Report No. 28911-EC Ecuador Creating Fiscal Space for Poverty Reduction A Fiscal Management and Public Expenditure Review (In Two Volumes) Volume I: Main Report November 17, 2004 Document of the World Bank and the Inter-American Development Bank FODESEC Fondo de Desarrollo Seccional (Fund for Sectional PAE Programa de Alimentacidn Escolar (School Breakfast Development) Program) FONDETEL Fondo para el Desarrollo de la Telefonia Rural PAHO Pan-American HealthOrganization (Telecommunications Development Fund) PA1 Programa Ampliado de lnmunizacidn en las AmCricas FS Fondo de Solidaridad (Solidarity Fund) (Immunization Program) FTSRL Law on Fiscal Transparency, Stabilization and PANN Programa Nacional de Nutricidn y Alimentacidn (National Responsibility Nutritional Program for Children) GDP Gross Domestic Product PEM Public Expenditure Management GOE Government of Ecuador PER Public Expenditure Review GTZ GermanTechnicalCooperation Agency PGE Presupuesto General de Estado (Central Government HDI Human Development Index Budget) HIPC(s) Highly IndebtedPoor Country (ies) PHRD Public HumanResource Development ICE Impuesto a 10s Consumos Especiales (Special PPA Power PurchasingAgreement Consumption Tax) PPC Programacidn Periddica de Caja (cash balance periodic IDB Inter-AmericanDevelopment Bank programming) IEOS Instituto Ecuatoriano de Obras Sanitarias (Ecuadorian PPS Project for Public Spaces Instituteof Sanitary Works) PRADEC Programa de Asistencia y Desarrollo Comunitario IESS lnstituto Ecuatoriano de Seguridad Social (Ecuadorian (programs of Aid andCommonDevelopment) SocialSecurity Institute) PRAGUAS Programa de Agua y Saneamiento para Comunidades IM Infant Mortality Rurales y Pequeiios Municipios (Project Appraisal IMF IntemationalMonetary Fund Document, Rural and Small Towns Water Supply and INEC (El lnstituto Nacional de Estadistica y Censos (Integrated Sanitation Project) System of Household Surveys) PROBE Public Report on Basic Education INECEL lnstituto Ecuatoriano de Electrificacidn (Ecuadorian PROMEC Proyecto de Modemizaci6n de 10s Sectores Ele'ctrico, Power Authority) Telecomunicaciones y Servicios Rurales (Power and INH Instituto Nacional de Higiene (National Institute of Communications Sectors Modernization and Rural Hygiene) Services Project) IPS interbank payment system PROST Pension Reform Options Stimulation Toolkit -developed IRS/SRl InternalRevenue Service/Servicio de Rentas Internas by the World Bank ISR personal income tax PRS Poverty Reduction Strategy ISSFA Military Social Security Institute RER Real Exchange Rate ISSPOL Instituto de Seguridad Social de la Policia Nacional RGP Referential GenerationPrice (Institute of Social Security of the NationalPolice) ROSC Reports on,the Observance of Standards and Codes m Oil Fields (Ishipingo-Tambococha-Tiputini) RUC Registro Unico de Contribuyentes (centralized taxpayers ITU International Telecommunications Union registry) NM InternalValue Measurement SAPYSB Subsecretaria de Agua Potable y Saneamiento Bdsico JASS Juntas de Agua y Saneamiento (Committee for Water and (Undersecretary of Drinking Water and Basic Sanitation) Disinfection) SELBEN Sistema de Identificacidn y Selecci6n de Benejiciarios de LAC Latin America and The Caribbean 10s Programas Sociales (Poverty Mapping and LMG Ley de Matemidad Gratuita y Atencidn a la lnfancia (Law StrengtheningDatabase) of Free Maternity and Attention to Infancy) SENATEL Secretaria Nacional de Telecomunicaciones (National LNG Liquefied Natural Gas Telecommunications Office) LOAFYC Ley Orgdnica de Administracidn Financiera y Control SENRES Secretaria Nacional Te'cnica de Desarrollo de Recursos (Organic Law of FinancialManagement and Control) Humanos y Remuneraciones del Sector Pliblico (National LSMS Living Standards Measurement Study Technical Office of Human Resources Development and MBS/STFS Ministerio de Bienestar Social (Ministry of Social Affairs) Remunerations of the Public Sector) MDGs MillenniumDevelopment Goals SIAN Sistema Integradode AlimentaciQ y Nutrici6n MED Ministry of Education SIEH Sistema lntegrado de Encuestas de Hogares (Integrated MEF Ministerio de Economia y Finanzas (Finance Ministry) Home Survey System) MEM Ministerio de Energia y Minas (Ministry of Energy and SIGEF Sistema Integrado de Genercia Econdnica y Financiera Mines) (Integrated Financial Management System) MIDUVI Ministerio de Desarrollo Urbano y Vivienda (Ministry of SIGOB Sistema de Gestidn de Gobiemo (Government UrbanDevelopment and Housing) Management System) MNL Multinomial Logit Model SINEC Sistema Nacional de Estadisticas y Censos (National M O Mendoza-Oviedo System of Statistics and Censuses) MPH Ministryof Public Health SNEM Servicio de Erradicacidn de Malaria (Malaria Eradication MYBF Multiyear Budgeting Framework Service) NBI NecesidadesBdsicas lnsatisfechas (Unmet Basic Needs) SOTE Sistema de Oleoducto Transecuatoriano (Trans- NFPS Non FinancialPublic Sector Ecuadorian Pipeline) NWFP North West Frontier Province SPA Subsecretaria de Proteccidn Ambiental (Undersecretariat OECD Organization for Economic Co-operation and Development on EnvironmentalProtection) OLS Ordinary Least Squares SPR Strategic Policy Research ORDs Regional Development Organizations ssc Seguro Social Campesino (rural social security) OSCIDI Ojicina de Servicio Civil y Desarrollo lnstitucional (Office sso Seguro Social Obligatorio (obligatory social security) of Civil Service and InstitutionalDevelopment) STFS Secretaria Te'cnicadel Frente Social (Technical Secretariat OSMERG Organism0 Supervisor de la lnversidn en Energia of the Social Front) (Supervising Agency for Energy Investment) STFS-SIISE Sistema Integrado de lndicatores Sociales del Ecuador OTECEL Bellsouth, Empresa Telefdnica (Integrated System of Social Indicators of the Social PACIFICTEL Pacific Telephone-Fixed line telecommunications Cabinet) operator inthe coastalregion SUF'TEL Superintendencia de Telecomunicaciones (Super- UNESCO United Nations Educational, Scientific and Cultural intendence of Telecommunications) Organization TELECSA EcuadorianTelecommunications-South America UNICEF UnitedNations Children's Fund UBN Unsatisfied Basic Needs V A T Value-Added Tax UDENOR Unidad de Desarrollo del Norre (Northern Development VOIP Voice-over-IP office) WFP World FoodProgramme UNDP United Nations Development Programme WHO World HealthOrganization UNE Union Nacional de Educadores del Ecuador (National Teacher's Union) IBm IDB Vice President: Davidde Ferranti Manager, RegionalOperation111: Ciro de Falco CountryDirector: Marcel0 M.Giugale Deputy Manager, RegionalOperationsDept.111: Mbimo Jeria Sector Director: EmestoMay Chief, CountryDivision5: Vladimir Radovic SectorLeader: Mauricio Carrizosa LeadEconomist: VicenteFretes-Cibils Senior Economist, Co-Task Manager JosCR. Mpez-Chlix Country Economist,Co-TaskManager: Albert0 Melo Ecuador Public Expenditure Review-Volume I Table of Contents ACKNOWLEDGMENTS PREFACE EXECUTIVESUMMARY................................................................................................................... Fiscal Policy FacesHistorical and StructuralConstraints......................................................... Fiscal Trends and Challenges................................................................................................... ...11..i Pro-Poor Expenditure and the Roomfor Additional Fiscal Space............................................ 111 Public Expenditure Management (PEM)and Other InstitutionalIssues ................................. v i x Conclusions and Policy Recommendations ............................................................................. x i A Selection of Key Policy Recommendations ........................................................................ xv 1 . FiscalPolicy ina DollarizedEconomy .................................................................................. 1 The EcuadorianEconomy: Some Historical Background......................................................... 1 The Role of FiscalPolicy ina DollarizedFramework.............................................................. 4 Structural Constraints on FiscalPolicy ..................................................................................... 6 2 . FiscalTrends and Challenges .............................................................................................. 11 11 Fiscal Performance Before and After Dollarization................................................................ The Volatility of FiscalVariables ........................................................................................... 12 The Challenge of Revenue Management ................................................................................ 13 Fragmented Tax Structure................................................................................................... 15 16 Earmarking: The Underminingof Budget Flexibility......................................................... Tax Exemptions and the Erosion of the Tax Base .............................................................. 17 Expenditure Trends................................................................................................................. 17 The FiscalTransparency, Stabilization and Responsbility Law ............................................. 25 Stabilization Funds.............................................................................................................. 26 Fine-TuningFiscal Rules........................................................................................................ 27 Issues in Debt Sustainability ................................................................................................... 29 Exercise 1. Fiscal Sustainability under a SuddenDrop inthe Price 31 Exercise 2. Fiscal Sustainability under Tax Revenue Volatility ........................................ of Oil, Surge inInterest Rate, or Capital Outflows ......................................................... 32 Policy Recommendations........................................................................................................ 33 3. Pro-Poor Expenditures and the Fiscal Space ..................................................................... 37 I s Social Expenditure Enoughfor Poverty Reduction?........................................................... I s Social and Basic InfrastructureExpenditure Pro-Poor? ...................................................... 37 39 Subsidies inBasic Infrastructure Services .......................................................................... Main Social Programs......................................................................................................... 39 45 How MuchFiscal Space I s Available for a Poverty Reduction Strategy................................ 47 Reversing Current Expenditure........................................................................................... 48 Making an Optimal Selection of Public Investment........................................................... 50 Assessing Defense Spending............................................................................................... 