and the Power Sector Superior Council; (ii) define the basis and creating the appropriateincentives to prevent an excessive turnover at the power companies'highest managerial levels; (iii) define and create adequate mechanisms to facilitate financial transferswithin the sector.

(e) financial planning and monitoring: acceleratingthe implementa- tion of the Power Sector InformationSystem (SINSE) and complement it simultaneouslywith the adoption of a Uniform Accounting System, to ensure the applicationof uniform operationaland accounting criteria, and to facilitateperiodic consolidationof sector finances. - 70 -

CHAPTER V

Health

A. Health Conditions

1. Colombia exhibits a pronounced dualism in health conditions. Two- thirds of the population has benefited from the impressive declines in mortality of -he past 33 years and now must deal with chronic diseases like heart disease, stroke, cancer and diabetes as the predominant health threats. The other third of the population, however, remains mired in the classic healt.i problems of underdevelopment--high infant and child mortality and high adult morbidity from infectious diseases--which, in turn, result from unsanitary living conditions, , poor health practices, and lack of basic health care. This third of the population, mostly rural but also living in peri-urban slums, does not have piped water or sewerage connections and, in the main, lacks effective coverage by the health care system.

2. As a matter of highest stated priority, the Government has declared its intention to extend basic health services to all as a central thrust of its "war on poverty". It intends to help the poor escape the worst health consequences of poverty and unsanitary living conditions by offering primary health care, including immunizations, oral rehydration therapy and instruction, and pre-natal and other maternal and child services. Specifically, the outcomes it would like to correct are the following: i) the high incidence of serious gastrointestinal and respi- ratory complications, ii) chronic malnutrition in one of five children, iii) the fact that fexwer than half of all children fail to receive BCG, DPT, measles or polio vaccinations, iv) resulting infant mortality that averages 100 per 1000 births in the poorest departments, v) rising malaria incidence of about 400 per 100,000 population threatening 18Z of the population, and vi) continuing and dengue problems in endemic areas.

3. The worst health conditions are concentrated in certain regions and groups, on the Pacific and Atlantic coasts, in the tropical regions and among the rural and peri-urban poor. Infant mortality is more than double in the poorer departamentos and territories than in more developed ones. The Government is attempting to improve basic health services in these areas by means of the Bank-financed Health Integration Services Project (HISP), initiated in 1985. The least well off group in the project area is the dispersed rural population. About half the rural population as a whole but only about 1OZ of the dispersed rural population have piped water. Among the latter, only 15X have an adequate system for excreta disposal.

4. It is estimated that 25Z of the Colombian population, most of the dispersed rural population included, does not have access to adequate basic health services. Thus those who are most vulnerable to poor environmental conditions--especially the absence of safe water, sanitary excreta disposal, and satisfactory liquid and solid waste management--are not provided with the treatment and health education services that could assist them in dealing with the problem. Lack of family planning services also contributes to higher fertility for rural, native Indian and the peri-urban groups, whose fertility rates are well above the national average. These - 71 -

high fertility rates, reflecting overly close spacing of births, contribute, in turn, to high rates of maternal and infant mortality. Substantial reductions in infant and child mortality would be possible through proven basic health measures, such as improved maternal and child care, family planning services, health and nutrition education and environmental conditions.

The Health System

5. The percentage of GDP presently spent on health in Colombia is not known with any precision, because of a lack of data on the amount of expenditures on private providers.1 One guesstimate is that the latter may be about 40Z of the total.2 If so, health expenditures as a percentage of

Table 5.1: PUBLIC HEALTH EXPENDITURES, 1985-1988

1985 1986 1987 1988-/ 1989

Per Capita (1987 US$)b/ 49 44 '1 43 - As Percent of GDP: Government Subectorc/ 1.4 1.4 1.5 1.6 1.8 Social Security Subsector 1.5 1.4 1.3 1.3 1.3 Total Public 2.9 2.8 2.8 2.9 3.1 As Percent of Central Government Budget: Government Subsectorc/ 4.6 4.4 5.0 5.1 5.5

a/ Estimated. b/ Including health expenditures by both the government and social security subsectors. c/ Not including government expenditures on child care (ICBF) and water and sanitation programs for towns (INSFOPAL).

Source: Departamento Nacional de Planificaci6n, Ministerio de Salud.

GDP were about 5.2Z in 1987, since expenditures on the public health sector, the other 60Z, are known and are estimated at 3.12 of GDP in 1989 (Table 5.1 above). The public health system is composed of i) the government subsector, operated for the general population by the Ministry of Health (MS), and ii) the social security subsector, operated for

1/ Private providersare importantsources for health consultations,laboratory analyses, drugs and, to a limitedextent, hospitalbeds.

2/ The guesstimateis that of a health sector specialistinterviewed by the mission. Private expendituresin 1978 were estimatedat 33X of the total, based on data from the NationalHealth Survey, 1977-1980 (WorldBank, ColombiaHealth Sector Review, 1982, p.28). A new survey is underway,which should provide data on the levelsof householdexpenditure on private provision of health services. It started in 1987, and three out of five survey regions were completedby the end of 1988. - 72 -

salaried employeesby various social security organizations. The share of social security in public expenditureon health has continuouslydeclined from 52% in 1985 to 42Z in 1988, reflecting the increasinglyserious financial problems faced by the Colombian social security system (see paras. 13-16 above).3

B. Structureand Strategy of the Health System

Coverage

6. According to the conventionalanalysis of coverage in Colombia, approximately15Z of the population, those wealthy enough to opt out of the public system, patronize private institutions;another 15% are involuntary contributorsto and recipientsof health services from,the social security subsector; and the remaining 70% are in theory covered by the government system.

The Government System

7. The structure of the governmenthealth system reflects a service delivery model that is appropriateto the health situationof Colombia. In theory, people are to be provided with Pasy access to health posts and centers staffed by auxiliarynurses and health promoters who provide basic preventive, curative and emergency care on an outpatient basis. These basic care facilities are designed to combat the infectious and parasitic diseases that constitute the major health problems of children under five who account for a third of all deaths. Health promoters, besides having curative functions, are also assigned health education responsibilitiesin their geographicalareas. In an effort to extend their reach, the Government in 1984 developed a "health sentinel" program in which high school students attempt to educate parents in child survival techniques. Health centers are ambulatoryfacilities like health posts but are also supposed to offer the services of a full-timephysician and facilities for minor surgery and delivery. Both are under the supervisionof local hospitals, which in turn, are under one of the 107 regional hospitals, each supposed to have at least eight specialties. Pstients theoreticallymove up in the treatment Dyramid by referral,ultimately reaching the specializedand univrersityhospitals for advanced treatment.

8. In fact, the government system is still far short of achieving either complete or effective coverage of the 70% of the population for which it has responsibility. Firstly, as noted, 25% of the population (36% of the group the government system would like to serve) live at too great a

3/ Althoughexpenditures in the governmentsubsector were smaller than in the social security subsector,the former had the largesthospital establishment measured in beds--638hospitals with 80,000 beds. This compareswith approximately50 hospitalswith 4,200 beds operated by the largestsocial securityorganization (ISS) and 190 hospitalswith 9,000 beds in the private sector. The governmentsystem has more beds but a lower total expenditurethan the social security subsector,because of a much lower expenditureper hospitalizedpatient (patientspay for all drugs), underutilizationof capacityand the fact that the social securityorganizations contractwith the governmentsystem (and the privai.esector) for hospitalcare. - 73 -

distance from governmentfacilities to be consideredas (formally)covered by them. Secondly, the lower levels of the system--healthposts, health centers and local hospitals--are,in fact, underutilized. This is because large numbers of those eligible for treatmenthave opted for improved services in two directions--fromhigher levels of the government system (especiallythe specializedand universityhospitals) and from private providers--orhave forgone treatment altogether.4 What is disturbing is that the services provided, though free, in many cases are so clearly inferior that they are not requested. Most of the 107 regional hospitals are also underutilizedfor the same reason of poor service quality.

9. At present, the governmenthealth system is a basically centralized system,modified by significantpolitical input at the departamentolevel. The Ministry of Health (MS) operates the government's health and facilitiesthrough 33 depart!.rento-leveloperating organizations (SSSs)5 composed of national emplo,>es 6 and headed by boards whose membership is shared between nationa.ly appointedheal.th functionariesand those appointed by the Governor of the Departamento7 , a non-elected official appointed by the national government. There is also citizen representationon the board. The Governo- s power is basically the power to influence appointmentsrather than to make policy directly; the most important health appointmenthe makes is the director of the SSS.8 Consistentwith Colombia's broader decentralizationprogram which, starting in 1986, has required elected mayors, a new law, entitled the Reorganizationof the National System of Health (RSNS) is currently under considerationby Congress under a certificationof urgency by the executive. It would decentralizeresponsibility for basic health services to the local level and would be an attempt to address the management problems of the governmenthealth system by eliciting community involvement. Community involvementwould presumably bring greater patient power to bear and thus improve service quality. How much this would actually change things is uncertain. MS power of policy would remain predominant,municipal funds would probably still account for an only minor

4/ Certain private prcviders,it should be noted in order to interpretthe large private share correctly,are complementaryto, not competitivewith, the governmentsystem, because while treatment is generallyfree in the system, drugs and many diagnostictests typicallyare not.

6/ Servicios Seccionalesde Salud.

6/ In the cases of certain formerlyprivate hospitalsat the local level,the employeesstatus as governmentemployees has been uncertainbut will be regularizedunder the Reorganizationof the National System of Health legislation.

7/ With the exceptionof Bogot6, MedellVnand Cali. Governor chairs SectionalHealth Boards which propose to the Ministerof Health a list of 3 candidatesto head 'SSs.

8/ That the traditionaldiscussion of decentralizationhas taken place ;n a highly centralized context can be seen in the fact that the four semi-independentnational a under .S are called 'decentralized'agencies. If anything,they are mere centralizedthan MS in their operations. - 74 -

fraction of the financing (Table 23 in the Statistical Appendix), and municipal-level involvement would not necessarily reduce the technical and managerial deficiencies of political appointees.

10. The basic problems of the government system are i) underfinancing of recurrent costs of government facilities, resulting in understaffing, undersupply and undermaintenance, and ii) inadequate management, resulting in poorly selected and trained personnel and costly but technologically inferior services. The Health Integrated Services (HIS) Project represents a major attempt to addrass both problems in a project area containing 19 of the poorest health regions (out of 07 regions), but it has been stalled since 1986 by the reluctance of the Govfrnment to fund the increased staffing of paramedics in the health posts and centers established or

strengthened under the project. - The underlying problem with the government health system is financial: tax-based sources of finance are unlikely to expand sufficiently to accomplish the goal of universaL effective coverage without major structural changes in the system. Doable changes within the present system could help improve performance, however, correct underspending on basic health care to a certain extent, and prepare the system for structural changes in the longer run.

11. Even the changes that are doable within the present structure would require strong political commitment. New, effective management at all levels would have to be appointed to carry our the necessary redistribution of funds from personnel expenditures at the system's secondary, tertiary and MS levels to personnel, transport, equipment and supplies at the basic level and effective maintenance programs at all levels. Maintenance programs have been crippled by a combination of poor management, unqualified personnel appointed through patronage, and diluted institutional responsibility. The unsatisfactory experience *with the HIS project to date illustrates the types of problems that wculd have to be solved if effective system-wide changes are to be miade.

The Social Security Subsector

12. The social security subsector in Colombia is complex, with three main subdivisions: i) the Social Security Institute (ISS), a public entity loosely supervised by the Ministry of Labor, which handles retirement, workman's compensation and health services for workers almost entirely in the private sector; ii) about 350 other provident funds (cajas de previsi6n), public bodies which do the same for various categories of mostly public workers (CAJANAL1 0 being the largest); and finally iii) about 70 family assistance funds (cajas de compensaci6n), private bodies that provide some health services to the families of employees of large employers.

9/ The project would assist the Governmentin expandingcoverage of maternal and child health care, immunizations,family planning,oral rehydration,and malaria control to the lowest income groups in the project areas. PreviousBank projects (a nutritionand two rural development projects)also financed basic health care extensions,mainly the expansion of primary health units (UPAs). This is the first health project per se.

10/ NationalProvident Fund. - 75 -

13. The ISS health services operation, which covers 15Z out of the 20% of the population covered by social security schemes, provides 85Z of its services in the three largest cities, Bcgot6, Medellin and Cali. It is funded by a payroll tax on private workers' salaries in the formal sector. Ninety percent of the contributors (together with their employers) pay 7X (in addition to contributions for retirement and workman's compensation) and recei.e comprehensive medical care covering the worker, maternal health care for workers' wives and care for children under 1 year. Ten percent (and their employers) pay 12% and receive comprehensive family health coverage.1 1 The 'cajas de prevision' covering public workers operate under a multitude of financing schemes and coverage plans; little systematic data is available about them, however, or about the "cajas de compensaci6nt. The latter are an interesting competitive element in the health picture. They are private contractors who compete to manage the funds generated by a 4% payroll tax paid by large employers for family assistance. The employers select among the cajas on the basis of their packages of family allowance, health, recreation and other service offerings.1 2 From a fiscal point of view, however, a case can be made to finance them from general taxat±on instead of payroll taxes (see Chapter III) and to increase government fiscalization and control (see Chapter II).

14. How one views ISS's problems depends on how one views the role of ISS in the health system of Colombia. The point of view that has determined the structure of ISS to date, is that the ISS is essentially a health maintenance organization financed by compulsory member contributions. There is a certain element of transfer from better paid urban salaried workers to more poorly paid ones, because the equal- percentage payroll tax m.eans different monetary contributions from people who receive the same service. Nevertheless, the system does net involve large average transfers and the contrib"ltionsare mainly regarded as user fees. However, retiremene-benefits are not covered according to a strict insurance perspective, i.e., the system. does not have in actuarial solvency comparable to that followed by private insurance companies. Instead, those in the labor force now pay to cover the benefits of the retirees, contributing to an intergenerational transfer. Insofar as each worker's contribution to the system looses the link to the benefits that he/she receives, the benefit principle of taxation is lost (see Chapter III) and it would be preferable to finance the system out of the general budget, through a better, less distortionary tax. In addition, the system has clearly underfunded its retirement obligations and it is depending each day more and more out of general budget transfers.1 3 A financial crisis of the whole organization is expected in the near future which could threaten its

11/ Employerspay 2/3 and employerspay 1/3 for both worker and family plans. Choosingbetween the plans is not now voluntary,but depends on the region.

12/ The existenceof many 'cajas de previsi6n'covering government workers is usually regardedin discussionsof the health sector in Colombiaas evidenceof chaotic fragmentation. These scajas",however, could be used to add a competitiveelement from the demand side, if they were free to select health providersfor their largeor small worker groups.

13/ The Ocajas de prevision"for public employeesare in worse shape than ISS, since the pension liabilitiesof most are fully unfunded. - 76 -

health part and either require a governmentbailout or force the health program to suffer the consequencesof a financial crisis in another part of the organization.14 A similar problem involves the financing of the full extension of coverage to family members; will the 12? payroll tax be sufficientto finance the extended coverage? Since the benefits of the system are only loosely related to the benefit principle,and since as time goes financing depend more and more out of the general budget, why not review the overall system to finance it from a more neutral tax, without the unemploymenteffects, regressiveness,and anti-exportbias of payroll taxes?

15. The arguments above open the door to viewing the ISS and other social security organizationsas an integral part of the National Health System. This opens up an entirely different set of problems. To regard the social security subsectoras part of the NHS has been governmentpolicy since at least 1975 and is the philosophyof the current governingboard of the ISS. The RSNS law would grant ISS additional authority and allow it to enroll approximatelysix million campesinosand self-employednon- agriculturalworkers, with average contributionssignificantly lower than those of the current membership of salariedworkers. Whether the organizationcould finance the subsidy to this large number of workers (and their families, if family coverage were to be provided) is in doubt. Raising the payroll tax to finance the subsidy would run into strong oppositionfrom employers and employees. Cost saving through an inc-eased emphasis on preventivemedicine, while useful, could not be expected to provide more than a small part of the necessary funding.15 Perhaps the problem is more academic than real, since there is also a probably disabling problem on the demand side. The problem on the demand side is that ISS may not be able to attract very many low income workers in the informal seccor even at reduced but for them still high contributionrates. As an indicator of this, ISS currentlyhas had the authority to enroll clergy and maids for several years but has only enrolled very few.

Integrationof the Public Health Sector

16. The implementationof the National Health System has revolved around the question of integrationof the social security subsectorwith the government system. Mostly, the motivationhas been to access the earmarked funds of the social security subsectorto finance subsidizedcare for the poor. The motivation for integrationhas also been to avoid wasteful duplicationof facilitiesand services. For instance,contracting for curative services in governmentsystem hospitals by ISS and other cajas has been sought by both systems as a way to increase efficiency in departamentosthat have few ISS members.16 The RSNS law would move

14/ Although ISS receivedbudgetary support from the Governmentin the late 1970s and early 1980s, it appears not to have in any essentialway in the last two years.

16/ The success of the ISS scheme to move to basic health CAB centers remains uncertain.

16/ There is a pricing problem here. Should the ISS pay marginalor full cost of the servicesfor which it contracts? - 77 -

significantlytowards integration-as-redistribution,however. ISS would be "integrated"into the government'ssystem of primary health care by being required to pay part of the cost. ISS members would be eligible to use the government system facilities,but given the different quality standards between the two systems and the differentplaces of residence of those covered, it is doubtful that ISS members would make much use of the currently deficient governmentsystem.

17. The most important policy question involved in the question of integrationis the equity concept desired. A clear decision on equity concept would be more useful for the integrationdebate than vague notions of improved cooperationbetween MS/SSS and ISS. If the governmentwould like to retain a system as at present with different qualit7 tiers in the public health sector--amiddle-quality tier for middle--class salaried workers and a lower-qualitytier for persons in the lower-incomeinformal sector,--itwould in effect support the equity concept of "governmentas safety net" and 'integration"would continue to be a question mostly of efficiency.17 A more redistributionistequity concept, "parity in health care... regardlessof region, rural-urbanresidence or income class", wculd naturally require a single-tiersystem. The "integration"question would then be the problem of how to achieve equality of access, uniformity of quality, and fairness in coinsurance18 and sources of finance.

C. The Government'sProgram Goals Compared to Implementation

The OBasic Health for All' Program

18. The present governmentcame into office in 1986 stating that basic health care is a fundamentalright of the people and announcinga universal basic health program entitled "Basic Health for All". In support of this program, the Plan de Economia Social 1987-1990 included an ambitious plan to expand basic health expenditure (Table 24, StatisticalAppendix).1 The key plan taitet was for an approximatequadrupling of the real annual expenditure on basic health from Col$37 billion in 1987 to Col$137 billion (constant1987 $) in 1990. Real expenditure on other levels in the governmenthealth system and on the social security health system would also increase but by a much lower percentage--26Zto 27Z in total over the four years. Basic health expenditureswould increase from 162 of the

17/ If a third tier financed by voluntaryhealth insurancepurchased by low-incomeparticipants in between these two tiers is established,the system would be a three-tiersystem, since there would necessarilybe a still lower-qualitytier for members of the informalsector who opted not to purchase the voluntary insurance.

18/ *Coinsurance*refers to the percentageof a given treatmentthat has to be paid by an insured patient or his or her family. In the governmentsystem at present,If servlce is thought to be providedthrough a free health insurancepolicy, coinsuranceis 0% for most curativeservices but 100M for drugs and some diagnostictests.

19 The targets given in the plan were aggregatedover the four-yearplan period but have been convertedto a time series using the working documentsused by DNP to prepare them. - 78 -

expenditures of the whole public health system in 1987 to 352 in 1990. Real expenditures on the government health system and on the whole public health system would increase by 1042 and 65X, respectively. DNP working documents show that the major push was to occur in 1988 (Table 24, Statistical Appendix).

Actual Performance v. Planned Expenditure Targets

19. One indication that the plan would not be realized on schedule is that there are no physical targets in the Plan document. 'None still existed as of early in 1989.) Planned expenditures in 19b7 for the government health system20 were to be Col$117 billion (Table 24, Statistical Appendix), and actual expenditures of Col$115.2 were very close to this (a real increase of 7.2Z over 1986 - Tables 25 and 26, Statistical Appendix.) It is not known whether the planned target for basic health care was attained in 1987, since MS data are not tabulated to give data on basic health care (defined institutionally as expenditure on health posts, health centers, local hospitals and programs such as malaria control).2 1 Planned expenditures in 1988, embodying the major thrust to a new level of basic health care, were certainly not realized. Total government health expenditures, which were to increase by 69.9Z, actually decreased by an estimated 3.42 (Table 26, Statistical Appendix). There is some evidence, furthermore, that there was an even greater decrease in basic health services than in the total. First, real expenditures on departamento a.ad local services as a whole (including those for basic health) declined bv an estimated 3.72. In the context of reportedly severe financial stringency affecting the entire hospital system, regional hospitals by anecdotal evidence reportedly secured increased funding at the expense of basic health programs.2 2 Secondly, salaries crowded out non-salary expenses in the government health system in 1988. Salaries increased by a real 17.1% in 1988 (following an ear'lier real increase of 8.1% in 1987). Within the decreasing operating cost total, non-salary operating costs decreased by 30.0% altogether to compensate for the increase in saiaries, and the "other' category of expenses, which includes supplies, maintenance and transportation expenses that are important for basic care, decreased by

20/ Health programs only (not includingthe ICBF child care/nutritionprogram.

21/ A recent World Bank publication,Colombia, Social Programs and Poverty Alleviation: Assessment of Government Initiatives,Report No. 7271-CO. Sept. 18, 1988 (p. 46) stated, 'It is estimatedthat within the Ministry-administerednational health service, primary health care receives less than 10% of total funding'.

22/ 'Massivehospital deficits are commentedon by observers,and a Ministry of Health tabulation of August 22, 1988 indicatesthat referralhospitals had projecteddeficits for the year totalling ColI216.5billion and regionaland local hospitals,Colf12.3 billion. Since projected deficits, which are subsequentlycovered by budget "adicionales',are a normal part of the (abnormal)budgetary process, further analysiswould be requiredto de4.erminewhether the 1988 deficits constitutedan unusuallyhigh degree of financial stringency. - 79 -

45.7% (Table 27, Statistical Appendix).2 3 Based on this evidence, it is reasonable to surmise that expenditures on basic health care and also its effectiveness declined significantly in 1988 rather than increasing rapidly as planned in a push toward "Basic Health Care for All".

Actual Performance v. Planned Financing Targets

20. The Plan de Economia Social included a financing plan in addition to a plan for expernditures. Financial resources would expand to finance the estimated 65Z increase in expenditures. It was not planned that health care expansion under the plan would require increased budgetary support from the ordinary revenues of the national government. Instead, sale of services was to become the major source of financing for basic health (up sixfold in 1990 over 1987 in real terms), and local funds from earmarked taxes on alcoholic beverages and gambling were to rise by 111Z (Table 28). Precise nature of the sales of services was not defined in the plan document,but most of these funds were to come from ISS, the other social security funds and, ultimately, newly established private plans.2 4 Under the Plan's financing schedule, sale of services was to rise from 5.0% of total sources to 21.1Z. How ISS would obtain its share of this increase--whether by an increase in the payroll tax or by alterations of its own health programs--was not specified. In any case, real revenues from payroll taxes used to finance social-security-operated health programs were slated to rise by 26.3Z over the plan period.

21. Actual financial performance in 1987 and 1988 fell far short of the targets set forth in the Plan. Total real finance for the public health sector2 5 increased by only 2.3% in 1987 and 5.62 in 1988. Sale of services, which was specified in the Plan as the major source of financing for basic health, actually declined in both years (by 5.8X in 1987 and increased by only 6Z in 1988), as did reveniues allocated to on social security programs.26

23/ 'Other' expenseswere also crowded out by hospital constructionexpenditure, in additionto the crowding out by salaries. This must have damaged the program effectivenessat ail levels. This crowding-outoccurred because an arrangementwas made with the German Government (KFW) to use funds originallyprogrammed for sLppliesand maintenanceto finish hospitalconstruction. All but two hospitals (of the uncompletedones) will be finisnedby mid-1989. No conscious adjustment in the MS budget is evidentlyplanned to take account of the increasedrecurrent costs associatedwith the new hospitals.

24/ The relativelysmall envisagedamount of coinsuranceand other cost recoveryobtained directly from patientsat all levels of the governmentsystem is also includedin 'sale of services'.

25/ Not includingthe ICBF payroll tax.

26/ The part of the value added tax to be allocatedto municipal social constructionis scheduledas follows: 1988 (SX), 1989 (10%),1990 (20%), 1991 (26%)and 1992 on (30%). - 80 -

22. In the national budget, there was a significantand successful effort to arrest the long-termdecline in real national budgetary support for health.27 In contrastwith the recent past, budgetary support for health increased significantlyin 1987 and 1988, 16? and 4.3Z, respectively.

Planning for Basic Health Expansion

23. Although the Plan period is half over, planning to effect a large increase in basic health care is not very far advanced.Z8 The Ministry of Health has not taken the lead, perhaps because there have been three ministers of health in the last two years (followingfive ministers in the previous four years under the previous government). Nevertheless,the RSNS legislation,the soon-to-be-issuedMS regulationsgoverning basic health care programs, the upcoming evaluation of the Bank-financedHIS Project, a major analysis of the health sector recently issued by a prominent Colombian research institute,29 and, finally, a Bank-financedhealth sector review,whose terms of referencehave been drafted, will all set the stage for basic decisions on the direction that a major increase in basic health care expenditurecan and should take under prevailing constraints.

24. One major decision that needs to be taken and refined is the scope and mechanism of desired universal care. It is likely to be shaped crucially by financingconstraints. In the Plan de Economia Social, the universal care to be provided is 'basic health care' and is defined as a broad list of services. The UNISAF30 plan being developed by the Presidencia is a plan for universal "familyhealth care", a somewhat different and not fully defined concept, which is to be financed by voluntary health insurance and overseen by new institutions. The scheme would cover informal-sectorworkers and campesinos and their families. The Government would pay all or part of the insurancepremiums for the poor. Basic preventive and curative care would be provided through community hospitals. Not fully thought out in this or in the original Plan proposal is the possibility that induced increased demand for secondary and tertiary curative services,which would be generatedby an increase in the quality and quantity of primary health care, might be larger than the decrease in the demand caused by the population'simproved health. In that case, extra financingwould have to be found for higher levels of care as well. An alternative plan with support within the Government and the ISS is to

27/ If the earmarked revenuesof the situado fiscal are included(as they shouto be, real budget allocationsdeclined at an annual rate of 2.2X between 1980 and 1986).

28/ The data base for effective planningalso does not exist. Program accountingdats on basic health programsdo not exist. The conventionalaccounting data that do exist ore availableonly in the SSSs,andare said to contain significantinconsistencies in accountingpractices. The Ministry also has no indicatorsof primary health care effectiveness. iJ FEDESARROLLO.

86/ Unidades Integralesde Salud Familiar. - 81 -

extend ISS coverage to family members of currentlycovered workers and to informal workers, campesinos and their families at different rates of contribution.31 Each also requires a different mix of ordinary government revenues,payroll taxes, and patient payments for direct care and insurance.

25. The central constraint is financial. Financing constraintswill probably be more influentialin determiningthe characteristicsof the plan ultimately decided on than equity considerations. While the current Government has been able to maintain the share of health in the national budget, there is general agreement that this share is unlikely to expand markedly. There is also the continued constraint of institutional weakness. One point of view is that there is little point in trying to expand coverage by more building or more training of paramedics unless the present system can be made to be more effective. The first task in this view is to make the present system work better by institutional strengtheningof both the governmentand social security systems and, in the government system, by shifting recurrent spending from hospital to community-basedcare. Another point of view holds that, taking likely political constraints into account, little can be done to either reform the present health care system or extend it without the infusion of major new funding. In this view, financing is constrainingboth the reform and extension problems and releasing it would allow work to proceed on both simultaneously.

D. The Financing of Expanded Health Care

Earmarked Taxes and Revenues in Health

26. Earmarked taxes are popular in Colombia (see Chapter III) and have been used as a legislativedevice to shelter long-run social spending against the short-runcompeting demands of other ministries and sectors in times of financial stringency. The non-payroll taxes that are earmarked for health in Colombia are either "sin" taxes on gambling and alcoholic beverages or taxes on notarizationand registration. Receipts from these taxes are passed directly to departamentoactivities rather than through the regular budget process. The major technical disadvantageof earmarking is that it can introduce a rigidity leading to wasteful spending when an activity receives greater resources than it otherwisewould. 3 2 It does not appear to be the case, however, that departamento-levelhea..th activities necessarily receive more than they otherwisewould from these

i&/ Otheralternatives would be to try to move towardCanadian or European-stylenational health modelsor the CostaRican model, which would have the MS provide illbasic health services and the ISS all publicsecondary and tertiaryhospital services.

L2/ Earmarkingalso has the potentialof increasinginstability in the appropriationsfor activities not benefittingfrom earmarkingif it increasesstability in thosebenefitting from it. The questionof earmarkingsin taxes is separatefrom the questionof theirbeing inequitable since they fall most heavilyon the incometvof the poor. Taxeson cigarettesand alcoholcan also be thoughtof as healthinsuranc- payments for alcohol-and tobacco-relateddiseases. - 82 -

taxes. The activities used 84-912 of total expenditures in recent years (Table 30), and received only 22-24Z of government-system revenues from these taxes (Table 23). The major distortion in departamento-level expenditures, in fact, is the relative overfinancing of regional hospitals compared to basic health activities, but earmarking funds for departame-ito- level activities does not restrict officials from using them for local hospitals and other basic health activities.

27. Earmarked revenues under the situado fiscal (see Chapter III) are by law allowed to range from 152 to 25% of the ordinary revenues of the national government and are shared between government health and education services at the departamento level according to a formula that gives 74Z to primary education and 26Z to health, and allocates among departamentos and territories so as to give a somewhat greater share to poorer, more rural ones. Despite the restrictions of the situado fiscal, the Government is free to flexibly determine the health or education budgets within most departamentos and territories by allocating regular budget resources so as to produce the desired total allocation. It can also select a convenient level for the situado fiscal between the 152 and 25% limits (in 1988, the Government set it at 18Z of ordinary revenues). Moreover, the base the Government uses to apply the percentages to is a projection of ordinary revenues rather than the revenues themselves, and the Government has some control over projection methodology.3 3 Finally, the Government can (and has) found ways to avoid strict adherence to the situado fiscal formula and make allocations of situado fiscal funds according to other criteria.

28. Because of these abilities, the earmarking of the revenues does not seem to either reflect legislative priorities at the margin nor constitute a major source of inefficiency, since few programs or jurisdictions appear to be allocated greater funds than they otherwise would by the executive. Under these circumstances it would not seem to be a major problem for MS management per se. In fact, it does have one management advantage in that, in contrast to regular budget funds, those from the situado fiscal are known at the beginning of the year and are not subject to discretionary delays. Much of the criticism directed at ea.-marking of taxes and revenues in the health sector appears to be, in fact, criticism of poor, overly politicized financial planning in MS. No detailed flow of funds projections are available to planners in preparing SSS and hospital budgets, and developing these should be a major objective.

The Financing of Expanded Health Care

29. How much would universal basic and/or family health care cost? The Plan estimate for the public health system with universal basic care of Col$388 billion (1987 $) in 1990, is not based un the costing out of physical targets. It may be conservative, considering the induced demand for secondary and tertiary services. There is also no official specification of what universal family health care would entail, nor any estimate of how much it would cost. Without greater specification of detail, no credible cost figure can be given; the 65% increase in 1990 over 1987 in total public health expenditure should be considered a lower bound.

33/ It calculatesit. Projectionas the differencebetween independentlyprojected total revenues and earmarkedtaxes. - 83 -

30. Six major new sources of finance have been proposed for major extensionsof the health system: i) increasedordinary national budget appropriations,ii) new, non-payroll earmarked taxes and increased tax effort for the old ones, iii) reform of the health system resulting in increased efficiency and redirectionof resources from hospitals to basic and/or family health, iv) cost recovery and reduction in demand through sale of services directly tc patients including any coinsurance,v) transfer of resources from the social security subsectorand ultimately from the payroll taxes paid by workers in the formal sector, and vi) voluntary health insurancepayments by rural and urban informal-sector workers. Each has serious problems as a source for a significantextension of the health system; each is likely to be either inefficient,innequitable or insufficient,with the latter being dominant.

31. Ordinary National Budget Appropriations. It is conventional wisdom in Colombia that increased ordinary budgetary resources (including the situado fiscal earmarked revenues)will not be available for health beyond the increased share recently achieved. In fact, since all other major possible sources of finance have even more serious problems, increased ordinary budget should not be eliminated from considerationas part of a package. Furthermore,the greater efficiency and/or equity of general tax revenues should not be ignored.

32. Earmarked Taxes. New earmarked taxes are another source pointed to by proponents of an extended health system.34 The new national lottery, which might eventually raise Col$30 billion (1988 $), would be a significantbut still minor source of funds for an extension of the health system.35 The extent of other earmarking opportunities--eithernew taxes or the earmarking of revenues currently not earmarked--isnot known but as a practical matter is probably not large. There is also no reason to suppose that the potential inefficienciesof any new earmarkingwould be counterbalancedby better social policy. What is clear is that as the proportion of earmarked funds in the health budget increases, so does the danger of inefficiencycaused by earmarking. Earmarked taxes, therefore, should not be considered as the major source of new funds.