50 How Much Off-Budget Fiscal Space Can BeFoundfor aPoverty Reduction Strategy.........52 Freezingor Reducing Budget Earmarking.......................................................................... 52 ReducingTax Expenditure.................................................................................................. 53 Towards a Results-Oriented Budget: Attaining the MillenniumDevelopment Goals............ 54 Making Transparent and Intergrating Off-Budget Public Investment ................................ 55 Policy Recommendations........................................................................................................ 59 To Shift Public Expenditures Toward a Pro-Poor Focus .................................................... 60 To Re-Target Public Subsidies............................................................................................ 61 Trade-offs BetweenIn-and Off-Budget Fiscal Space ....................................................... 63 To Reach SelectedMDGs................................................................................................... 63 4. Performance of Public ExpenditureManagement............................................................. 65 The PEMProcess and its Recent Performance ....................................................................... 66 Budget Management Review inthe Central Government....................................................... Budget Formulation ............................................................................................................ 69 70 Budget Management Review inSocial Programs and Subnational Governments..................72 Execution ............................................................................................................................ 75 Provincial and MunicipalGovernments.............................................................................. Social Programs .................................................................................................................. 76 82 Budget Transparency, Accountability, and Participation........................................................ 87 Policy Recommendations........................................................................................................ Are There Sufficient Conditions for a Multiyear BudgetingFramework inEcuador?...........85 88 APPENDIX A An Estimationof the PotentialOutput . And the Structural FiscalBalanceinEcuador..................................................... 95 APPENDIX B Summary of the BudgetProcess.......................................................................... . 101 APPENDIX C Major Budgetary IssuesinPrioritySocial Programs ....................................... . 105 APPENDIX D Budgetary Framework of DecentralizationinEcuador.................................... . 115 APPENDIX E Budgetary Framework of DecentralizationinEcuador.,.................................. . 123 STATISTICAL APPENDIX ........................................................................................................... 129 BIBLIOGRAPHY ............................................................................................................................ 197 Tables Table ES.1 Potential Sources of Fiscal Space and EstimatedAnnual Impact............................ ... v111 Table 2.1 Volatility of NFPS Variables. 1993-2003 ................................................................ Table 2.2 Ecuador: Actual and StructuralFiscal Balance. Fiscal Stance and Fiscal Impulse...11 Table 2.3 Total Revenue of the Non-Financial Public Sector and Central Government..........13 15 Table 2.4 The Structure of the Tax System............................................................................... 16 Table 2.5 Selected Tax Administration Indicators. 2001-03 .................................................... 16 Table 2.6 "Optimal" Fundvs FEIREPFlows for DebtReduction ........................................... 27 Table 2.7 Evaluation of Compliance with the Fiscal Rules ...................................................... Table 3.1 Ecuador: Social Expenditure of Central Government as a Percentage of GDP........28 38 Table 3.2 Pro-Poor and NonPro-Poor Social Expenditures ..................................................... 39 Table 3.3 Social Expenditure and Energy Subsidies, by Income Quintile................................ 41 Table 3.4 Basic Services Subsidies by Expenditure Quintile ................................................... 46 Table 3.5 Rigidity of the Central Government Budget. 2001-04 ............................................. 49 Table 3.6 Ecuador: Personnel Involved inDefense and Security ............................................. 51 Table 3.7 EstimatedTax Expenditures of the InternalVAT Exemptions................................. 53 Table 3.8 55 EstimatedFiscal Cost of Attaining Key MDGs and Social Targets ......................... Progress by Ecuador inMeeting the MillenniumDevelopment Goals..................... Table 3.9 58 Table 3.10 Potential Sources of Fiscal Space ............................................................................. Table 4.1 Basic Elements of Public Expenditure Management: The "Three-level Analysis" ..62 65 Table 4.2 Allocation of Public Spending by Tiers .................................................................... Table 4.3 Review of Ecuador's Public Expenditure Management and Benchmarks................66 Table 4.4 Ranking of Ecuador's PEMinRelationto Peru. Bolivia.and HIPC Indicators........68 70 Table 4.5 Budgetary Assumptions and Actuals ........................................................................ Sources of Government Financing............................................................................ 71 Table 4.6 72 Table 4.7 Level of Budget Executionof Sector Expenditures .................................................. 73 Table 4.8 Changes inthe Budget Share between Executedand Approved............................... Table 4.9 Forms of Local Participation within Selected Priority Social Programs................... 74 76 Table 4.10 Budget of Priority Social Programs Groupedby Social Sector ................................ 77 Table 4.11 Budget Approved and Executedfor Priority Social Programs ................................. Table 4.12 Budget Execution and Tracking Survey of Transactional Delays within PSPs........78 80 Table 4.13 Positive Qualifications per Country .......................................................................... 86 Figures FigureES 1 . Oil Prices. Growth. and the FiscalDeficit .................................................................. .. 11 Figure ES.2 Debt Sustainability underthe Fiscal Rule................................................................... v Figure ES.3 Incidence of Social Expenditure and Energy Subsidies. 1999.................................. .. v11 Figure 1.1 Total Public Sector Net Work ..................................................................................... 6 Figure 2.1 Non-Financial Public Sector Balances...................................................................... 12 Figure 2.2 Tax and Non-Tax Revenue of the Central Government and Important Tax ReformEfforts 1964-2004 ............................................................... 14 Figure 2.3 Total and Primary Expenditures................................................................................ 17 Figure 2.4 18 Public Expenditure by Selected Functions................................................................ Trends inNFPS and Central Government Expenditures .......................................... Figure 2.5 19 Figure 2.6 Composition of Wages by Selected Sectors.............................................................. 21 Figure 2.7 Institutional Composition of Public FixedInvestment ............................................. 25 Figure 2.8 27 Base Case for Debt Sustainability............................................................................. "Optimal" and Actual Stabilization Accumulation Scenarios .................................. Figure 2.9 31 Figure3.1 Incidence of Social Expenditure and Energy Subsidies............................................ 40 Figure 3.2 Telephone Penetration by Household Income Quintile ............................................ 46 Figure 3.3 Rigid Compared to Flexible Public Expenditure ...................................................... 49 Figure 3.4 Military Expenditure. 1998-2001 ............................................................................. 51 Figure 3.5 Decomposition of Budget and Off-Budget Financing Sources................................. 52 Figure 3.6 Chile: Fondo Concursable........................................................................................ 60 Figure 4.1 Changes in Shareof NFPS Expenditures. by Government Tier ................................ 66 Figure 4.2 Emergency Decrees................................................................................................... 72 Figure 4.3 Stock of Arrears ........................................................................................................ 75 Figure 4.4 Budget Execution of Social Programs ...................................................................... 79 Figure 4.5 SeasonalPatterns inBudget for Social Programs ..................................................... Figure 4.6 Planned Compared to ExecutedTransfers of the 15-Percent Law. 1997-2003 ........81 82 Figure 4.7 Transfers and Payments of the 15-Percent Law to Municipalities............................ Figure 4.8 Transfers of the 15-PercentLaw and the New Subnational Budget Cycle...............83 84 Figure 4.9 L A C Index of Budgetary Transparency .................................................................... 85 Figure 4.10 Revenue Base for the 15-Percent Law with/without Oil Revenue............................ 92 Boxes Box 1.1 The Costs and Benefits of Dollarization ..................................................................... 5 Box 1.2 Ecuador's Political Economy...................................................................................... 8 Box 2.1 The Civil Service and the Wage Unification Law: A Stepinthe Right Direction.................................................................................... 21 Box 2.2 26 The Base Case Scenario............................................................................................ Ecuador's Oil Funds.................................................................................................. Box 2.3 30 Box 3.1 Teacher Absenteeism inPrimary Schools................................................................. 43 Box 3.2 47 56 The ChileanCompetitive (Concursable)Fundfor Public Programs........................ Main Conclusions of the World Bank Ecuador Poverty Assessment ....................... HouseholdExpenditures on Water: The Case of Machala. ElOro........................... Box 3.3 Box 3.4 61 Box 4.1 The Six Stages of a Comprehensive Multiyear Budgeting Framework.................... 87 Box 4.2 Main Recommendations of the IMF-ROSC ............................................................. 