33. Reform of the Health Systen. In theory, reform of the health system by redirectionof resources from the secondary and tertiary levels of the hospital system to the basic health componentcould be a significant source of funds to expand the latter. In practice, three major problems limit the Government'sability to do this: i) the vested interest groups of personnel and patients at the higher levels, ii) the increasingcost of modern hospital care, and iii) the financial crisis now gripping the government health system. The reality of the financial situation is that given its existing complement of personnel and its managementpractices, the whole government system is financiallystarved. Except at the tertiary

34/ There is also an opportunityto increasethe yield of the present ones by strengtheningand standardizingtheir administration.

35/ Funds would be distributeddirectly to municipalities. Criteriawould be population,tax/income per capita (within4-6 DNP categories),and cofinancingefforts. - 84 -

level, its subsidizedservice is not even of high enough quality to induce its clients to patronize it fully, and there are many problems at the tertiary level as well. The desire to redirect funds from secondary and tertiary hospitals to expand the basic care component runs aground on the reality that the higher levels appear to be making successfulruns on the financing of the basic component. Just to rectify the current underfinancingof the basic component so that it could provide adequate service to those supposedlynow covered would require significantpolitical effort. Thus reform to make the present system work would not be an easy task. Achieving such a degree of reform that resourceswould be freed to expand the system seems improbable.36

34. Cost Recovery. Cost recovery from patients (as distinguishedfrom the sale of hospital services to social security)would result in funds both from the costs recoveredand from the diminutionof demand that would come from charging a non-zero price for services. Although institutional care is often free to the patient now, much of the cost of treatment is already covered by the patient -virtuallyall drugs and many diagnostic tests. These present expendituresby patients on drugs and tests serve as a rough and ready coinsurancepayment by patients receivingwhat amounts to a free health insurancepolicy. If cost recovery from patients raises patient costs further, the below optimal utilizationrate of the lower levels of the system may decline further, as the poorer patients forgo treatment altogetherand the better off patronize private sector alternatives. Some decrease in the productionof underpriced services may increase system efficiency,but the overall system does not appear to be in a position to raise large net resources through cost recovery. Cost recovery opportunitieswill probably be limited until the value of the service provided increases and effectivehealth insurancebecomes widely available to finance the cost of drugs and tests.

35. Transfers from Social Security. Transfers of resources from the social security subsector,in this case labeled "sale of services",was the major source of funds contemplatedby the Plan to finance the extension of basic health care. The formal idea is that the ISS and other funds would purchase basic health services for their members from the government system, just as they now purchase hospital services. In actual operation, however, only a portion of the transferredfunds would benefit fund members, and a significantcross subsidy was envisaged.

36/ Conceptually,the reform needed is to drasticallyprune the present employe, list and use th- proceedsto purchasesupplies, medicines, spare perts, fuel, equipmentand vehiclesand to hire key paramedical,technician, maintenance and consultantpersonnel from outside the system. Improvedmanagement, which wouhd be part of the solution,wo4ld use the system's newly reconstitutedresources to provide more basic care and better service within the whole system. In the process, largenumbers of hospitalswould probablyhave to be closed or downgraded. Large amounts of politicalcapital would undoubtedlyhave to be expendedto achieve this solution, and further largeamounts would have to be used to retainthe increasednew non- personnelexpense In the budget in the face of budget pressurefrom other ministries. - 85 -

36. There are already significantcross subsidies in the ISS. Cross subsidies from richer members now benefit poorer members, but in extending social security health services to present members' families and to informal sector workers and their families at lower contributior.levels, the amounts of cross subsidy would increasemarkedly. All these cross subsidieswould entail higher payroll taxes and probably falling service quality for present members (modifiedby whatever increases in efficiency, if any, were simultaneouslyobtained). Increased resistanceto cross subsidieshas recently been expressedby labor unions, and it is doubtful that greatly increased cross subsidy transfers are politically feasible. Even if they were, they would tend to raise relative labor costs in the formal sector and, therefore,i) cause higher-than-otherwiseunemploymei.'t in the formal sector as labor costs rose, ii) increase the dualism of the economy as labor in the formal sector became more expensive relative to capital and informal-sectorlabor, iii) give an incentive for payroll tax avoidance and evasion, iv) cdecreasethe average returns to investmentin education and technical training,and v) cause negative macroeconomic effects on inflation and the real exchange rate. Both the political and economic disadvantagesof a large increase in the payroll tax argue against financing big health subsidies for campesinosand informalworkers with it and thereby cast doubt on its use as the predominant financing source.

37. Voluntary Health Insurance. In appreciatingthe difficultiesof finding enough financing from the above sources, some analysts in Colombia have put predominant emphasis on voluntary health insurance to be purchased by campesinosand informal sector workers and provided free or at reduced cost to the poor by the government. This financing source has two great benefits: it would constitutea measure of cost recovery and would subject health financing and services to market tests. If the low-income purchasers do not believe the insurance is worth the premium, they will not buy it. Like unsubsidizedproducers of low-incomehousing, unsubsidized health providers and insurers would have an incentive to design products that would sell, but again like in low-incomehousing, in the existing economic and health regulatoryenvironments in Colombia, even the best of entrepreneursmight not be able to develop health and insuranceproducts that would satisfy governmentregulators and also sell to the public. Although one would hope that the voluntary health insurance approachwas promising for Colombia, little convincingevidence exists to suggest that it is. Extended-term,)ilot projects will be necessary to validate concepts in this difficult area. ^t could be years before any large scale trials are held and a further long period before successfulmodels, if any, could be implemented.

38. Tlis discussionof the major potential sources of financing for the expansion of ends with the following tentative conclusions:37

(a) Both major potential sources of additionalfinance in the medium run--increasedgeneral budgetary resources for the government health system and cross subsidies from payroll taxes routed through the social security funds-- are severely limited in how much might be obtained from them.

L7/ Health sector finance Is one of tho key topics of the health sector study currentlybeing undertakenby US with fi i - 86 -

(b) The other potential sources are either relativelysmall or, in the case of voluntary insurance,premature and thus small in at least the medium run.

(c) This implies that financial sourcesmust be eclectic to be substantialin total amount and that financingwill therefore be the major structuringinfluence on the expanded system.

(d) Moderate extension of the ISS (in particular, to families of current members) will probably be possible as long as the membership goes along with both the extension and the increased payroll taxes. Increasingthe payroll tax will have ecoliomic costs, however. Whether the subsidy to the governmentbasic health system will be forthcomingin the future (as it has not been to date) is doubtful.

(e) Solutions to the present problems of the governmentsystem, including the extensionof the basic health component,are at least partially financeableby a combinationof the minor sources (efficiencyand earmarkedtaxes), but a significantincrease in general revenues will probably also be needed, unless greater- than-expectedfunds are diverted from the social security subsector.

(f) Financing imperativeswill probably make the public health system that emerges from a major effort at extensicA and reform resemble the present two-tier one, hopefully expanded ar.Arationalized.

(g) A major voluntary-insurancesystem for campesinosand informal sector workers is unlikely to emerge unless it is heavily subsidizedfrom general tax revenues, and even then, it will probably be many years in coming.

E. Impact of Past Deficienciesin the Budgetary Process on Health Programs

39. In Colombian public finance, the budget process has created serious inefficienciesin most areas of government. Although many of these deficienciescould be corrected after implementingthe new Organic Law of the Budget, it is useful here to illustratethe inefficienciescreated by the old budgeting system. Agencies have been fpced with extremely difficult flow-of-fundsproblems in both investmentand recurrent Pxpenditures. It is not just that funds are not availablewhen needed, but there is also uncertainty as to both their amount and timing. The budgetary gauntlet,which most agencies must run, starts with the budget appropriationlaws. Using haalth investmentexpenditures (mainly constructionservices and equipment purchases)as an example, it can be seen that in 1987 and 1988, 69.4Z and 78.5% of total appropriationswere still unlegislatedat mid-year (Table 31).38

38/ The mid-yeartiming occurs because the executivedecree contairingthe pro-mid-year appropriationsmust be timed to occur before Congress convenes in July and at least partially depends on the receipt of international commercial-bank funding. - 87 -

40. Even when the funds have been legislated, agencies cannot obligate them until tine appropriation is released by Ministry of Hacienda (ME) and paymcnt is approved for o)ligation by the Contraloria. Since 1 funds must be officially obligated by December 31 or they are lost, agency heads and financial officers expend much political and bureaucratic effort to be able to obligate the funds before December 31. This results in a misdirection of effozt by key executives and a bunch;ng of non-personnel expenditures in the last three months of the year. Health agencies in 1988 received MH releases for only 42.9Z of investment appropriations as late as October 31.

41. After MH release of the appropriation, a series of further bureaucratic hurdles must be gotten over prior to DecemDer 31. Nevertheless, nothing is quite as difficult as it seems in this process, because time-F.onoredways have been devised to make it operate irnmost cases, albeit with some damage to agencies, missions For instance, only limited carryovers of obligated funds are allowed to the next year. Lut funds transferred outside the government arlk considered spent, whether or not the activity has actually started. Agency heads have used this as a loophole in recent years to "spend" a large proportion of thel.r funds on master contracts to certain non-governmental institutions, which then release the funds in the next year tu Lhe real contractors.3 9

Table 5.2: RELEASES AS PERCENT OF HEALTH AGENCY INVESTMENTAPPROPRIATIONS 1988

Total Total Health Agency Investment Releases 2 Appropriations as of 1988 October 31

(billions of Col$)

Ministry of Health 8.5 2.8 32.6 Instituto Nacional de Salud 0.5 0.2 34.1 Fondo Nacional Hospitalario 5.2 2.4 46.6 Servicio Erradicaci6n de la Malaria 2.1 1.6 78.2

TOTAL 16.3 7.0 42.9

Source: Departamento de Planeaci6n Nacional and Fondo Nacional Hospitalario.

39/ In the case of health and education,the Fondo para la EducacidnSuperior (FES) is conventionallyused. Starting in 1988 FES is also allowed to contractwith municipalities. - 88 -

42. To make this process work administratorsmust have finely tuned political and bureaucratictalents, constantlymonitor the situation,and take timely action. Political novices, disorganizea iministratorsor those with enemies in the wrong places may find their agencies seriously hurt in the process. The process is time consuming and, of course, leaves no time for long-termplanning or thinking. Recurrent expendituresin health, which are predominantlyfor personnel, are affected less than investment expenditures. Personnel costs are routinely included in the initial budget law (but without the inflation adjustmentand all fringes). Non-personnelcosts, however, are delayed and squeezed by the budget process, seriously damaging the ability of health care institutionsto deliver tecinologicallycompetent care. The administratorsmust usually expend great effort to get inflationadjustments and appropriatedfunds for non-personnelexpenditures released and paid.40 The process deleteriously affects investmentprojects not only through its impact on the investment budget but also through recurrent cost problems. Supervisionof projects is difficult, for instance, if vehicles are not operable and per diems not allowed. Serious cost overruns in hospital projects are attributed to the difficult budgetary process in both expenditurecategories. 41

43. One inefficiencycaused by the budget process is particularly serious in the governmenthealth system. Some hospital administratorshave learned in the last two years to adopt a risky strategy. The hospital that spends early and much (and thus has a large "deficit"to plead about) is reputed to get the largest budget additions. To adopt this strategy,the hospital needs to mobilize temporary financing for its deficits from supplier credit and commercial bank loans. MS would like to move to curtail these practices, but few hospitals have cost control systems that vou'd allow it to monitor and control them.

P. Conclusions

44. The examinationof public expendituresin the health sector leads to the foilowing conclusions:

(a) Colombia exhibits a pronounceddualism in health cond4tions. For two-thirdsof the populationthe predominanthealth threats are chronic diseases like heart disease, stroke,cancerand diabetes. The other third of the population,mostly rural but also living in peri-urban slums, has to deal with the classic health problems of

40/ As an example, it was assertedby a former administratorof the InstitutoNacional de Salud that the intricatebudget process reduced the productionand resultedin a shortageof vaccinesand oral rohydrationsalts under his successorwho did not know how to manipulateit sufficiently well.

41/ One example cited is the CartagenaHospital building, which was scosted at ColS400 million in 1984. ColS6,000million has been spent as of 198e, and it is projectedto cost Co186,00 when completed in 1989. In nominal Coll, this is a 1400% increase;in real terms (adjustedby the GNP deflator) it is a 366% increase. - 89 -

underdevelopment--highinfant and child mortality and high adult morbidity from infectiousdiseases--which are result or unsanitary living conditions,malnutrition, poor health practices and lack of basic health care.

(b) It is estimated that 257 of the Colombian populationdoes not have access to adequate basic health services. The worst health conditions are concentratedin certain regions and groups,on the Pacific and Atlantic coasts, in the tropical regi.onsand among the rural and peri-urbanpoor.

(c) Despite economic growth over the 1985-88 period, public health expendituresdeclined somewhat as a percentage of GDP. This decline in the public health sector expendituresresulted entirely from a decline in the social security subsector,while the government'sshare in financinghealth increased to about 407 of the total.

(d) The basic problems of the government system are: i) under- financing of recurrent costs of governmentfacilities, resulting in understaffing,undersupply and undermaintenance;ii) inadequate management, resulting in poorly selected and trained personal and costly but tech..ologicallyinferior services; iii) non- availabilityof health services to 25Z of populationbecause of a distance from governmentfacilities.

(e) In its Plan de Economia Social 1987-1990,the Governmentplanned an approximatequadrupling of the real annual expenditureon basic health from Col$37 billion in 1987 to Col$137 billion (constant 1987 $) in 1990. Real expenditureon other levels in the governmenthealth system and on the social se-.urityhealth system would also increase (26 to 27Z in total ove~.the four years). There are indications,that the plan will not be realized--no physical targets are set in the plan document,and instead of a planned increase in total governmentexpenditures on health by 69.9Z there was actually decrease by an estimated 3.4X.

(f) Six major new sources of finance have been proposed for major extensionsof the health system: i) increasedordinary national budget appropriations,ii) new, non-payrollearmarked taxes and increased tax effort for the old ones, iii) reform of the health system resulting in increasedefficiency and redirectionof resources from hospitals to basic and/or family heplth; iv) cost recovery and reduction in demand through sale of services directly to patients includingany coinsurance;v) transfer of resources from the social security subsectorand ultimatelyfrom the payroll taxes paid by workers in the formal sector; vi) voluntary health insurancepayments by rural land urban informal-sectorworkers. Each has serious problems as a source for a significantextension of the health system; each is likely to be either inefficient, inequitableor insufficient,with the latter being dominant. - 90 -

(g) Past deficienciesin the budgetaryprocess have had a large impact on health programs. Many of these deficienciescould be corrected after implementingthe -newOrganic Law of the Budget. - 91 -

CHAPTER VI

Education

A. The Educational System

1. Over the past 25 years, Colombia'seducation system expanded dramaticallyat all levels, and within both the public and private sectors. This continued a long-term educationaleffort that resulted in an overall literacy rate of 88Z of the populationover 10 by 1985. While a third of the children lacked standard access to primary education in 1960, the correspondingfigure today is about 10%. Since the same date, primary enrollmentmore than doubled, secondaryenrollment increased six-fold and higher edacation enrollment rose 15 times. Private education,offered mainly in urban localities,accounted in 1988 for 15.2% of primary enrollment,37.6% of secondaryenrollment and 57.6% of higher education enrollment.

2. Primary education (comprisinggrades 1-5 in Colombia) enrolls 4.2 million students (almostequally divided between girls and boys) (Table 15). This level of enrollmenthas remained amazingly constant for a decade,as the steadily increasingnumber of parents or school-agechildren steadily lowered their fertility. Some 65.1% of those enrolled are in urban areas and the remainder in rural areas (Table 32, Statistlcal Appendix). Over 130,000 teachers are employed in primary schools. On average for the country as a whole, in 1988 about 90% of the relevant cohort (age 7-11) was enrolled in the system, both public and private.1 The period 1980-1988 still saw an urban primary enrollment rate 2.5% below what it had been in 1980, indicatingthat providing schooling for the peri- urban slums is a serious problem.2

3. Secondary education enrolled 2.2 million students in 1988, 48Z of young people in the post-primaryage group, and employed some 100,000 teachers. It was almost entirely an urban phenomenon,with 93.8% of the students in urban schools. The remarkablefertility stasis of the late 1970s and the 1980s has not yet caused the secondary school-agepopulation to level off (althoughthis will happen in the early 1990s). The expansion of secondary enrollmentby 29% in the 1980-1988period was mainly to accommodatethe growing numbers of urban children,who increased in number by 26.0%. The urban enrollmentrate which was relativelyhigh at the beginning of the decade increasedmodestly from about two-thirds to about three-fourths. The measured rural secondaryenrollment rate remained very low at about 7 per 100 children of the age group. This low measured level

1/ Figureson the averagerate for 1987and 1988 werenot available,but thispercentage has been between29-31 per 100 age 7-11 since1980. Assuming30 for 1988,this resultsin an enrollment rate of 90 In the age groupwhen subtractedfrom the grossenrollment rate of 120 in 1988 (Table 33, StatisticalAppendix).

2/ The likelyundercount of the populationin peri-urbanareas also probablyresults In an understatementof the measuredurban enrollment rate decline. - 92 -

probably understates the demand for rural secondary schooling,however, since significantnumbers of rural children commute to secondary schools classified as "urban" or migrate to urban areas to live with relatives so that they can attend secondary school (nct to mention rural-to-urban migration of familieswhere gaining access to secondary schooling is a major motivation) and are classifielas urban enrollment. Even as measured, in the 1980-1988period, rural secondary enrollment increased by 70.3Z and the rural secondary enrollment rate increasedby 43.0Z. This latter may indicate rapidly growing demand for rural secondary education, but further analys4.swould be necessary to be sure.3 Overage primary enrollment should also be consideredwhen assessing the rural demand for secondaryeducation.

4. Why the urban secondary enrollment rate increased relatively slowly during the 1980s (by only 12.6Z) is not clear, but it might be because the private rate of return to general secondary schooling seems to be low. There are a number of other reasons to explain the growth that did occur (higher expectationof entry intc higher education, increased demand for the non-pecuniarybenefits of education due to higher incomes, increased need for daytime child care, increased spillover of rural demand into the urban statistics,etc.). But it is not clear that the underlying per-age-group-memberdemand for terminal urban secondary schooling increased at all. Technical education,which do not appear to suffer from the same low rate of return, is neverthelesscomparatively costly and inaccessible. It is concentratedin diversifiedsecondary schools (INEMs), which offer academic and vocational streams at the upper secondary level and account for only 5Z of secondaryenrollment in all streams.4 Outside the formal school system, the national vocational training system (SENA) is seen as having proved effective in training young workers. It was recently given financial and operating responsibilityfor formal technical education and may in the future increase its accessibility.

5. Higher education served an enrollment of 4se,000 students in 1988, the majority in private institutions. Enrollment in public universities was 183,000, 40.3Z of the total (Tables32 and 33, StatisticalAppendix). Higher education grew rapidly during the 1980-1988 period in response to the same demographicfactors that caused secondaryeducation to expand, and this demographicpush will continue for the next decade before diminishing. Unlike secondary education, growth also occurred in response to increases in demand. Enrollment grew by 67.OZ over the period, and the enrollment rate grew by a large 27.1Z to 7.82 of the relevant age group (Tables 33 and 34). Low-cost programs in business and the other social sciences accounted for most of the increase in enrollments. Part of the increase in demand for higher educationwas due to a price effect. Real average tuition declined during the period because of the increasing share of low-tuition public institutionsand the declining rea. tuition that many of these institutionscharged. By 1988, the high and growing enrollment in higher

3/ The rapid growth of rural secondaryschools (up from 509 in 1980 to 929 in 1988) might simply have resultedfrom the reclassificationof some former students in Ourban' schools as 'rural' studentsbecause they now attend a 'rural' institution. The true increasein the enrollment rate of rural residentswould then be lower.

4/ See source of rate of return estimatesin precedingfootnote. - 93 - education had produced a large population of degree holders, who are now experiencingsevere educationalunemployment. This is probably an indicator of lower returns to non-technicalhigher education than in the 1970s and will probably serve to somewhat dampen the demographically impelled growth ir enrollmentsexpected in the next decade.

Resource Mobilizationfor Education in Colombia

6. Government education expendituresin 1988 were about US$29 per capita (1987 $) and amounted to 2.3Z of GDP (Table 6.1 below). Combined with private expenditureson education (mainlyhigher education) in the range of 1.5-2Z of GDP, total educationalexpenditure in Colombia reached the level of about 4% of GDP. Public expenditurewas below other Latin American countries and private expenditureabove that of similar countries. Educational expenditureshave been declining since 1985 both in real per capital terms and as a percentage of GDP. Real per capita expenditure declined by 17% (from US$33 to US$29) since 1985; as a percentage of GDP the drop was from 2.7Z to 2.3Z. This decline extended to education'sshare of the central government budget, which dropped from 22.8Z to 19.1Z.

Table 6.1: GOVERNMENT EDUCATIONEXPENDITURES, 1980, 1985-1988

1980 1985 1986 1987 1988*

Per Capita (1987 US$) 44 35 33 30 29

As Percent of GDP 2.4 2.7 2.7 2.5 2.3

As Percent of Central Government Budget:

Recurrent 32.7 33.0 34.1 30.9 29.8 Investment 4.6 4.4 4.5 4.9 2.9 Total 22.3 22.8 24.1 22.9 19.1

* Estimated

Source: DepartamentoNacional de Planeaci6n and Tabie 35.

B. The Government'sPrimary Education Go Is Compared to Implementation

The Gap in Primary Education

7. The dualistic pattern referred to in health repeats itself in Colombian education. Elements on both the supply and demand sides result in the low enrollment rates experiencedin some rural areas and contribute to the situationwhere only 20Z of rural students entering Grade 1 complete - 94 -

Grade 5 (compared to 60Z for urban students), and only one-quarter of these continue to secondary education. Rural children pass only 1.7 grades on average compared to a still low 3.5 grades for urban children. On the supply side, many children do not have standard access to the full primary program. Only 60Z of rural sch-Dls offer the full five-year cycle. Supply-side factors in internal efficiency are also significant. Promotion policies have been unduly restrictive and are only now changing with the Government's new 'automatic promotion" policy, introduced during the 1988-89 school year. Poor quality instruction also lessens the value of the educational product. On the demand side, some parents of rural children have not been willing to "buy' even at a n'low price'. Low family income and positive opportunity costs for child labor in agriculture end services make these parents (and their informal sources of liquidity) unwilling to afford the small cash costs of attendance and likely to take their children out of school before completion of the primary grades. In an effort to keep the cash price low, teaching materials, traditionally the responsibility of parents, have to all intents disappeared from rural schools that have not benefitted from the Government's Escuela Nueva program. Teachers and administrators attempt to keep the price low for students by not enforcing parental provision of texts.

The Escuela Nueva Program

8. The Escuela Nueva program is an ongoing project assisted by the Bank, which supports the introduction of a revised rural curriculum for the primary grades in small 1-2 classroom rural schools. Among othet things, it provides mcst learning materials free of charge. It was first introduced in the mid-1970s and is now a well-tested program for smaller schools. It utilizes a learning package that enables children to progress individually or in small groups and teachers to cater to different learning and grade levels in the same classroom. Students progress from unit to unit on the basis of their mastery of core skills rather than end-of-year assessments. The program has allowed many small rural schools, which would not otherwise have done so, to provide the full five-year primary cycle. A 1987 assessment confirms that student achievement in Spaaish and mathematics is higher, and repetition rates lower, for Escuela Nueva students than for students attending traditional rural primary schools. By late 1987, the Escuela Nueva mGdel was used in full or part in about half of Colombia's 26,000 rural public schools.

9. Non-teacher inputs are critical for the Escuela Nueva program, especially: i) the semi-programmed teaching and learning materials, which substitute for teacher time by allowing students to work independently and thus support higher pupil/teacher ratios; and ii) transportation inputs which allow effective supervision by departamento centers and teacher travel to demonstration schools. Similar reforms in curriculum to the 'Escuela Nueva' were supposed to have been introduced since the mid-1980s in other rural and urban primary schools, but they have had limited coverage due to administrative and financial problems. Related to these reforms was a 1983 law requiring that 1Z of the education budget be devoted to textbooks, but this also has not been implemented. - 95 -

The 'Basic Education for All' Program

10. The new GoverrLnentassumed office in September 1986 announcing a "war on poverty" to raise the productivityof the poor. This "war"

Table 6.2: EDUCATIONEXPENDITURES

Four-Year Plan Targets 1987-1990 Compared to 1983-1986

Plan Average Targets Program 4 Years Yearly Avg. 1983-1986 1987-1990 Increase

(billionsof 1987 Col$)

RECURRENT Basic Education for All 85 105 24.4 Secondary & Vocational* 56 88 57.2 Higher 43 53 23.0 Total Three Programs 183 245 34.1

INVEST1ENT Basic Education for All 1 9 825.0 Secondary & Vocational* 3 1 -75.8 Higher 2 10 314.9 Total Three Programs 6 20 206.2

TOTAL Basic Education for All 86 115 33.8 Secondary & Vocational* 59 88 50.2 Higher 45 62 38.2 Total Three Programs 189 265 40.0

* Includes national schools

Sources: Tables 18 and 22. inciuded a program for universal primary education entitled "Basic Education for All", which would increase the quality and accessibilityof primary education for rural and peri-urban children. Highest priority in the Plan is assigned to raising the overall quqlity of primary education and improvingaccess to enable all schooi-agechildren to enroll in the full primary cycle. In support of this program, the Plan de Economia Social 1987-1990 includes a significantincrease in basic education expenditures,both recurrentand investment. Annual recurrent expenditure targets for primary education are not given in the plan, but the four-year target of total expendituresis for an increase of 33.8% over the previous 96 -

four-yearperiod (1983-1986)(Table 6.4). Consideringthat primary education siffered a large real cut in the 1985 and 1986 budgets, compared with 1983-1984,the effort described in the Plan is effectively a pledge to restore the real level of primary education in the earlier budgets in the first two years of the plan and then to add a significantamount of additional resources in the last two.

11. The increase in primary educationexpenditure in the Plan was not designed to be the result of a reallocationof expenditureswithin the educationbudget, however. Total real expenditureson secondary and higher educationwere also targeted to increase by similar or somewhat larger percentagesthan primary education (Table 6.1 above). In investment,a relativelysmall part of education expenditures,a significantIncrease from a small base and some reallocationwas planned in line with the availabilityof external credits. Investmentwould expand in all areas except secondary education,which would be subject to a virtual freeze for the Plan period. The reallocationaway from secondary educationwould free up counterpart funds to be used in investmentprojects in primary education and in higher education.

Actual Performancevs. Planned ExpenditureTargets

12. Actual expendituresduring the first two years of the Plan period have fallen far short of those programmed. Using annualizedestimates as an approximationof what the Plan calls for in each year, total real expenditureswere tc have increased from Col$221.9 in 1986 (Table 35, StatisticalAppendix) to 'ol$238 billion and $269 billion (1987 $) in 1987 and 1988, respectively.5 Instead they declined in both years, to Col$218 and $216 billion, respectively,and both recurrent and investment expenditureswere less in 1988 than in 1986 (Table 35, Statistical Appendix). At mid-Plan, therefore,the planned allocationof increased resources tc education has not been implemented. Primary education sh.red the stagnationof education as a whole. Total real expenditureson primary education, in the Col$85-86 billion range in 1987 and 1988, did match the 1983-1986 average and surpass those of 1986, but they were well below planned expenditures,measured by either the annualizedPlan targets of Col$95 and Col$112 billion for 1987 and 1988 (Table 36, Statistical Appendix) or the average annual target of Col$115 billion (Table 5.1 above).

13. There was some progress in increasinginvestment in primary education during 1987 and 1988, but nowhere near the amount planned (Tables 36 and 37, StatisticalAppendix). Investment expendituresdoubled to Col$3.8 billion (1987 Col$) in 1988 from 1986. Neverthelessthey remained low in absolute terms and well below the 1987 and 1988 Plan targets of Col$6 and Col$13 billion, respectively.

14. There is some evidence that the reform process was actually reversed in 1987 and 1988, if only temporarily. Shifts among expense categories--awayfrom crucial non-personnelexpense--probably worked significantlyagainst the planned education reforms. Despite recognition

6/ Not includingSENA. See precedingfootnote and footnotebelow Table 18 for an explanationof how annualizedexpenditures were derived. - 97 -

of its importanceby the Government,real non-salary expense in the education budget as a wiholeappears to have suffered a large decrease in 1988, according to preliminaryfigures (Table 39, StatisticalAppendix). It had already declined by half between 1983 and 1986, from Col$4 billion in 1983 (1987 $) to Col$1.9 billion in 1986 and had reached a very low level. There was no progress in 1987, and then in 1988 non-salary expense dropped a glaring further 45.9%--downto a mere 0.72 of totel salary plus other expenses.6

Actual Performancevs. Planned Financins Targets

15. Financing in the education sector has been conceptuallysimple since 1975 when the central governmentassumed unitary responsibilityfor finance,which had previously been shared with departamnentosand municipalities. Virtually all of the sector's funds are now derived from the central governmentbudget. Although the Plan hoped for increasedcoCSL recovery in higher education,its financingplan assumed that 912 of the financing for formal education programs7 would come from the national budget (with 4Z from external credits and 52 from all other sources). This being the structure of financing, the shortfall in the Plan's education expenditurescan be immediatelytraced to a shortfall in the amount provided by the national budget. These fell short of planned amounts for education in both 1987 and 1988. In 1987, the shortfall from what the Plan intended was only about Col$20 billion, using the annualized estinates,but in 1988 the shortfallwas large Col$53 billion.

Table 6.3: EDUCATION FINANCE 1987-1988,COMPARED TO PLAN*

1987 1988 Education Expenditures Amount % Amount z (billion Actual (billion Actual 1987 Nat'l 1987 Nat'l Col$) Budget Col) Budget

Actual 218 22.9 216 19.1

Planned: Annualized** 238 24.9 269 23.8 Annual Plan Average 278 29.1 278 24.6

* Plan de Economia Social, DNP. ** Methodology described below Table 18.

Source: Tables 18 and 21.

!/ Colombia is well below both the developingcountry and Latin Amer'canaverages of spendingon instructionalmaterial: for LDCs, the average was 3.4% of recurrentexpenditures (41 countries) and, for Latin America, it was 3.9X (16 countries).

7/ Not IncludingSENA^ - 98 -

C. The Government'sNew InvestmentProgram for Primary Education

16. In 1989 the Government is embarking on a Bank-assistedinvestment and educational policy program to correct some of the worst flaws in its primary education system. It is designed to finance curricular reform and school upgrading for needy rural and urban primary schools. In rural primary education, the program will finance the completion of the introduction of the Escuela Nueva materials to 4,000 schools, upgrade the furniture for 13,000 schools and upgrade the classrooms for 4,300 schools. For urban primary education, texts, classroom materials and teacher training are to be provided for schools in marginal urban areas that do not currently have them, and furniture for 2,100 such schools. Two-thirds of the total amount spent on these items is to go for rural schools and one- third for urban schools. The project also finances components to strengthen MEN's resource use and mobilization policy and its management.

17. The nqw program reverses direction from that of the Plan de Economia Social, which it replaces in the area of primary education. While the Plan justified and made provision for a significant increase in national budgetary resol ces for primary education, the new program backs away from this commitment mainly due to limited implementation capacity of the executing agencies. Therefore, the Plan's already rather modest primary-education investment program was stretched out and reduced with major parts of the program scheduled to be implemented as late as 1995. The program is expected to generate only US$13 million in net new annual recurrent costs by 1994 (about Col$2.2 billion in 1987 pesos).8 The major changes the program expects to produce--an increase in rural age 7-11 enrollment from 70% to 90X and an increase in certified grade 5 completion from 20% to 60Z--are expected to be obtained with better efficiency: (1) major resources can be mobilized within primary education and transferred to needy schools to cover most of their increased recurrent costs, and (2) improvements in educational technology and promotion policy will economize on needs for these resources so that the resources transferred will be sufficient. The program expects to 'De able to hire 5000 additional rural teat.ners in this way--by increasing the average urban student-teacher ratio and thus decreasing the required urban-teacher payroll.

18. If there is no large compensating decrease in demand (or if the urban teacher slots are not actually transferred to rural areas due to resistance from the urban communities concerned), the demand for new rural teachers could be greater than anticipated and require large new budgetary resources to prevent the program from being stillborn. The bottom line is that major educational progress cannot be made for an extended period of time with few new resources.

D. Domestic Sources of Finance in Education

19. Besides ordinary budget appropriations, the only quantitatively important source of financing for Lhe formal education system is the situado fiscal, a portion of ordinary national revenues earmarked for

8/ World Bank, Staff AppraisalReport, Colombia: Second SubsectorProject for Primary Education, November2, 19b' p. 19. - 99 -

health and education. Ordinary revenues and the situado fiscal together account for 95Z of the domestic funds used in the formal public education system. The other 5X comes mainly from earmarked value added tax revenues, and cost recovery in higher education. Within the public education system, funds for particular programs are also generated by internal reforms and reallocations. The private higher education system satisfiesdemand for university places that would otherwise probably be financedwithin the public system and, in this sense, is a source of funds for it. Outside the Zormal system, the National Training Service generates finance equal to 10% of what the formal system receives through an earmarkedpayroll tax.