89 Box 4.3 Ecuador-Institutionalizing a Virtual Poverty Fund: A Look at Best-Practice inUganda........................................................................... 90 Annex Tables Annex Table A.1 Structural Fiscal Balance ............................................................................. 98 Annex Table C.1 106 108 Budget Executionat LMG. by Financing Source...................................... Budget Executionat PAI. by FinancingSource......................................... Budget Execution at BDH.by Financing Source....................................... Annex Table C.2 Annex Table C.3 110 Annex Table C.4 Budget Executionat PRADEC. by Financing Source ............................... 112 Annex Table C.5 113 Annex Table D.1 Budget Execution at PAE. by Financing Source ....................................... Current and Capital Spending of Municipalities and Provincial Councils ............................................................................. 118 Annex Table D.2 Priorities of Subnational Investment Spending.......................................... 119 Annex Table D.3 Capital Spending of Municipalities and Provincial Councils inthe Health and Education Sectors.......................................................... 121 Annex Table D.4 Administration of Schools. Teachers. and Students in The Education Sector. School Year 2000-01 ............................................ 122 Annex Table D.5 Public SpendinginHealth Sector .............................................................. 122 Annex Table E.1 Subsidy Distribution by Consumption Deciles.......................................... 123 Annex Table E.2 Distribution of the Cooking-Gas Subsidy by Ethnic Group ...................... Annex Table E.3 Effect of Various ReformScenarios for the Cooking-Gas Subsidy...........124 124 Annex Figures Annex Figure A.1 Actual and Potential GDP............................................................................ 97 Annex Figure A.2 Primary Balance........................................................................................... 98 Annex Figure B.1 Flow of Budgetary Funds........................................................................... 101 Annex Figure C.1 Organizationof the BDHOperatingin 14Provinces ................................ 106 Annex Figure C.2 Organizationof PA1OperatingNationwide .............................................. 107 Annex Figure C.3 Organizationof LMG Operatingin40 Municipios ................................... Annex Figure C.4 Organizationof PRADEC Operating in443 Juntas Parroquiales..............109 111 Annex Figure C.5 Organizationof PAE Operating inthe Coast and Sierra............................ 112 Annex Figure D.1 IntergovernmentalTransfers. 1996-2003 .................................................. Annex Figure D.2 Sources of Revenue for Transfers to Subnational Governments ...............115 117 Annex Figure D.3 Own Revenue and Expendituresper Level of Government and Veritcal Fiscal Imbalances. 1997 and 2002......................................... 119 Annex Figure E.1 Distribution of the Gas Subsidy Retargeted with SelBen .......................... 125 ACKNOWLEDGMENTS This report is ajoint effort of the World Bank and the InterAmericanDevelopmentBank. JosC R. Lopez-C6lix (WB) and Alberto Melo (IDB) are the task managers responsible for completing the whole task. Its initial preparationwas coordinated by Sara Calvo (WB) and Carlos Elias (IDB). The report i s a collective effort of several authors of chapter sections and background papers presented intwo volumes. Volume I. Chapter 1was prepared by Alberto Melo, with inputs from Jonas Frank (WB-party politics). Chapter 2 was prepared by Alberto Melo, Elaine Tinsley (WB) and Sara Calvo (WB) including inputs from Jorge Shepherd (consultant-Fiscal Trends), Daniel Artana and Cynthia Moskovits (consultant-Fiscal Policy), Carlos Diaz and Alejandro Izquierdo (IDB-Fiscal Sustainability), Jeffrey Rinne (WB-Civil Servive) and Jonas Frank (WB-Tax Revenues). Chapter 3 was prepared by JosC R. L6pez-Calix based on inputs from Michael Geller (WB-Military Expenditures), Rob Vos, Juan Ponce, Mauricio Le6n (consultants-Social sector). Chapter 4 was prepared by JosC R. L6pez-CAlix with inputs from Elaine Tinsley (WB-Budget Management), Jorge Shepherd (consultant-Social Expenditures), Jonas Frank (WB-Decentralization), and Carolina Sanchez-Paramo (Appendix E-reform of cooking gas subsidy). Vicente Fretes-Cibils (WB), Bruce Fitzgerald (WB) and McDonald Benjamin (WB) provided invaluable oversight and comments on all the chapters. The Statistical Appendix was prepared by Elaine Tinsley with inputs from Jorge Shepherd, Jonas Frank, and Rob Vos. The report also benefited from the useful contributions and complementary analysis on budget management by Diego Mancheno (MEF), Roberto Salazar (MEF), Hugo Muiioz (MEF), Maria de 10s Angeles Rodriguez (MEF), Lenin Parreiio (Banco Central), Fabian Carrillo (MEF), Javier Game (MEF), Roberto Iturralde (MEF), and Paula Suarez (MEF). The draft also benefited from comments collected during a joint IMF/IDB/WB seminar held in Quito on May 20-21, 2004, and another during a joint MEF/IDB/World Bank workshop held in Quito on November 4-5, 2004 on a preliminary version of this report. More in particular, valuable suggestions were received from Patricia McKenzie (WB), Gonzalo Afcha (IDB), Julio Viiiuela (IMF),Raju Jan Singh (IMF),Trevor Alleyne (IMF)andEstebanVesperoni (IMF).Peer reviewers were Yasuhiko Matsuda (WB), Amanda Glassman (IDB), and Carlos Elias (IDB), who provided very helpful comments. Valuable contributions were also received from Francesca Castellani (IDB), Rodrigo Suescun, Carolina Sinchez-Piramo, Emmanuel James, Franz Drees, William Doritinsky, Daniel Dulitzky, Emesto May and Mauricio Carrizosa (all WB) during all process. This Volume was edited by Diane Stamm and Chris Humphrey (Executive Summary) and assembled under the general production support by Michael Geller. Volume 11. Contributing authors include: Carlos Diaz and Alejandro Izquierdo (Fiscal Sustainability); Daniel Artana and Cynthia Moskovits (Fiscal Rule and Oil Stabilization Fund); Rob Vos, Juan Ponce, Mauricio Le6n (IDB-Health, Education); Rafael Rofman (WB-Pensions); F. Halsey Rogers, JosC R. L6pez-Cblix, Nancy Cbrdova, Michael Kremer, Karthik Muralidharan, Jeffrey Hammer, Nazmul Chaudhury (WB, CEDATOS and Harvard University-National Teachers Tracking Survey); Maria AngClica Sotomayor (WB-Water and Sanitation) with support from Franz Drees (WB); Horacio YCpez (Consultant-oil) with support from Eleodoro Mayorga (WB-Oil); Carlos G6mez and Eloy Vidal (WB-Telecom); and Fernando Lecaros (consultant-Electricity) with support from Philippe Durand (WB-Electricity). The Volume was edited by John Moody and Maria Antonieta Gonzalez and assembledunderthe general production and editing support by Rosalia Rushton. The report benefited from the overall guidance and support of Marcel0 Giugale (WB-Country Director), Ana Maria Rodriguez (IDB), Vladimir Radovic (IDB), Vicente Fretes-Cibils (WB-Lead Economist), Mauricio Carrizosa (WB-Sector Manager), Emesto May (WB-LCSPR Director), Javier Game (consultant-IDB), Margarita Andrade and Alexander Shapleigh (SALTORJSAID), and David Yuravlivker (IMF). This document also benefited from the excellent administrative and production support of Francisco Irias, and Carolina Torres. Alexandra del Castillo, Ana Maria Vicuiia, Lucy Vargas, Cynthia Guzmin, Vinicio Valdivieso, and Ana Maria Villaquirin provided critical, timely and qualified operational support to field research inQuito. The team would also like to express its sincere gratitude and appreciation for the cooperation and contributions of the Ecuadorian authorities throughout the process. We are particularly grateful to Mauricio Pozo, Gilberto Pazmiiio, and Messrs. Gal0 Viteri, Femando SuQez, Fausto Herrera, Roberto Salazar, Hugo Muiioz, Mauricio YCpez, Leopoldo Biez, Ramiro Galarza, Ivonne Baki, Christian Espinosa, Diego Martinez, Polibio Cbrdoba, Jenny Guerrero, Paula SuQez, and Liszett Torres. In the same vein, we would like to recognize the valuable inputs provided by the ROSC/CFAA/PER seminar participants held inQuito on May 20-21, 2004. Participants at the first seminar included: Mauricio Pozo, Gilberto Pazmiiio, Roberto Salazar, Diego Mancheno, Hugo Muiioz, Fabian Carrillo, Luis Benalcizar, Maria de 10s Angeles Rodriguez, Jorge Recalde, Stalin Nevirez, Eduardo Lbpez, Milton Ordbiiez, Fausto Solbrzano, Estuardo Peiiaherrera, Ra61 Baca Carbo, Carlos Pblit, Xavier Ledesma, Patricio Acosta, Roberto Passailaigiie, Te6filo Lama, Mauricio Lebn, Genaro Peiia, Elsa de Mena, Leonard0 Viteri, Juan Abel Echevem'a, Jorge Martin,Jaime Estrada, Mauricio YCpez, Leopoldo Biez, JosC Maria Borja, Luis Palau, Dora Currea, Mauricio Valdez, David Yuravlivker, Lars Klassen, Julio Viiiuela, Gabriel Montalvo, Carlos Pimenta, Javier Game, Paula SuArez, Rick Garland, Maria Lorena Correa, Alexander Shapleigh, Margarita Andrade, Wistano Saez, Rafael Donoso, Victor Acosta, Marco Varea, Esteban Vesperoni, Patricio Rivera, Fausto Herrera, Virginia Velasco, I v i n Leiva, Giovanni Coronel, Singh Raju, Alexandra Lastra, Alicia Guzmin, Juan Garcia, Nelly Molina, Emesto PCrez, JosC Samaniego, Pablo Lucio Paredes, and Alfredo Mzaga. Participants at the second seminar inQuito on November 65,2004 were: Mauricio YCpez, Javier Game, Ramiro Viteri, Renato Valencia, Roberto Iturralde, Fausto Herrera, Patricio Rivera, Jenny Guerrero, Nelly Molina, Ver6nica Poveda, Paula Suirez, Ver6nica Lojin, Diego Mancheno, Alfredo Astorga, Nelly Molina, Pa61Carrillo, Nelson GutiCnez, Alfonso Tique, and BaudouinDuquesne. PREFACE The last Public Expenditure Review for Ecuador (PER) was produced in 1993. More than a decade later, the purpose of this report i s to provide the Ecuadorianauthorities with the World Bank's and Inter-American Development Bank` s joint comprehensive account of their diagnoses and recommendations in the area of fiscal management and public expenditure. As the GutiCrrez Administration moves forward in its second year inoffice, it i s hoped that the content of this PER will be useful for Ecuador bothto deal with the formidable fiscal challenges it faces and to take advantage of the existing opportunities inits development agenda. This report consists of two volumes. Volume Iexamines whether, and how, the core goals of public expenditure management, Le., balanced fiscal aggregates, resource allocations to strategic sectors, and equity and microeconomic efficiency of public spending are met in Ecuador. Volume I1presents sector studies on fiscal sustainability, the fiscal rules, education, health, pensions, the results of a national teachers tracking survey, water and sanitation, electricity, telecommunications and oil. The report does not provide full coverage of all areas affected by public expenditure. It focuses on the main themes that are critical for Ecuador's fiscal consolidation and poverty reduction following dollarization. In most cases, it does provide choices to key policy questions that are likely to occupy Ecuadorian policymakers over the remaining of the Government, like defining FEIREP proceedings, budget allocations, or social programs prioritization. Thus, it provides an independent analysis of the selected areas where bothBanks are specially involved, and a set of possible recommendations to address them. This report reflects policy developments through May 31,2004. According to the Ecuadorian Authorities, the PER i s "an important contribution from the World Bank and the Interamerican Development Bank to public policy. Volume I,in particular, correctly identifies fiscal vulnerabilities in the new context of dollarization, and proposes an adequate fiscal management that increasesexpenditure flexibility, develops budget management reform, increases public (social) investment, and brings transparency to public expenditure. All this i s supported by an implicit proposal for a Fiscal Pact for Poverty Reduction. For its part, Volume I1deals with sectoral policies, and their link to fiscal management. It identifies the most efficient and cost-effective interventions in the social sectors, while making an optimal use of the reduced and available fiscal space. The study also recognizes the importance of political constraints, and the difficulties of setting steady rules in a non- cooperative game among national political actors that are particularly reflected in budget allocations. It correctly emphasizes the need to bring full transparency of information on the management of public accounts among all domestic actors as starting point for sectoral reform. The report has a global view and suggests positive steps. Somehow, it should contribute to align fiscal and institutional policies in the social and basic infrastructure sectors, and to strengthen them in the context of the ongoing negotiations for a Free Trade Agreement with the US., while preserving difficult domestic equilibria on the development agenda.'' ECUADOR: CREATING FISCAL SPACEFORPOVERTY REDUCTION FISCALMANAGEMENTAND PUBLICEXPENDITURE REVIEW EXECUTIVE SUMMARY 1. Since Ecuador adopted full dollarization in early 2000, its fiscal performance has significantly improved. The new exchange rate regime i s underpinned by sound fiscal policies and structural reforms. Following a difficult transition, the GutiCrrez administration strengthened the dollarization framework with its Program of Economic Restructuring and HumanDevelopment. The results achieved thus far are encouraging: Ecuador is one o f the best- performing economies inLatin America. 0 Growth has resumed and i s expected to reach above 5 percent in2004. 0 Inflation fell to single digits in late 2002 and i s projected to continue declining below 3 percent in 2004. 0 The Non Financial Public Sector (NFPS) primary and overall fiscal surpluses in 2003 are among the highest in the Latin American region (4.7 percent of GDP and 1.7 percent of GDP) and are projected to reach similar levels in 2004, reinforced by the new Fiscal Responsibility, Stabilization, andTransparency Law (FRSTL); 0 The current account deficit has halved, most arrears were cleared and public indebtedness was lowered by about 5 percent o f GDP in 2003. These outcomes are also projectedto further improve in 2004. 2. The Ecuadorian economy, however, remains vulnerable. External factors, particularly oil prices, have a strong impact on the economy (FigureES.l), as do shocks such as sudden stops of capital flows, rising interest rates, falling remittances or natural disasters. Shocks cause stress in the fiscal accounts, depreciate the real exchange rate, and threaten fiscal sustainability. The effect of these shocks could be augmented or alleviated by the Government's use of fiscal policy, which is the principal macro-economic policy tool available in a dollarized economy. 3. Poverty reduction is critical to sustain the country's stability in the medium term. As a result of the triple-banking, debt and exchange-crises of the late 1990s, poverty has increased. The national poverty rate increased from 40 to 45 percent between 1990 and 2001, and the number of poor increased from 3.5 million to 5.2 million, with a markedconcentration o f new poor in the urban areas. President Gutitrrez has committed to designing and implementing a Poverty Reduction Strategy (PRS) to reduce this high level of poverty and to achieve the Millennium Development Goals (MDGs). Reversing poverty trends and improving living standards is a sine qua non for maintaining the country's stability, while mitigating macro volatility. i Figure ES.1. Oilprices, Growthandthe NFPSFiscalDeficit 10.0 5.0 0.0 -5.0 -10.0 0 Source: World Bank staff's estimates. 4. To strengthen the economy's resistance to shocks, reduce the high rates of poverty, and achieve the MDGs, this report highlights the need for public policy to focus on three maingoals: (a) strengthened fiscal sustainability; (b) increased fiscal space for pro-poor efficient and equitable public spending; and (c) improved budget management for results-oriented service delivery. These three objectives are closely interrelated. Given external vulnerability and dollarization, fiscal sustainability i s a pre-requisite for poverty reduction, as nothing hurts the poor more than an unstable macroeconomy. However, meaningful poverty reduction also requires fiscal space, understood as the amount o f non-wage expenditure devoted to poverty reduction, and resources shifted toward pro-poor priority programs, executed with efficiency and equity considerations. Furthermore, in a context of scarce resources, sound budget management i s essential to eliminate waste andrigidities and improve service delivery. Fiscal Policy Faces Historical and Structural Constraints 5. To consolidate fiscal discipline, Ecuador must overcome the inherited effects of three decades of predominantly misguided fiscal policies before dollarization. Public sector net worth declined between 1970 and 2000, and it has remained flat since then. Fiscal revenues are volatile and pro-cyclical. Past expansionary spending resulted in high deficits financed with mounting debt. In addition, budget rigidity due to earmarking repeatedly provokes serious liquidity problems for the national Treasury. These shortcomings are compounded by the political economy constraints. In Ecuador, the prevailing political-economy regime was formed inthe early 1970sunder the influence of the oil boom of those years. Its main features are those common in societies where rent-seelung i s pervasive: acute competition for oil rents; conflict- prone social relationships; a fragmented political system where social and political actors face seemingly insurmountable difficulties to reach consensus; a state captured by privileged groups and always on the verge of becoming just the institutional locus where decisions on rent 11 distribution are made and clientelistic favors and privileges are purveyed. As a result, governance and institutions are weak, the efficiency and efficacy o f public administration i s severely impaired and the opportunities for arbitrariness and corruption in the exercise of power multiply. Throughout the last three decades, the internal connection between this political- economy regime and the practices and institutions that frame the conduct o f fiscal policy i s particularly visible in four systemic features of fiscal policy and public finance, namely: (i) the large size of Government and its role as a producer and provider of non-public goods and services, frequently at subsidized prices; (ii)the existence of a set o f tax expenditures, whose function i s to channel (potential) government revenue to the private sector; (iii) the fact that policymakers' incentives tend to be biased toward a short-term horizon; and (iv) the extended use of the deeply flawed institution of earmarking. 6. To face these shortcomings, Ecuador chose a very demanding institutional framework for the conduct of economic policy. While dollarization eliminates the risk of a currency crisis and the hyperinflation it entails, policy tools for demand management are severely restricted, and the buffer of the nominal exchange rate i s no longer available. Moreover, for all its advantages for financial stability, dollarization in the context of an open capital account needs to be supported by a robust, well-regulated financial system. Fiscal Trends and Challenges 7. The first task is to consolidate the current trend toward fiscal sustainability. Fiscal revenue i s close to 25 percent o f GDP, high by Latin American standards. This implies that while, non-oil tax receipts should be raised, expenditure adjustments rather than revenue increases must be the principal means to achieve a sustainable fiscal path. The current administration must deal with the fact that currently too much i s spent on rigid and non-priority goods and services, and too little on pro-poor programs. 8. The Government i s not using high oil prices as an excuse for expansionary fiscal policy, despite the highest oil prices in more than two decades. This restrained fiscal policy recognizes that Ecuador i s still in a fragile fiscal position. Structural estimates of the fiscal stance confirm the prudent management of fiscal policy after dollarization. The average fiscal stance (the difference between the actual and "structural" budget balance) was a surplus of about 1percent of GDPbetween 2000-03. 9. Revenue management has improved significantly under the reformed Internal Revenue Service (IRS), but faces difficult structural constraints. Since revamping the IRS inthe late 199Os,for the first time in30 years tax revenue has becomeroughly equivalent to non- tax revenue as a share o f GDP. The IRS has a centralized taxpayers registry with crosschecking systems, and applies sanctions to non-compliers. However, the tax system remains segmented in a myriad of nuisance taxes (84 overall), and i s burdened by extensive earmarking and a multiplicity of tax exemptions. Both earmarlung and exemptions are costly (each accounting for about 4 percent of GDP in 2004) and increasing. Tax earmarking severely undermines budget management, since it promotes an inefficient and inequitable use of resources, and constrains the authorities' ability to reduce expenditure when needed. 10. Although fiscal policy has been prudent in macro-economic terms, the existing spending structure is not conducive to poverty reduction: public wages and pensions have increased at the cost of cutbacks in public investment and social outlays. Since 2000, non- 111 ... financial public sector (NFPS) primary spending has been on an expansionary trend. This i s mainly propelled by wages and salaries, and contrasts with the constant trend maintained during the pre-dollarization period. This expansionary pattern indicates that the gains achieved by cutbacks in capital and social outlays, as well as in savings in interest payments, have been used to pay for the growing salaries and pensions. 11. The rapid increase in payroll spending is not mainly due to the size of the civil service, but to increases in wages. The size of the Ecuadorian civil service i s about average when compared to other Latin American countries. However, the rate of growth of the public payroll has been high. Inreal terms, the payroll grew 21.3 percent in2001, 35.4 percent in 2002 and 19.5 percent in 2003. Not surprisingly, the share in current expenditure going to wages and salaries almost doubled in the last four years, rising from 25 percent in 2000 to 45 percent in 2003. 