Ordinary Budget Appropriationsand the Situado Fiscal

20. Starting around 1970, the Colombian Congress developed a complicated revenue earmarking device, the situado fiscal, in an attempt both to protect health and education spending against the short-run competing demands of other ministries and sectors and to ensure that all departamentosand territoriesreceived their proportionateshare (Chapter 3). In the 1980s, any binding structures this may have once imposed on the budgeting process, however, have been almost fully relaxed in practice, according to Colombian educationalanalysts. The situado fiscal, in theory, delivers most of the primary and secondarybudget of the Ministry of Education according to a rigid territorialdistribution of funds, 9 but, despite the formal restrictions,the Government can flexibly allocate the health or education budgets at the margin to departamentosand territories by a number of mechanisms, includingthe following: i) selecting a convenient level for the situado fiscal between the legal.152 and 25Z limits (18Z in 1988), ii) applying the statutorypercentages to well- crafted budget projectionsof ordinary revenues rather than to actual magnitudes, iii) finding ways to literal strict adherenceto the situado fiscal formulas, and iv) allocating regular budget resources so as to produce the desired total allocation.

2'1. The main impact of the situado fiscal at present is not in its allocation formula but in its more automatic than ordinary budget delivery mechanism and in the fact that the amounts distributedare timed more predictably. The Ministry of Finance regularly releases predicable amounts of situado fiscal funds directly to MEN's departamento-levelfinancial units (FERs) rather than according to the more discretionarymethods it

9/ Earmarked revenuesunder the situado fiscal may by law range from 16% to 26% of the non- earmarked revenuesof the nationalgovernment and are shared between governmenthealth and education servicesat the departamentolevel accordingto a formula that gives 74% to primary and secondaryeducation and 28% to health. It also allocatesfunds among departamentosand territoriesso as to give a somewhatgreater share to smaller,poorer, and more rural ones. The distributionformula distributes30% of the situado fiscal equally among the departamentosand territoriesand the other 70% r.cordingto populationas counted by the out-of-date1964 census. The lattergives greater emphasisthan current populationcounts would justify to the slower growing departamentosand territories,which also tend to be the poorest rural areas. - 100 -

uses to release the regular budgetary allocationspassing through the MEN budget. As a practicalmatter, this practice may be more the result of a modus vivendi with the teacher corps than a necessary part of the situado fiscal mechanism, since situado fiscal funds have not always been distributed in a non-discretionaryway. In the 1970s and early 1980s, situado fiscal funds, and the teachers salaries they financed,were often delayed for significantperiods. The result was teacher militancy and frequent,ultimately instantaneousstrikes. In recent years, particularly automaticityand predicting,and, as a consequence,teacher strikeshave been virtually eliminated.

Earmarked Taxes and Other Earmarked Revenues

22. In contrast to health, earmarked local taxes (and local taxes, in general) have little role in financingeducation. Local taxes did have a role before 1975, but in the next five years, the national budget assumed almost complete responsibilityfor financingpublic education. The one significantearmarked tax in education is at the national level. It is a payroll tax on salaries (22 or.private salaries and 0.52 on the government payroll) to support the National Training Service (SENA).10 Not only does this tax support the SENA's extensivevocational training programs outside the formal education system,which are reputed to have enrolled 152 of the urban work force at one time or another, but it is now used to finance the vocational education program at the secondary level.

23. Besides the situado fiscal, there is a second earmarkingof revenues in favor of education from the proceeds of the nationally administeledvalue added tax. The drive toward decentralizationin recent years has resulted in greater sentimentfor restoring local tax bases. In line with this, part of the value added tax has recently been earmarked for municipal social investmentand placed at the disposal of the newly elected mayors.11 In the opinion of observers,the mayors are likely to apportion most of these funds for water, sewerage and health facilities,with educationalinvestment a somewhat lower priority. Nevertheless,if the scheme netted even one billion pesos annually for investmentin primary and secondary educationalinvestment, it would be sizable relative to the national funds currently allocated to it. Even a modest ability of mayors to service bottleneck educationalneeds ignored in national budgets and by the FERs could be very valuable.

10/ There are a number of minor earmarkedtaxes at the national level: 1) a 0.5% payrolltax on governmentsalaries to finance the Higher EducationInstitute for Public Administration(ESAP), and ii) the Fund for Service to Teachers (FSD),which was estimatedto raise ColS100 million in 1988.

11/ The part of the value added tax to be allocatedto municipalsocial constructionIs scheduledas follows:1988 (6X0, 1989 (10%), 1990 (20X),1991 (25%) and 1992 on (30%). - 101 -

Reform of the Education System

24. As in any large organizationwith relativelystable cash flows (in this case, MEN and its FERs), changes in operationspresent opportunities to redirect cash flows from certain activities to others, and these changes thus represent sources of funds for the gainers. A particularlygood example was given above, where the Government'snew primary education investmentprogram plans to free up funds for rural schools by increasing the urban student-teacherratio. The redirectionof funds away from teachers' salaries to non-personnelinputs by the impositionof a percentage requirementoffers a similar opportunity. The attempt to benefit one subsectorby reform or retrenchmentin another sometimesfalls victim to the unanticipated (or the anticipatedbut unprovided for) increase in demand within the second. For instance,without a change in admissions policy, the demographicallydriven increase in the demand for public higher educationwould probably absorb any resources derived from cost recovery within the subsector. The unanticipated (by some) leveling off in the demand for secondary education, as the stable primary age cohorts fill up the secondarygrades, is an example of this possibility. Will the freed-up resources be used to expand secondary educationor used elsewhere in the system?

Cost Recovery and Privatizationin Higher Education

25. There are two major financingquestions in public higher education that affect the whole public education system to one degree or another--costrecovery and scale of operations. Theoretically,cost recovery in public universitiesthrough tuition and charges for researchi product could 'ssistin financingthe institutionsand thereby free up budgetary resources for lower levels of the educationalsystem. This redirection of resourcesto other levels might be done on both equity and efficiencygrounds. 12

26. The equity issue is not as simple as it seems. On the surface, students in higher education have little claim to receive public resources on equity grounds in preferenceto poor, young rural and slum children, because compared with their age peers and regardless of their family background,they already have a high probabilityof achievingmiddle class status. This is the case because: i) they have demonstratedsufficient character and intelligenceto complete secondary school; ii) their secondary credentialsthat will allow them to get relativelywell-paying clerical and industrialjobs in the modern sector job even without further schooling;and iii) most of them will complete higher education and thereby receive still higher incomes. Despite this, it is arguable sound public policy or.equity grounds alone to subsidize students from poorer backgrou:idsin university studies, so that they will not be deterred from seeking social mobility through higher education and the country will not be saddled with a rigid class system.

12/ The Bank-assistedprimary educationproject has provisionfor studies of the financingof higher education,which,among other things, will investigatethis possibility. - 102 -

27. The efficiency issue is clearer. The country will be better off if its talented children become talented,productive adults. This said, the question becomes how much, if any, subsidy should be given to higher education in order to promote the developmentof their human capital, independentof any equity considerations. If fewer than the optimal number of children go on to higher studies, there may be some grounds for subsidy, particularlyin technical fields that are in shortage and have high social rates of return. In many fields there can be significantspin-offs for society beyond the incomes received by the skilled people involved.

28. Nevertheless,the chief beneficiariesof higher education--both students from poor backgrounds and those in technical studies--arethe students themselves. At present, most studentsmake use of intrafamily transfers to pay tuition or depend on state subsidies. What is needed is an effective student credit program, which will allow students to tap their own future income streams to pay for their investmentin human capital.

29. If such a system should come into existence,the role and scale of public universitieswould need to be reexamined. State support could not be justified for student costs in that case, except for a relativelysmall number of students from disadvantagedcommunities and regions (ua equity and efficiencygrounds) and for a relativelysmall number of students studying in key, shortage fields (on efficiencygrounds). Whether these studentswould study in public universiviesor private universitieswould become a second order question. If there were no general subsidies to students,the case for a large scale public university system would decline.

30. These considerationslead to the identificationof major ways in which policy changes in public higher educationwould enable the Government to mobilize public resources for reallocationto lower levels or to improving quality in public universities.13 The first major way would be to freeze enrollment in public universities(or curtail it) with the objective of redirectingenrollment growth to private universities. Sixty percent of the enrollmentin Colombianuniversities is presently in private institutions;there is thus a strong private sector in higher education, which, on past performance,would be able to satisfy the demand for universityplaces. Without a policy to further privatizehigher education, public universit,enrollment would undoubtedlycontinue to increase rapidly. It increasedby 81.6Z between 1980 and 1988 (Table 34, StatisticalAppendix) in response to strong demographicpressure, low tuition and the Government'swillingness to find places in public universitiesfor an increasingshare of the total (Table 32, Statistical Appendix).

13/ This is an option and does not prejudgethe wisdom of spending any freed-upfinancial resources within the university subsector to improve the quality of facultiesand research. - 103 -

31. Continuing strong demographicpressure in the universityage group in Colombia sets the stage for continued large increases in public enrollmentin the next decade.14 If budgets are accommodative,this could easily result in total annual costs increasingby one-third-roughlyCol$15 billion (1987$) per annum--in ten years time. If the enrollmentgrowth were diverted to the private sector, however, this Col$15 billion would be available for other uses in the public sector.

32. Cost recovery through higher tuition within public institutionsis the second major way in which policy changes in public higher education would enable the Government to mobilize public resources for reallocation to lower levels or to improvingquality in public universities. Students would be expected to finance the higher tuition out of family resourcesor educationalloans. The Government'sstated policy from the outset has been to move in the direction of higher real student fees, but this has not been implementedat most public institutions,including premier National Universityof Colombia (UNC), perhaps because student fees have been a politicallytricky question. UNC matriculationfees (assessedon an out- of-date sliding scale of family or personal income dating from 1976) even declined between 1987 and 1988. In the latter year, larger percentages of students began to pay the trivially small minimum semestral fee of about US$1 (table). Other public universitieshave more flexible tul-ion systems based on multiples of the minimum wage, but overall, tuition recovered only 5Z of the total costs of these institutions(compared to 80Z for private universities). If tuitions were quadrupledin real terms, not an easy thing to do rapidly, resources equal to 15Z of current university expenditures--roughlyCol$7 billion (1987) per annum--wouldbe generated.

33. Depending on the pace of implementation,freezing public university enrollmentat present levels and decreasinguniversity subsidies in line with the increase in student fees could together probably generate financing equal to Col$22 billion--abouthalf of current annual real universityexpenditures--by the end of the next decade. While impressive in relation to higher education,this should not be regarded as a panacea for financingelementary and secondaryeducation, since even this large percentagewould be only about 1OZ of current total expenditureson formal public education. Increasingthe quality of public higher education and research would, moreover, be another strong claimant for these funds. One reason for this disappointingconclusion--disappointing because being able to mobilize funds already in the sector and transfer them to activities that would better serve equity and efficiencygoals is in general a highly practical and widely applicable strategy--isthat Colombia has not in the past expanded its public higher eoricationsystem in a grossly dispropor- tionate manner. Thus, the benefits from correctingunwise overexpansion are less.

14/ ContinuingGNP growth will boost this basic demand but saturatingmanpower markets and whatever rise in real tuition occurs will dampen it. - 104 -

Table 6.4: TUITION PER SEMESTER,ENTERING STUDENTS, NATIONAL UNIVERSITYOF COLOMBIA, 1987-1988

Semester Tuition (Col$) 1987-1 1987-2 1988-1

(percentof students)

Below Col$250 28.3 77.2 73.5

Col$251-2000 4.5 0.5 0.4

Col$2001-8000 18.9 3.4 1.65

Over Col$8000 48.2 18.9 24.5

Total 100.0 100.0 100.0

Source: Departamentode Registro y Matricula, UniversidadNacional de Colombia.

E. Conclusions

34. The following conclusionshave emerged from this examinationof public expendituresin the education sector:

(a) There is a serious gap in the accessibilityand quality of educationalopportunity for rural and poor urban children in Colombia.

(b) In its Plan de Economia Social for 1987-1990, the Government sketched a significant (but not large) program for Basic Education for all, which would be financedmostly by national budget resources. These resourceswere to be 402 higher than those of the preceding four-yearperiod and reverse that period's decline.

(c) Actual expenditureson education at all levels have fallen well behind planned levels, to the extent that the Plan f.n education) could be characterizedas not having been funded.

(d) Non-personnelexpenditures, which have been historicallylow, and which are critical to curricularreform and educationalquality, continued to be hard hit.

(e) Aside from the non-constrainingearmarking of certain revenues to education, finance for formal educationalprograms in Colombia is almost entirely dependent on the national budget. - 105 -

(f) Cost recovery and privatizationin higher education are worthwhile in their own right for allocative efficiency and equity. They should not be regarded as a major source of funds of other educationalprograms, however. Quadruplingof tuition and freezing of enrollment in public universitiesat the 1988 level would only free up about 1OZ of the education budget for other uses by the end of the next decade.

(g) The most straightforwardway to enable speedy gains toward univer- sal basic education to be achieved would be to restore education's share of the national budget and fund and strengthen sector management. - 106 - ANNEX 1 Page 1 of 16

Financing of the InvestmentProgram 1987-2000

1. The analysis is presented for two periods: (a) short term 1987-1990, the main focus of the commitments for the Power Sector Adjustment Program supportedby the PSAL and a period for which a comparisonmay be made of the actual results and current expectationswith those of 1987 when the PSAL was approved; and (b) long term 1987-2000 for which ISA has prepared detailed financial statementsas part of its studies of various alternativesto determinethe least cost-expansionplan. In the latter respect, the financialprojections corresponding to the expansion plan through 2000 approied by ISA in December 1988 (para. 43) are used. Where pertinent, reference is also made to the views of DNP as expressed in a document, dated January 11, 1989, which covers the period 1987-1996 and which evaluates the macro-economicrestrictions which may affect the investmentprogram and financial situationof the power sector in the medium term.

Short Term tPSAL Period)

2. In June 1989, FEN provided updated financial data for 1987-1990 for the consolidatedColombian electricity sector, includinga comparison with the data used in 1987 for the PSAL. The actual figures for 1987 and 1988 are based on audited statements for individual companies to the extent available and preliminary figures for the others. They are subject to important adjustments which may be needed in the data for the ICEL group for which certain items without sufficient supporting actual data have been estimated. Adjustments may also be needed to reconcile intercompany accounts, which are eliminated in consolidation to avoid double counting, between ISA and the companies wlhich are its shareholders. Sales figures to final consumers may also need to be adjusted.

3. The forecasts for 1989 and 1990 are based on assumptions authorized by the Government with respect to macroeconomic indicators for inflation and exchange rates, external sources of financing, and the investment program. They also reflect the targets for reducing losses and accounts receivable set forth in the recently instituted Emergency Plan, as well as targets for reducing accounts payable to suppliers and contractors established by FEN and DNP. The impact of the recently agreed program for increasing tariffs of EEEB is not reflected in the consolidated projections for 1989 and 1990.

4. The comparison in US dollar terms, of these updated figures and the corresponding ones elaborated in 1987 for the PSAL, as presented by FEN, is reproduced in the following tables:

(a) Table 4.1, showing macro-economic assumptions, GWh sales and average price per KWh, and summary income statement figures;

(b) Table 4.2, detailing the sources and applications of funds; and

tc) Table 4.3, summarizing the financial plan. - 107 -

ANNEX 1 Page 2 of 16

The figures used by FEN for those elaborated in 1987 differ from those which were presented in President's Report No. P-4676-CO for the PSAL. For the most part, these differences are minor and unimportant. For convenience, the comparison piepared by FEN is used in this analysis.

5. Comments on FEN's June 1989 financial projections and on the principal differences in the financial situation of the power sector as forecast in 1987, and as has actually occurred in 1987 and 1988 and as now forecast for 1989 and 1990, are presented below.

Sales, Tariff and Income StatementData for 1987-1990 (Table 4.1)

6. As compared to the 1987 forecasts, electricity sales (GWh) were actually lower by about 1Z for 1987 and 1988, and are now forecast to be 2.4% higher in 1989 and 1990, or a net increase of 0.84Z in total sales for the four years. In current US dollar terms, the average price per kWh is now expected to remain virtually unchanged at about 4 US$ cents, whereas an increase from 3.8 cents in 1987 to 4.4 cents in 1990, or about 4.6% per year, was expected in 1987. To some extent, this is attributableto a greater real devaluation of the compared to US$ than was predicted in 1987. It is significantto note, however, that in constant 1987 peso terms, the average price per kWh declined from 9.6 in 1987 to 9.1 in 1988 and is estimated to recover to about 9.5 in 1989 and 1990. Including the impact of the recently agreed increases in EEEB tariffs, the average sector price per kWh in 1990 would be 9.6 in constant 1987 pesos. This indicates that the increases in real terms in average peso prices for electricity expected under current tariff policy (JNT resolution086) are not being achieved.

7. The income statementcomparisons in US$ terms are very unfavorable. Net income for the four year period is now estimated to US$1,033.5 million, or 86Z, less than in 1987. However, US$524.5 million of the decrease is accounted for by an increase in interest charged to operations,which as explained in para. 54, is offset by a reduction in interest charged to construction,with the result that there ,s very little difference in total interest charges. Nevertheless,the rmnaining differencesare very significantsince they account for a reduction of US$413 million, or 13%, in gross funds generated internally. This is the result of reductionover the four-yearperiod of US$112 million, or 3%, in revenues from sales to final consumers,US$136 million, or 42, in other operating revenues, and US$240 million, or 144Z, in net non-operating income, offset by a reduction of US$75 million, or 6%, in cash operating expenses. The increase of US$78 million, or 7X, in the provision for depreciationand other non-cash charges for the four years does not affect the generation of funds.

8. As indicated earlier, to some extent the poorer US$ results reflect a greater real devaluationof the peso than anticipated in 1987. Other contributingfactors are the inabilityto achieve a real increase in - 108 -

ANNEX 1 Page 3 of 16 the average price of electricityand slower progress than originally planned in reducing losses. The very significantreduction in other operating revenues is attributedto over optimisticprojections of sales of surplus electricityat special concessionaltariffs for new industrial consumerswhich have not materialized. No reasons have been advanced to explain why significantdeficits rather than profits are being realized from non-operatingactivities. This deserves further scrutiny.

FinancinR Plan and Sources and Applicationsof Funds. 1987-1990 (Tables 4.2 and 4.3)

9. As stated in para. 51, the comparisonof income statement figures indicates that the estimate of gross internallygenerated funds for the four year period is now US$413 million less than in 1987. This is also indicated in the comparisonmade on the first line of the Funds Statement (Table 4.2). As presented in the comparisonof Financing Plans (Table 4.3), however, smaller amounts are used for gross funds generated internallyand the differenceis a decrease of US$514 million for the four year period. This differencearises because an item shown on the Funds Statement as "other uses' (in addition to changes in working capital) does not appear on the Financing Plan table but is implicitlydeducted from gross internallygenerated funds. In the presentationmade in 1987 in President'sReport P4676-CO, the item 'other uses' does not appear on the Fund's Statementbut was apparentlyincluded as part of Operating Expenses. No explanationhas been made for this substantialincrease of about US$100 million in "other uses', and this should be clarified.

10. The comparisonof debt service requirementsis affected by a change the Colombian authoritieshave made in the basis for the allocation of total interest charges which has resulted in a substantial increase in the amounts charged to operationsand a substantialdecrease in the amounts charged to construction. This affects the amounts shown as net internally generated funds which reflect deduction of interest charged to operations and amounts shown as constructionor investmentrequirements which include interest charged to construction,but it does not affect total funds requirements. Including interest charged to construction,there is only a small change in total debt service requirements,which have been about US$98 million lower in 1987 and 1988 and are estimated to be about US$124 million higher in 1988 and 1999, or an increase of US$26 million for the four years compared to the 1987 projections. This is not a material change in the flow of funds for the power sector.

11. There are differences in the amounts shown for borrowingsand Governmentcontributions on Tables 4.2 and 4.3, with larger amounts for borrowingsand smaller amounts for Governmentcontributions shown on the Funds Statement than on the Financing Plan. This differenceis due to the treatment of the US$300 million proceeds of the PSAL, which are shown as part of foreign borrowings on the Funds Statementbut under a caption referring to Governmentcontributions and loans on the Financing Plan. - 109 -

ANNEX 1 Page 4 of 16

These proceeds were also shown as part of the Government'sequity contributionsto the power sector in the President'sReport for the PSAL. For purposes of this analysis, the amounts for borrowings and Government contributionsshown on the Financing Plan table will be used. In addition, in order to reconcile the amounts shown on the two tables for the projections elaborated in 1987, it is assumed that the Funds Statement overstatesthe amount of fore'.gnborrowings for 1987, as projected in 1987 by an amount equivalentto the US$52 million surplus shown thereon for that year. This brings the figures for the borrowing program expected in 1987 much closer to the amounts detailed in President'sReport P4676-CO.

12. One additionaladjustment is needed both to the Funds Statement and the Financing Plan to understand properly how the power sector's financing requirementsare being met in the 1987-1990period. In order to meet their financial requirements,the Colombianpower companies are borrowing significantamounts from the Fondo de Exportaciones (FODEX). On the consolidatedbalance sheet for the power sector, the amounts of these borrowings at year end are shown as US$332 million for 1987, US$416 million for 1988, USS440 million for 1989, and US$319 million for 1990. These amounts are not explicitlyindicated on either the Funds Statement or Financing Plan tables because they were consideredto be short-term borrowings; instead, they were included as part of net changes in working capital. For purposes of this analysis they will be consideredpart of long-termborrowings.

13. Summarizedbelow are the principal differencesin the sources and applicationsof funds and financingplan for 1987 - 1990 as projected in 1987, and as has actually occurred and is now forecast in the June 1989 FEN report:

- The cost of the investmentprogram, exclusive of interest during constructionand without any change in its scope, has increasedby US$199 million, or about 10%, to a total of US$2.26 billion. For the two years 1987 and 1988 the increase is reported to be US$41 million; and it is estimated to be US$158 million for 1989 and 1990.

- Changes in net working capital, instead of increasingand representinga requirementof funds, have on balance been negative and representeda source of funds. This is the case even treating the FODEX borrowings as long-term rather than current liabilities. On this basis, changes in working capital have been reduced by US$132 million for the two years 1987 and 1988 and are expected to be reduced by US$47 million in the two years 1989 and 1990, compared to what was projected in 1987. For the four years, instead of a.a increase in working capital totallingUS$223 million as projected in 1987, the increase is now expected to by only US$44 million, or a reduction in this requirementof US$179 - 110 -

ANNEX 1 Page 5 of 16 million. About US$91 million of this reduction may be attributed to higher increases than expected in 1987 in current liabilities for items such as interest payable and accounts payable to contractors than increases in the components of current assets such as accounts receivable and inventories. The other factors which would explain the reduction in working capital requirements are not apparent from the limited comparison of balance sheet data elaborated in 1987 with the actual and present estimated balance sheets provided in the June 1989 FEN report.

- Gross funds generated internally have declined by US$514 million over the four year period, with US$236 million of this decline occuring in the two years 1987 and 1988, and US$278 million expected to occur in the next two years 1989 and 1990.

- The comparison of net internally generated funds is distorted by the change in the basis for allocating total interest charges to operations and construction. As previously explained (para. 54), the US$26 million ircrease in total debt service (which included a US$15 million increase in total interest charges) for the four year period has a minor impact.

- Government contributions are now estimated to be US$298 million over the four years, a decline of US$92 million from the US$390 million projected in 1987, with US$5 million of the decline occuring in 1987 and 1988, and US$ 87 million anticipated for 1989 and 1990.

- Consumer contributions, a relatively small factor in the flow of funds for the power sector, are now estimated to total US$ 77 million, an increase of US$14 million over the US$63 million projected in 1987.

- Foreign and local borrowings, exclusive of those from FODEX, were about US$436 million lower and US$ 70 million higher, respectively, in the two year period 1987 and 1988 and are expected to be about US$873 million higher and US$186 million lower, respectively, in the two year period 1989 and 1990; over the four year period total foreign borrowings are now expected to be about US$437 million higher and total local borrowings US$116 million lower, or a net increase in total borrowings of US$321 million.

- To meet the sector's financial requirements, reflecting the needs arising from the large reduction in gross internally generated funds, the increase in physical costa of the investment program, and delays to date in arranging and drawing down expected foreign :redits, "short term" loans - 111 -

ANNEX 1 Page 6 of 16

from FODEX have been provided and are expected to be utilized as a very significantand new source of sector financing (para. 56) which was not anticipated in 1987. Including this source, total borrowings (foreignand local) were US$50 million higher, rather than US$ 366 million lower, in the two year period 1987 and 1988; and are expected to be US$24 million higher than the US$687 million increase previously indicated for the two year period 1982 and 1990.

Total requirementsfor amortizationof debt (exclusiveof FODEX borrowings)have been US$23 million lower in the two year period 1987 and 1988, and are expected to be US$ 34 million higher in the two year period 1989 and 1990, or TTS$ll million, or only 1/2 of 1%, higher for the four years than projected in 1987. Including the anticipated reductionof US$ 121 million in FODEX borrowings anticipated for 1990, total amortizationfor the four years would be US$132 million, or 62, higher than the 1987 projection.

The net increase in debt (exclusiveof FODEX) for the four year period is now projected to be US$827 million, or US$ 311 million more, than the US$516 million projected in 1987. For the two year period 1987 and 1988, the increase in net debt would be only US$38 million more than the US$380 million projected in 1987. For the two year period 1989 and 1990, the increase would be US$789 million more than the US$135 million projected in 1987. Including FODEX, the net increase in debt for the four years is US$630 million greater than the 1987 projection,with US$73 million of this occuring in the first two years and US$557 million in the last two years.

14. Since the two years, 1989 and 1990, are for the most part still prospective,it is useful to consider them separately. As indicatedby the analysis presented herein of FEN's June 1989 forecast, the very large increase in external resourcesto meet the power sector's financial requirementsduring this period, in comparisonto what was projected in 1987, is needed because of: - 112 -

ANNEX 1 Page 7 of 16

Increases in - US$ Million

Physical cost of investment program 158 Total interest costs 90 Debt amortization (including US$121 million reduction in FODEX debt in 1990) 155

Decreases in - Gross funds generated internally 278 Governmentcontributions 87

TOTAL 768

To meet these needs, additionalinternal resourceswould provide about US$46 million from reductionsof working capital and US$10 million from consumer contributionsfor construction. The balance of US$712 million would be raised by additional borrowings. As indicated above, to some extent the larger borrowingsplanned for 1989 and 1990 permit a partial repayment of the FODEX 'short term" borrowings as well as a reduction in the amount previouslyplanned to be provided in 1989 and 1990 as Government contributions. The larger borrowings are needed, however, principallyto offset the very large drop in gross funds generated internallyand the significantincrease in the physical cost of the investmentprogram.

15. In addition to the tables provided in FEN's June 1989 report for the consolidatedColombian power sector comparing the 1987 projectionswith actual and current forecasts for 1987 - 1990, which have been reproduced herein as Tables 4.1, 4.2 and 4.3, the report alsc contains consolidated income statements,funds statements,and balance sheets, in Colombian pesos and US$, for 1987-1990. The data in US$ in these income and funds statementsis identical to that reproducedas wactuals' in Tables 4.1 and 4.2. It is surprising,therefore, to note that the balance sheet amounts for outstanding debt indicate that total outstandingdebt would increase from year end 1987 to 1990 by only US$222 million based on the amounts of borrowingsand amortizationshown on the Funds Statement and Financing Plan whereas this analysis indicatesthat total outstandingdebt would increase by US$762 million over this period. This should be explained.

Long Tert (1987-2000)

16. The financialprojections for the period 1987-2000prepared by ISA are contained in a report issued in December 1988. The figures for 1987 are historical while those for 1988-2000 are forecasts. The ISA report contains a set of financial statementsproviding pertinent annual data for Colombia's consolidatedpower sector for each of the expansionplans evaluated. These include: - 113 -

ANNEX 1 Page 8 of 16

- Source and Applicationof funds (in US$ millions)

- Financial Ratios, plus details of the borrowing program (in USS millions)

- Income Statement (in Col$ millions)

- Source and Application of Funds (in Col$ millious) - Balance Sheet (in Colt millions)

- FinancialRatios, plus some signific.antvalues (in Col$ millions)

The first, second, and last of these statementsfor the alternativewhich was determinedto be the least-costexpansion plan are attached as Tables 4.4, 4.5 and 4.6. Although the informationcontained in these statements for the period 1987-1990 does not correspondto the more current actual and forecast figures set forth in FEN's June 1989 report, the ISA forecasts are broadly indicativeof the results which may Le achieved and the problems which may be confro.-tedin the intermediateand long term.

Sales, Tariffs and Income StatementData 1987-2000

17. The GWh sales figures used by ISA for the period 1987-1990 are higher than those used by FEN by about 3.52 for 1987, 21 for 1988, 1.5Z for 1989 and only marginally for 1990. More significant,they are consistently higher than those used for the referencedemand forecast (para. 18) by 3.5% to 41 for each of the years 1987-1989and about 1.252 on average for each of the years 1990-2000. Although adjustmentof sales figuresmay be needed to reflect actual results and current short run expectations,the average annual growth rate implicit in the GWh sales figures for the basic long run financial forecast should be consistentwith the referencedemand forecast used to determine the least cost expansion plan.

18. The average revenue per kWh used by ISA, in Colt, compared to that in the June 1989 FEN report, is 3.42 lower in 1987, but higher by 2.7Z in 1988, 1.8% in 1989, and 3.8% in 1990. In real terms, the ISA forecast assumes annual increases in average revenue per kWh in Col$ of about 2.7Z in 1988-1990, 0.52 to 0.7% in 1991-1994,and 0.2 to 0.1% for 1995 to 2000. Using the exchange rate and inflationassumptions applied for the ISA report, the average revenue per kWh in constant (1988) US$ would be 4.01 in 1988, increase to 4.31 in 1992, 4.37 in 1996 and 4.38 in 2000. Implicitly, therefore, the ISA forecast expects very little progress in raising average tariffs to LRMC. This should be reexaminedsince it should be feasible to achieve significantlymore in this respect over the period covered by the forecast. - 114 -

ANNEX 1 Page 9 of 16

19. The progress expected to be made in reducing losses as a percentage of total demand is another significant factor whizh should be reviewed. Under the recently agreed Emergency Plan now being ;implemented, targets have been set which are somewhat lower than ISA assumed, namely a reduction from 22.7Z in 1989 to 18.52 in 1992 compared to a decline from 23.1Z to 19.6Z under the ISA forecast. ISA also assumed subsequent gradual annual reductions down to 17.62 by 2000. Since more substantial reductions in losses should be achievable over the period to 2000, ISA's forecast of total sales and revenues appears to be understated in this respect.

20. The ISA forecast makes allowance for annual increases in operating expenses, exclusive of fuel costs and provision for depreciation, which are equivalent on the average to 4.62 per year in real terms (constant 1988 Col$). This is less than the average annual increase in demand of 5.72 expected under the reference demand forecast, but it may be adequate. A careful review of this question, however, would be advisable. Fuel costs are a relatively small component (about 4.5Z) of total operating expenses of the consolidatad power sector. About two thirds of fuel costs are for natural gas which is highly subsidized in the sense that they are less than economic costs. The ISA forecasts make no allowance for the possibility that policy changes would be made to increase gas costs to eliminate this subsidy. The depreciation provision is equal to about 3% of average gross plant in service which is reasonable.

Financing Plan and Sources and Applications of Funds, 1987-2000

21. The financial situation of the consolidated power sector, as presented in the ISA forecast, is very constrained until about 1995. Thereafter, the situation become very comfortable, with substantial surpluses projected for the three year period 1988-2000, which would accumulate to US$585 million by the end of 2000.

22. Two key and complementary indicators of this are the line on Table 4.4 showing net internally generated funds (i.e. gross internally generated funds less total debt service) and the line showing total debt service coverage on Table 4.5. Until 1994, the latter is negative or marginally positive and increases to 1.21 in 1995 and 1.36 in 1996. It is not until 1997 that the debt service coverage ratio reaches what is usually considered an adequate level (at least 1.5) and thereafter become very ample. Parallel to this, the average rate of return on average net revalued fixed assets in service, which was about 6% in 1988. remains below 8% (except in 1991) through 1993, and then increases from 8.5% in 1994 to 12.3% by 2000.

23. The borrowing program relies heavily on foreign sources until 1991. Throughout this period, total foreign borrowings are substantially greater than the estimated foreign component of physical construction costs, and they are also substantially in excess of the total foreign costs of the construction program including capitalized interest until 1990. For - 115 -

ANNEX 1 age 10 of 16

1992-1994, foreign borrowingsare about the same amount as the foreign component of physical constructioncosts; thereafterthey are in a lesser amount. Four categoriesof foreign borrowing are identified:existing commitments;future; the expansion plan for which borrowingsare projected in the same amount as the estimatedphysical foreign costs of the expansion plan until 1996 and thereafterin smaller amounts; and deficit. The provision for foreign 'deficit'borrowings ends in 1994. No explanationis given as to how the amounts for future and deficit borrowingswere determined. It would be helpful to know this.

24. Local borrowingsare a significantsource of funds throughoutthe period, in annual amounts averagingUS$138 million in 1987-1990,US$377 million in 1991-1996,and US$125 million in 1997-2000. There are three categories of local borrowings:existing for 1987 and 1988; future for 1988-2000; and deficit for 1989-2000. Until 1997, the amounts of local borrowings attributedto the "deficit"category greatly exceed those for the "future' category. It is not apparent how these amounts were determinedor what the relationshipis of those in the "future"category to the local component of physical costs of construction. This should be clarified. The feasibilityof local borrowings in the amounts forecast, particularlyin relation to the capacity of the local capital market, also should be reviewed (see para. 71).

25. In addition to inadequatetariffs and inefficiency,the constrained financial situationof the power sector is due to the inappropriateshort grace and repaymentperiods for foreign and local borrowings,which result in an excessive burden for repaymentand amortizationof debt. The "deficit"borrowings, both foreign and local, may compensate for this. Instead of an ad-hoc approach it would be preferable, in this respect, to have a systematicand well designed plan for refinancingforeign and local borrowings so that grace and repayment periods are more appropiiatelyrelated to constructionperiods and the lives of the assets which will generate earnings to service debt. If the forecasts,however, are correct in projecting substantialsurpluses in 1998-2000,the borrowing program should be reduced, rather than continue to provide for "deficit"loans.

26. Allowing for amounts required for repaymentof debt, the ISA forecast indicates that the power sector would increase its total outstandingdebt until 1992. Thereafter,net borrow ngs would become negative. The annual increase in total net long-term debt would average US$281.5 million in 1987-1990 and US$177 million in 1991-1992;the annual decrease in total net long-term debt would average US$153.5 million in 1993-1996 and US$260 million in 1997-2000. On this basis, the debt/equity ratio which was estimated to be about 50/50 at the end of 1989, would be about 40/60 by 1992, and would decline further to about 25/75 by 1996 and 12/88 by 2000. Such large declines normally would not be essential and would be remarkablesince debt/equityratios of 50/50, or even 60/40, are considered prudent for the electric power industrywhen its earnings - 116 -

ANNEX 1 Page 11 of 16 provide a reasonablereturn on invested capital and repayment terms on debt are reasonably related to constructionperiods and the period over which investmentin fixed assets generates earnings.

27. The reduction in net borrowings forecast by ISA would occur principally for foreign borrowings,which as indicated in Table 4.5 would become negative by US$64 million in 1991 and remain negative in annual amounts ranging from US$177 million to US$452 million, and averaging US$265 million, in 1992-2000. Net local borrowingswould not become negative until 1997. During 1991-1996,when the ISA forecast projects heavy reliance on local borrowings (para. 68), annual net local borrowingswould increase by an average of US$214 million, compared to an average increase of US$64 million in 1987-1990 and an average reduction of US$33 million in 1997-2000.

28. ISA's forecast allows for Governmentcontributions totalling US$344 million, or an average per year of US$86 million, in 1987-1990, and US$307 million, or an average of US$51 million, for 1991-1996. No Governmentcontributions are expected in the period 1997-2000. Consi- deration should be given to the feasibilityof ending reliance on Government contributionsearlier than 1C97. This would most likely depend, however, on the rate of progress which may be achieved in improving the financialperformance of ICEL, CORELCA, and their subsidiaries.

29. The constructioncost estimates should be carefully reviewed to check whether they are adequate with respect to:

- the translation into current terms of the constant cost estimates used in determiningthe least-costexpansion plan

- sub-transmissionand distributioninvestments

- expendituresfor the next phase of expansion for commissioningbeyond 1996.

Total constructioncosts subsequentto 1996 appear to be lower than would be required to keep pace with the expected growth in demand as well as to allow for higher cost generationoptions than previously developed. The other points mentioned also deserve further scrutiny.

DNP's Views Concerning the Financial Situation in Medium-term (1989-1996)

30. In a report issued in January 1989, DNP evaluated the financial situation of the electric power sector in the medium term, 1989-1996. Its purpose was to establish appropriatemaximum limits on the electric sector's share of total public investmentand public debt consistentwith management of the economy and the needs of other sectors. A table summarizingits preliminary financialprojections, covering the years 1987-1996, is reproducedin Table 54. This was characterizedas a general guide - 117 -

ANNEX 1 Page 12 of 16 for determining limits on the electric sector's investmentplans and borrowing programs and within which the electric sector should operate.

31. Based on historical relationshipsand estimated yearly increases in total public investment and the total outstandingamount of public debt, DNP suggested that the electric sector's share should average about 23% and 282, respectively,of these two e.ements. On this basis, and in terms of current US$, the electric sector's investmentprogram for 1989-1996would be limited to US$5,650 million and its share of external public debt would increase by US$1,177 million from US$4,339 million at the end of 1989 to US$5,516 million by the end of 1996.

32. There are significantdifferences in DNP's preliminary financial projections for 1989-1996and those of ISA for the same period. ISA's forecast for implementingthe least-cost expansionplan provides for total investment expenditures(exclusive of capitalizedinterest) of US$6,503 million or 15% more than DNP. ISA also foresees a much higher need for "other applications",principally for working capital requirements,than DNP; this requirementis estimatedby ISA to total US$896 million, or US$756 million more than DNP's estimate of US$140 million. ISA's forecast of gross internally generated funds, totally about US$11.2 billion, is US$2.9 billion, or 352, higher than DNP's US$8.3 billion. Total borrowings (foreign and local) are estimated by ISA at about US$6.4 billion and Government contributionsat US$470 million, ccnpared to DNP's US$6.9 billion and US$534 million, respectively. Estimated total interest requirements(including capitalized interest) are virtually the same at about US$4.8 billion. However, ISA's estimate of about US$6 billion for total reqvtirementsto amortize and repay debt are about US$800 million, or 16% higher than DNP's estimate of US$5.2 billion; this reflects the fact that DNP explicitlydid not allow for amortizationof some US$2 billion of borrowingsneeded from sources still to be determined. Under ISA's scenario,net debt (foreign and local) would increase by US$400 million, or US$50 million per year, over this 8-year period, while under DNP's net debt would increase by US$1.6 billion, or US$200 million per year.

Conclusions

33. As recognized in the DNP report of January 1989, it is imperative that every effort be made to improve the internal generation of funds by the electricity sector in order to lessen its demands on borrowings from external sources and on support from the Government. This requires vigorous implementationof efforts to reduce losses, improve collection of accounts receivable,pay amounts owed to contractorsand suppliersmore promptly, and improve management, operationsand finances of ICEL, CORELCA and their affiliated utilities.

34. New intermediateand long-term financialforecasts for the individualcompanies and the consolidatedsector should be prepared which would reflect the best current estimates of all pertinent factors. These - 118 -

ANNEX 1 Page 13 of 16 would serve as a basis for evaluatingthe sector's abiLity to implement the proposed least-costdevelopment program in a manner consistentwith the requirementsof other sectors and the country's ability to raise the funds required, internallyand externally,to support its public investment program and meet its other needs. If it becomes apparent that it is not feasible to implement the least-costdevelopment program because the required resourceswould not be available,a second-bestapproach, such as relying on less capital intensive gas-turbinesto meet part of the growth in demand, should be considered. Fechade leoreikOnf29/06/89 Archivo:CONPROTE.101

29-Jun-89

Table 4. 1: COMPARACIONAPDRYECIONES FINAKCIERAS ELABORADAS EN 18?7 9 ACIUALES

lN1lkones de USOI

DIFEREKIA31991-0Dt/1998) EtAD06DASEl I967 ACIUALES 1987 1838 1989 1990 I987 1988 1989 1990 IS97 1989 1989 1990

D0IDSNACROECOKOHICOS 26.0 24.0 14.0 -Iasa de Inf1icLdo 24.0 .4.9 4..) 4.0 24.0 241.4 2S99. S16.1 466.9 -lasa de C^AbaoPros. 240.0 291.6 346.1 410.9 335.9 422.5 515.9 -lasa de camaba)in de mao 265.8 318.9 318.5 449.3 26t.1

2S497.0 25574.0 27542.0 305 1la -501 -135 -Ventis (6.63 22212.0 23615.0 25067.0 26817.0 21907.0 0.039 0.040 0.040 -0.00111 O.0J5434 0.00206310.0039227 -luaita deola 0.038 0.0N 0.042 0.044 0.04)

ESTADODE RESULTADOS MC98. 1105.7 -12.7 17.5 S1.5 75.6 -Ventastonsusidor Final 854.6 939.1 1050.3 1181.4 961.4 922.2 44.9 36.6 46.2 20.3 27.1 41.9 44.7 -Otras Inqresos 65.D 72.0 30. 90,9 45.0 961.1 1l55.1 1151.9 7.6 44.5 75.4 12A.4 Totallogresos de Exploaticdo 9S0.0 l011.7 lI .O. 1272.3 Q12.1

291.2 301.7 321.0 40.3 31.7 14.3 8.9 -Gastes de Explotacnlo 297.1 302.9 316.1 129.9 256.8 (Sindepreckaci6n na otros no calal 302.9 300.7 309.2 -28.5 -27.7 -14.7 -7.2 -Depreciacion y otros lo Cali 252.9 2S5.? 286. 1*02.0 201.4 538.1 594.1 602.5 630.2 11.8 -16.0 -0.1 1.1 lotal6astos de Enplotacaon 550.0 5?8.I 602.1 631.9

373.0 452.6 521.8 -4.2 60.5 75.8 118.' Total Ingreso Keto de Egpi. 310.4 433.5 523.4 640.5 374.2

11.0 -55.9 -16.3 -12.4 17.8 340.5 60.2 21.1 IngresoslErenos ujenos a Explota 28.6 84.6 43.9 9.3

267.6 342.4 394.5 473.8 -15.9 -78.4 -162.9 -28S.3 Gistos financieros de Operacaon 251.7 264.0 231.6 18e.5

311.6 -25.2 41.8 35.5 29.5 279.4 298.9 425.7 LItthldad/Perdida leta 147.1 254.1 340.7 461.2

O > crD1

.. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- Table 4.2: EGlARACHMPOVEECeIOKES FINANCIERAS ELA8ORAIIAS ENI"?7 9 ACIUALES

IRiiloan de USSI

ELAICANDASEl It?8 ACIUALES DIFEAlEIIAIMS8M-1IC/91198 I"8 19e 1989 190 M1 M9e8 1989 M99O IS81 floe Im It"

(S8DOK FUENIESI DiSCS

FUENTES

Funtn latetnas

km. Inter,oula 651.7 793.4 858.4 951.8 666.6 620.1 131.1 818.5 -34.9 17;.3 121.3 131.2

AaortiatioLsi Deuga 486.1 496.1 541.1 611.7 442.3 517.1 598.6 592.2 43.1 -20.9 -S7.5 23.5 IntenrtseskOda 251.7 264.0 231.6 188.5 265.2 314.2 331.4 416.A -13.5 -50.2 -99.8 -227.7 PagsfoAdes 0.0 0.1 0.0 0.0 Interests Fodet 2.5 26.7 63.1 31.6 lotl Sericia Deud. 731.8 760.2 772.8 804.3 710.0 858.2 993.2 1)066.1 27.8 -91.0 -220.4 -261.8

fien. IriternaANt& -86.1 33.2 85.7 147.5 -43.4 -238.1 -256.1 -247.6 -42.7 27. 341.71 395.

FuentesEsternas

AottresPresepuusto iOil. 113.9 11.3 3.8 0.0 106.9 84.8 11.3 6.4 6.9 -11.6 -13.5 -6.4 o Cr8ditoEsterno 654.3 688.6 640.6 553.0 J78.4 405.3 947.6 1012.4 275.9 283.1 -307.1 -459.4 6ros fo4de 0.0 0.0 0.0 0.0 0.0 0.0o .0 0.nn CrtddtoInterne 15t.0 76. 9I 308.4 115.4 181.1 12.7 1.2 35.6 -105.0 84.4 101.2 Otros 19.2 14.2 14.4 15.4 39.8 36.9 23.3 16.9 -0.6 -2.7 -8.9 -1.4

lotat FuentesEaternus 9S8.3 790.2 '55.8 676.8 620.5 688.1 1000.9 1042.8 317.8 142.1 -245.1 -366.4

TotalFuentes 852.2 82R.4 841.5 824.3 517.1 450.0 744.9 795.3 275.1 313.4 96.7 29.0

Costos Construccadn

6WstosFnancieros de Const. 205.1 195.6 2;2.0 283.7 148.5 85.3 96.0 60.6 56.6 110.5 136.0 221.2 x0 InversiihU.E. 261.9 291.9 280.4 226.1 399.2 382.9 213.4 269.7 -137.3 108.9 7.1 -43.5 D InversionU.L. 224.0 243.7 25".5 219.0 333.8 349.9 28.0 290.! 90.2 -106.2 -28.5 -71.1 Iq Estudios 4.4 4.2 4.1 4.4 2.6 9.1 11.7 8.5 1.7 -4.9 -7.6 -4.3 :>C PlantaSeveral 33.2 10.0 12.0 33.1 5.3 9.3 18.0 1l.3 5.7 0.8 -5.0 -4.3 n I Total Costosde Construction 706.6 745.5 718.9 144.3 689.6 636.3 680.1 646.2 16.9 109.3 300.9 98.2 0 Otros Usos

Yariacth. Cap.Trabalo 76.0 20.9 37.2 80.9 -148.9 -29S3.0 42.9 1.4.? 224.9 .22.7 -5.8 -43.8 Otr0shaOs 11.6 48.a 23.4 0.0 35.4 101.4 21.9 24.3 -18.8 -58.6 t.6 -24.3 ;otalOlros Uises 93.6 71.7 S60.6 80. -t12.s -180.3 64.8 149.. 2Oo.! 24.1 -4.2 -68.2

Fotal Usos 900.2 82n.2 843.5 825.?2 57.1 450.0 744.9 7915.2 221.1 373.2 96.7 30.0

,"I*rflt tlov t'It S?.n ki ? t(.k .1 A.08 0.1 o A I' ?. rl?1) n1. -t! O SECIUEtECTUSO CLULaMIU

Pft* FINAICIEWOI91-190 IIIS Niloanl

Table 4.3: cwPcU PRaOECciKslES FI3I3CE3*S EltuI4S El I"31 ACTltES

ELIDAS Ell 197 ACTUMtES OIFEREKIAICUlY-ACTWUMESI 196 19SU 1989 1990 1917 198 IS8 M99O 3997 1985 1989 1990

ECESIIAKS

?la$ de Inversion" SO 350 549 463 541 551 584 586 -40 -I -SS -12S laternes de Construtckoa 205 196 2S2 282 149 85 96 61 5? 1II IS 221 Capitalde Irabao 7A 29 31 81 -149 -294 4S 125 225 323 -6 -44 iai.L 783 774 81S 825 543 343 123 M1 242 432 95 54

RANFIKACIERO

SnerarcionInterm ruta 634 745 835 952 630 333 715 794 4 232 120 158 boas:Intersns de Operacion 252 264 232 18S 265 314 331 416 -t3 -S0 -100 -229

A. FlUoNIbe 382 481 603 763 365 It8 384 378 17 212 220 ;G5 8. Aportesde particulares v otros 19 14 14 15 20 11 23 17 -1 -3 -9 -t C. Aportesdel obisrnoa ICELy 150 130 100 10 107 168 17 6 43 -S8 83 4 CORELCAy creditos (11 D. Iotal 3A*5tC) 552 625 718 789 492 383 424 401 60 242 293 387

E. Credites 716 646 641 6b 494 504 960 1020 222 142 -3S9 -619 -PrFnt. Wiletn lit. y nt.l 483 318 207 I50 494 266 484 465 -11 S1 -216 -315 -CredttoL.Kai Futuro(FEIl 98 16 97 308 0 0 13 7 98 76 84 101 -Oiras hstnts lotales S5 2 1 43 0 35 2 7 II -CreditoSectorfal (23 300 250 3S0 350 0 237 464 548 tO 33 -114 -190 F. 1nos Aaortizacio,ns 486 496 541 616 442 517 599 592 44 -21 -58 23 6. Credito Neto (E-F) 230 I50 100 35 53 -11 363 429 179 143 -261 -392 U. IknosfIterests FODEl -2 -27 -63 -58 2 21 63 58

1. SUiHiTAL(0464143 782 774 818 824 341 343 723 771 241 431 95 53

0_ _

(3) s Para 3988inclupe 335683.0 sbillon del credito 83RF-Sectlorual. 121:En lascitras actualtn, los denbaolsosde Credito sectorial i.cluyen F- USS400.0 sillones del credito FEM-Chaiengr eq US 150.0silaons delcredito b6bierno-Cballeoger tlS 200.0skilones dticredito FEN -Concorde tS 141.0millones del credits Eusis-FE US3211.0 cuilons del credito 11IF-Se torial US5141.0 .allones de otors creditos en snoat eilrnna queestan en procesv treatacion STATISTICAL APPENDIX (i)

TABLE OF CONTENTS OF STATISTICALAPPENDIX

Table 1: Major Sources of Earmarking in Colombia Table 2: Total Public Revenue and Incomes National, Departmental and Municipal Level, 1980-87 Table 3: Earmarkingof Revenue, National, Departmental,Municipal. 1970-87 Table 4: Potential ReallocationsUnder Law 55 (1985) Table 5: Reallocationof Earmarked Funds Under Law 55/85, 1986-89 Table 6: Share of Various Traditionaland Non-Traditional Earmarking Sources in Total Earmarking, 1980-85 Table 7: Share of Tax Allowance in National Government'sRevenue Table 8: JurisdictionalShares in the .ax Allowance, 1980-87 Table 9: Share of Education and Health in Divisional/Territorial Tax Allowance, 1980-87 Table 10: Distributionof Situado Fiscal, 1980-87 Table 11: The Distributionof Sales Tax Revenues, 1985-92 Table 12: Fondo Nacional del Cafe: Income and Expenditures, 1979-87 Table 13: Fondo Vial Nacional (F.V.N.)Total Expenditures Table 14: Fondo Vial Nacional (F.V.N.)Total Revenues Table 15: Fondo de Promocion de Exportaciones (PROEXPO):Total Expenditures Table 17: Legal Contributionsto Social Security by Program Source and Fund in Colombia: 1988 Table 18: Servicio Nacional de Aprendizaje (SENA): Total Expenditure,,1980-87 Table 19: Servicio Nacional de Aprendizaje (SENA: Total Revenues, 1980-87 Table 20: Instituto Colombiano de Bienestai Familiar (ICBF): Total Expenditures Table 21: InstitutoColombiano de Bienestar Familiar (ICBF): Total Revenues Table 22: Revenues - Expenditures,1980-87 Table 23: Public Health Sector Financing,1980-1988, Not Including ICBF Financing Table 24: Health Expenditure Targets Table 25: Expenditureson Health Services Table 26: Government Subsector 1980-1988,Rates of Change, By Substitution,Constant Values Table 27: Government Subsector 1978, 1985-1988,By Cost Category, Constant Values Table 28: Health Expenditure,Source of Financing Table 29: Public Health Sector Financing,1980-1988 Rates of Change, Constant Values Table 30: Government Subsector 1980-1988,By Institution Table 31: Amounts and Timing of Budget Appropriationsfor Investment Table 32: EducationalEnrollments, 1980-1988, Percentage DistributionPrimary. Secondary and Higher Education Table 33: EducationalEnrollments and EnrollmentRates, 1980-1988, Primary, Secondary and Higher Education Table 34: EducationalEnrollments and EnrollmentRates, 1980-1988, Rates of Change Primary, Secondary and Higher Education (ii)

Table 35: Government Expenditureson Education, 1980-1988, Constant Values Table 36: Education ExpenditureTargets Table 37: Education ExpenditureTargets, Percentage Distribution and Rates of Change Table 38: Primary Education Physical and Financial Investment Targets Table 39: Public EdLcation Budget, By Cost Category, 1981-1988 Table 40: Bank-AssistedPrimary Education Investment Project Table 41: Power Sector - Statistics Sector 1987 Table 42: MacroeconomicVariables Growth Rates (Z) in Constant Values Table 43: Tariff Growth Rates Table 44: Growth Rates of Prices for Electricity Substitutes Table 45: DemographicVariables Table 46: Electric Service Co-erage Targets Table 47: Scenarios for ResidentialConsumers Growth Table 48: Growth Rate Scenarios for Commercial Consumers Table 49: Recovery of Losses, Own Consumption,and Sector Losses (X) Table S0: Reference Demand Forecast Table 51: Demand Forecast SensitivityStudies Table 52: Investment Program 1989-1996 Table 53: 1989-1990 InvestmentProgram by Category of Investment and Power Company Table 54: Preliminary Financial Projections Table 55: Load and Sales Forecast 1987-1996 - 122 -

Table 1 MaXor Sources of Earmarking in Colombia

Atme L.941 ft"strussAts Ton-vew $_* s, Tan N*oterF

A. boo.e. She'ioe

1. Tot Allswsce Law 46/1971 (itel_ti.9 0w%;Stinl C.. et Re.ewe 131 (1973) 0f treitf.,. 701 an is. of (S; tve FtceI) Artiele 162 of 1968 virnistIr"i, ei b 141 (194A) 069It.6e sod 301 $-*i'd csotit tio) 11s (thereafter) .8,811, iO_g doi'U,ts.

District ofSgoea Of seq.vt reseOvd by oath ., ad ct *- 741 foe t. goo- Ed.cat *- (Foe) *r4 261 ftr Set% *or

2. Sle TomTrantfer Low13/1966 Sales %as 'eveos. 101 (I9n9) M..oi'Pel set'''|s, *eceo- Cueo w las o0%4i) 20 ('$9O) rd tie sct_d.te, on The Ws (1171) pe_ on *ds. .vtea.t 9 v.*5

by Los t2/1973.Lit 41/19"S s-d

Law 12/1916 Viiw. s66. tam .eto,w"_ 20.(1966) Fee O.o.cptl ict. viti. *ot *. 12.0 (1967) .i;c..el t.w*. A.-% to -o 34.6(1901) "St..*I tep,to'esw S4eco^* -a O* 37.t (1969) *tcutpty *.0ds, co4 tow ,-t t.to 41.0 (1980) G.egetfieC A9.tt. Codejt (::;AC i 46.0 (1911) Esc*ls Sop.-. de Adif it-scme 10.0 (1992) publ[C (ESAP) C. Nit;ool Level

S 'SWO (Foo de Dt. 444/1967 (Artzcle 229) CIF sot tt uorchodite tof 1.61 i04*crOtf1 promteCe I ca*c PUetoO de i;seepte of *MPe'tt. Eapi.uteisoe) Ot. 2366/1974 (Article 6) CIF .lu,* of aeelhiodi.e 3.79 in"boti (61 Pone 191)

Lao 76/19" CIF -ivel of _ptelhodie 61

4. iq;9o,y F (Foods ue of (1967) at'fi.ey ptice Poe 96110. 1141 sow ".6 NigAwy ;"W ettasot *o..9.- tiA 1CS Viet Widsiel) sd Decree of easel; ends ACPU .imcttvWly s.og to N ineov F.d fo. hog.rp o.hod 2962 (1964) It544 (FPais Wegi.sil de C4s,v*S Vetiseimo so PWCV); 101 to, ootsotl1 roiliyws (Feoriocariloe MNt;o^.SS Li So (1"2) RefliM.y Pries per falls" C01.13.SO de Coela;*); sod 61 fee .boa of gasl le amd AC30 per gel l* to-s tsoo;t s i (Coepeossis Finsaniers de Tooteste" or CFT)

S. SW (ie11l,leleael Ot_s 118/$iN7 payrsllo f priuat1 o 4t lfy Pgtast" il of teetoI sitSills toi;A. .2 prpo laejs) (Artl.lo 8) toteo)pos"

Low 63/1141 (Arsl.lzs 2 Payrolls of pubile sad i) iodjLotoig *od *_erisl 21 s**swlIJ .oteop.e;sees od *.Sod

Gaidipe i No

pr#Ost. loteropises 11 ao*eI

Prolla of Centeol. 0.61 5504517y depattemot, taoeiteoi s. Oi4iP;si i%sQ sod Sogota Da qowep~t.. - 123 -

'Co' r-4) Table1: Maior sources of Earonarking in Co-!nbia

lacuerha od/ Apogaa LOPI Iiag96.etea lose To* Oatl aDo

S. Ic (teatitito oI Gile Lee ?$/lail (Artlele ) Sle *pigs *of sot 121 ,,.. of *'t4e*eA of de 6l.Wosse pop;Il-p) Ch,Idr.. a*re.* ftA

La. 7/II'S (Acisle a,) payretls of PwAlirt *ft%t*Piws 21 naaalaly Pepr.llg of Peitat ateepeise 61 *Sie ISy

Las01/1905 Pe)'1lieof ppI6to Sad pa,blta; eaterpe" as SARVI I

t. C_y&eptieo rm (CSJ* de ceoaoie eamiliar * y,e*lliof Private omtep.riseo 45 Paafl17 allawooteecoo sac A.

; Aceo"fpet~t tt .ta. . Dr..- t

F.PAM (Foods AerbtstiCO La St%/" Sa *eIotleno trtoraeioeAl pesesei,ee 13SS Cpeetrb, O b- 'tat o. atioaal) 1012/1112 * _ o oor *ad EAlsegenot *4 A

t1"W1.06 1310/1M tAterntionel peosteXr tGl 8s°° per

. IF! (Zastit" de fa_ta Lo 6w m $sadeCr_ C.t.F gave Of .Meerts 0.81 (401 of Cas,talitatlea Of 'pI :'*1 I,dutGiel) 25/1606 2?) coh.0'e eo of .a.eotAants 04. dots ae. Ce)

10. Cjej Agreia Law "/1M Bad rreo C.Z.F. Gl.* of ;seets 0.6n (401 ef Caeviul.zot.a of te. 0nr 623/16 2X) *wSordi5. to th I,., 13/16'

and 604/112 end #or the ll.. 4 fe. W,'al dwellings (Leo 20/:19W

1 Ceppertieol Naional Lao 20/(179 (Crtlolo 24) tariff on hote l dis 5x Planiang, prwitiaa and fvaOAC ag do 1Trim of the tooriat iadaatec

12Fed lGe leal de, Cero L'a 61/175 ( I ve *) te*dtto of 2. ( o f Finanolo *ragrin of ie,alo'evto. ef s a tI e a, ati SI) e.p tatieto, , aranaand er.ializat;io of coal.

13. Poc4 (COaaI6 do pez La 6U/I26 (Atolole 16) at Volve of *lA SStiellets 1 _aogueaotand deelIe~t of Cloamoe"f lao) af the OiM"foah. ;adas,t' in Celoabie (6.62 foe tho Frd to fteoveege f IO waliAl sad 6 foe*., , e e,.,, datcoten ca eibiti;oa of aetefl e*o,.). 14. Caja bolomoede Le 4do/IS (hotele 1) Valvoof bopcells for th Col.$0.la W Fer aw iel eeewigae Pe'alea aolal govOsPtos(Aaaioal, O"eh6100 d__rwso"l. woal.lpel, ar fre.le toccitWeela an4 aiol thereof dletelet), a.l io off 1a, Om daes"traligod lostita.tiaaa - 124 -

Table 1: Malor Sources of Earmarking in Colombia

£a#fteeP.4Pwg a - < Legal lostruawss r.a/wo.e..e, ftee Too n*o p.'ea

Is. Ieelttee_mesi d I.e 60/1IN (Article I) pu 1blic de4aeetade d' W ledeswont ;a As of im, t toi be ."w *s noiemst* De"1 3140/1C and ietpie, oe.at.t. i. 8*4 i 65s/1s (AMtme- 1a) e.dei of j.dc.a eCovte sod je- le.

1i. per m.u.ea I d Cafe Volvo of geftm a_ fla do of *apors SabiIizetieft of *P0CeP .8me*. Volve. iu,eme_,e a eaainq .84 see.l

*pejetA ;* Coffee gpoo_ 8.041. 11160tlee Ge f;nwwisl few .e,tt *f.equepato. Sart W*ve y*9io08 rvoo Set'..ti;0 :e.g. qil ity I.8tteI. oeme ePe_t me. OI so ;forl.o* I~

C. 049emea,tal

it. e * case.. De,a_ IUI/X, c"W.. i * et .me ;" ,,I Col. o. Ow Ceetn."'im. "'mtae-'e. **4 volnel,s (Ara.gIe I) ¢ 0 s.ms Co".aliliP bottl.e rasie of local *oe.

so. snis t _tA ew i .. e,. Ii , *;^ O, - isfU jan * itme Litme, - SU1 Ulee * Jini

to. Smatimi "mof" gee.,em a.O. variosem g leg goofLea NeeltS ooo4i0,mm S" am fei1385 A4iW Oif laso em" (bo ;41404")

20. l%ioma# S 8.8. CIge Pd welepoe Aaemlutes toe me. Of C_iee ae (Jitof " ceC-l.3e-1s ad Velp_e ,e . w £etieIn,vefm de O" mee (dam-s saeme Celamiee leatesse %ei S" Cel .,0.20 of fe,_* me SeeeS pm past (imter) (C.BU?TU) aseeN is fiee, so" teefe son mA Colomba*lee tiut ateo Calfr (COcUTUR)I

n. s_jiste S" wlew N..t L18_ __1itsm nole 410*ft fart bl;ie wI_ 4$Iead 1

S. 50Ielmm

U2. YeOe.18Iast, we fee LeMto/lIf leSfemd Volvo of p,.peeti Ano equa to toewmtot as,eelew "Wiewo amiepal goes et Iag=6 omelipsi lpeoeqam-e atleilee sm NW _me.- (O.g. _ef. e_te. 1e54 _ _m son"). (Continued) Table 2: total Public Revenue and Income National,Departmental and Wunicipel Level, 1960-87 (Millions of Colombian Current Pesos

Total Public Revenue National. Departmental. Municipal Levels 1970 1975 1980 1981 1982 1983 1984 1905 1986 1967

111. Capital Cities Municipal Income 81

A. Central Administroaton Municipal Level 1,550 3.238 11.589 15,506 19.828 25.096 34.893 44.402 1. tax Revenues 1.179 2,458 42.623 n.e. 8.750 11.644 15.211 20.100 29.063 37.062 a. Industry and Coomerce 193 624 42.623 n.e. 2.968 4.105 5,551 7.026 12,533 16.618 b. Lend Tax 369 860 2.56? 18.281 n.s. c. 3.426 4.502 5.772 7.310 10,011 12.399 Valorization tax 91 261 461 2.067 2.583 n1. d. 3,291 5.309 5'937 141 6,083 141 6.967 141 n.e. Other 356 513 1,148 1.530 1.867 1,993 3.283 4,~30 4,976 n.*. 2. Mon-Tax Revenues 371 80 2.839 3.862 a. 4,617 4,99c 5,832 7,340 n.. n.u Total except Bogota 166 264 1.276 1.717 b. 2.095 1,807 2.348 3.254 ne. n.. Bogota 101 205 516 1,563 2,145 2,522 3.169 3.484 4.086 n.e. n.s. S. Social Security n.a. n.a. 679 988 1,289 2,098 2.618 3.716 n.o. n.e. C. Operating Surplus Departmental Enterprises II1 n e 2,220 8.370 14,374 20,012 24.5l9 32,064 56.517 121 77,'00 131 D. Total Current Revenes NunilIpal na. n.e. 20.638 30.868 41.129 51,733 69.577 104.635 n.e. n.e. IV. Total Government Revenues All Levels n.e. n.a. 294.286 400,178 529,715 616.825 734,372 1.016.558 n.e. n

Notest 1. Gross operating income of national decentrslized agencles such as ICEL (electricity). security aegnclee. TELECOM(telecomzications). and ICJ (housing) and the social 2. ISS. CAJANAL. Fondo Mlctonal del Ahorro (NationalSavings Fund). CAPRLCOH.Hilitary 3. Excludes transfers from central Forces Retirement Fund, MationalPollce Retire ent Fund government and Interdepartmental transfers. Includes oil royalties. 4. Includes transfere from the distillaries. 5. Social welfare agencies,lotterles, distilleries net operating income. 6. Soe departmental Infomntion missing. 7. Some transfersto other departmental agencies Included. S. Transfers from higher levels of government excluded. 9. Taxes collected directly by the agencies. 10. Beer end tobacco taxes in Bogota. 11. Public utilities and Bogota lottery. 12. Bogota.* edelline, and Cali. 13. Bogota and Medellin. 14. Estimate. Table 3: Earmarkin of Revanue. National Departental. Municipal. 1970-87 (Millionsof Colombian Current Posos)

1970 1976 1980 1981 1982 1983 1984 1985 1988 1987

1. Earmarked Funds, National Level

A. Tax on C.I.f. Value of Imports 72 768 a,881 7,904 11,595 10,867 17,395 22,826 35,337 54,847 1. 2X Tax for: 0 0 0 0 0 0 4,845 8,086 8,299 11,847 a. Caja Agraria (40%) 0 0 0 0 0 0 2,340 3,018 4,147 6,872 b. Industrial (IFI) (40%) 0 0 0 ° 0 0 2,306 3,047 4,162 s,976 for Proexpo c. 2% 72 758 8,681 7,904 11,696 10,057 12,749 18,780 27,038 43,200