12. The changes in the sectoral composition of government expenditure also constraints social outlays. Social expenditures remain low and have slightly declined in the last two years. In2004, education andhealth spending accounts for about 4 percent of GDP, about halfthe L A C averages of 7.5 percent of GDP. Social assistance, an important element for preserving a safety net on vulnerable sectors, accounts for about 1 percent o f GDP. At the same time, military spending, at about 3 percent o f GDP, i s twice as high as the Latin American average of about 1.5 percent of GDP, and on the rise, following the increased military activity in neighboring Colombia. 13. To deepen its fiscal consolidation, Ecuador implemented an oil stabilization fund together with well-defined fiscal rules. The FRSTL created an oil fund (FEIREP) with the objective of stabilizing fiscal revenues, repurchasing debt and saving some funds for education and health. Quantitative rules were introduced on the growth of the Central Government real primary spending (3.5 percent a year), non-oil deficit reduction (0.2 percent a year), and debt ratio reduction (toward a ceiling 40 percent o f GDP in the medium term). The Law also introduces constraints on subnational debt and rules on fiscal transparency. The creation of the fundis positive for preserving government's net worth. However, estimates show that the current rules might force the government to save too much at certain times and not enough at others; that resources allocated to its stabilization (countercyclical) component are less than optimal for providing full insurance; and that if its design were improved, the mandatory reduction of the non-oil deficit at an annual pace of 0.2 percent of GDP might be eliminated once public debt reaches 40 percent. In the short term, however, making the fund comply with its mandated role i s critical, and any change to its rules for political reasons would damage its credibility. 14. A preliminary assessment indicates that compliance with fiscal rules has been reasonable so far. In 2004, the rules governing planned real primary spending and the non-oil deficit were met. The scheduled reduction of public debt-to-GDP ratio was also on track in 2003. On the execution side, however, performance was mixed: the executed real primary expenditure, instead of increasing, fell by almost 1 percent o f GDP in 2003, and the non-oil deficit reduction was not achieved in 2003, after two consecutive years of reduction. However, the reduction in the public debt-to-GDP ratio of about 5 percent of GDP in 2003 was consistent with high primary surpluses devoted to reach the mandated ceiling target of 40 percent o f GDP in2006 (Figure ES.2). iv Figure ES.2 Debt Sustainability under the FiscalRule 15. Despite this positive 10.0 performance, debt sustainability g 2000 remains vulnerable to shocks, as a 8.0 debt sustainability simulation 2ii \ 6.0 illustrates. A 50 percent drop in the price of oil from US$24 to US$12 per 4.0 ?? barrel would lead to a drop in export g 1999 2.0 proceeds and to a 33 percent 'C a I 1997 8z depreciation of the real exchange rate. O.O I Tradable goods would become more -2.0 expensive relative to nontradable 0.0 20.0 40.0 60.0 80.0 100.0 120.0 goods and this will lower the value of Debt / GDP output-that has a large nontradable component-thus leading to an Source: World Bank staff's estimates. Iincrease in the debt-to-GDP ratio. The 50 percent drop in the price of oil would also require an increased primary balance from 4.5 to 5.2 percent of GDP, that is, almost 1additional point of GDP, just to sustain 2003 levels of debt. And if this shock were accompanied by an increase in200 basis points ininterest rates and a fall o f 1 percentage point in GDP growth, then the required primary surplus for debt sustainability would rise to about 7 percent o f GDP. These numbers illustrate potential risks for the future. Pro-Poor Expenditure and the Roomfor Additional Fiscal Space 16. The Government recognizes that the present fiscal stance is inconsistent within the framework of a poverty reduction strategy. Its objectives are to take advantage of the process leading to the PRS to improve the amount and quality of public expenditures: level, composition, andtargeting. To help assess these objectives, this section addressesthe following issues: 0 Trends in social outcomes, particularly ineducation andhealth 0 Trends inpro-poor expenditure 0 Findingfiscal space to increase spending for poverty reduction Trends in Social Outcomes 17. Educational outcomes continued to improve during the 1990s and into the new millennium: 0 There has been continuous growth inthe average level of schooling since the 1970s: in 2001, the average adult had completed 7.3 years of schooling, up from 6.7 years in 1990. This level i s above the Latin American mean, and i s about the same as East Asia. 0 B y 2001, the gender gap had practically been closed: 7.5 years for males compared to 7.1 years for females. Educational levels of the female population have risen much faster than that of males, such that, in terms of net enrolment rates, girls already outperform boys at all educational levels. 0 Net enrolment in primary education increased from 88.9 to 90.1 percent between 1990 and 2001, approximating the MDGof primary education for all. V 18. Health indicators have also improved: 0 Life expectancy at birth increased from 48 years to 72 years between 1950 and 2000. This upward trend was sustained duringthe 1990s, adding another 5 years to life expectancy. 0 Parallel declining trends are found in child and infant mortality rates. The overall mortality rate dropped from 13.8 per 100,000 inhabitants in 1960 to 4.5 per 100,000 inhabitants in 2001. This rate did not change much during the 1990s. In contrast, since 1970, the infant mortality rate fell by 70 percent inEcuador, which i s an impressive achievement. The infant (aged 0-1) mortality rate has followed an almost linear trend since 1950, reaching 33 per 1,000 live births during 1995-2000, down from 140 during 1950-55. Child (aged 1-5) mortality rates follow similar trends (WHO 2003). 0 The drop in infant mortality coincides with a long-term decline in fertility rates. Fertility dropped from almost 7 per woman in the 1950s and 1960s to 2.8 during 2000-05. During the 1990s, fertility dropped faster in rural than in urban areas, but the rate is still 1.5 times higher for rural women. 19. Important concerns, however, remain in the education sector. The transition rates from primary to secondary education and from secondary to tertiary did not improve in the 1990s. Significant disparities remain, particularly affecting rural, indigenous, and black populations. The average level o f schooling of the rural population i s less than half the one of the urban population, and the gap i s even larger for the indigenous and black populations. This i s also the case for illiteracy rates. Overall education quality i s poor, with math and language test scores worsening between 1996 and 2000 and starting from an extremely low baseline (Ecuador scores lowest for the Latin America region). Internal efficiency indicators, measured by desertion and repetition rates, have also worsened, with the number of years pupils need to complete primary education increasing from 6.7 years to 6.9 years between 1995 and 2001; and higher dropout by girls in secondary rural schools and by boys in urban schools, seemingly for economic reasons. Finally, retention rates and education quality also appear affected by the high rate (14 percent) of teacher absenteeism and frequent teacher strikes inEcuador 20. Similarly, concerns appear in the health sector that will put pressure for additional financing. The decline in fertility and the increase in life expectancy are changing Ecuador's demographic profile. The causes of mortality are moving away from traditional child diseases (malnutrition, respiratory and infectious diseases) toward diseases associated with higher levels o f economic well-being and urban lifestyle (cardiovascular and cancer health risks). Preventable diseases remain the main causes o f child (1to 5) and infant (0 to 1)mortality. The prevalence of A D S has increased, with 10 times more cases reported in 2002 than in 1990. Finally, malaria trends remain closely associated with the occurrence of the El Nifiu effect. 21. Selected MDGs in education and health are within reach, however, especially if supported by additional social expenditure, well targeted, effective, and financed by low- cost programs explicitly linked to specific outcomes. Positive educational and health outcomes have been obtained despite low education and health budgets, apparently poorly functioning education and health systems, a significant amount of non-pro-poor spending in both sectors, and a high incidence of malnutrition. Continued overall improvements in education, urbanization, fertility rates, and sanitary conditions explain these seemingly paradoxical vi outcomes. However, no linear extrapolation guarantees that these trends will continue. This i s why a pro-poor shift insocial spending i s desirable. Trends in Pro-Poor Expenditure 22. About half of social spending and all subsidies to basic services are not pro-poor. In a context of fiscal adjustment, making better use of resources i s essential to reduce poverty. Government spending on education and health could be better spent to achieve improved educational and health outcomes and greater equity. In education, while primary education and to a lesser extent secondary education are pro-poor, spending in tertiary education i s heavily skewed against the poor. This i s all the more worrisome considering that spending in primary and secondary education has had a constant share in the total sector budget since 1995, but higher education has received about a 33 percent increase over the same period. 23. Incidence analysis also FigureES.3 Incidence of Social Expenditure,1999 points out to significant non pro- poor spending in social expenditure. Takingthe difference Schd breakfacl between the richest and poorest m wschmi quintiles, it appears that the school breakfast, primary school spending, BO"0 and the Bono cash transfer, in that Healthcare SSC order, are the most pro-poor programs (Figure ES.3). This Healthcare MSP happens despite significant secondaryEchml targeting problems in their implementation and poor nutritional Cmiunggac impact on their beneficiaries, specially in the school breakfast. Heathm e E S S At the bottom of the classification, UDlvenlIy IYllim university education, IESS health care and the cooking gas appear to 0 10 20 M 40 50 60 70 80 90 Source: Table 3.1. be the most non-pro-poor outlays. 24. Furthermore, none of the three subsidies to basic services-water, telecom, and electricity-caters to the poor, and they are highly distortionary from an efficiency perspective. These subsidies represent a big drain for the government resources: about 1.3 percent o f GDP. The implicit subsidy for telephone service i s the most unequally distributed, followed by the water subsidy. Electricity i s the largest subsidy. While tackling the telecom and electricity subsidies i s a priority for the central government, dealing with the water subsidy requires collaboration with subnational governments. 