8. Payroll taxes 148 147 29,799 41,857 66,213 77,181 99,215 123,435 188,824 210,804 1. SENA 0 0 3 ei8 6,193 7,938 90ose 12,187 14,423 18,258 I.C.B.F. 23,196 2. 140 132 3,584 4,872 8,484 8,631 10.788 13,887 17,831 23,534 3. Cajanal a 15 122 98 138 103 158 143 1.079 964 4. Retirement and Severance psymeqts na. n.e. 22,277 31,894 41,875 59,439 76,102 95,002 131,868 182,921 C. CasolineTax 771 1,837 18,973 22,234 24,808 24,081 27,499 39,840 48,493 59,600 1. Highways Fund (762) 679 1,228 12,730 18,878 18,eo8 18,081 21,364 28,878 32,792 43,765 2. Small Roads Fund (10%) 77 184 1,697 2,223 2,481 2,408 2,468 6,288 b,480 8,334 National 3. Railways (102) 77 184 1,897 2,223 2,481 2,408 2,468 5,286 5,480 8,334 4. Transport Fund (6X) 39 82 849 1,112 1,240 1,204 1,229 2,832 2,741 3,187 D. Tax on International Travel 40 n.e. 897 1,631 2,886 2,628 3,117 3,808 3,426 4,428

E. 4X Coffee exportstax 1,034 2,389 3,808 3,387 3,736 4,062 5,638 8,758 17,886 1. Coffee Fund 14,006 (80X) n.a. n.e. 801 684 822 876 939 1,469 2,981 2,334 2. Coffee Private Federation, Rural Zones (20%) na. n.s. 3,006 2,823 3,113 3,377 4,899 7,299 14,906 11,,%71 F. SX Tourism Tax 0 132 346 612 738 023 241 213 327 542 C. Movie Tax 0 0 0 0 0 0 0 0 221 606 H. Participations and Contributions 151 406 1,225 1,018 2,186 3,038 3,843 8,101 9,084 11,063 1. Banks Rogulatory Commission 81 184 650 129 974 1,242 1,584 2,481 3,471 2,940 2. Privat, firms Regulatory Commission 36 84 290 394 662 743 983 1,410 1,672 2,736 3. Contribution Contraloria Ceneral 66 161 330 483 839 776 1,089 1,890 2,312 3,098 4. Family Funds Regulatory Commission 0 0 0 0 0 0 217 172 190 348 6. Industry and CommerceRegulatory Commission 0 0 80 0 0 98 140 6. Revenue Sharing Aeronautics Fund 0 0 49 17 9 275 10 106 108 61 7. Superintendencisde Notariado y Registro 0 0 0 0 0 0 0 2,284 1,316 1,760 (Continued) Table 3: Ermarking of Revenue, Nstional D*partmental.Municip*l, 1970-87 (Mi o of Colombion Current Pesos)

1970 1976 1980 1981 1982 1983 1984 1986 1988 1987

1. Special Funds 6 90 6d 90 74 673 1,127 3,208 4,968 6,212 1. National Defense Fund 6 72 18 0 1 438 878 933 1,166 1,327 2. ConmunicationaFund 0 0 1 49 26 89 198 227 38? 404 3. Educational Services Fund 0 0 0 0 0 0 0 0 0 0 4. Tax Education Fund 0 18 48 41 49 46 66 88 71 86 6. Coal Fund 0 0 0 0 0 0 0 832 1,986 2,996 8. Agricultural Promotion Fund 0 0 0 0 0 0 0 360 274 87 7. Rice Fund 0 0 0 0 0 0 0 171 261 240 8. Cocoa Fund 0 0 1 0 0 0 0 300 418 667 9. Cereal Fund 0 0 0 0 0 0 0 187 211 364 10. Stockmarkets Regulstory Commission 0 0 0 0 0 0 0 60 36 23 11. Fondo Rotatorio de Advance 0 0 0 0 0 0 0 70 160 140

J. Transfert and Revenue Sharing 349 8,608 29,206 39,744 66,871 65,794 87,941 110,786 131,428 190,294 1. Tax Allowance (Education, Health) 0 4,303 19,438 28,912 38,886 47,219 68,628 69,600 76,838 109,862 2. Sales Tax 349 2,306 9,788 10,832 17,986 18,676 29,313 41,186 66,590 80,432

K. Reallocation Law 66/86 (Several agencies) 0 0 0 0 0 0 0 0 3,164 1,167

L. Other 62 630 1,441 2,100 2,868 3,271 3,639 4,266 6,419 6,906 1. Electricity Regional Company 12 66 378 643 649 1,168 1,426 1,826 2,409 1,617 2. Sports Promotion Agency 0 0 633 748 1,038 1,022 696 820 482 803 3. Artx Promotion Institute 0 0 13 8 18 19 26 34 60 81 4. Regional Corporation 49 103 323 444 681 680 933 1,260 1,830 2,184 S. Valorization Fund 0 7 32 34 40 17 0 40 0 201 8. Languages National Institute 1 2 2 0 4 0 0 3 1 10 7. Agricultural Technology Promotion Agency 0 383 181 326 348 476 660 683 847 1,009

M. Operating Surplus Non-Financial DecentralizedAgencies 2,969 8,339 7,976 8,617 33,690 29,299 70,684 82,383 81,482 170,791

Total Earmarked Funds [Al National Level 6,689 19,013 98,194 128,893 195,233 221,373 320,239 407,612 608,028 729,963 (Cont.inued) Table 3: Earmarkingof Revonuo: National.Departental. Municipal. 1970-87 (Million.of ColoVbianCurrent Pesos)

1987 1970 1976 1980 1981 1982 1983 1984 1986 1986

II. Earmarked Funds D-partmental Level 2,633 6,434 19,133 24,390 28,878 37,681 47,916 81,016 78,461 n.n. A. Taxes 3,468 2,960 6,2tB n.o. n.s. 8. SocialSoecurity n.s. n.x. 1,476 2,233 2,721 C. OperatingSurplus NonFinancial DopartmentslDecentralized 7,492 9,469 11,067 10,224 n.o. Agenci-s na. 2,600 4,043 4,629 6,760

TotalEarmarked Funds 48,621 60,344 78,301 n.n. n.n. DepartmentalLovel tB) n.o. n.a. 24,651 31,162 37,349

III. EarmorkedFunds Municipal Level

A. CentralAdministration 6,967 n.o. OD 468 977 3,830 4,728 6,813 8,498 9,421 10,169 MunicipalLevel 6,309 6,937 6,083 6,967 n. 1. Volorizotion 261 461 2,067 2,683 3,291 2. Non-TaxRevenue 3,484 4,085 n.o. n.s. (BogotaMainly) 206 616 1,663 2,146 2,622 3,189 2,098 2,618 3,634 n.n. n.o. B. SocialSecurity n.s. n.n. 679 988 1,299 C. Net OperatingSurplus 24,639 32,064 68,617 77,700 n.o. MunicipalEnterprises n.s. 2,220 e,3?D 14,337 20,012

Total EarmarkedFunds Municipal 36,136 44,103 70,320 n.o. n.n. Levol (C] n.o. n.s. 12,679 20,063 27,114

Total Non-DiscretionarySources 306,029 424,686 666,233 n.o. n.o. of Revenues [AI.[B]+.CJ n.n. n.e. 136,624 180,098 269,698

Reporton EarmarkedRevenues; various agency reports. Sources: ContralorisCenoral Informes Financieros; Manrique and Associates, - 129 -

Table 4; Po:r.al r'aoca: : .a .' :

Or;an.zation or lec:p.ent: o: _r AaenCY Earn-arkedSource Reallocation

1. National Highvay Fund Tax on gasoline Construction; maintenance and National Fund for and valorization and repair of national Neighborhood Roads tax on national highvays and local roads: public works and improvemoets in river nav:ga::_

2. National Aeronautical Stamp tax on sales Airport activities and those Funds aboard and airport tax. activities listed in 1 above.

1. PROEXPO 6S tax on CI? value Funding of CERTS (tax refund of imports certificates and other instruments to promote esports) .

. SI Tazxon cif value Small-scale industry. tLmber of imports industry, and fisherios.

5. National Tourism Tourism tax Preservation and restoration Corporation of national monuments and parks: promotion of cottage industry; and the financing of tourism developmenz certificates.

6. National Coal Fund Coal production tax Finance of m.ning invoetI- gation and exploration activities.

7. Notary and Registration Duties on the reSistration FLnance of investments for Superintendency of public instruments construction: improvement and issuance of notary and equipping of judicial certificates offices and prison fac&- lities. a. Motion Picture Promotion A nev 161 tax on einema *.51 of the tax mount to ticket pricee be paid into the Motion Picture Promotion Fund to be managed bv :hr Motzon Picture Development Company and used to fiaunce cultural and artistic programs. The remaining 7.51 shall be paid to producers, distributors, and exhibitors of Colombian films in proportiono esta- blished by government with greater azounts going for full longth features.

9. SENA (National Appentice- Payroll tax Training geared to jobs. ship Service) technical and small industry and agriculture extension. SENA to assume responsibility for financing. in full or in par industrial schools and technical institutes. colleges. egricultur training schools, and computer a telemetries programs.

10. ICSF (Colombia Tamily Payroll Taxes Child and family protection Welfare Institute) programs: and prevention health progrms; food assistance programs for India comunities; public homes and also the maintenance of children's hospitals and construction of wvter supply infrastructure in comaunity with lose the 60.000 inhabitants (based on 1973 census).

Source: LEY 55 Par medio de la cual so dictan normas tondientes sl ordenamiento de las Finanrss delEstado v diCtaw p . A ---- -e - 130 -

Table 5: Reallocationof EarmarkedFunds Under Law 55/065 198-89 (millionsof Colombian current peoo*)

1986 1967 1988 1989 Donor Budget Actual Budget Actual Budget Budget

PROEXPO 6,861.0 6,408.0 10,603.0 20,744.0 31,120.0 42,000.0

IC8F 3,100.0 2,636.0 3,300.0 1,145.0 0.0 19 ,3 16.08

SENA 2,286.0 2,449.0 2,980.0 2,984.0 0.0 19 ,5 0 0 .0a

FAN 990.0 600.0 1,000.0 332.0 696.0 860.0 SupernotariadoY Rogistro 910.0 677.0 1,610.0 1,315.0 1,760.0 1,750.0 Fondo RotatorioDo Aduanas 0.0 56.0 369.0 70.0 360.0 670.0 Corturismo 278.0 242.0 643.0 643.0 404.0 682.8 Focino 35.0 146.0 32.0 164.0 185.1 Fondo Nacional Del Carbon 6E.0 139.0 776.0 297.0 649.0 926.2 Ga tolinnAcpUb 9,883.0 9,937.0 17,666.0 17,680.0 25,492.0 40,324.0

Reallocationsamong various agencies (LEY 65/896) 3,463.0 3,164.0 1,246.0 1,167.0 1,66S.0 2,969.0

Ecopetrol (Decroe 399/86) 6,120.0 5,120.0 1,600.0 6,600.0 6,500.0 InstitutoDo Fomsnto Industrial 970.0

TOTAL 33,027.0 30,362.0 46,517.0 52,099.0 67,680.0 129,942.1

a. ICBF will reallocatethis amount within its budget to a program that is a governmentpriority (Children'sWelfare Homes). SENA will reollocatethis within its budgot to special government programs. b. Estimates based on Law 66/86 percentagoe. c. Soveral agencies included. Source: ContraloriaConeral do Is Republics for actual figures and DepartmentoNacional do PlaTneacionfor budget figures. - 131 -

Table 6: Share of Various Traditional and Non-traditonal Earmarking Sources in Total Earmarking. 1980-85 l/

1980 1981 1982 1983 198

National 72.5 .1.-6 5.2z 2 5 Traditional Earmarking 44.6 43.7 40.6 37.6 35 _ Tax on Imports C.I.F. 4.9 4.4 4.5 3.5 4. Gasoline Tax 12.5 12.3 9.6 7.9 6 5 Tax on International Travel 0.7 0.9 1.0 0.9 0 7 Coffee Export Tax 2.7 1.9 1.4 1.3 1 3 Tax on Tourism 0.3 0.3 0.3 0.3 0.1 2 v Tax on .{ovie Tickets 0.0 0.0 0.0 0.0 0.0 ; ' Contributions and Participations 0.9 0.6 0.8 1.0 0 Special Funds 0.0o.0 0.0 0.2 C 3 Revenue Sharing 21.6 22.1 21.9 21.6 2C Tax Allowance 14.3 16.1 15.0 15.5 13.8 .2.5 Sales Tax 7.3 6.0 6.9 6 ! Other 1.1 1.2 1.0 1 0 9

Son-Traditional Earmarking 27.9 27.9 ILA 34 9 0 - Pavroll Taxes 22.0 23.2 21.6 25.3 23.4 Social Security 16 4 1' 6 16 0 19 5 i1 9: SENA, ICBF, etc. 5.6 5.6 5.6 5.8 5.5 5 Operating Profits of Non-Financial Public Enterprises 5.9 4.7 13.0 9.6 16 6

Departments 18. 17.3 14.4 15.9 2 Traditional Earmarking 14.1 13.5 11.1 12 3 11.3 : Son-Traditional Earmarking Social Security 1.1 1.2 1.0 1.1 0.. ; : Operating Profits of Non-Financial Public Enterprises 3.0 2.6 2.3 2.5 2.2 '

Municipal 4 11.1 10.4 11.5 10 a 1 Traditional Earmarking 2.7 2.6 2.2 2.8 2.2 1 5 Non-TraditionalEarmarking Social Security 0.5 0.5 0.5 0.7 0.6 0 6 Operating Profits of Non-Financial Public Enterprises 6.2 8.0 7.7 8.0 7.6 '0 2

Total 100.O 10 100.010Q0. 100.0 1Cc 0 Traditional Earmarking (All Levels) 61.4 59.8 53.9 52.7 49.0 49 1 Non-TraditionalEarmarking (All Levels) 38.6 40.2 46.1 47.3 51.0 50 9 a. See footnote to Table 8 for definitions. Table 7: Share of Tax Allowance in National Government's Revenues (millions of pesos; percentages)

1975 1980 1981 1982 1983 1984 1985 1986 198'

Current Revenues 41,819 168,247 221,792 279,955 296,960 330,884 464,676 690.435 964.176

Less: Earmarked Revenuesa 5,325 35,935 44,120 60,051 60,371 88,174 127,761 173,302 231,634

Equals: Ordinary Revenues 36,494 132,312 177,672 219.904 236,589 242,710 336,738 517,133 732,542

Tax Allowance 4,303 19,438 28,912 38,886 47,219 58,628 69,600 75,838 109,862

Tax Allowance + Ordinary Revenues X100 11.8 14.7 16.3 17.7 20.0 24.2 20.7 14.7 15.0

a. Included In earmarked revenues are the taxes on imports C.I.F., gasoline, international travel, tourism, cinema tickets, and after 1984 the ad valorem tax on coffee exports; also Included are contributions and participations, the special funds, and the sales tax (or IVA) transfer.

Source: Information supplied by Contraloria General de la Republica. - 133 -

Table 8: :urisdict:onal Sh.ares in the "ax Allowance. 1980-87 (percentages)

Formula 1964 Share Division/ Population in Tax Territory Share Allovance' 1980 1981 1982 1983 1984 1985 190- 1987

Antioquia 14.2 10.85 11.61 10.92 11.33 11.08 11.08 11.20 11.06 11.10 Atlantico 4.1 3.78 3.13 3.70 3.45 3.74 3.76 3.41 3.45 3.36 Bogota 9.7 7.70 8.51 10.32 9.47 8.36 9.37 9.04 8.83 9.32 Bolivar 4.0 3.71 3.26 3.47 3.30 3.59 3.31 3.25 3.41 3.44 Boyaca 6.1 5.18 4.77 4.44 4.96 4.84 4.99 5.17 5.11 5.28 Caldas 4.1 3.78 3.66 3.77 3.89 3.76 3.60 3.56 3.77 3.41 Cauca 3.5 3.36 3.16 2.99 2.98 3.08 2.87 2.73 2.85 2.88 Cesar 1.5 1.96 1.89 1.78 1.95 2.17 1.99 1.92 1.96 1.85 Choco 1.0 1.61 1.67 1.73 1.72 1.90 1.92 .1.82 1.79 1.88 Cordoba 3.4 3.29 3.58 3.65 3.62 3.68 3.18 3.53 3.61 3.43 Cundinamarcs 6.4 5.39 7.12 6.67 6.67 6.37 6.57 6.84 6.85 6.91 Guajira 0.8 1.47 1.23 2.21 1.42 1.44 1.34 1.34 2.51 2.58 Huila 2.4 2.59 2.85 1.55 2.76 2.73 2.94 2.76 1.48 1.44 Kagdalena 3.0 3.01 2.82 2.97 3.05 2.98 2.75 2.97 2.86 2.98 Heta 1.0 1.61 1.48 1.49 1.34 1.64 1.61 1.46 1.61 1.56 Narino 4.0 3.71 3.91 3.39 3.68 3.77 3.95 3.70 3.61 3.70 Norte de Santander 3.1 3.08 3.42 3.14 3.53 3.04 3.36 3.74 3.59 3.55

Quindlo 1.8 2.17 2.15 1.91 1.89 2.04 1.83 1.75 1.69 1.74 Risaralda 2.5 2.66 2.45 2.40 2.38 2.70 2.46 2.50 2.46 2.54 Santander 5.8 4.97 5.19 6.06 5.47 4.96 5.22 5.98 5.84 6.04 Sucre 1.8 2.17 1.97 2.02 1.94 2.27 2.00 2.01 1.97 1.94 Tolina 4.9 4.34 4.59 4.18 4.44 4.34 4.88 4.09 4.29 4.18 Valle del Cauca 10.0 7.91 7.31 6.97 6.71 7.62 6.79 6.95 6.78 6.82 Arauca 0.1 0.98 0.79 0.79 0.79 0.82 0.82 0.81 0.91 0.81 Caqueta 0.6 1.33 1.15 1.16 1.06 1.36 1.07 1.33 1.38 1.43 Casanare n.a. 0.912 0.93 0.92 0.92 0.80 0.98 1.12 1.29 1.17 Putumayo 0.3 1.12 0.95 0.98 0.88 1.06 0.95 0.99 1.16 0.98 San Andres 0.1 0.98 0.77 0.76 0.75 0.73 0.75 0.71 0.82 0.69 Amazonas 0.1 0.98 0.74 0.74 0.74 0.73 0.74 0.65 0.57 0.53 Guainia 0.0 o.9lb 0.73 0.73 0.71 0.69 0.71 0.63 0.66 0.61 Guaviare n.a. O.9 1 b 0.70 0.72 0.71 0.69 0.72 0.63 0.64 0.76 Vaupes 0.1 0.98 0.75 0.74 0.73 0.71 0.75 0.68 0.49 0.47 Vichada 0.1 0.98 0.75 0.74 0.73 0.62 0.72 0.69 0.71 0.63

TOTAL 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

al According to the sharing forrula, the ith jurisdictions share in *situado fiscal' is equal to .3/33 + .7 Pi/Pop where Pi/Pop is a constant since it is based on the 1964 census.

b/ Jurisdictions vith virtually no population in 1964 or which did not exist in 1964 are entered as sharing in the non-population component--i.e. 1133 of 30Z.

Source: Data supplied by Contraloria General de la Republica and mission estimates. Table 9: Share of Education andiHealth ln Dlv Tat Allowance,wionaljTerrltoral 1980-87 (percentages)

Divislon/ 1980 1981 1982 1983 1984 1985 1986 1987 Territory E H E H E H E H E H E a E H E N

Antioqula 80.96 19.04 78.30 21.70 77.59 22.41 78.49 11.51 78.27 21.13 75.80 24.20 71.85 28.15 78.34 21.66 Atlantico 75.41 24.59 73.36 26.64 77.03 22.97 75.17 24.83 75.48 24.52 70.01 29.99 68.13 31.87 72.55 27.45 Bogota 81.50 18.50 78.94 21.06 81.91 18.09 77.81 22.19 78.47 21.53 71.88 28.12 72.53 27.47 75.76 24.24 Bolivar 76.97 23.03 72.13 27.87 72.45 27.55 74.01 25.99 72.22 27.78 68.37 31.63 67.23 32.77 69.39 30.61 Boyacs 79.20 20.80 77.18 22.82 80.03 19.97 78.60 21.40 79.17 20.83 76.74 23.26 73.93 26.07 77.21 22.79 Caldas 79.07 20.93 78.97 21.03 78.78 21.22 78.55 21.45 78.78 21.22 76.38 23.62 71.80 28.20 75.49 24.51 Cauca 78.50 21.50 75.20 24.80 73.64 26.36 77.62 . 22.38 70.72 29.28 67.88 32.12 63.75 26.25 69.08 30.92 Cesar 78.85 21.15 73.93 26.07 74.64 25.36 71.73 28.17 76.42 23.58 69.36 30.64 69.52 30.48 12.36 27.64 Choco 79.94 20.06 76.25 23.75 69.61 30.39 77.49 27.51 77.39 22.61 71.74 28.26 72.27 27.73 78.50 21.50 Cordoba 81.47 18.53 73.53 26.47 78.48 21.52 77.14 22.86 78.41 21.59 73.95 26.05 72.26 27.74 74.00 26.00 Cundlnamarca 84.54 15.46 81.73 18.27 81.88 18.12 87.83 12.17 81.45 18.55 79.82 20.18 77.32 22.68 81.45 18.55 Guaj3ra 75.42 24.58 84.84 15.16 66.24 33.76 66.86 33.14 65.64 34.36 60.43 39.57 82.31 17.69 85.54 14.46 Hulla 81.59 18.14 63.25 36.75 79.20 20.80 72.77 27.33 78.16 21.84 76.64 23.36 53.93 46.07 56.50 43.47 Magdalena 78.14 21.86 66.08 33.92 67.99 32.01 73.31 26.69 69.98 30.02 70.22 29.78 70.18 29.82 70.94 29.06 Heta 78.40 21.60 74.25 25.75 73.51 26.49 73.84 26.16 71.88 28.12 67.55 32.45 70.03 29.97 75.95 24.05 Narlno 80.53 19.47 76.02 23.98 73.39 26.61 73.43 26.57 73.72 26.28 67.93 32.07 68.77 31.23 70.71 29.29 Norte de Santander 81.83 18.17 75.14 24.86 81.20 18.80 77.55 22.45 77.46 22.54 71.87 28.13 72.03 27.97 75.06 24.94 Qulndlo 79.67 20.33 75.77 24.23 75.37 24.63 75.78 24.22 72.93 27.07 65.11 34.89 65.29 34.71 71.31 28.69 Rlsaralda 77.73 22.27 76.77 23.23 76.30 23.70 75.80 24.20 73.70 26.30 69.50 30.50 68.78 31.22 71.61 28.39 Santander 80.65 19.35 80.66 19.34 78.53 21.47 76.56 23.44 77.17 22.83 73.13 26.87 74.46 25.54 74.78 25.22 Sucre 77.55 22.45 71.53 28.47 72.75 27.25 73.35 26.65 73.93 26.07 67.12 32.88 67.78 32.22 73.06 26.94 Tolima 80.96 190.4 78.15 21.85 74.62 25.38 78.67 21.33 79.92 20.08 74.03 25.97 69.16 30.84 75.37 24.63 Valle del Cauca 78.10 21.90 76.14 23.86 75.17 24.83 76.83 23.17 72.47 27.53 66.65 33.35 65.38 34.62 70.19 29.81 Arauca 74.51 25.49 73.51 26.43 73.14 26.86 75.82 24.18 71.61 28.39 73.49 26.51 72.94 27.06 71.83 28.17 Caqueta 76.79 23.21 76.12 23.88 75.18 24.82 70.65 29.35 72.16 27.84 74.30 25.70 72.93 27.07 79.47 20.53 Casamare 74.03 25.97 73.68 26.32 73.54 26.46 84.13 15.87 75.22 24.78 79.92 20.08 77.18 22.82 78.08 21.92 Putumayo 75.54 24.46 74.65 25.35 72.97 27.03 69.50 30.50 73.17 26.83 70.74 29.26 76.17 23.83 73.17 26.83 San Andres 74-00 26.00 73.30 26.70 74.06 25.94 71.18 28.82 72.44 27.56 72.23 27.77 65.76 34.24 68.68 21.32 Amazonas 73.43 26.57 73.83 26.17 73.96 26.04 70.14 29.86 73.32 26.68 70.64 Z9.36 57.97 42.03 59.86 40.14 Guainia 73.76 26.24 72.86 21.14 74.01 25.99 71.69 28.31 73.25 26.75 71.30 28.70 65.59 34.41 66.17 33.83 Guaviare 70.80 29.20 73.91 26.09 74.01 25.99 '71.04 28.96 71.53 28.47 70.34 29.66 64.32 35.68 73.18 26.82 Vaupez 73.79 26.21 73.71 26.29 74.04 25.96 71.64 28.36 70.84 29.16 72.21 27.79 51.35 48.65 54.86 45.14 Vichada 73.79 26.21 73.71 26.29 73.94 26.06 82.19 17.81 73.18 26.82 72.56 27.44 67.22 32.78 66.57 33.43

TOTAL 79.58 20.42 76.51 23.49 76.79 23.21 76.45 23.55 76.16 23.84 72.37 27.63 70.80 29.70 74.52 25.48

Sources: From data supplied by Contralorla General de )a Republica. Table 10: '384 8!'s4ttL* Q7

43. 420 to2" 2234 gas 2472 3337 gay May7 4404 33126 4108 8223 1412 0067 4499 7454 407y MSa 236 so2n 636 2041 464 12319 Al I-3t- Ma 4o0 11lo *oa MIS toio 209 ion2 mass *as 3226 11184 840 1602 2202 763 3640 2271 38a I763 717 lot8 2897 s6ot

W-. Po 1334 344 a- 2354 246 644 sags 2601 o,s a072 114.16 ties 43 S s.494 1ie, 4623 *io Is"9 eAM m1a9s1 2a2 rig. 3024%

3I.- 146 464 824 26 722 30014 24 531 3268 443 1240 364y 834 1401 *440 7* ,1414 2264 646 3740 216 JIB7 2628 87m 192 738 920 293 "I3 1264 268 1843 146 464 1940 2208 430 2336 292 637 7763 27.4 *030 2664 6674 1223 44 r6 8797 (.4. 44 54s 732 226 Sao soot 321 3it2 Isla 238 13248 377 440 3482 231* 56 3042 2477 O07 2088 2642 436 2627 6748

132 462 634 214 644 0613 200 662 Its7 323 3177 3482 411S 139* Is6" 410 211s9 369 T714 3276 2383 6fl 2168 3360 C.... I ? 201 244 *24 260 834 332 848 781 290 724 1020 216 693 it" 430 26 3836 488 I= 3966 643 1474 20317 fh-. gs8 264 2124 It. 202 510 202 468 6 220 6ON a36 256 673 1120 846 29" 1296 06 40 i33" 443 3617 20400 1-.624 847 644 27v 716 304 203 3300 3406 847 1340 2371 403 1444 1647 64 16t7 2487 760 3460 7740 "63 274 87721 (-A ...... 234 3370 3364 262 873 3927 470 2324 25044 846 2462 600 7*4 A386 *68 64 I 3603 4762 11r6 4017 6346 S406 4162 764

C- . 3 , ,. 84o l61 240 9l OAS 640 34 265 683 224 44 660 2041 834 703 a68 642 420 667 3IN6 Iwo 40 2423 26310 . t230 452 884 M6 264 *ee 222 844 307 801 920 1224 676 38441 3328 448 3470 19l38 S16 go4 3320 660 900 192 9.43. 20 429 846 203 64 0a" D60 607 3367? 6ig 0300 3400 484 3320~ 1612 46361 340 2064 44? 3608 217 440" 28341 8266

33.3. 6~~~2 226s 267 lot 320 433 1326 30 873 202 B7D 776 204 460 94 820 my7 3017 868 M68 323 412 3203 31132

1468 612 16o 226 748 460 235 l083 3432 478 387M 7M 60I 3706 2837F M2 3760 2176 4645 im2 2726 1142 2077 40610

'..-.4- 32 41 "4a 226 ess 90 780 3334 172 222 3312 1134,4 444 is"6 3470 738 3676 2406 742 348 27724 472 26201 664 0-.4. 06 233 46 1324 434 863 361 864 726 253 72 4412 293 704 3036 426 791 3236 446 66 3268 680 367 loll 8 -... 4. 108 370 476 163 832 642 234 70 424 204 448 1277 179o 3002 34443 823 3230 1741 663 329 3I"1 19 2000 2723

3*.*,. 46 631 100 23" 1414 3763 461 3412 2324 849 31423 2842 a"4 23861 2042 131s 60)44 4336 33690 824411 4420 I6is 4443 6434

64 297 363 344 437 88S 206 Ag0 780 286 761 3072 we6 648 3170 462 948 3400 483 1012 11492 678 3884 21384 1Ito,37 722 042 264 944 3200 436 266 3720 427 3632 2044 878 226 266.3 740 21*0 200 647 2256 32SS 1121 8443 4462

C.. 333 I304 34120 "81 1638 2016 846 to442 2610 0S4 2766 884 3094 264 8463 132 am2 44134 Irm0 not 13 32234 a2m 74*4 24 li4 333 s0 Is? 227 023 220 309 t0o 279 266 132 143 479 146 4)3 882 307 604 oil 243 440 6I

C-t. 5~2 372 224 60 268 336 030 232 438 309 488 644 134 403 426 227 -f 9"a 254 ISIS 344 WS 7230 3678

C...... 47 334 l63 70 *96 243 40 204 ADS s0 236 5761 134 434 877 384 623 7T" 222 7014 el? 26* 3003 12012 -~ 43 124 304 72 232 204 e3 263 3144 140 sit 489 330 404 689 203 484 "8 200 648 cry 239 7? t077 3..4-4 39 lot 180 64 3862 223 70 237 241 300 247 847 321 330 434 326 364 491 234 431 626 2386 22 760 *...... 26 300 *42 26 IS* 234 71 238 280 303 242 146 its 81 43) 333 820 462 182 213 423 284 84 64S

*... 3? 304 143 07 163 230 72 200 277 92 222 a28 III 203 430 320 We8 49 172 S"5 447 I1 440 S"8 C ~~~ ~~4047 132 84 SOLD 207 72 M0 277 98 233 226 123 804 428 124 No6 420 372 830o 462 223 834 829 26 307 483 Dal 357 232 24 233 26s 9S 240 8138 120 8ll 420 322 "4S 478 360 340 870 238 260 034 1-.O.4 so t0y 346 64 18? 233 74 230 264 02 240 242 134 Sal 420 322 149 483 377 S62 64 282 442 4444

1.3 3464 38446 34438 4690 22118 20406 402 2464 266W 33323 34046 47234 12476 44480 684420 34223 50247 41960 22346 03642 To 712 27441 63490 3049 ~.() 204 74e 306oo0 226 7Fs5 300 0 22 2 760 300D0 234 76lo4 300 0 226 702 100oo0 2774 774 too 0 29 2 706 3too0 2SIS1 186 3000o

6n.. .1. -,Of* i Op C.3,3.,.C..,3 * I - 136 -

.able The Distrib:ion of Sales Tax Revenues, 1985-92 (as t of IVA collection)

Recipient 1 98 5a 1 986 b 1987 1988 1989 1990 1993 ;992

All municipalitiesC 25.0 25.8 25.9 26.4 27.0 27.5 28.0 28.5 Small municipalitiesd 0.4 1.8 3.8 6.0 9.0 12.5 16.8 National territoriese 1.5 0.7 0.6 0.6 0.5 0.5 0.5 0.5 Social securityf 3.5 3.5 3.5 3.5 3.8 3.8 3.8 4.0

1GACg 0.1 0.1 0.1 0.1 0.1 0.1 0.1 ESAph 0.1 0.1 0.1 0.1 0.1 0.'

Total: 30.0 30.5 32.0 34.5 37.5 41.0 45.0 50.0

Notes: a. As set out in Decree 232 of 1983 (with the distribution in part based on 1973 population). b. Thus allocation actually only takes effect only as of July 1, for the first six months. 1985 formula applies. c. Includes Bogota. and after mid-1986, municipalities in national territories. Note however, that 30% of allocation to cities between 100,000 and 500,000 population and 50% of allocation to cities over 500,000 in fact goes directly to FERs. d. To be allocated in accordancewith fiscal effort formula among municipalities of less than 100,000 inhabitants. e. After mid-1986, this amount goes directly to national territories for their own use, without restriction. f. This share goes to the sectional social security funds (including those in the territories) and is intended mainly to cover social security payments for teachers. g. Instituto Geografica Agustin Coddazi. To be used for revaluation in smaller municipalities. h. Escuela Superior de Administracion Publica. To co=ence as of 1987.

Source: 'IBRD Colombia - A Review of Recent DecentralizationMeasures, Washington. D.C.; April 6, 1987, p. 36. Table 12: Fondo Nacional del Cafe: Income and Expenditures, 1979-s7

1979 1980 1981 1982 1983 1984 1985 1986 1987 11

1. Revenues 109,863.2 155,837.3 67,941.4 91,777.4 106,967.4 190,281.3 264,049.9 627,881.4 512,246.8

A. Operational 74,609.5 116,792.4 49,571.2 57,094.0 67.618.4 116,230.5 155,190.1 409,159.4 220.032.0

Coffee Sales 74,609.5 116,792.4 49,571.2 57,094.0 67,618.4 116,230.5 155,190.1 409,159.4 n.a. 1. Coffee for exports 72,246.1 114,052.2 47,000.4 54,547.9 59,599.7 102,549.3 123.517.4 326,806.9 n.s. 2. Coffee for exports 830.8 618.7 919.2 762.3 1,365.4 2,139.3 3,305.6 6,996.6 n.s. 3. Coffee for Retentlon 0 0 0 0 0 0 0 0 n.s. 4. Coffee for Domestic Consumption 1,524.6 2,121.5 1,651.6 1,783.8 2,095.6 2,453.3 3,157.8 5,402.9 n.s.

B. Other operating revenue 1,430.7 2,393.3 3,536.8 3,138.2 3,939.5 4,230.9 6,527.3 10,189.8 10,295.2 1. Other goods sales 1,210.6 2,082.9 3,102.9 2,602.8 3,302.1 3,470.4 5.492.7 8,956.4 10,295.2 2. Fees reimbursements (Freights) 220.1 310.4 433.9 1535.4 637.4 760.5 1,034.6 1,433.4 n.s.

C. transfers 32,314.6 34,121.2 12,419.9 26,643.8 32,639.1 66,440.5 88,906.1 189,049.0 205,870.5 1. From the Central Government 2,468.4 3,005.0 2,823.0 3,112.6 3,377.0 4,898.5 7,298.7 14,905.2 11,671.1 Ad valorem Tax 2,468.4 3,005.0 2,823.0 3,112.6 3,377.0 4,698.5 7,298.1 14.905.2 11.671.1

2. From the Private Sector 29,846.2 31,116.2 9,596.9 23,531.2 29,262.1 61,742.0 81,607.4 174.143.8 194,299.4 a. Retention-in-kind 29,453.7 31,098.2 9,588.8 23,510.5 29,146.5 61,539.9 80,821.1 172,567.8 n.s. b. Retention-in-value 392.5 18.0 8.1 20.7 115.6 202.1 786.3 1,576.0 n.s.

D. Financial Revenues 1,453.8 2,447.2 2,354.8 2,321.9 2,738.9 2,840.9 12,420.0 17.608.0 51.100.4 E. Other Revenues 54.6 83.2 58.7 4,579.5 31.5 538.5 1,006.4 1.675.2 24,940.7

11. Expenditures 97,221.1 150,340.2 74,556.6 92,237.0 100,857.5 158,902.3 175,400.8 407,250.7 495,414.0

A. Operational expenditures 39,757.3 65,304.2 38,911.2 45,190.2 45,364.7 73,887.9 64,214.4 142,214.4 246.700.8 1. Coffee Purchases 39,75).3 65,304.2 38,911.2 45,190.2 45,364.7 73,887.9 64,214.4 142,214.4 246,700.8

B. Other expenditures 4,935.6 5,998.1 6,600.2 8,157.0 9.842.2 13,582.9 17,125.2 25,369.5 208,130.0

1. Commercial expenditures 3,414.6 4,279.7 4,152.9 4,810.5 5,904.3 8,344.0 11,157.8 17,725.1 138,089.0 a. Commisslon 497.6 698.0 919.9 1,169.5 1,290.5 1,393.9 3,336.3 2,656.9 n.a. b. Freight changes 2,304.7 2,753.3 1,952.5 2,178.7 2,637.4 2,861.3 3,386.6 4,131.5 ft.. c. Advertising 142.7 123.4 723.4 953.7 1.094.3 2,061.5 3,042.5 5,984.1 n.s. d. Foreign promotion 0 0 0 0 727.4 968.0 245.0 1,784.0 n.a. e. Packaging 505.9 705.0 557.1 508.4 836.1 1,054.6 1,147.4 1,786.5 n.m. f. Quality changes (36.3) 0 0 0.2 (681.4) 4.7 0 782.3 n.a. (Contlnu.,d) 12: I.OOTaleFindo Nar ional del Cafe: Income and Expenditures. 1979.-8

19)9 1980 1981 1982 1983 1984 1985 1986 1987 1i

2. Coffee Handllig 1,521.0 1.118.4 2,44Z.3 3.346.5 3,937 5,238.9 5,967.4 7.644.4 54.733.5 a. Trashing expenditures 562.6 499.0 461.0 692.6 194.3 870.4 1,288.7 1,565.7 n.s. b. Storage 828.9 1,051.8 1,530.0 2.247.2 2,878.8 3,745.1 4.401.6 5,901.7 n.9. c. Ports expenditurrs 128.Z 161.4 113.6 134.9 131.1 145.5 152.J 170.5 n.t. d. Interest payments 3 0 340.7 211.8 133.7 477.9 124.2 6.5 n.a. e. Insurance 0.8 2.2 0 0 0 0 0 0 0

3. Other Goods 1.380.9 2,427.8 3,460.6 3,07.1 3,917.4 4,298.0 6.646.5 12.571.1 15,307.4

S. Other GoiLs Purchases 1,019.9 I.Z76.6 2,946.6 2.501.1 5.224.3 3,463.0 5,225.7 10,088.0 11,694.1

b. Other Goods Expenditures 361.0 651.2 514.0 576.0 693.1 835.0 ',420.8 2,563.1 3,613.3

C. Pr ire different ials Adjustments 6,316.1 15,773.2 5,860.4 3,574.1 4,900.1 3,002.6 4,963.1 49.614.6 n.s. D. Transfers 37.003-5 43.647.2 13.359 8 22,103.5 21,333.1 40,959.0 55,046.1 146,283.9 n.s. E. Administration 4,617.8 6.620.5 3,011.9 3,989.9 4.911.5 7,161.2 10,933.5 13,823.0 17,424.7

1. FPederafe Administration 1,583.8 2.4Z10. 2,810.1 3,919.4 4,903.3 7.671.5 10.926.7 13.597.2 17.163.'j 2. Contraloril 5.6 51.4 56.1 70.5 8.2 89.7 6.8 225.8 260.9 3. Exthange rate losses (tertlFicates) 3,028.4 4.098.4 145.1 0 0 0 0 0 0

F. Integest payments 2,746.1 2,259.0 2,894.0 5,482.9 9,972.1 14,042.9 11,039.5 7,019.7 17.785.4 C. Other expenditures 463.2 8,310.2 458.5 662.3 616.4 1,367.8 5.432.5 10,354.5 5,3173.1

11,R-sultq

A. lnfloqw minus outflows total (I * t1) 12,642.1 5,497.1 (6.615.2) 1,540.4 6,109.9 31,379.0 88,649.1 220,630.J 16,832.8 linus: Exchange Losses 0 0 0 904.1 4,424.1 5,517.7 11,740.5 9.04J.2 0

B. Sutplu.zi or Deficit 12,642.1 5.497.1 (6,615.2) 616.3 1,685.8 25,861.3 76,908.6 229,6Z7.9 16,812.8 Hinus: Reserves 0 205.4 0 0 0 0 0 0 3,954.9

C. S.lrplus or Deficit 12,642.1 5.291.I (6,615.2) 636.3 1.685.8 25,861.3 76,908.6 229,677.9 16,832.8 D. lnti.mnal cash lltw 31,141.3 41.883.4 7,596.6 6,885.0 16,351.0 32,990.6 80,317.8 251,965.3 (224,503)

1t 198: Somne of the clats are not avalblale. The ssource in FNC but differs from 1980-86.

-.ur,e: Fondo Maclonal del Cafe. Actual dsta. T;ab)le 13: Fond. Viol Neclon.l (F.V.N.) Totol E'npoditur*v

(MilIlonn of Colomblon 2)

Programs and Projects 1960 (S) 1981 (M) 1992 (3) 1933 (6) 1984 (6) 1966 (6) 1906 (1) 1in? (6)

1. Hlighry construction, repairIrg end maintenanc- 10,390.0 63.3 16,112.0 66.2 13,127.0 43.6 13,800.0 46.4 16,616.0 44.0 22,468.6 50.3 20,079.0 49.4 42,900.9 60.2

11. TIchnical Service. and Special Studies 280.0 1.3 624.0 2.8 531.0 1.2 SO2.0 1.2 392.0 2.6 2,022.9 4.6 2,467.6 4.7 3,631.2 4.2

111. Equipment Invostment 307.0 1.9 609.0 2.2 2,967.0 11.0 1,682.0 6.7 342.0 2.4 I,617.3 *.4 909.7 1.7 614.8 0.7

0.3 4S3.0 1.4 946.4 2.1 U41.2 0.6 602.4 0.7 IV. Highways S.f.t y Control 46.0 0.3 49.0 0.2 60.0 0.2 94.0

V. Othor Roado Milntenanco . and District Admini- C,) atratlon 3,666.0 23.2 6,662.0 24.0 3,106.0 10.3 10,`84.0 36.4 11,368.0 32.6 11,306.4 26.6 13,696.9 26.3 16,933.1 22.1

VI. Maritime and Rivorvays TrenSportation 626.0 3.8 1,342.0 6.3 2,019.0 7.6 2,194.0 7.4 4,271.0 12.9 4,148.7 7.0 3,371.2 7.3 8,091.0 4.6

VIl. Oebt Servicing 103.0 0.7 79.0 0.3 69.0 0.3 1,006.0 8.4 1,167.0 8.0 2,691.1 6.0 5,602.2 10.4 14,363.1 17.4

1. Esternal 0.0 0.0 0.0 0.0 17.0 0.1 76T.0 2.6 1,146.0 8.3 2,290.1 C.1 4,727.1 9.0 10,001.4 12.6 2. Internal 103.0 0.7 79.0 0. 62.0 0.2 238.0 0.6 211.0 0.6 401.0 0.9 776.1 1.5 4,0f0.9 4.3

Total Gmato. 16,313.0 100.0 23,167.0 100.0 26,996.0 100.0 29,461.0 100.0 34,023.0 100.0 44,670.7 100.0 62,76u.3 100.0 35,483.0 100.0

Source: MOlPT- Oticino de Plon-scion y Direcclon Corciel y Financlra (Eatado. Flnsnel.ros). Informs Financiero. do Controloril General d. Is Republics. - 140 - Table 13.1

COLOMBIA: GASOLINEPRICES AND TAXES (1968)

Gosol. Casol. Fuel Jet Motor LPG Cocinol Averge Reg. Extra Oil Fuel Diesel Price: (USJ/bbl)

Prices

,. Retell Price 22.04 27.51 11.72 20.50 22.16 12.56 1.81 21.2 2) CIF Price 20.88 22.98 10.54 17.17 17.72 16.66 20.88 19.7 3) EcopetrolPrice 13.23 18.14 11.59 20.19 14.36 7.73 1.09 13.6 4) Retail lese Ecopetrol 8.81 9.37 0.13 0.31 7.82 4.83 0.72 7.7

Taxes and DistributionCost Average Taxes

6) Rood Tax 6.71 5.71 0 0 5.71 0 0 4.86 6) Other Taxes 1.65 1.60 0.13 0.31 0.21 0.13 0.05 1.08 7) DistributionCost 1.66 2.06 0 0 1.89 4.70 0.67 1.76 8) Total Taxes A Dist.Cost 8.81 9.37 0.13 0.31 7.81 4.83 0.72 7.68

TOTAL Volume (bbl/day) 101307 9s6o 2384 8947 33699 11469 2819 170 176

Revenues (USJ million)

At retail price 815.0 97.0 10.4 68.9 272.0 52.6 1.86 1315.6 At CIF price 772.1 81.0 9.2 66.1 217.3 e6.6 21.6 1222.7 At Ecopetrolprice 489.2 84.0 10.1 65.9 176.1 32.3 1.12 838.7

Taxes (ec. definition) 268.46 26.79 0.11 1.01 72.72 0.64 0.06 368.6 (USJ million) ConsumptionTax (Retail- CIF) 42.89 15.98 1.03 10.87 54.70 -12.97 -19.62 92.8 ProductionTax (CIF- Ecopetrol-Distr.) 225.6e 9.81 -0.91 -9.36 18.03 13.51 19.67 275.7

Taxes (actual) 268.46 2 .90 0.11 1.01 72.72 0.64 0.06 368.6 (US9 million) Road Taxes 211.14 20.16 0 0 70.06 0 0 301.3 Other Taxes 57.31 6.65 0.11 1.01 2.67 0.64 0.06 87.2

DistributionCost 67.6 7.3 0 0 23.2 19.7 0.7 108.4

Source: Colombia,Energy Pricing Study, October 1989, Volume II. Annex 2.2 and 2.3. - 141 - Table 13.2

COLOMBIA: Gasoline Prices and Taxes (1989)

Gssoline Gasctine Fuel Jet Motor Average (2) Gasolire Average (3) Regular Extra Oil Fuel Diesel LPG Cocinol Prices Extra Prices

Prices

Retail Price 22.13 27.59 12.52 22.72 22.26 12.56 2.86 21.07 27.59 21.44 CtF Price 24.73 (1) 12.76 20.26 21.33 18.55 24.73 23.16 23.29 23.17 Ecopetrol Price 13.31 18.22 11.55 22.38 14.41 7?9 1.13 13.43 18.22 13.7 Retail less Ecopetrol 8.82 9.37 0.97 0.34 7.85 4.66 1.73 7.64 9.37 7.74 ...... ------......

Taxes and 0istribution Cost Average Average ...... ------Taxes Taxes

Road Tax 5.74 5.75 0 0 5.73 0 0 4.85 5.75 4.91 Other Tax 1.54 1.57 0.13 0.34 0.21 0.13 0.04 1.05 1.57 1.08 Oistribution Cost 1.54 2.05 0.84 0 1.89 4.53 1.69 1.74 2.05 1.75 Total Taxes & Dist Cost 8.82 9.37 0.97 0.34 7.85 4.66 1.73 7.64 9.37 7.74

Total Total

Volume (bb/day) 105359 0 (1) 2406 9000 35282 11700 2700 166447 10054 176501

Revenues (USS million)

At retail prices 851.0 0.0 11.0 74.6 286.7 53.6 2.8 1280.1 101.2 1381.2 At CIF prices 951.0 0.0 11.2 66.6 274.7 79.2 24.4 1407.0 85.5 1492.7 At Ecopetrol prices 511.8 0.0 10.1 73.5 185.6 33.7 1.1 815.9 66.9 882.6

Taxes (USS Miltion) 280.0 0.0 0.1 1.1 76.8 0.6 0.0 358.4 26.9 385.9 *-.-- (Econanic Classification) Consumption Tax/subsidy *100.0 0.0 -0.2 8.1 12.0 -25.6 -21.8 *127.0 15.8 -111.5 Production Tax/subsidy 379.9 0.0 0.3 -7.0 64.8 26.1 21.6 485.4 11.1 497.3

Taxes (USS Million) 280.0 0.0 0.1 1.1 76.8 0.6 0.0 358.4 26.9 385.9 ----. (Actual classification) Road taxes 220.7 0.0 0.0 0.0 74.0 0.0 0.0 294.7 21.1 316.3 Other taxes 59.2 0.0 0.1 1.1 2.7 0.6 0.0 63.8 5.8 69.6

Distribution Cost 59.2 0.0 0.7 0.0 24.3 19.3 1.7 105.7 7.5 112.7

(1) Not reported in the Energy Pricing Study (2) Gasoline extra incLuded (3) GasoLine extra included

Source: Colombia, Energy Pricing Study, October 1989, Volume It. Amex 2.2 and 2.3 -142 -

Table 13.3

COLOMBIA: Gasoline Prices and Taxes (January1990) (1)

Gasoline Gasoline Average (3) Regular Extra Diesel LPG Cocinol Prices

Prices

Retail Price 25.12 31.25 25.24 14.45 3.27 24.38 CIF Price 26.5 27.14 22.83 17.04 26.9 24.98 EcopetrolPrice 15.29 20.83 16.55 9.7 1.34 15.29 Retail Less Ecopetrol 9.83 10.42 8.69 4.75 1.93 9.09

Taxes and DistributionCost Average Taxes

Road Tax 6.47 6.47 6.47 0 0 5.87 Other Tax 1.76 1.81 0.25 0.17 0.04 1.27 DistributionCost 1.6 2.14 1.97 4.58 1.89 1.95 Total Taxes & Dist Cost 9.83 10.42 8.69 4.75 1.93 9.09

Total

Volume (bb/day) 99125 10448 37043 12394 2520 161530

Revenues (USS million)

At retail prices 77.2 10.1 29.0 5.6 0.3 122.1 At CIF prices 81.4 8.8 26.2 6.5 2.1 125.1 At Ecopetrolprices 47.0 6.7 19.0 3.7 0.1 76.6

Taxes (USS Million) 25.3 2.7 7.7 0.1 0.0 35.8 ----- (EconomicClassification) Consumption Tax/subsidy -4.2 1.3 2.8 -1.0 -1.8 -3.0 Production Tax/Subsidy 29.5 1.4 4.9 1.1 1.8 38.8

Taxes (USS Million) 25.3 2.7 7.7 0.1 0.0 35.8 ----- (Actual classification) Road taxes 19.9 2.1 7.4 0.0 0.0 29.4 Other taxes 5.4 0.6 0.3 0.1 0.0 6.4

DistributionCost 4.9 0.7 2.3 1.8 0.1 9.8

(1) Prices of Jet Fuel and Fuel Oil not included. Table 14: Fond. Vlol Nscional (F.V.N.)Total Revenue (MIllions of Colombiant)

Dsecriptlon 1900 () I9lo (U) 1982 (S) 193 ( 1)9194 (U) IveI (S) 1966 (S) 97 (S)

1. Current Revenues 14,290.0 so.? 16s,7iS. 91.7 22,201.9 99.2 24,411.2 63.6 20,296.3 *2.? 82,848.5 67.r 40,393.3 92.5 02,701.9 77.5 A. Opet tional Incoeo 503.0 3.6 630.0 4.1 1,377.1 0.1 1,605.2 e.S 8,044.0 9.0 4,906.0 13.4 8,571.4 1.0 1,3U02.1 11.3

1. Sal. of Services 2.4 0.0 2.7 0.0 3.6 0.0 17.2 0.1 84.1 0.3 27.2 0.1 36.4 0.1 70.C 0.1 2. Other 2/ 500.0 3.5 633.3 4.1 1,313.3 0.1 1,668.1 0.4 2,969.9 9.3 4,90.6 13.8 I6,532.3 15.0 12,291.0 15.2

e. Transform. Revenue Sharing, Txes. 13,133.0 66.2 17,922.0 97.0 20,904.8 93.1 22,526.9 77.2 23,251.8 73.2 27,002.5 74.1 83,021.6 77n. 50,39.8 62.2

1. R.O.E. 3/ 12,730.0 79.0 10,070.0 61.5 18,00.0 62.8 16,001.0 01.9 21,354.0 67.2 20,675.7 1t.7 2,0791.6 1S.1 43,765.1 54.1 2. Budget transfers 4/ 1,003.6 6.2 1,240.6 0.1 2,290.6 10.2 4,404.9 15.3 1,697.3 0.0 906.6 2.7 1,030.4 2.4 6,574.7 6.1

11. Copltal Income 1,626.6 1I.$ 1,101.7 6.3 179.8 0.6 4,77a.0 16.4 5,494.6 17.3 4,609.6 12.0 a,29a.2 7.C 10,390.4 22.5 Total Revenues 16,123.2 100.0 20,480.3 100.0 22,461.2 100.0 29,164.0 100.0 31,760.1100.0 87,216.3 100.0 43,0".1 100.0 6o,m.3 100.0

X R.D.E./Current Income 09.0 89.0 63.6 74.0 91.2 61.? 81.2 69.9

1/ Actual Figuren. I/It Include, toll revenue and revnues from grr_nts with other *gencl-e. 3/ Gasoline and ACFMtaxn. Proced.o arm used for debt servicing and Investment. 4/ Other budget tranaferm for investmont. Sourc-: Contralorie C*nerol de la Republice,Intorm.o Flnencleroo. FV14fInncial reports. Table 15: Fonda d- Promlon de Esportecloeo (PROEUPO):Total Erpedltvres (Millons of Coloebian 5)

I290I (e I"9 (3) 29 ) 1963 (s) 94 () 1906 () D- cription 160 (1) 1962 (s) t

Operatienal Ep9ndituroe 1,T1,2.2 .1 1.,485.2 0.O 0. 2s9.6 0.1 46S.1 0.6 e13.1 0. 680.9 0.6 1. Peresonl Serwlcee 208.6 0.1 244.4 0.4 492.9 0.6 660.? 0.6 644.0 0.5 1iO.6 0.4 136.6 o.a 16t.0 0. 2e9.7 0.4 t7t.9 11. Cenorl Expenditures 700.0 0.6 4,641.9 4.0 7,509.6 1.0 379.1 0.2 250.2 O?.?0.6 O. 122.6 0.2 1,614.6 2.1 211. Tranetfre 5,914.6 6.0 6,61.2 9.6 2,601.6 1.1 2.0 491.8 1.2 191.1 1.0 2,806. 8.0 1,139.7 1.2 SMbt,tel 564.2 3,172.6 1.6 5.7 1,170.0 2.1 1,926.0 2.1 1,971.2 2.0 2,621.6 2.5 IV. Debt Servicing 1,512.9 4.5 1,569.6 8.6 2,243.1 5634.6 6.8 Ut.1.Operational 4,019.2 6.5 3,656.8 8.6 1946.6 1.5 122,96.? 111.1 Espendituro. 1,008.1 0.5 2,067.4 5.0 2,656.2 4.1 Inwestent Expendilture 504.0 0.2 Tr arde Fairs 106.1 0.1 151.8 0.2 122.6 0.1 270.2 0.2 Participatlon 49.9 0.2 65.4 0.2 40.0 0.1 4.4 1U,057.9 18.0 69,45.9 6.1 524.5 0.2 339.4 1.2 2,61.09 5.6 8,391.0 6.6 2,306.1 .7 4,106.6 It. projects 71.4 120,376.0 72.8 _ 56.2 6,0916.6 59.9 52,121.1 16.1 716,60.2 16.4 6,140.7 111. E.port Crodit Lens 25,912.1 61r.6 56,621.1 n6.0 51,14.7 12.4 > IV. BSuidinoe *nd Other 120.1 O.0 2,306.4 5.0 621.2 0.6 55.4 0.3 22,243.2 Axdt 12,2n9.0 4.4 1,066.5 2.6 2,60.2 2.5 0.0 0.0 2,96.3 8.0 6,406.0 4.9 20,144.2 11.6 V. Lee 66/1965 !/ 0.0 0.0 0.0 172,990.5 96.1 Total l2vet.ent 12664.5 94.1 60,106.1 90.1 92,366.4 02.1 0,645.9 n. E.pond;turo 21,596.4 93.6 89,460.6 96.0 S1,26.9 906.5 100.0 100.0 92,021.4 100.0 200,84.2 200.0 122,541.0 200.0 176.,265. Total E.pendltur.s 29,401.6 200.0 42,527.9 100.0 0,064.41 200.0 76,083.6 74.6 60.5 96.0 92.1 I1.1 61.1 X Inversion 91.9 95.5 redemption. Lew 66 reallocatlons .1ll be excluelvely for CERTand eoter expertIncontiwe 1/ In 19611CERT pereitt Included(Tho CERTor* export promotiontee exemptionnotes). Source: ContrelorloCenrel de Ie Republice, Infor-ee Financl.roe. - 145 -

Table 17: Legal Contributionsto Social Security by Program. Source and Fund in Colombia: 1988 (as a percentage of wages)

Fundl Active Prcgram insured (a) Employer State Total

ISS (private sector) Pensions 2.17 4.33 d 6.5 Health-Maternity 2.33/4.0 (b) 4.67/8.0 (b) 7.0/12.0 (b) OccupationalRisks 0 1.5 (c) 1.5 TOTAL 4.5/6.17 10.5/13.83 15.0/20.0

CAJANAL (public sector) Pensions (e) 0 3.0 f 3.0 Health-Maternity 5.0 5.0 10.0 TOTAL ,5.0 8.0 13.0

CAPRESUB/CAPRECOM (public sector) Pensions (e) 0 8 g Health-Maternity 5.0 g g 5.0 TOTAL 5.0 5.0

MILITARY Pensions (e) 8.0 h h 8.0 Health-Maternity 0 i i 0 TOTAL 8.0 8.0

POLICE Pensions 5.0 h h 5.° Health-Maternity 0 i i C TOTAL 5.0 5.0

OTHER PAYROLL TAXES 0 9.0 (j) 0 9.0

(a) In ISS. the voluntarily insured self-employedpays a percentage over income equivalent to the combined percentage paid by the salaried insured and his employer. (b) Among those covered by family health program. tc) Average premium; the premium should vary from 0.28 to 7Z according to risk but the classificationof risks is not yet in effect. (d) Initially the state had to contribute but never actually did and such obligationwas abolished in 1971. (e) There is no distinction between treatment of conon and occupational risks and the contribution to pension and health cover such risks. In any case the employer is considered directly responsible. (f) The state is responsible for absorbing most of the resulting deficit. (g) The employer (state) pays the rest. (h) The state pays all the resulting deficit. ti) The employer (state) pays all costs. (j) .4Z goes for Family Allowances (Cajas de Compensacion).2Z for SENA, and 32 for ICSF.

Source: Current Legislation. Table 18: Sroricio W.cionl d. Aprendiz z SA: Total E.penditure!, .L -80 (Millions of Colombian 3)

Da.cription 19B0 (t) 19BI (11) 1902 (2) 1902 (2) 19B4 (X) I905 (2) 1996 (1) 197 t9)

iT 6;i'c Ing

1 Personal Services 2,164.S 47.6 2,623.3 41WS 8,270.9 44.6 3,910.1 31.7 ,617.1 10.6 1,370.9 9.6 2,387.4 1131 2,75.6 10.1 f C.ne.rl E(pense 969.9 21.a 1,094.0 13.2 1,034.1 22.2 2,081.9 19.0 999.2 6.6 1.136.3 6.2 1,22S.9 6.3 3,657.6 6.0 i"' I stfer 394.3 3.7 969.0 16.9 1,484.2 20.2 1,690.3 16.3 2,207.? 14.3 1,063.7 5.6 1,70.9 7.1 3,311.3 6.6 Subtotal 5,629.0 77.4 4,677.1 77 t e,397.2 *7.1 7,839.3 76.0 4,824.0 al.3 4,100.9 213. 4,770.S 24.4 6,205.4 9. IV. Debt Serviclng 19.6 0.4 IT.6 0.3 16.0 0.2 16.3 0.2 14.6 0.1 7.3 0.0 0.3 0.0 6.0 0.0 Total Operational Espens., and Dabt Servicing 3,643.S 77.9 4,694.9 71.9 6,413.6 87.3 7,864.1 76.7 4,838.6 31.4 4,317.7 21.6 4,77`7 24.6 6,29. 4 23.0 !nestment/Programa I Educational Buildings Construction *nd Maintenanco 612.9 13.4 950.0 16.3 612.2 *.0 2,025.S 19.6 3,492.4 22.6 2,160.6 13.3 697.5S .6 3,337.2 0.7 II. Rural Development Program l3376 4.1 206.7 3.4 204.6 2.a 361.0e .4 567.0 3.7 407.3 2.1 647.2 3.6 737.0 1.9 l11. Co_unity Psrticlp.tion Program 34.6 1.9 70.3 1.2 60.4 1.2 126.9 1.2 135.3 0.9 0.0 0.0 0.0 0.0 13.5 0.0 IV. Food Prograso 10.0 0.2 108.2 1.8 24.2 0.3 12.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Human Resources Infra- structure ond Resa-rch 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1,242.1 6.2 2,310.1 12.1 2,135.0 10.9 2,970.5 10.9 VI. Scientific and lechnology Research and Prmotion 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4,762.1 33.0 6,246.3 32.7 7,474.s 39.3 30,668.2 36.2 Vii. Educational Materit a"nd Equipment 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 85.6 0.0 1,026.1 5.4 S67.4 4.4 1,032.7 83. VIII. Eaploypee Social telfare 0.0 0.0 0.0 0.0 0.0 0.0 C.0 0.0 0.0 0.0 1,209.9 0.3 642.7 3.3 1,229.0 4.6 IX. Industrial Sector Promotion 0.0 0.0 0o0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 216.4 .1 2 eea 1.4 200.0 0.7 X Financial Inveatmant 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 50.3 1.3

and Planning 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 SOO.0b 2.0 480.0 2.4 2 0 0 .3 ' 0.7 XII. Los 65/19R6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4 0 c'9 4.6 1,61S.c 3.09 ,5681.64 S.9 XIII. Others 113.4 2.6 0.0 0.0 4.4 0.1. 0.0 0.0 283.0 1.t 0.0 0.0 0.0 0.0 *0.1 0.9 Total Investment Espenditures 1,000.6 22.1 1,334.6 22.1 931.7 12.7 2,516.0 24.3 10,567.4 66.6 14,966.6 7n.4 14,743.7 7t.5 21,104.9 n7.0

Totol E.ponditure, 4,667.1 100.0 6,029.7 100.0 7,346.5 100.0 10,369.1 100.0 16,426.9 100.0 19,084.3 100.0 39,6261. 100.0 27,893.3 100.0 a. It Included In 1906 3130 Ililiona (Low 66/65) agraemet Sona-Art,esania do Colombl; 1906: 3161 miIlions; 1907: 9200 millions. b. Low 5g: 366 million. and 1987: 3200. millions wore reallocated Sena-Centrol LetInnosmriceno de Informatlca agr-e nt. c. Low 56 reallocated funds are plaed through agreements wlth different gsncies. d. Lsa 66 agreemants: Pacific Coast urban development plan; 3.ool.6 millions tor Sens-Fonlo MEN (Lay b5/86). a. In 1967 it belong, to Co_munity Development Progras. (Training and counity organization) Source: ontrlaorla Coneral de Is Republics, In0form Financiero*. Table 19: Servclois cIonal d. AgrEndlale SENA): TotalRev nu-e. 1960-67 i/ (Millions of Colomblan *)

Oescription 1060 (2) 1963 () 19032 (1) 1903 (2) 1914 (2) 1986 (X) 19t9 (2) 1967 (2)

1. CurrnetRevenues 4,490.9 96.9 6,60S. 6 100.0 8,679.6 97.4 13,060.6 87.4 14,819.7 63.0 16,662.1 67.1 20,363.9 9.6 26,431.3 93.4 A. Operstional 304.? 7.9 426.9 6.6 316.9 4.2 1,040.8 0.2 1,711.9 10.3 2,094.6 10.9 1,661.1 7.7 2.709.1 9.6 1. Selo of Services 73.3 1.7 99.0 1.6 106.1 1.2 119.3 0.9 149.9 0.9 172.6 0.9 197.6 1.0 4.7 0.0 2. Agrements 7.0 0.2 4.6 0.1 2.0 0.0 16.6 0.1 14.2 0.1 167.9 0.6 162.1 0.9 239.6 0.6 3. Financial Revenues 106.6 2.3 290.0 4.4 236.6 2.6 686.0 6.0 1,636.4 6.9 1,617.6 6.6 900.6 4.9 1,429.6 6.1 4. Oth.r 173.6 3.6 36.7 0.6 31.0 0.3 40.6 0.3 69.4 0.4 146.3 0. 191.0 0.9 1,036.6 3.7

a. Transfersand 1s.. 4*,126.2 609.0 6,177.6 93.6 6,302.6 93.1 10,016.2 79.2 12,647.0 72.0 14,666.2 76.1 16,792.4 92.1 2,721.60 13.6 1. Budgettransfers 2/ 310.2 6.7 286.9 4.3 125.6 1.4 326.8 2.6 380.8 2.2 146.6 0.6 263.6 1.4 816.2 3.1 2. (FIC) Industry Construction Fund 0.0 0.0 698.3 30.6 241.6 2.7 602.7 4.6 646.9 4.9 464.6 2.6 263.1 1.4 3116.2 1.1 3. 9.0.E. 3/ (ROE tames) 3,836.0 62.3 6,193.4 76.6 7,936.6 69.0 9,067.7 71.9 11,321.1 06.6 13,936.0 72.3 10,266.6 69.1 23,194.6 62.0 ft. CapitalIncome 143.3 3.1 0.0 0.0 234.9 2.6 1,667.1 12.6 2,928.1 17.0 2,474.6 12.9 40.0 0.2 I,666.6 6.6 A. Balance Surplus Rs..nu*s 4/ 143.3 3.1 0.0 0.0 234.9 2.6 1,667.1 12.6 2,920.9 16.9 2,489.4 12.9 20.9 0.1 i,6s6.6 6.6 O. Other Capital R-v-nu-s I/ 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.2 0.0 6.1 0.0 19.1 0.1 0.0 0.0 Total R-e-Du.s 4,634.2 100.0 6.806.6 100.0 8,914.4 100.0 12,043.9 100.0 17,247.8 300.0 19,137.2 100.0 20,393.9 100.0 26,302.1 100.0

X ROE/Current Income 0.0 86.0 0.0 78.6 0.0 91.4 0.0 82.2 0.0 79.1 0.0 63.6 0.0 69.7 0.0 67.7 i. ActualFigures 2. Covernme*nt pecial programs: DRI, PAN, IPC. (Rural development, foods programs CommunityProgram) 3. It Includes Law 66/86 reallocated revenues. 4. ID. 1910 surplus for 6376.4 millions to not Included In 1901. 8. Fi.ad sseatssales proceeds and so" Funds surpluses (Employee Funds). Source: Contraloris Ceneral de Is Republics,Infor_s Flnancleros. Table 20: IstItwto Coloxbisno do SlI_xter Familiar (JID)F: t.telEaopndituro (millions of Colombian 3)