25. The efficiency of expenditure on basic infrastructure also has significant shortcomings. There are high losses in the power sector due to theft and inappropriate billing; the water and sanitation sector has massive shortfalls o f resources in non-wage and maintenance expenditures due to significant cuts in transfers to municipal governments, and the telecom vii sector has low operational surpluses due to low and distortionary rates against the poor. Telephone penetration was almost 10 times lower in the poorest quintile and only 1 out of 20 rural people has telephone service. Finding Fiscal Space to Increase Spendingfor Poverty Reduction 26. Rigid expenditure limits fiscal space for poverty reduction. Central government expenditure has become increasingly rigid, leaving almost no fiscal space for development needs. In 2004, wages and salaries account for 32 percent of total spending, earmarkedtransfers, inertial services and investment are 34 percent, and debt service i s 30 percent, which adds to 96 percent of the total budget, up from 86 percent in 2001. This means that the non-rigid and non- inertial spending barely represents 4 percent o f the total budget, leaving a very small residual fiscal space-no bigger than half a percent o f GDP-for freely allocated public investment. Should this pattern continue, and the residual fiscal space that could potentially be allocated to investment inpoverty reduction would fully disappear by 2006. Reversing this situation points out to the urgent need to explore possible sources for creating additional fiscal space within the fiscal aggregate ceiling allowed by the fiscal rule. Table ES.l Potential Sourcesof Fiscal Space and EstimatedAnnual Impact Measures Percentof GDP On-budget Curb capital spending ratio toward its "structural" level (budget reallocation) 0.2 Interest savings from debt repurchase 0.2 Reduce defense spending to end-1990s level I.o Make optimal use of public spending (Competitive-based Fund) 0.2 Off-Budget Reduce selected off-budget earmarking o f oil revenues 0.6 Incorporate all oil-subsidies to budget (gas, diesel, and electricity) 2.3 Eliminate25 percent o f overall tax exemptions 1.o VAT 0.7 Internal 0.3 External 0.4 Income 0.3 Firms 0.2 Individuals 0.1 Rationalize spending o f ORDs NA Integrate 10percent o f subnational spending with national priorities 0.4 Allocate resources from Solidarity Fundto the PRS 0.1 TOTAL 6.0 Source: World Bank staff calculations. 27. The Government could create fiscal space of up to about 6 percent of GDP. This i s a considerable margin, considering that only 1-2 percent of GDP would be needed to achieve selected MDGs in education and health (Table ES.l). These estimates assume that the Government decides to maintain a constant tax burden, curbs expansionary spending in the payroll and pensions, and improves the pro-poor content of expenditure policies. The additional fiscal space would also contribute to a further reduction of non-priority expenditure required to compensate for any revenue loss emanating from a fall in oil prices or another external or ... V l l l domestic shock. In addition, the Government could create fiscal space for increased pro-poor spending, especially in the education, health, and social protection sectors, by: (a) revising the current allocations of social expenditure and programs through the development of a competitive-based fund; and (b) malung better use of the available targeting instruments, like SELBEN, to unify criteria and consolidate programs. Notice however, that whatever solutions are adopted, they will have to be accomplished within the annual fiscal ceiling mandated by the FRSTL. The existence of this ceiling, and the little room for tax reform implies that the main effort will have to come from expenditure shiftingactivities. 28. A n input-output model developed as part of this report identifies the main inputs that determine achieving key MDGs. It is possible to identify inputs required for reaching at least the three MDGs of universal primary and secondary education enrolment and reduced infant mortality. These can be achieved with four cost-effective programs: teacher training and the expansion of the Bono de Desarrollo Humano for primary and secondary education, and expansion of the coverage o f the immunization andFree Maternity programs for infant mortality. Public Expenditure Management (PEM)and Other Institutional Issues 29. Developing an effective poverty reduction strategy for Ecuador requires, as a precondition, an overall reform of the budget process and, more broadly, of all levels of PEM. A sound PEMis the key policy instrument that articulates the country's fiscal ceilings and rules with, on one hand, priorities reflected in the budget and, on the other hand, improvements in public sector performance and service delivery. Hence, PEM reform requires an enhanced performance o f the budgeting system, rapid upgrading o f its budget and financial management procedures, a complete overhaul of budgeting procedures by social agencies in charge of priority social programs and of provincial and sectional governments receiving transfers, transparent information access at all levels o f government to allow results-oriented budgeting in the future, and, only when these reforms have gained ground, a multi-year budgeting framework (MYBF) that would allow aligning expenditure inputs with expected social outcomes. At present, Ecuador i s not ready for a MYBF. 30. Since 2003, Ecuador has taken steps to improve its overall PEM. The passage of the Fiscal Law in 2002 set quantitative rules for budget formulation, laying the groundwork for multiyear budgeting, and requiring subnational entities to submit monthly revenue and expenditure reports. The country has been prudent in its budget formulation and assumptions; while attaining with an acceptable aggregate level of budget under-execution (below 5 percent); It has reinitiated the extension and modernization of the coverage of its integrated financial management system (SIGEF) with the goal of producing consolidated balances for 90 percent of the central government by 2005; A single database for central government-financed public investment has been built; An inter-bank payment system (IPS) of public employees located at the Central Bank has started to develop a central registry database for all government employees at SIGEF; and ix 0 CONTRATANET, an electronic public procurement system, has been set up on a pilot basis, initially as an informational-not yet transactional-system. 31. Despite these improvements, a standard global assessmentcarried out as part of this report shows that the country ranks poorly in all but one of the 16 international PEM benchmarks. Benchmarks refer to overall budget formulation, monitoring, execution, control and reporting procedures. A survey developed by this study shows that the country ranks in the bottom tier, even when compared to heavily indebted poor countries (HIPCs). The survey findings are confirmed by parallel studies in standard transparency rankings (ROSC) preparedby the IMF, and the Country Financial and accountability Assessment (CFAA) jointly prepared by the World Bank and the Inter-American Development Bank (IADB). 32. The most important weaknesses are: (a) poor and inertial budget planning, with a bias infavor of defense and security forces and against spending inthe social sectors; (b)the presence of significant off-budget funds; (c) poor Treasury management, reflected in arrears, cash rationing and long delays inthe transfer of resources to social programs; (d) absence of a results- oriented framework; and (e) an outdated integrated financial management system that does not allow timely and reliable reporting on budget execution, which affects monitoring and limits transparency, control and public oversight of fiscal accounts. 33. Poor performance of social expenditure in Ecuador is closely linked to PEM shortcomings. Recent reviews o f international experience with poverty reduction strategies have concluded that in many countries, the practice of PEMi s an obstacle to the achievement of poverty reduction objectives. Ecuador i s no exception. Failures in the budget process and institutional bottlenecks systematically lead to underexecution of social programs. These shortcomings result in underbudgeting or in long interruptions and delays in the channeling of budgeted resources. Perhaps the most important failures are unrealistic budget planning, wide variations in deviations between budgets approved and executed-with a bias in favor of defense and security forces and against spending in the social sectors-a lack o f effective interventions resulting from budget fragmentation through a myriad of overlapping social programs and the presence of significant off-budget funds, and delays in the actual transfer of resources, arising from cash rationing and poor execution capacity at the level of line agencies. Arrears have been declining since 2000, but a sizable financial gap of about US$548 million still remained for 2004 by mid-year. 34. Weaknesses in the budget processes and institutional bottlenecks play a major role in the poor performance of social programs and municipal spending. Recent reviews of international experience with poverty reduction strategies have concludedthat in many countries, the practice of PEM i s an obstacle to achieving poverty reduction objectives. Ecuador i s no exception. A review of budgeting procedures in both selected social programs and subnational governments done as part o f this report also identifies important shortcomings. 35. In the priority social programs, shortcomings are multiple. A lack of effective interventions and budget under-execution result from budget fragmentation through a myriad of overlapping social programs, which have grown increasingly fragmented and disorganized. There are 45 social programs and some are duplicative. Long interruptions, delays and deviations exist in the channeling of budgeted resources to priority social programs. Cash constraints are particularly acute in the first semester o f the year. Excessive bureaucratic controls also play a role in delaying compliance with budget allocations. Most performing social X programs are those which have their own budget execution capacity (and financing mechanisms) and little intermediation from ministries. 36. In provincial and municipal governments, budget procedures replicate similar weaknesses observed at the central government. This is the case in terms of inertial budgeting, low predictability o f transfers, absence o f national directives, and poor reporting. However, this report finds that since 2003, MEFhas over-complied with committed transfers to provincial and sectional governments, which reflects that fact that Treasury's cash rationing has been unequally applied in the public sector. Unfortunately, this has not been accompanied by increased accountability and responsibility, thus weakening the framework for fiscal discipline at the subnational level and opening the door for wasted and unreported resources transferred to subnational governments and for irresponsible subnational borrowing. 