Deceription 10o (S) 106 (3) 3962 (1) 1906 (1) 1994 (X) 1906 M 1o" ) 19S7 (

Ikrt...on0el PersonalServices 205.1 6.0 are.7 *.t 629.6 7.1 190.0 2.1 321.9 2.3 310.0 2.7 440.1 2.1 610.6 2.1 CGneralExpenses 155.6 2.9 179.0 2.1 229.2 5.1 01.1 1.0 112.0 1.0 186.8 1.2 233.4 1.1 248.0 1.0 Transters 94.7 1.3 101.6 1.1 152.6 2.0 929.1 9.1 2,739.9 24.2 8,027.6 21.0 3,039.0 IS.$ 4,n71.1 16.0 TotalCongel Eap.*dlturos 513.4 9.0 603.I 10.0 011.4 12.2 I,O06.e 12.1 3,100.4 28.0 8,670.2 25.5 4,518.2 21.5 5,56e.2 22.1 Investment Eafenditxre

Children AsssltancePrograms 3,86.0 67.6 4,706.6 72.8 4,905.g e8.9 s,m9. s .s A.m. 0.0 10,123.0 72.2 13,103.6 64.69 6,,O1.0 11.0 Adult. Ameletence Programs 163.9 14.3 621.? 12.4 1,130.1 36.6 1,403.0 16.3 n.m. 0.0 0.0 1,460.0 7.0 221.0 0.9 PhgeicelInfrastructure - Hoelth 99.0 1.9 200.1 3.2 35.1 1.1 336.7 1.3 n.*. 0.0 90.2 0.6 1,213.7 6.1 1,106.0 4.9 HumonResoureo Trrainin and Retearch 0.0 0.0 27.2 0.4 IS. 0.2 31.1 0.3 S.c. 0.0 22.9 1.6 113.7 0.0 104.5 0.4 oo Clobal Planning, Social Ovelop.ont 4.3 0.1 114.0 1.7 0.0 0.1 0.0 u.s. 0.0 0.0 1".6 0.1 '0.0 0.0 I Food and Rural Developmnt Progra. 343.0 6.4 0.0 23.09 3.6 403.0 4.4 *.a. 0.0 17.0 0.1 0.0 7.3 0.0 lotal Investaent Expenditures 4,671.2 90.4 ,0963.6 90.0 6,660.7 37.0 ,036,.9 7.6 3,13.9 1/ 12.0 10,462.1 74.1 16,467.6 71.5 19,010.1 77.9

Total Cat.to 6,369.6 100.0 6,011.9 100.0 1,472.1 100.0 9,143.7 100.0 11,309.3 100.0 14,022.3 100.0 20,970.8 100.0 2S,160.3 100.0

1/ Budget estimates. Actualdata not found. Source: Contralorla loneraldoe IaRewbblIc., trforme Flnancl.ros. T2ble 21: Instituto Colo blo. de eI..tet F Il ll5COFlF:Total 11w n (ilIlions of CoI..bieo )

10 (3 1r (s) O eription I9O (S) 10 (3) 1982 (3) 190I (S) 194 () 1995 ( t)

167,24.0 2.5 24,095.1 67.4 I. Current Revenu*e 4,140.4 61.6 5,837.9 16.4 0,703.4 63.6 6,737.0 60.0 111,14.0 69.5 14,210.0 n0.1 1.7r 113 0.4 A. Oper*tIonal 25.2 0.4 75.r 1.0 32.0 0.4 39.0 0.4 51.7 0.8 6I0.2 1.2 $90.0 B.Bdget Transtfer, 67.0 T.... 4,705.2 57.1 5,162.2 75.4 0,730.0 03.4 6,896.6 66.0 11,132.8 69.0 14,024.4 go.5 16,148.1 60.6 23,962. 17.0 0.1 9.6 0.1 IO. 0.6 1. Tr*nmfs.r 1,121.2 13.0 90.2 11.7 293.5 3.3 167.9 3.7 344.6 2.6 60.r 28,623.0 0.4 2. R.0.D. 3,564.0 43.6 4,672.0 63.6 6,464.3 *0.1 6,530.7 66.9 N0,7M7.5 6. 14,007.4 66.4 1,114.6 10.r 1,636.6 10.3 $1.6 17.5 8,464.2 12.0 11. Capital Ie_Cos 3,500.1 42.5 1,199.6 23.0 1,306.2 16.2 1,078.1 11.0 1,34.18 300.0 2,7.6.4 100.0 f Revenue, 0,240.6 100.0 7,637.7 300.0 6,066.6 100.0 9,615.7 100.0 12,06S.3 100.0 15,049.2 100.0 22442.0 Total - F

S R.D0E./Current Incoe 75.6 63.5 9s.6 97.6 si.0 6.6 97.6 96.6

Source: ContrsloriaConral de Is RepublIce., Informe Fianiel.es. Table 22: Revenuet - Expenditures, 1980-87

1984 1985 1986 1987 Revenues 1980 1981 1982 1983

Revenues

61,445 75,586 .. 171,770 Total .. .. 47,680 21,866 29,080 n.a. 51,258 1/ Taxes 6.979 9,348 12,457 17,048 53,720 n.a. 93,073 2/ 120.512 21 Commercial Operation n.a. n.a. 35,223 44.397

Expenditures 21,866 n.a. n.a. n.a. Total 6.979 9,348 12,457 17.048 11,196 .. Subsidy in Cash 3,278 4,664 6,204 8,828 358 n.a. n.a. n.a. Subsidy in Kind n.a. .. 152 246 4,021 n.a. n.a. n.e. Subsidy in Services n.a. 1,702 2.439 3.270 6,290 n.a. n.a. n.a. Other 31 n.a. 2,981 3,662 4,704

Notes:

1. Sixty-nine Family Funds. 2. Forty-Five Funds. 3. Administrative expenses.

Source: Manrique and Associates, Years 1980-84. Magazine Negocios (Business), No. 59, 1988. - 151 -

Table 23 Official Health Sector Costs Financing (Coll millions)

Sources 1980 1986 1988 1987 1988 1989 1990

1. National Budget 41,769 47,099 51,012 69,187 61,737 79,424 76,689 1.1 Operating Costs 30,107 38,422 40,284 46,633 48,694 60,127 56,013 - Earmarked revenues 15,770 28,862 27,470 37,091 38,526 43,716 43,364 - Other 14,337 9,669 12,814 8,542 10,169 18,412 11,859 1.2 Investment 11,662 8,677 10,728 13,634 13,042 19,297 21,876 1.2.1 Ordinary revenues 11,652 4,902 6,880 5,903 9,965 13,e69 16,710 1.2.2 External credits 0 3,774 4,848 7,631 3,078 6,e28 4,988 -FOCOEX I - 3,774 3,721 1,400 - 3,174 2,674 -IBRD - - 1,127 298 2,382 1,306 497 -Dresdner - - - 5,936 - - - -KfW - - - - 674 270 - -Dutch Agreement - - - - 21 - - -Providers(CSNS) - - - - - 878 1,00e -Equipment Rec. - - - - - 0 888

2. Other Earmarked taxes 28,916 36,199 41,281 43,004 48,970 49,427 49,344 2.1 Taxes 14,694 19,100 22,490 25,320 29,261 28,760 28,987 Liquor 3,487 3,950 6,926 9,444 10,830 10,648 10,732 Beer 3,852 5,303 6,846 6,406 7,141 7,021 7,076 Lottery & Welfare 7,339 6,182 6,706 6,307 6,732 6,618 6,671 Gambling - 1,437 1,2e8 1,446 1,737 1,709 1,722 Registry 18 3,248 2,74e 2,718 2,811 2,764 2,786 2.2 Dept., municipal. & others 1,e84 1,268 801 742 1,766 2,339 1,88e Dept. Assembly funds 1,390 915 709 6s5 ees 934 763 Municipal council funds - 102 11 29 1,008 1,332 1,074 Others 294 261 81 56 92 73 59 2.3 Sale of services 12,638 15,831 17,990 18,942 17,963 18,32e 18,472

3. Own rents of institutions 1/ 179 369 e61 309 421 634 389

4. Total Health 70,864 83,e66 92,908 102,480 111,129 129,386 128,4'J2

National Budget shows definitive appropriateness. Earmarked taxes are estimates. 1/ Fondo Nacional Hospitalario (FNH), Instituto Nacional de Cencerologia (INC), Instituto Nacional de Salud (INS). It doesn't include ICBF. - 152 -

Table 24 OfficialHealth Sector Cost Financing (Coll millions)

Sources 1980 1985 1988 198' 1988 1989 1990

1. Notional Budget 10,616 31,938 41,131 59,167 79,097 128,334 163,684 1.1 Operating Costs 7,651 26,612 32,481 45,633 62,387 97,164 110,247 - Earmarked revenues 8,971 19,233 22,149 37,091 49,369 70,636 86,882 - Other 3,610 6,379 10,332 8,642 '13,029 20,618 23,635 1.2 Investment 2,934 6,784 8,e60 13,634 16,710 3,,180 43,437 1.2.1 Ordinary revenues 2,984 3,268 4,741 6,903 12,767 22,087 33,486 1.2.2 External credits 0 2,616 3,909 7,631 3,943 9,093 9,951 -FOCOEX I - 2,616 3,000 1,400 - 5,129 5,169 -IBRD - - 909 296 3,052 2,111 995 -Dresdner - - - 6,936 - - - -KfW - - - - 864 437 - -Dutch Agreement - - - - 27 - - -Providers(CSNS) - - - - - 1,416 2,017 -EquipmentRec. ------1,790

2. Other Earmarked taxes 7,281 24,130 33,285 43,004 62,741 79,864 98,886 2.1 Taxes 3,700 12,732 18,134 26,320 37,477 46,471 58,089 Liquor 878 2,633 6,684 9,444 13,875 17,206 21,506 Boer 970 3,536 4,713 e,406 9,149 11,346 14,181 Lottery & Welfare 1,848 3,441 4,601 5,307 8,826 10,694 13,386 Gambling - 958 1,022 1,440 2,226 2,761 3,461 Registry 4 2,165 2,214 2,718 3,602 4,466 6,683 2.2 Dept., municipal. A others 424 846 646 742 2,262 3,779 3,779 Dept. Assembly funds 360 61o 572 e6s 862 1,609 1,509 Municipal council funds - 6s 9 29 1,292 2,162 2,152 Others 74 167 es 65 118 118 118 2.3 Sale of services 3,157 10,663 14,606 16,942 23,002 29,614 37,017

8. Own rents of institution 1/ 46 239 496 309 540 883 780

4. Total H1alth 17,841 55,766 74,912 102,480 142,378 209,061 253,849

NationalBudget shows definitiveappropriateness. Earmarked taxes are estimatcs. 1/ Fondo NacionalHospitalario (FNH), InstitutoNacional de Cencerologfa(INC), Instituto Nocionalde Salud (INS). It doesn't include ICBF.

Source: ComptrolltrGeneral's Office, Ministryof Health and NationalPlanning Department, Health Division. - 153 -

Table 26 OfficialHesith Sector Cost Finsncing PercentageVariation

Sources 1986/80 1986/85 1987/88 1988/87 1989/88 1990/69 1990/86

1. NntlonslBudget 2.4 8.8 18.0 4.3 28.7 (3.4) 10.7 1.1 Operating Costs 6.0 4.8 13.3 6.7 23.5 (8.6) 8.1 - Esrmarked revenues 12.8 (4.8) 36.0 3.9 13.6 (0.8) 12.1 - Other (7.8) 33.9 (33.3) 19.1 81.4 (29.0) (2.3) 1.2 Investment (6.7) 23.6 28.2 (3.8) 48.0 12.3 19.2 1.2.1 OrdinaryRevenues (16.9) 1S.9 0.4 e6.8 37.2 22.2 29.8 1.2.2 Externalcredits - 28.4 67.4 (69.7) 82.9 (11.8) 0.6 -FOCOEX I - (1.4) (62.4) - - (18.9) (8.8) -IBRD - - (73.7) 704.8 (46.2) (82.0) (18.5)

-Dresdner ------KfW - - - - (69.9) _

-DutchAgreement ------Providers (CSNS) - - - - - 14.9 _ -EquipmentRec. ------

2. Other Earmarkedtaxes 4.8 14.0 4.2 13.9 0.9 (0.2) 4.8 2.1 Taxes 6.4 17.8 12.6 15.6 (1.7) 0.8 6.6 Liquor 2.6 76.3 38.4 14.7 (1.7) 0.8 11.6 Beer 8.6 10.2 9.6 11.6 (1.7) 0.8 4.9 Lottery A Welfare (8.8) 10.6 (7.0) 28.9 (1.7) 0.8 4.0 Gambling - (11.8) 14.1 20.2 (1.7) 0.8 8.0 Registry 189.8 (16.6) (1.0) 3.4 (1.7) 0.8 0.4 2.2 Dept., municipal.A others (6.6) (36.8) (7.4) 137.9 32.5 (19.4) 28.9 Dept. Assemblyfunds (8.0) (22.5) (7.2) 1.1 40.4 (19.4) 1.5 Municipal council funds - (89.1) 169.8 3,377.3 32.1 (19.4) 213.2 Others (3.1) (67.8) (31.8) 67.5 (20.7) (19.4) (7.6) 2.3 Sale of services 4.8 13.6 (6.8) 8.0 2.1 0.8 0.7

3. Own rents of institution1/ 14.9 71.8 (49.8) 38.4 26.7 (27.1) (10.8)

4. Total Heolth 3.4 11.1 10.3 8.4 18.4 (2.3) 8.0

NationalBudget shows definitiveappropriateness. Earmarked taxes are estimates. 1/ Fondo NaclonalHospitalario (FNH), Instituto Nacional de Cencerologra(INC), Instituto Nacionalde Salud (INS). It doesn't includeICBF. - 154 -

Table 26 Official Health Sector Cost Financing Participation Percentage

Sources 1985/80 1986/85 1987/8e 1988/87 1989/88 1990/89 1990/86

1. National Budget 58.9 6e.3 64.9 67.7 66.6 61.4 60.7 1.1 Operating Costs 42.5 45.9 43.4 44.6 43.8 46.5 43.6 - Earmarked revenues 22.3 34.5 29.6 36.2 34.7 33.8 34.3 - Other 20.2 11.4 13.8 8.3 9.2 12.7 9.2 1.2 Investment 16.4 10.4 11.6 13.2 11.7 14.9 17.1 1.2.1 Ordinary Revenues 16.4 6.9 6.3 5.8 9.0 10.6 13.2 1.2.2 External credits 0.0 4.6 5.2 7.4 2.8 4.3 3.9 -FOCOEX I - 4.6 4.0 1.4 - 2.6 2.0 -IBRD - - 1.2 0.3 2.1 1.0 0.4 -Dresdner - - - 6.8 - - - -KfW - - - - 0.6 0.2 - -Dutch Agreement - - - - 0.0 - - -Providers (CSNS) - 0.7 0.8 -EquipmentRec. ------0.7

2. Other Earmarked taxes 40.8 43.3 44.4 42.0 44.1 38.2 39.0 2.1 Taxes 20.7 22.8 24.2 24.7 26.3 22.2 22.9 Liquor 4.9 4.7 7.6 9.2 9.7 8.2 8.6 Beer 6.4 6.3 6.3 6.3 6.4 5.4 6.6 Lottery & Welfare 10.4 6.2 6.1 6.2 6.1 6.1 5.3 Gambling - 1.7 1.4 1.4 1.6 1.3 1.4 Registry 16.9 3.9 3.0 2.7 2.6 2.1 2.2 2.2 Dept., municipal. A others 2.4 1.5 0.9 0.7 1.6 1.8 1.5 Dept. Assembly funds 2.0 1.1 0.8 0.6 0.8 0.7 0.6 Municipal council funds - 0.1 0.0 0.0 0.9 1.0 0.8 Others 0.4 0.3 0.1 0.1 0.1 0.1 0.0 2.3 Sale of services 17.7 18.9 19.4 18.6 16.2 14.2 14.6

3. Own rents of institution 1/ 0.3 0.4 0.7 0.3 0.4 0.4 0.3

4. Total Health 100.0 100.0 100.0 100.0 100.0 100.0 100.0

National Budget shows definitive appropriateness. Earmarked taxes are estimates. 1/ Fondo Nacional Hospitalario (FNH), Instituto Nacional de Cencerologfs (INC), Instituto Nacional de Salud (INS). It doesn't include ICBF. - 155 -

Table 27: Government Subsector 1978, 1985-1988 Bv Cost Categorv, Constant Values

1978 1985 1986 1987 1988#

(billions of constant 1987 ColS) Operating Costs: Salaries 46.2 54.0 57.1 61.8 72.3 Medicines 8.5 6.9 7.6 9.9 9.7 Food & Beverage 7.2 3.6 4.8 6.3 6.1 Others 5.5 24.2 26.5 28.0 15.2 Total Operating Costs 67.4 88.7 96.1 105.9 103.3

Investment: Construction & Equipment Purchases 13.3 8.0 10.8 NA NA Other4 4.8 19.4 21.7 NA NA Total Investment 18.1 27.4 32.4 34.3 39.6

Residual$ 21.9 (0.7) 1.6 0.0 0.0

Total Cost 107.4 115.4 130.0 140.2 143.0

Growth Rates 85/78 86/85 87/86 88/87 Avg.

(percent per annum) Operating Costs: Salaries 2.2% 5.8% 8.1% 17.1: Medicines -3.0% 9.9% 30.2% -2.3% Food & Beverage -9.8% 33.4% 30.1% -2.3% Other5 21.2% 9.8% 5.5% -45.7% Total Operating Costs 3.9% 8.3% 10.3% -2.4%

Investment: Construction & Equipment Purchases -7.3% 35.2% OtherQ 20.0% 11.4S Total Investment 6.0% 18.32 5.7% 15.6%

Total Cost 7.4% 12.72 7.8% 2.0%

& IncludingColombian institute of FamilyWelfare (ICBF). * -timated. * Table constructed by combining data on distributionof government health cars allocations (vhich give th siplit betweenoperating costs and investmuent) and Mpsnditure dta (which give wzp.editures in objectcategories). The 'Other" estimatesare the differences betveenthe stated object coat categoriesand the total of operating costs or investment. TSh 'residual" is the difference betweeu total allocations and spanditures. - 156 ^

Table 28: Health Expenditure,Source of Financing

PLAN DE ECONOMIASOCIAL, 1987-1990,CONSTANT VALUES

Sourceof Financing 1987 1988 1989 1990 1987- 1990

(billionsof constant 1987 ColS) Basic Health for All OrdinaryBudget 0 1 2 2 5 Sale of Services& 12 72 73 82 239 EarmarkedLocal Taxes 25 33 45 53 156 Total 37 106 120 137 400 OtherLevels of Care Ordinary.Budget 61 66 61 62 250 EarmarkedLocal Taxes 19 26 35 40 120 Total 80 92 96 102 370 SocialSecurity Health Programs PayrollTazes# 118 127 136 149 530 Total 235 325 352 388 1,300

?ate of Increase 88/87 89/88 90/89 90/87 p.a. p.a. p.a. four yrs

OrdinaryBudget 9.8% -6.0% 1.6% 4.9% Sale of Services& 517.9% 1.5% 11.3% 598.3% EarmarkedLocal Taxes 34.1S 35.6% 16.3% 111.42 PayrollTaxes# 7.6% 7.1% 9.6% 26.3% Total 38.6% 8.32 10.02 65.2:

PercentageDistribution 1987 1988 1989 1990 1987- 1990

OrdinaryBudget 26.0% 20.6% 17.9% 16.52C 19.6% Sale of Services' 5.0% 22.2% 20.8% 21.1% 18.4% EarmarkedLocal Taxes 18.7% 18.1% 22.7% 24.0% 21.2% PayrollTaxes# 50.3% 39.0% 38.6% 38.42 40.8% Total 100.0% 100.0% 100.0% 100.0% 100.0% souac: Plea d Zcoumia Social, Agoasto1987 and DNPworking documnts. & Inelua Paymnts by iss, generated ultimately by payroll taxes. Payroll taxrevenues used to finance health programs of social security organizatiou only. - 157 -

Table 29: Public Health Sector Financing, 1980-1988 Rates of Change, Constant Values

(rate of change - percent per annum)

85/80 86/85 87/86 88/87 avg.

GovernmentHealth Subsector:

NationalBudget -2.2 2.5 11.3 19.7

InitialRegular Appropriation -15.9 20.8 0.4 -30.8 AdditionalAppropriations -13.3 55.9 101.7 43.7 Subtotal -15.4 27.6 24.6 -2.0 Earmarked(Situado Fiscal) 10.9 -10.3 1.6 39.1

EarmarkedTaxes 11.0 9.1 6.2 2.4 [TotalWithout ICBF PayrollTax] -4.8 1.4 3.6 -1.8

LotteryShare -7.6 4.1 -7.3 -11.1 Other GamblingTaxes 438.4 -15.6 -14.0 -10.7 Liquor Tax -5.3 59.0 77.9 -7.1 Beer Tax 8.0 3.7 11.3 -9.0 Registration& Notary Taxes -17.7 -22.5 -0.6 ^S S ICBF PayrollTax NA 10.9 -4.8 13.3

Departamento,Municipal & Other -13.2 2.1 7.7 -2.4

DepartamentoAssembly Funds -4.3 4.6 -7.5 0.5 MunicipalCouncil Funds NA -89.5 159.0 -18.5 Other -443.3 NA -32.0 -28.4

Sale of Services -7.8 7.1 -6.1 -5.5

External Credits -15.4 5.0 73.8 -63.3

Carryover from Previous Year -28.2 221.9 -5.0 -22.6

Total Government Health Subsector -2.1 12.7 7.8 2.0 (TotalWithout ICBF PayrollTax] -6.6 13.2 10.9 -0.5 3 Social SecuritySubsector 10.1 -0.7 -4.2 -1.6 Total PublicHealth Sector 4.6 4.4 0.7 0.0 LTotalWithout ICBF PayrollTax] 3.0 3.8 1.2 -1.1

SO=CZ5: Departanto Plaaaacion Racional. World Bank. Realth Sector ReviW, Dec. tS, 1962. Note: See footnotes at Table 9. - 158 -

Table 30: Government Subsector 1980-1988, Bv Institution (percent)

1980 1985 1986 1987 1988'

Health ProgramsNot IncludingChild Care (ICBF) CurrentPrograms: Ministryof Health 12.3% 4.6Z 4.1% 7.2% 11.42

Departamento& Local Services 74.8% 90.5% 90.0% 88.0% 84.1% Hospitals 72.2% 67.8% Other 18.2% 22.2% NationalInstitute of Health 8.2% 2.4% 4.1% 2.5% 1.7:

MalariaErradication Service 4.6% 2.5% 1.9% 2.3% 2.82

Total CurrentPrograms 100.0% 100.0% 100.0% 100.0: 100.0%

HospitalInvestment: NationalHospital Fund 9.0% 0.4% 2.7% 9.6: 5.1%

Total Health Programs 109.0% 100.4% 102.7% 109.6% 105.1'

Health ProgramsIncluding Child Care (ICBF) CurrentPrograms: Ministryof Health 9.3% 3.8% 3.3% 5.4S 8.5:

Departamento& Local Services 56.6% 73.4% 72.4% 66.0% 62.3: Hospitals 58.62 54.6% Other 14.8% 17.9% NationalInstitute of Health 6.2% 1.9% 3.3% 1.9% 1.2%

MalariaErradication Service 3.5Z 2.0% 1.5% 1.7% 2.1%

Total CurrentPrograms 75.6% 81.2% 8U,.5% 75.0% 74.0%

HospitalInvestment: NationalHospital Fund 6.8% 0.3% 2.11 7.2% 3.8%

Total Health Programs 82.42 81.5% 82.61 82.2% 77.8% Child Care Colom. Inst. of FamilyWelfare 17.6% 18.5% 17.41 17.8% 22.2% Total Health and Child Care 100.0% 100.0% 100.01 100.0% 100.0%

SOURCES: D.partamautoPlafoecioa Nacional,World Bank,Health Sector Review, Doc. 15, 1982. # Estimated. - 159 -

Table 31: Amounts and Timing of Budget Appropriationsfor Investment

MINISTRY OF HEALTH AND SPECIALIZEDAGENCIESV, 1987-1988,CURRENTVALUES

Cumulative Cumulative Percent Amount Percent Unlegislated Budget Law Month (bil ColS) Percent Legislated Prior to Law

1987D

Basic Budget Law 6.8 30.6% 30.6% 100.0%

Decree 1350 12.1 54.5% 85.1% 69.4%

Law 29 0.3 1.4% 86.5% 14.9%

Law 51 3.0 13.5% 100.0% 13.5%

Total 22.2 100.0%

1988&

Basic Budget Law December 87 2.8 17.2% 17.2% 100.0%

Dscree 400 March 0.7 4.3% 21.5% 82.8%

Decree 1230 June 8.9 54.6% 76.1% 78.5%

Law 33 August 1.2 7.4% 83.4% 24.0%

Law 44 September 2.7 16.6% 100.0% 16.6%

Total 16.3 100.0%

SORCZ: DeWtamuto da Plameation Nacional. * ltntituto Rac. de 5alud, Yondo Nac. mospitalario, Servicio Errs. ds MaUlia * Also includes Iast. amiliar de Coloob. andINSFOPAL. & Does not include Inst. Familiar de Colob. and INSFOPAL. - 160 -

Table 32: EDUCATIONALENROLLMENTS, 1980-1988, PERCENTAGE DISTRIBUTION PRIMARY,SECONDARY AND HIGHER EDUCATION

1980 1981 1982 1983 1984 1985 1986 1987 1988

(percent) Primary

Rural 35.0 35.0 35.6 35.6 35.7 35.3 34.9 34.9 34.9 Urban 65.0 65.0 64.4 64.4 64.3 64.7 65.1 65.1 65.' Public 86.0 86.2 86.6 86.5 86.3 85.5 84.8 84.8 84.3 Private 14.0 13.8 13.4 13.5 13.7 14.5 15.2 15.2 15.2 0otal 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Secondary

Rural 4.7 5.0 5.8 5.8 5.9 6.0 6.1 6.2 6.2 Urban 95.3 95.0 94.2 94.2 94.1 94.0 93.9 93.8 93.8 Public 57.6 58.1 59.9 60.5 61.0 61.3 61.6 62.1 62.4 Private 42.4 41.9 40.1 39.5 39.0 -38.7 38.4 37.c 37.o Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

HigherEducation

Public 37.1 40.0 39.1 39.6 39.2 40.9 40.9 41.6 40.3 Private 62.9 60.0 60.9 60.4 60.8 59.1 59.1 58.4 59.7 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

5OUCl: T?14 IS. - 161 -

Table 33: EDUCATIONALENROLLMENTS AND ENROLLMENTRATES, 1980-1988 PRIMARY,SECONDARY A.D HIGHER EDUCATION

1980 1981 1982 1983 1984 1985 1986 1987 1988

Primary (millions)

Rural 1.4 1.4 1.4 1.4 1.4 1.4 1.4 1.4 1.5 Urban 2.7 2.6 2.5 2.6 2.6 2.6 2.6 2.7 2.7

Public 3.5 3.5 3.4 3.4 3.5 3.4 3.4 3.5 3.5 Private 0.6 0.6 0.5 0.5 0.5 0.6 0.6 0.6 0.6

Total 4.1 4.1 3.9 4.0 4.0 4.0 4.0 4.1 4.2 EnrollmentRate (Per 100 Population7-11) Rural 101 101 100 103 104 105 105 106 110 Urban 130 128 122 123 123 124 125 126 127, Total 118 117 114 115 116 116 117 118 120

OverageRate 29 30 30 31 31 . NA 31 NA NA Secondary (thousands)

Rural 82 88 113 116 121 126 130 135 139 Urban 1.S1 1,680 1,835 1,879 1,932 1.967 2,006 2,053 2,096

Public e^O 1,027 1,167 1,207 1,253 1,284 1,317 1,358 1,396 Private 735 741 780 788 800 809 820 830 840

Total 1,733 1,768 1,947 1,996 2,053 2,093 2,136 2,188 2,235 EnrollmentRate (Per 100 Population12-18) Rural 5.1 5.3 6.7 6.7 6.9 7.0 7.1 7.2 7.2 Urban 67 67 73 73 74 75 75 76 76 Total 43 43 46 47 47 47 47 48 48 Higher Education (thousands)

Public 101 127 132 140 149 159 165 181 183 Private 171 191 205 215 230 230 238 254 271 lotal 272 318 337 355 379 389 402 435 454 EnrollmentRate (Per 100 Population18-25) 6.2 7.0 7.1 7.3 7.5 7.4 7.4 7.8 7.8

SO*C: DOtcpaexnto de Plan*&cion N4acional. - 162 -

Table 34: EDUCATIONALENROLTTNTS AND ENROLLMENT RATES, 1980-1988, RATESOF CHANGE

PRIMARY,SECONDARY AND HIGHER EDUCATION

81/80 82/81 83/82 84/83 85/84 86/85 87/86 88/87 88/80 (percentper annum) (% 8 yrs)

Primary

Rural -1.0 -1.8 1.4 1.0 -0.6 -0.6 1.7 1.1 1.0 Urban -1.0 -4.1 1.3 0.5 1.1 1.1 1.6 1.1 1.5 Public -0.7 -2.9 1.3 0.4 -0.4 -0.4 1.7 1.1 0.1 Private -2.5 -5.7 1.8 2.1 6.2 5.8 1.3 0.9 9.8

Total -1.0 -3.3 1.4 0.7 0.5 0.5 1.6 1.1 1.4

Enrollment Rate . (Per 100 Population7-11) Rural -0.1 -1.0 2.2 1.8 0.2 0.2 1.1 3.5 8.2 Urban -1.5 -4.6 0.8 0.0 0.6 0.6 1.1 0.6 -2.5 Total -0.9 -3.2 1.4 0.7 0.5 0.5 1.1 1.7 1.6 Secondary

Rural 7.6 28.0 3.4 4.4 3.5 3.6 3.4 3.3 70.3 Urbaa 1.7 9.2 2.4 2.8 1.8 2.0 2.3 2.1 26.9

Public 2.9 13.7 3.4 . 3.8 2.4 2.6 3.1 2.8 39.9 Private 0.8 5.3 1.1 1.5 1.2 1.2 1.2 1.2 14.2

Total 2.0 10.1 2.5 2.9 1.9 2.1 2.4 2.2 29.0

EnrollmentRate (Per 100 Population12-18) Rural 5.3 25.2 1.2 2.1 1.3 1.3 1.2 1.1 43.0 Urban 0.4 7.7 1.0 1.4 0.5 0.6 1.0 0.7 13.9 Total 0.3 8.3 0.8 1.2 0.2 0.3 0.7 0.5 12.6 Higher Education

Public 26.4 3.5 6.5 5.9 7.1 3.3 10.0 1.2 81.6 Private 11.8 7.4 4.6 7.3 -0.2 3.5 6.6 6.7 58.4 Total 17.2 5.9 5.4 6.7 2.7 3.4 8.0 4.4 67.0

EnrollmentRate (Per 100 Population18-25) 13.2 2.3 1.8 3.2 -0.8 0.0 4.4 0.9 27.1

SOURCE:Table 15. - 163 -

Table 35: GOVERNMNT EXPENDITURESON EDUCATION, 1980-1988,CONSTANT VALUES

(billionsof 1987 ColS)

1980 1981 1982 1983 1984 1985 1986 1987 1988

RECURRENT

Primary 73.9 81.2 87.7 88.0 89.2 31.1 80.0 82.6 81.1 Secondary0 41.9 46.4 50.1 49.9 56.8 55.5 60.4 64.4 64.0 Universities 37.6 39.0 40.3. 42.8 45.3 40.6 42.1 42.9 43.o Other 7.0 13.5 13.2 19.6 25.9 23.8 25.5 13.7 17.5

Total 160.3 180.0 191.3 200.2 217.2 201.0 207.9 203.6 203.1

INVESTMENT

Primary 1.7 1.8 0.7 0.9 0.5 1.0 1.6 3.0 3.8 Secondary 0 3.0 2.7 1.0 2.1 5.7 3.0 1.6 1.0 0.' Universities 1.8 0.6 2.0 1.0 2.5 2.6 3.3 4.0 5.1 Other 6.7 4.5 3.3 3.5 8.6 8.2 7.4 6.6 3.8

Total 13.2 9.7 7.0 7.5 17.3 14.7 13.9 14.6 13.1

TOTAL

Primary 75.6 83.0 88.4 88.9 89.8 82.1 81.7 85.6 84.8 Secondary# 44.9 49.1 51.1 52.0 62.5 58.5 62.0 65.4 64.4 Universities 39.4 39.6 42.3 . 43.8 47.8 43.2 45.4 47.0 45.7 Other 13.7 18.0 16.5 23.1 34.5 32.0 32.9 20.3 21.3

Grand Total 173.5 189.7 198.3 207.8 234.5 215.8 221.9 218.2 216.3

souacz; DepartamentoNacional de Planeacion. e Includes national schools. - 164 -

Table 36: EDUCATIONEXPENDITURE TARGETS

PLAN DE ECONOMIA SOCIAL, 1987-1990, ANNkALIZED#

(billionsof constant 1987 ColS)