37. Fiscal discipline and implementation of an effective poverty reduction agenda could be facilitated by a reform of all levels of PEM. Sound PEM i s the instrument that articulates the country's fiscal ceilings and rules with, on the one hand, priorities reflected in the budget and, on the other hand, improvements in public sector performance and service delivery. To ensure the fiscal discipline needed to support dollarization, the Government should first set the annual ceiling o f fiscal balances. Then, based on such a constraint, a shift to pro-poor allocations of expenditures would be easier to implement with an improved expenditure management process that is strategic in focus, feasible in terms o f available fiscal space, and results-oriented with proper monitoring and evaluation mechanisms. Conclusion and Policy Recommendations 38. Ecuador's impressivefiscal performance of 2003 is encouraging, but fragile. Several structural bottlenecks could impede fiscal discipline and recovery, which i s a pre-condition to develop a poverty reduction agenda. Tax eannarlungs and exemptions and an expansive payroll and pensions bill have reduced to a minimumthe available fiscal space for development needs. Reversing poverty trends i s critical for the country's stability, and this can only be achieved with well-targeted, effective and efficient pro-poor programs. The status quo i s not an option for poverty reduction. 39. Preserving a sound fiscal position and deepening positive social outcomes is well within reach. Among the country's many strengths are: a prolonged oil windfall; the existence of and compliance with fiscal rules; decreasing arrears that should fully disappear in 2004, substantive progress on social outcomes despite decreasing budgets; and a series of on-going reforms on budget management. Last September an important test was the Government's capacity to successfully resist short-term election-motivated pressures for amending the fiscal rule in order to misuse FEIREP resources. The lessons from international experience on the implementation of poverty reduction strategies suggest three guiding principles 40. First, the GOE needs to articulate the message that its fiscal management reforms are designed to help the poor. If reforms are to succeed, they have to be pro-poor. Ecuador's fiscal stress and poor budget management i s deeply rooted in a governance system benefiting the elites, be it reflected on pro-rich subsidies, especially on basic infrastructure; off-budget operations that prevent transparency and foster corruption, or regressive transfers to subnational governments explained by party politics. The challenge for the Government is to provide more effective, efficient, sustainable and equitable assistance to the poor. xi 41. Second, the reform o f the fiscal management agenda needs to be designed and implemented with a medium-term view and national consensus. Piecemeal, short-term reforms can only bring short-term, often not long lasting gains. For example, the creation of Contratanet allowed the surge o f an informational system about public procurement that has improved its transparency, but the more difficult task to converting it into a transactional system still remains to be undertaken. The establishment o f a commission to draft such bill and the commission's decision to consult with civil society on the draft are steps in the right direction. 42. Third,the implementation of the PRS has to be monitored inan transparent way. Sharing reliable and timely information is as critical as the strategy itself. Inthe absence of transparency, the strategy loses credibility. This requires a combination of several steps including the development of benchmark indicators, not only fiscal, but especially social-inputs, outputs and outcomes. These should be designed in such a way that they can be monitored on a regular basis and reported before the Legislature and civil society. In addition, client surveys could be commissioned to assess the quality o f service delivery. All reports should be made public. Policy Recommendations 43. To address the challenges faced by the Government in its fiscal policy, this report recommends an agenda of policy actions that would promote the three key objectives of fiscal stability, pro-poor spending, and budget management. While fiscal reforms face formidable political-economy and institutional obstacles in Ecuador, increasingly large segments of public opinion may be won over to the pro-reform camp if a close connection between the need for fiscal adjustment and the creation o f the fiscal space needed for a poverty reduction strategy i s spelled out to the citizenry. 44. The medium-term objective of fiscal policy in Ecuador remains to strengthen and preserve the sustainability of the fiscal accounts. This could be supported by an implicit or explicit Fiscal Pact on the following general goals: deal with the public sector's insolvency risk; eliminate the structural bias toward expenditure expansion in the management of public finances; address the issue of the Treasury's short-term liquidity problems; and comply with the fiscal rule requirements in the short tem (thereby buildingcredibility). The fiscal rule could be amended in a few years to make it a sharper, more coherent, and more powerful instrument for fiscal consolidation. More specific recommendations in this area include the following: Ecuador needs to lower its insolvency risk by producing and preserving the high primary surpluses needed to gradually reduce the debt-to-GDP ratio to the sustainable levels of 40 percent of GDP in the period 2006-2007. An important caveat i s that while the FRSTL sets 40 percent as the goal, and this is a reasonable achievement, an additional 5 percent reduction would be desirable for unexpectedcontingencies and shocks. The Government should persevere in the pursuit o f a comprehensive tax reform in the medium term, especially once oil prices start returning to their historic level. This would create additional fiscal space. As the adoption of revenue measures is politically constrained in the short-term, curbing the bias towards expenditure expansion, especially on the payroll and pension benefits, should be the top key priorities o f fiscal policy. xii Regarding FEIREP, in the short-term strict compliance with the mandatory use of 70 percent of FEIREPfunds to repurchase the most expensive debt first (particularly in global bonds), i s needed to reduce interest payments. If the country complies with projected debt reduction repurchases, it would save an average 0.2 percent o f GDP, which would be available for pro- poor programs. Inthe medium term, legal amendments to the FEIREPand fiscal rules would improve the fund's effectiveness. Possible amendments include the redefinition of the reference price for crude oil that i s included in the budget. Another i s to gradually approach the criterion of the 3.5 percent real growth rate for primary expenditures with reference to the executed budget of the previous year, as opposed to the approved budget. A third is the elimination o f the 0.2 percent of GDP mandated non-oil deficit reduction, once the country reaches the 40 percent o f GDP ceiling and oil proceeds for stabilization purposes can be increased. Given the large country risk premium on public debt, and findings from the debt sustainable analysis, oil revenues should reduce public debt to 40 percent of the GDP and, later, accumulate financial assets that would eventually allow the country to have a small non-oil deficit, even after oil reserves are depleted. The external debt buyback should be accompanied by additional fiscal space to buffer the impact of future shocks and streamlined expenditures. Debt buyback i s no substitute for expenditure rationalization. Besides, by following the proposed debt strategy, it may be more difficult and costlier to borrow internally if Ecuador faces a shock, because developing countries, in contrast to developed countries, cannot borrow commercially when they suffer a shock. The Treasury's short-term liquidity difficulties badly affect the authorities' credibility and the country's reputation. Given the specific composition of the public debt, closing the liquidity gap is critically dependent on the Government's willingness and ability to design programs of substantial structural reforms. This would be the basis for an agreement with the IMFand would enable the country to obtain rapidly disbursing and freely disposable funds. This is also the road to enhanced credibility for the medium-term debt reduction plan. 45. Shifting public expenditure toward a pro-poor focus involves actions on several fronts. Basic Infrastructure Subsidies: (a) reduce the total electricity subsidy providedto consumers below a maximum amount of electricity consumption, since the actual ceiling of residential consumers below 300 lulowatt-hours i s too high to target the truly poorest households; (b) reduce tariffs for public telephones, which are 10 times higher than tariffs for residential users, and eliminate cross-subsidies through completion of the tariff rebalancing between domestic and international rates approved by CONATEL in 2003; and (c) reduce and make transparent cross-subsidies in the highly decentralized water and sanitation sector, linked to operational performance, while defining the amount of subsidy allocated per connection considering the size and income level of the population. Expenditure on social services: (a) freeze or reduce subsidies to university tuition to finance access for poorer groups, for instance, to secondary education; (b) increase the very small budgets o f pro-poor programs, like primary education, the school breakfast or the Free maternity Law; and (c) introduce results-oriented budgeting to all pro-poor programs, b y definingmonitoring indicators, undertaking regular monitoring and evaluation mechanisms, and allowing strong civil society participation. ... Xlll o Link and protect budwet support to MDG goals and improve its performance: (a) provide additional budget resources (for about 0.1-0.2 percent of GDP) to primary education and infant mortality; (b) focus additional resources on secondary education, child malnutrition, basic health, and child care, which would raise additional budget needs to about 0.8 percent of GDP in 2004, 1percent of GDP in 2005, and almost 2 percent of GDP in 2007; and (c) define a set of performance indicators that would allow their progress monitoring. Indicators should result from a combination of a consensus-building exercise and international expertise. An important conclusion of this approach is that not all sector and program budgets need to be linked to performance indicators, but only the ones that are critical for achieving the goals of the PRS. 46. A reform of public expenditure management is essential to accompany poverty reduction. This includes implementing a budgeting system that reverts inertial expenditure; rapidly upgrades the budget and