ExpenditureTargets 1987 1988 1989 1990 1987- 1990

RECURPRENT#

Basic Education for All 89 99 110 123 Secondary & Vocational3 70 80 93 107 35 Higher 46 50 55 60 210 Science & Technology * * * * 1 Culture, Sports & Recreation 5 5 5 .5 20 Subtotal MEN 210 235 263 295 1,302 SENA 5 5 6 6 22 Total 215 240 269 3C1 1,024

INVESTMENT

Basic Education for All 6 13 11 8 347 Secondary& VocationalQ 2 0 0 0 3 Higher 13 14 7 4 39 Science & Technology 5 4 5 5 ' Culture, Sports & Recreation 3 3 3 3 11 Subtotal MEN 29 35 26 21 113 SENA 19 20 22 24 85 Total 48 55 48 44 195

TOTAL

Basic Education for All 95 112 121 130 458 Secondary & Vocational8 71 81 93 108 353 Higher 59 64 62 64 249 Science & Technology 5 5 5 6 21 Culture, Sports & Recreation 8 8 8 8 31 Subtotal MEN 238 269 289 315 1,112 SENA 24 26 28 30 107 Grand Total 263 295 316 345 1,219

SOURCE:Plan de EconomiaSocial, Departamento Nacional de Planeation, Aug.- 1987. # Annual recurrent targets for the three levels of formal education were appr. i5ated year by year by calculating the constant grovth rates in recurrentexpenditures from 1986 that would result in tbh four-yeartargets being achieved. The resulting annual targets are added to the publishedannual investmenttargets to get the figures presented here. (Averageannual recurren: expendituresvera used for the recurrenttargets of other two MEN categories,and all SENA's t rgts ware distributed annuallyusing the average1987-1990 MEN growth rate.) I lacludesnational schools. Less than 1. 165 -

Table 37: EDUCATIONEXPENDITURE TARGETS, PERCENTAGEDISTRIBUTION AND RATES OF CHANGE

PLAN DE ECONOMIA SOCIAL, 1987-1990, ANNUALIZED#

PERCENTAGEDISTRIBUTION 1987 1988 1989 1990 198-- 1990

TOTAL EXPENDITURE

Basic Education for All 40% 41% 42% 41% - 41% Secondary& VocationalQ 30% 30% 32% 34% 32. Higher 25Z 24% 21% 20% 227% Science& Technology 2% 22 2% 2% 2%. Culture,Sports & Recreation 3% 32 3% 2% 3%

Subtotal KEN 100% 100% 100% 100% 1vC7e SENA 10% 10% 10% 9% '32 Grand Total 110i 110% 110% 109% 110%

RATES OF CHANGE 87/86 88/87 89/88 90/89 90/86 (of constan'ColS magnitudes) p.a. p.a. p.a. p.a. 4 vrs

TOTALEXPENDITURE BasicEducation for All 16.0% 18.0% 8.2% 8.0% 59.82 Secondary& Vocational$ 15.3% 13.1% 15.3% 15.3% 73.4% Higher 30.4% 8.4% -3.9% 3.7% 40.9%

Subtotal 19.2% 13.9% 7.4% 9.5% 59.71%

Science& Technology -13.6% 12.92 8.5% Culture,Sports & Recreation - 2.0% -1.7% -2.62 SubtotalIEN 12.9% 7.2Z 9.2% SENA 7.2% 7.22 7.2%

Grand Total 12.4% 7.2% 9.0%

SOUICZTabl and Table for 19S6 expenditure dta used in rate of change ealculations. Footets in Source. $ Includes national schools. - 1E6 -

Table 38: PRIMARY EDUCATION PHYSICAL AND FINANCIAL INVESTMENTTARGETS

PLAN DE ECONOMIA SOCIAL, 1987-1990

1987- Annual 1987 1988 1989 1990 1990 Average

PHYSICAL TARGETS (units as indicated) RURAL PRIMARY SCHOOLS Texts ('000 Schools) 14.0 15.7 6.2 7.43 3.4 10.8 Libraries ('000 Schools) '5.0 7.0 - - 21.0 5.3 Furniture ('000 Schools) 3.5 7.5 7.7 - 18.7 h , ClassroomsUpgraded ('000) 1.8 4.0 4.0 4.0 13.9 3.5 Teachers Trained ('000) 6.8 20.5 13.7 7.4 '-8.4 12.1

URBANPRIMARY SCHOOLS Texts (million) 0.6 0.7 1.2 1.7 4.1 1.0 Schools Upgraded ('000) - 1.2 0.6 0.6 2.4 0.6 Teachers Trained ('000) 10.0 21.0 21.0 - 52.0 13.0

CORRESPONDING.FIANCIALTARGETS (billionsof constant 1987 ColS) RURAL PRIMARY SCHOOLS Texts 1.1 0.8 0.3 - 2.2 0.6 Libraries 0.5 0.4 - - 0.9 0.2 Furniture 1.0 3.5 2.8 - 7.3 1.8 ClassroomsUpgraded 1.6 5.0 5.0 5.0 16.7 4.2 Teachers Trained 0.2 0.6 0.4 0.2 1.5 0.4

Total Rural 4.4 10.4 8.5 5.2 28.6 7.1

URBAN PRIMARY SCHOOLS Texts 0.3 0.3 0.6 0.8 2.1 0.5 Schools Upgraded - 0.8 0.4 0.4 1.7 0.4 Teachers Trained 0.3 0.1 0.1 - 0.4 0.1

Total Urban 0.6 1.2 1.1 1.3 4.2 1.0

Total Rural & Urban Primary 5.0 11.7 9.6 6.5 32.8 8.2

CONTINUING& ETHNO EDUCATION 0.6 0.9 0.9 0.9 3.4 0.8

TOTAL 5.6 12.6 10.5 7. 36.2 9.0

SOURC£. Plan de EconomiaSocia:, DepartamentoNacional de Planeation,August 1987. 4 Physicaltarget not matched by a financial target in this year. - 167 -

Table 39: PUBLIC EDUCATION BUDGET, BY COST CATEGORY, 1981-19883

CURRENT VALUES 1981 1982 1983 1984 1985 1Q86 1987 1988

(billionsof current ColS)

Personal Services 38.0 52.0 60.2 69.9 76.0 95.5 113.4 '92.3 Other Expense 1.1 1.2 1.6. 1.3 1.3 1.6 '.9 1.3 Contributions 6.5 8.8 9.0 11.1 12.9 15.5 18.3 27.8 Total 45.6 61.9 70.9 82.3 90.2 112.6 133.6 221.4

PercentageDistribution 1981 1982 1983 1984 1985 1986 1987 '.83

Personal Services 97.2% 97.8% 97.4% 98.2% 98.3% 98.4Z 98.3% 99.3% Other Expense 2.8% 2.2% 2.6% 1.8% 1.7% 1.6% 1. 0O.7 Subtotal 100.0% 100.0% 100.0% 100.0% 100.0Z 100.0% 10O.C.% 00C. Contributions 16.8% 16.6% 14.6% 15.7Z 16.6% 16.0% 15.94 ' Total 116.8% 116.6% 114.6% 115.7% 116.6% 116.0% 115.9% 1% .4%

CONSTANT VALUES 1981 1982 1983 1984 1985 1986 1987 1&88

(billionsof constant 1987 ColS)

Persrnal Services 139.0 152.5 146.7 139.4 121.4 118.9 113.4 151.4 Other Expense 4.0 3.4 4.0 2.5 2.0 1.9 1.9 1.1 Contr;'tions 24.0 25.9 22.0 22.2 20.5 19.3 18.3 21.9 Total 166.9 181.8 172.8 164.1 144.0 140.1 133.6 174.4

Rate of Change 82/81 83/82 84/83 85/84 86/85 87/86 88/87

(percent per annum)

Personal Services 9.7 -3.8 -5.0 -12.9 -2.1 -4.6 33.5 Other Expense -14.1 17.4 -36.9 -18.8 -5.5 0.9 -45.9 Subtotal 9.0 -3.3 -5.9 -13.0 -2.2 -4.5 32.2 Contributions 7.8 -14.7 0.8 -7.6 -6.0 -5.2 19.8 Total 8.9 -5.0 -5.0 -12.2 -2.7 -4.6 30.5

SOURCE:Departanento Nacional de Planeacion. I Temporaryfootnote: This sarias is not consistentin total and in the 1987/1988 rate of increasevith other expenditure data presented. A consistentseries should be obtained from the Government,i9 possible. - 168 -

Table 40: BANK-ASSISTEDPRIMARY EDUCATION INVES XNT PROJECT

FINANCING PLAN, CALENDARYEARS 1989-1995

1989 1990 1991 1992 199, 1994 1995 "ota' Average

(USS millions)

Government 3.0 4.4 7.3 9.0 10.6 11.3 11.0 56.6 3.

Departamentos 0.0 0.2 0.3 0.4 0.5 0.6 0.5 2.5 a.4

Municipalities 0.0 0.8 1.4 1.7 2.0 2.1 2.1 10.1 1.4

World Bank 4.6 8.1 13.0 16.0 18.8 19.9 19.6 100.0 14.3

Total (Current USS 7.6 13.5 22.0 27.1 31.9 33.9 33.2 169.2 24.2

(1987 US$)8 6.9 11.7 18.1 21.2 23.8 24.1 22.5 128.3 18.3

(ColS billions) (1987 ColS)* 1.7 2.8 4.4 5.2 5.8 5.8 5.5 31.1 4.4

SClRC: StaffAppraisal Report, Colombia: Second Subsector Project for PrimaryZducaticc, ktveber Z, 1988,Annex 6, Page 3 and missionestimates. 0 F& chis amalyuis, assums US$ inflation of 5S per annum. f Convertedat the average ebhangerate for 1987of 242.61 ColS/USS. - 160 -

Table 41 POWER SEC'TOR - STATISTICS SECTOR 1987

Utility Number of Sales per Number of Losses Aveersg.. Customers Customer Employees k.988) Trriff (OQOs) (kWh/yr) (%) (Colt/kWh)

1. Empress do Energa El6ctrice de Bogota (EEEb) 996 5832 3771 25.0 13.27 2. Empress Publicas do MYdellfn (EPM) 462 929t 2350 20.1 7.98 3. Corporeci6ndel Valle del Cauca (CVC) S51 6050 942 21.0 9.13 4. Empress*Municipoles d- Cali (EMCALI) S. Corporeci6nAutonoma Regional de Guindlo 20 2096 S7 8.61 6. EmpresesPublics de Pereira 59 4711 152 7.33 7. Er-proesPublicas de Armenia 38 3425 70 8.45 8. Empress Publics do Cirtago 21 3235 6.28

ICEL Group

9. Empr-eo Antioquenrde Energfa 229 3883 998 6.07 10. Electrificadoradel Boyace 144 3870 829 9.64 11. Central Hidroel6ctricade Caldao 176 4028 1199 6.98 12. Electrificadoredel Caqueta 16 2871 as 9.17 13. Central.sElectriced del Cauca 74 2747 587 7.98 14. Electrificadoredo Cundinamerca 90 4361 520 10.76 15. Electrificadoradel Choco 14 4068 87 6.48 10. Electrificadoradel Huil 99 2489 384 8.16 17. Eloctrificadoradel Mote 47 3414 105 9.80 18. CentralesE1Ictrices de Marino le 2390 S15 7.30 19. CentrelesEl6ctricas del Norte do Santander 134 3616 402 7.59 20. Electrificadora de Santander 217 2692 1015 9.68 21. Electrificadoredel Tolima 137 3122 964 7.96

Total ICEL Group 1so8 3356 7870 25.1 8.06 - 170 - Cont. Table

Ut;lity Number of Soles per Number of Losse. Average Customers Customer Employees (1986) Tariff (000S) (kWh/yr) (S) (Co ll/kWh)

CORELCA Group

22. Electrificadoradel Atlintico 213 $692 1223 11.09 23. Electriflcadorado Bol var 132 8309 791 10.36 24. Energra El6ctricsde Maganque 18 2690 92 25. ilectr;ficadorsdel Cesar 56 2752 227 10.71 76. Electrificadorade Ctidobs 80 3056 476 9.98 27. Etectrificadorade la Guusira 33 3423 236 9.19 28. ElectrifIcadoradel Wagdal.ns 81 3611 368 11.04 29. Electrificadorade Sucre 6o 3123 303 9.48 30. Electrificadorade Son Andres y Provincis a 9319 182 10.50

Total CORELCA Group 658 6906 3888 23.6

31. Interconexi6nEl1et-ica SA. (ISA) 6 1273 3.0 32. Central Hidroel6ctricaBetanis

Total (Conso1idated) 4108 5314 20173 24.5 10.02

Source. ISA: Sstema ElectricoColombiare. Balance EnergiticoWstSoro 197.-198' 08-E-21,Medellfn, Octubre 1988. - 171 -

Table 42: MacroeconomicVariables Growth Rates (%) in Constant Values

Pi8 PR1 VA VA VA INVERSION PIS AX0 TOTAL URSANO CONY SERV OFICAL INDUSTRIA INDUSTRIALPERCAPITA (KILL/$75)

1997 5.3 4.7 4.6 4.7 6.1 22.4 1958 4.6 4.2 4.3 4.9 4.0 8.8 3.06 1989 5.2 5.1 4.5 . 7.0 4.5 5.7 t.63 1990 4.9 4.9 4.4 5.0 5.3 7.4 3.37 1991 4.7 4.9 4.6 4.6 5.1 7.5 3.20 1992 4.6 4.8 4.5 4.5 5.2 7.3 3.12 1993 4.6 4.8 4.5 4.3 5.3 7.2 3.07 1994 4.6 4.7 4.3 4.4 5.3 7.8 3.07 1995 4.6 4.7 4.5 4.6 4.9 6.3 3.06 1996 4.6 4.7 4.5 4.4 5.1 7.5 3.05 1997 4.6 4.8 4.6 4.5 5.2 7.7 3.01 1998 4.6 4.8 4.7 4.6 4.9 6.4 3.05 1999 4.6 4.1 4.7 4.1 S.1 6.4 3.09 2000 4.6 4.7 4.7 4.1 S.1 6.4 3.10 2000-20074.6 4.7 4.7 4.1 5.1 6.4 3.10

PSA DIE ECE

1997-2000 4.66 4.'6 4.52 4.70 5.01 7.12 3.15 1997-1990 4.90 4.71 4.40 5.63 4.62 7.31 3.35 1990-19954.60 4.80 4.48 4.49 5.18 7.22 3.10 1995-20004.60 4.76 4.63 4.35 5.07 6.90 3.07 - 172 -

Table 43: Tariff Growth Rates

ES CEN ARIQ 1: SE SUN JO rT fESCR2 ESCARto I a E S I a t N C I A L INDUSTRIACaN-SERY RESIDENCIAL1002C?Pt EN _..IllKM.Mll AC1"K EEE 1998- INDUSTRIA EEEIi EEPPNCVC tICEaI COReCA m.CKAL

1986 1.45 6.0 1.34 2.8 1.31 0.7 2.9 1.45 0.97 119 1.91 5.0 2.70 2.51 1.4 2.72 - 3.25 (2.00) 1990 1.04 4.0 1.80 1.62 1.24 1.46 (1.011 3.25 (2.001 199 0.26 4.0 1.61 1.43 1.92 1.21 t1.o1l 31.5 (2.00) 19,2 0.31 4.0 1.51 1.39 1.34 M0.10) (1.011 3.25 (2.00) 1993 0.41 4.0 0.53 146 1.22 (0.10) (1.01) 3.25 (2.001 19q4 0.52 4.0 0.53 1IS 0.65 0.50 (1.01) 3.25 (2.00) 1995 0.94 4.0 0.53 1.16 0.65 0.40 (1.01) 3.25 (2.OQ) 1996 1.00 4.0 0.58 1.19 0.65 - (1.011 3.25 (2.001 1997 t.09 4.0 0.60 1.22 0.14 - (1.01l 3.25 0.50 1g99g t.T 4.0 0.60 1.26 0.45 0.60 (1.01) 3.25 0.20 1919 1.17 4.0 0.40 1.27 0.15 0.60 (1.01) 3.25 Ot5 2000-201 0.S - - 0.50 0.50 0.60 (1.01) 3.25 0.75

ESARIO 4: 02de crucimimnto - 173 -

Table 44: Growth Rates of Prices for ElectricitySubstitutes

ESCENARItIsSE6UN POLITICA 6LOSAL NINNINAS [SCENARIO2'

I NUS TRI AL RESIODNCIALDERIVADOS DERIVADOSGN6AS I CARBON I RESIOU0S I PETROLED PETRUOLEOATURAL HIDROEN.

19" (1.39) 1.71 - - (1.39) I,, - - - - 0.5 19M0 2.4 - 3.3 - 0.5 2.4 19m 3.0 2.5 2.4 - 0.5 3.5 1992 3.0 2.5 2.4 - 0.5 3.5 1993 3.0 2.5 2.4 - 0.5 3.5 1994 3.0 2.5 2.4 - 0.5 3.5 1995 3.0 2.5 2.4 0.5 3.5 19"9 3.0 2.5 2.4 - 0.5 6.0 19m 3.0 2.5 2.4 - 0.5 6.0 19, 3.0 2.5 2.5 - 0.5 6.0 199 3.0 2.5 2.4 - 0.5 6.0 2000 3.0 2.5 2.4 - 0.5 6.0 2000-20073.0 2.5 2.4 0.5 2.5

ESSCENARIO3s Los precios nocrucm tn tirsinos rsalus. - 174 -

Table 4R Demographic Variables

POBLACIONPERSONAS VIVIENDAS POBLACION PODURB/POB TOTAL VIVIENOA URBANA /TOTAL

1985 29.282.739 5.14 5.692.729.19.723.213 67.4

1990 32.040.686 4.91 6.529.754 22.181.335 69.2

1995 35.030.045 4.71 7.444.315 24.925.650 71.1

2000 38.110.571 4.53 8.410.445 27.872.195 73.1

2005 41.258.553 4.39 9.421.230 31.014.131 75.2 Table 46: Electric Service Coverage Targets

ESCENARI0 1 ESCENARIO 2

META A PREVISTO CUBRIMIENTO META A PREVISTO CUBRIMIENTO Ar;0 2000 ALCANZAR Ar"0 1995 Ar;0 2000 ALCANZAR AtnO 1995 A;iO 2007 AEI0 2007 83.97. 87.5% EEEB 907. 82.4'. 84.7% 10O. 94.4% 94.8-. EEPPM 95% 94.4% 94.86. 95% 89.5% 93.37. CYC 90% 86.3% 88.1% 100% 69.3% 78.2% t ICEL 75% 64.27% 68.6% 100% 67.8-. 74.7'. CORELCA 80% 65.3% 69.8% 1007. 75.4%' 82.0% NACIONAL 72.1% 75.7% 100% - 176 -

Table 47: Scenariosfor ResidentialConsumers Growth

METADIFERENCIAL METAl002

MRO INCREN. TAs DECREC NRO IKCRE. TASADE CREC

1987 3763008 ' 3743008 1968 3958974 195966 5.21 3959874 195864 5.21 1999 4155531 196457 4.97 4179731 220857 5.58 1990 4353614 190283 4.77 4404344 226614 5.42 1991 4553495 199681 4.59 4638440 232095 5.27 1m 4754393 200898 4.41 4075732 237292 5.12 1993 4954373 201990 4.25 5117933 242202 4.97 1994 5159346 202973 4.10 5344759 24u824 4.82 195 5363266 203920 3.95 5615934 251174 4.49 199 5U52654 199390 3.72 5945427 249493 4.44 1997 5762414 199960 3.59 6118620 252933 4.31 1998 5943205 200589 3.48 4374202 255942 4.18 1999 6164506 201302 3.38 4433047 258845 4.06 2000 6366628 202122 3.2 6894619 261572 3.94 2001 6543237 194609 3.09 7151730 257111 3.73 2002 6760539 197302 3.01 7410784 259055 3.62 2003 4958@49 198145 2.93 7671691 260905 3.52 2004 7157831 199147 2.84 7934381 262691 3.42 2005 7358144 200314 2.80 8198921 264439 3.33 2006 .7552353 194209 2.64 8456468 257847 3.14 2007 7747676 195323 2.59 8715750 259083 3.06 - 177 -

Table 48: Growth Rate Scenarios for CommercialConsumers

4. % - %

19S7 26l3824 6S 3324 '268:4 l9e6 2796Sa 2768S9 2742'_100 1 C's 290928 28 195 279684 199o:s .3:2657 29757i1 :685278 1991 ---t-'4850 :302564 290994

1q996 3-75Z311641 2968c--7 1 99_'4 07'9 321 )99 0 (0:2739 1994 -54471 -.rTi62f.) 0'18794 1 .5,WS '756 1314 5 874a 314970 1994 -'S67617 :5)754 : 1 27 0 1997 s..99077 - 61 277 Z27695 1993 415160) 72115 -`54249 1979 4-Z1892 B:6279 340:iw`4 Z0:00 4 4,; -t 744r .>4777 3 477r5S 20)o1 467404 406620o 3 47.06

436241 416819 61 :2 -. Cf)t1_505837 470 1384 :69(C)'

2-C( 4 262222 44 4._25 -76416 547429 457655 -18-747 569491 471 Z85 *391Uofb

2-.'{. - 59 442 4855Z6 994, - 178 -

Table 49: Recovery of Losses, Own Consumption, and Sector Losses (%)

RECUP. FPER. 5ONSUMO PERDIDAS NEGRAS FROP}O

'099' l)C.15 2.0)0) 2Z.97 1: (;0.41 1.81 24.52 1969 1.40 41 .86 25. 06 199) 2.20 1.84 21.84 1991, .0C0C 1.83 2' ) . 7- 1992 7 . 7'C: 1.89 1 9 . 62 i997 4. 10) 1.88 19.20: 1994 4.40 1.86 18. 9 1995 4.70 1.82 18.60 1996 5. 00 1.76 18.=2 1997 5.20 1.93 18.14 1996 5 .40 1.95 17.90.' 1999 5.6b0 11.87 17.6 20(u:)0 5.80 1 . 2 17.55 '0: l., S6.{0 1.79 17.41 -2C,2 6. 1 C0) 1.77 1 7 . 3:0 :0073 6.20 1.76 17.29 2'004 6. 20:) 1..76 1 7. 2 WX:10S 6.2( 1.76 17.23 2006 6.20 1.76 17.23 2:0(0'7 6.2X. 16 17.2' t'016 WI- lt* KrY U-e Y;IS U' w, It-t hie woz-Ow@ InI 11tlt S 11t--6 (1 . 7s ., 19 II Ot-566 K- 1-1- IK S0UIl *I-S 911 S6-S 05-9 OYY 001 SCl W-OSSI

33333R'1s! 34998 Sf101 9111 I $tilt UtS1 s1mIC lll a16s til (e3 001 S3Itl 610u0 11011 testl lIlt 91199 liii 19919 liii 1191 (III 00eW113 60Q "It 9O00 69131WT 10115 1111 9301o {113 61931 5131 t5 un 61 110U91101S 11 ot1001 10301 K111 9011 Slttl e61t S6ZUS lll061 ^SI 919 titsIfSI ll £001 61311 £19ti Ull 'tttS l0it 9ZO 1911 lOll tits Ul5 ¶930 I11t 9001 Ottet ;sliw t1ll SKIS It8t oo 01o1 f lS6 at lI sttil *19II too Mort 1P10011 19191 IllInst5 Ol fill Wot1 tilt UP11 90t1 stll 66 06901 06l5l 9101 91l65 eU01 0161 1161 19901 11,11 69 5911 lOl 11111 01931 6661 loll 96t05 tSU 11131 El, loll 1911 0Lilt SZSII (tIl t66t 01oot 11st ll601 10111 11 16l1t 3911 tOl8 36(t lots £16 *I1e9 1091 11St? UU1 1o1 3910 9tl 1sUt U991 U01P t5t £0t ttlI 61te11*UI 09t11 1tn0 631 31111 511t 8611 0t1l 616 61I toll 3103 911i0 5661 06I6K lOll "11 lint Mt lt 19161 111 I91 1901 Uh 1SU 000S0 1661 9111110911 1Il9111 1996 10t £1511161t1 0011131l 163199101 I1hIll 1111t IUt Its IUU Iti1 11tol 1016silt 1191°3t giefIfni 161111t6ll 0661t16 611t 891 659tt91t 10101lilt( fill 0I lilt UUf OSI 9IUtl IUI t9h1 ICON N1 itilt lo3l 160 SeIs IIOt OSU 616l1 6WI lit 19fic 59tl its iseU osst site lstil eu1

0011ttomi l1g191l SP5991 S OSt tuf t Itot t9013161split tS91 5sllfC19 loftsoot 911istt91 tilt58)9 Efilt3111 O63161 W61U 690 III 01911t to S0t ns tot t11 Loft OIlY 19I0 1t61

13101 OlniJ flUIDl 5913114*11110335 39133 eal03nd 3O3 310 s9ine33a s ) 111 55334 '3Al331 SWI03A Sl9 WOlnW W131A W539303Pl 19wl3sni5 l 1W1311538

3SeDaoId puemaa a ua.aJaa :OS alqel Table 51: DemaindForecast Sensitivity Studies

suscIiel. SUSCSI?IOES F6fa11i oraos Ca'aCIcs iMSfIhll8Aii,1 4153506(51 SkI4lCOMfMIcOS PFECIOS 06li5s10(N. CoL51A1ES

Is 12 ll If LSaI.5is UF O 3 4 S A 1 I I 10

* SUScliP1loES ,6SIf a tD l 0t * * * U 1.251 oc 3.;f Imlfi L.EL JiN * 50 3.251 1 1 D.1011

a . . .3 REcIo SUSL S 0.51 0 I

mEciasusv.tmX FtaH * = * * *.*.. * 1 21 * 21 21 SUiSCilF.ttlO 41 * a *

3 3 * VA4LOECO. AMP lot -. 51t IP *.51 w -.51 * 3 3 3 U 1991-2001 efIF 2 *' tlODIm of nu I * *** **

-- FE CISIIS FE CIS KS CIS -- -- FEO ------o1905 -- F-.

29493 29493 '29493 29493 29493 11M 29493 29493 29493 29493 29493 29413 29493 29493 29493 31149 35491 31143 31521 31315 31233 1990 31403 341 3154s56 356144 1 35ulWl59 544 31949 44311 6141 44418 4329, 41511 414814 1191 44430 41841 41408 44512 44349 44894 41318 48040 s9l95 40221 4005 51605 54190 2i30 t0ils 58193 42505 59110 44088 40841 4ills 43493 40026 14414 1039 2005 78349 14843 52049 14211 11424 11221 19341 82999 11229 14244 11114 71142

SiEW8(KIAS (RiSE[SU~NM305 luf-I

0 0 0 a a 0 0 lie? a 0 0 0 0 a 54 104 -140 10 225 310 8990 152 -151 0 42 -1II -143 -344 4P 12 Sal lOll 1114 1995 14 -118 98 241 .44 -528 -1430 253 132)1 38 s4t 2200 3921 2000 3822 -3890 o41 321 -23? -118 -2818 591 2123 411 821 353 1313 2005 3104 -3100 2312 945 145 -998 -4430 1140 Table 52: Inveatment Program 1989 - 1996 (US$ million)

TOTAL TOTAL TOTAL 1989 1990 1991 1992 1989-92 1993 1994 1995 1996 1993-96 1989-96

On-going plan and distribution 550 571 542 458 2121 294 273 294 306 1167 3287

Future expansion generation & transmission. 0 11 104 188 303 300 430 660 670 2060 2363

Total investment 550 582 646 594 703 954 977 3228 5650 (2 of public inv.) (22) (22) (23) (22) (23) (19) (21) (27) (26) (23) (23.)

Int. during const. 146 153 208 143 650 108 156 170 227 661 1311 >

Source: DNP. Situacion financiera del sector electrico a mediano plazo. Documento DNP:2408:UNIF, Bogota, Enero 11, 1989. -182 -

Table 53: 1989 - 1990 Investm-ntProgram by Category of Investment and Po4er Company

(US$ Million)

Company Generation Transmission Distribution Studies/Other Total %

CORELCA 1989 5.98 23.56 3.11 0.50 33.1S 1990 49.42 34.49 2.43 0.50 S8.54

EPM 1989 69.18 0.00 28.67 9.28 107.13 1990 62.19 0.00 12.79 11.16 76.13

CvC 1989 0.00 5.29 12.74 8.14 24.17 1990 0.00 4.88 20.74 0.67 28.19

ISA 1969 0.00 89.66 0.00 17.78 107.42 1990 2.49 139.70 0.00 9.38 151.67

ICEL 1989 9.80 29.61 6.43 1.96 46.70 1990 7.36 12.64 20.37 0.43 40.eO

EEEB 1989 162.47 0.00 62.39 3.57 228.43 1990 126.89 0.00 64.84 6.55 198.08

Electrificadoras/ICEL 1989 0.00 0.00 48.93 0.00 48.93 1990 0.00 0.00 68.19 0.00 68.19

Electrificadoras/CORELCA 1969 0.00 0.00 21.21 0.00 21.21 1990 0.00 0.00 19.08 0.00 19.08

TOTALS 1989 247.43 148.02 180.48 39.21 616.14 1990 238.35 191.71 198.24 28.68 e65.88

GRAND TOTAL 486.78 393.73 378.72 67.79 1272.02 - 183 -

Table 54: Preliminary Financial Projections

SECTORELECTRICO CONSOLIDADO 1587-1996 tPERACIONES EFECTIVAS DE CADA USS (Mi l ones Corr;cntos)

1987 1988 1989 1990 1991 1992 1993 1994 1996 1996 TOTAL89-96

FUENTES INTERNAS Se0 933 1068 1159 1273 1398 1538 1691 1869 2031 12016 VENTAS DE ENERGIA 809 867 952 1045 1165 1276 1409 1658 1718 1884 10994 OTROS INGRESOS 70 as l1 113 118 123 129 136 141 147 1022 CASTOS DE OPERACION 383 382 386 402 431 466 471 503 529 56e 3739

GENEPACION INTERNA BRUTA 517 661 682 767 842 943 1059 1108 1331 1475 8277

SERVICIO DEUDA 800 799 977 969 900 932 1058 1079 1075 1027 8007

GASTOS FINANCIEROS ME 240 le 224 234 171 230 269 210 187 145 l1d9 GASTOS FINANCIEROS FALTANTE 46 47 93 le 136 149 155 744 GASTOS FINANCIEROS ML 98 119 140 104 114 122 123 128 148 164 1043 AMORTIZACIONES ME 372 423 696 659 646 461 518 sea 566 523 4323 AMORTIZACIONES ML 90 69 19 1s 22 27 32 37 36 40 228

CENERACION INTERNA NETA -283 -248 -296 -202 - S8 11 1 109 256 449 270

FUENTES EXTERNAS 813 1196 1099 968 462 461 416 492 668 684 6218

APORTES DE PARTICULARES 14 29 0 APORTES GOBIERNO NACIONAL 129 161 347 187 0 0 0 0 0 0 53' PRESTAMOS VIGENTES M E 278 528 657 682 228 1se; PRESTAMOS M L 103 108 96 87 97 97 89 106 143 147 860 PRESTAMOS FUTUROS ME 0 0 0 127 364 327 387 625 537 2257 FODEX 293 369

TOTAL FiENTES 630 947 804 764 394 462 417 601 924 1132 6488

APLICACICNES 630 947 1268 8a1 893 809 712 869 1136 1214 7711

INVERSIOitES 530 842 698 735 864 787 702 859 1124 1204 6962

INVERSION IOTAL 431 499 650 682 646 644 594 703 954 977 565C

OTOS. FINAN. CONSoRUCCION 99 143 146 153 208 130 60 72 46 40 e8s CTOS. FINAN. CONSTRuC-TIONP.E 0 0 0 13 48 84 124 187 46e

Otras Aplicaciones 0 30G 572 76 39 22 10 10 10 10 749 FODEX 0 306 386 39 39 22 10 10 10 10 625 GCN-86 0 39 39 22 10 1D 10 10 140) 88 706 385 38S

QEUDAS COBIERNO 187 37 224 OTRAS

SUPERAVIT/DEFICIT ANUAL 0 0 -464 -67 -499 -347 -295 -268 -211 -82

SUPERAVIT/OEFICIT ACUMULADO 0 -464 -521 -1019 -1366 -1662 -1930 -2141 -2223

Jurce: Doportam.nto Nuiconal d Flaneacl6n. Situaci6n Financiern del S.ctor Electricidad En-ro 198° Table 55: Load and SaleX.Forecast 1987 - 1996

LOAD AND SALES FORECAST 1987-1996

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

Sales GWh 22630 23497 25574 27542 29625 31666 33575 35502 37557 39625 Average Rate US$IkWh 0.038 0.040 0.043 0.045 0.049 0.051 0.054 0.057 0.059 0.062 Sales US$ Million 869 945 1093 1234 1448 1560 1810 2008 2227 2460

Sources: ISA: Evaluaci6n Financiera del Plan de Expansi6n de Minimo Costo. Medeliln, Diclembre 1988. FEN: Proyecciones Financieras del Sector Electrico Colombiano Periodo 1988-1990. Bogota, Diciembre 198i8.