financial management system (SIGEF); overhauls budgeting procedures by both social agencies in charge of priority social programs and subnational governments receiving transfers; makes information access transparent at all levels of government to promote participation; and, only when previous reforms have gained ground, establishes a multi-year budgeting framework that align expenditure inputs with expected social outputs. More specifically, the following measures are suggested: o On overall public expenditure management: (a) create a Cash Committee at MEFcomposed of representatives of all offices that manage budget design and execution; (b) revert inertial budgeting through already adopted freezing o f the wage payroll; (c) gradually integrate off- budget activities into the Treasury's Single Account (Cuenta Unica), especially non- constitutional earmarkings and subsidies currently channeled through PetroEcuador; and (d) overhaul SIGEF to promote proper registration and timely information of budget execution. An important step inthis direction has been done with their publication inthe 2005 budget. o On budget management in social programs: (a) review the overall budget protection policy with an initial assessment of the number o f social programs and the amount of resources allocated to them effectively representing government priorities, with programs receiving the minimum amount of resources needed to achieve their goals; (b) merge or eliminate duplicate social programs to reduce resource waste following the example of the nutritional ones under the Sistema Znntegrado de Alimentacibn y Nutricibn (SIAN); (c) eliminate cash constraints in the first half of the year, a recurrent and severe problem in many agencies; (d) rationalize, simplify, and ifpossible automate, budget procedures and forms for requesting reimbursement of payments; and (e) consider the creation of "virtual" poverty fund. Obvious candidates for elimination are those programs that show a significant degree of low budget execution and/or poor targeting. o On budget management in provincial and sectional governments: (a) design a strong regulatory and institutional framework that clearly assigns expenditure responsibilities in line with subnational governments' administrative capacity; (b) condition delivery o f some or all transfers on timely and reliable budget reporting by subnational governments, as mandated by the FRSTL and following up upon recent MEF efforts for building a database on subnational fiscal accounts; and (c) promote responsible subnational borrowing by establishing further norms under which the central government can intervene in local governments when and if they violate the fiscal rules, and clearly excluding the possibility of a bailout. xiv A Selection of the Key Policy Recommendations Based on the analysis done inthis report, and among the set of recommendations proposed, the following sub-set of sequencedpriority actions are suggested: I. High Priorityfor thenear term (612 months) II. Priorityfor the medium term (1-3 years) 1. Obtain primary surpluses between 4.5-5 1. Amend the Fiscal Transparency, percent of GDP.This could be achieved in Stabilization and Responsibility Law, the short term by reducing spending and by ensuring that: (a) quantitative rules apply to complying with the fiscal rule. Current executed, not only approved spending, (b) primary spending should be curbed by resources for the anti-cyclical role of FEIREP preserving until 2005 the on-going freeze of are augmented, once the ceiling 40 percent the public payroll and maintaining pensions debt-to-GDP ratio i s attained; (c) additional at their 2003 level inreal terms. rules are introduced for subnational 2. Comply with mandated debt reduction govemments; and (d) provincial and municipal using70 percent of FEIREPproceedings for govemments comply with the transparency repurchasing of external debt. requirements contained inthe fiscal rule. 3. Announce a draft budget reform bill to: 2. Adopt decisions on policy alternatives (a) integrate off-budget activities, especially proposed for creating fiscal space in the subsidies paid by PetroEcuador; (b) freeze context of an implicit or explicit Fiscal Pact or reduce non-constitutional budget (see Table ES.l). This implies expenditure earmarkings; and (c) reduce central shifting activities. Tax reform, in particular, government 40 percent contribution to IESS. should expand the tax base by reducing tax 4. Create a Treasury Committee that exemptions, and continue improving tax manages and makes transparent current cash administration. strappedbudget execution. 3. Increase the level of education, health and 5. Ensure SIGEF overhaul by: (a) fulfilling social protection budgets, while raising i t s commitment to consolidate fiscal their quality and share devoted to pro-poor accounts; (b) producing timely programs. This implies to focus additional disaggregated reports in the MEF website; spending on (a) teachers' training and and (c) completing its re-design for moving secondary education, on health provision by to an internet-based system. the Free Maternity Program, and on the 6. Undertake a comprehensive review of revamped Bono; (b) pro-poor programs budget protection policy to: (a) weed-out (possibly supported by a Competitive-based non-performing priority social programs and Fund); (c) other priority public investment, select the most performing ones under a especially if it i s donor-financed, which would "virtual" poverty fund; (b) merge most of require close coordination through sector the remaining overlapping ones following approaches. the example of the nutrition ones under 4. Freeze or re-target non pro-poor subsidies SIAN; (c) guarantee a high level of and spending on non pro-poor social execution of their budget agreed for 2004 programs in nominal terms at their 2003 and 2005; (d) commit additional resources level, especially for higher education and required by compliance with selected MDGs pensions. Retargeting of the cooking gas, inthe 2005 budget. diesel, and electricity subsidy i s a priority. 5. Develop a strategic vision for electricity, water, sanitation and telecom sectors, accompanied by a time-bound implementation plan, to: (a) expand coverage; (b) improve service provision; (c) reduce regressive subsidies; and (d) allow for private competition for service provision among suppliers and accountability to users. xv Chapter 1 Fiscal Policy ina Dollarized Economy 1.1 On January 9, 2000, Ecuador decided formally to adopt the U.S. dollar as its national currency, thus converting fiscal policy into the centerpiece of its macroeconomic management. This was a major institutional change that created a new context for both government economic policy and private economic activity. The banking, currency and debt crises that precipitated the move to dollarization were triggered by a combination of exogenous (both external and weather-related) shocks. These shocks included: (a) the sudden stop incapital flows into Ecuador, (b) the fall in oil prices, and (c) the supply shock from the El Nifio phenomenon. These shocks also took place at a moment where expenditures were increasing rapidly in response to shocks. The decision to dollarize stabilized expectations, and economic activity began to turnaround. 1.2 Since dollarization the economy has recovered and macroeconomic balances, including the fiscal balance, have improved. The recovery i s due to highoil prices, low international interest rates, a trend toward a recovery in the financial system, growing remittances from expatriated workers, strong private investment (much of it associated to the construction of a new oil pipeline), an increase in real wages, and recent sound fiscal policy leading to the stabilization of expectations. Real gross domestic product (GDP) growth during 2000-03 averaged 3.5 percent; inflation declined from 91 percent in 2000 to 6.1 percent at the end of 2003; the capital account benefited from the foreign direct investment (FDI) flows associatedto the construction of a new oil pipeline; and after the surge inimports brought about by the pipeline project came to a halt (because of project completion), the current account deficit of the balance of payments dropped from 4.8 percent of GDP in 2002 to 1.9 percent of GDP in 2003. 1.3 Inlight of these positive developments, the issue arisesas to whether the country has finally taken a path leading to long-term, sustained economic growth and can at last feel assured of its economic viability and future stability. The short answer i s that the country i s making progress along the path toward improved stability, but the economy i s still fragile, major vulnerabilities remain, and there i s still some way to go to reach the goals of sustained growth and durable stability. As a first step to diagnosing the country's vulnerabilities, this chapter discusses historical and structural developments in fiscal policy and public expenditure. The chapter i s divided into three sections. The first section reviews the historical background that preceded dollarization; the second section discusses the main challenges the country faces in strengthening the institutional framework for dollarization; and the third section explores the structural features of fiscal policy in Ecuador, emphasizing the political economy determinants of fiscal performance. In doing this, the chapter claims that the ultimate root of Ecuador's fiscal predicament lies in the political-economy regime that emerged in the aftermath of the oil boom of the seventies. A. The Ecuadorian Economy: Some Historical Background 1.4 The discovery of oil and gas in the late 1960s and the oil boom of the early 1970s changed the landscape of Ecuadorian society.' The value of petroleum and natural gas output, which was virtually zero in 1971, jumped to 9.5 percent of GDP in 1973 and 23.0 percent of GDP in 1974, reflecting both the start-up of large-scale production and the fourfold increase in world oil prices. The newly found wealth gave rise to a rent-oriented, conflict-prone, and deeply fragmented political-economy regime that 1. The outline of developments in the 1970s and 1980s draws on World Bank (1984, 1991, and 1993); and Beckerman(2002). 1 would come to dominate public policy making in the next decades. A major feature of this regime was that fiscal income from oil was usedto finance an increaseinthe size of the State and to subsidize private spending through, among other means, low prices of domestic petroleum products, reduction in the taxation of non-oil activities, and a number of other subsidies. In light of the large amount of resources initially available, this strategy was successful at first. Economic growth in the 1970s was spectacular: the average annual growth rate of real GDP during 1971-80 was 9.1 percent.2 Average inflation was a moderate 14.0 percent. Growth of manufacturing output was also stimulated by the adoption of import- substitution policies, which shielded a buddingbut inefficient, domestic-market-oriented, manufacturing sector. 1.5 But, as the economy boomed, the seeds of future crises were planted by the authorities' fiscal behavior. Despite large revenues from oil, Ecuadorian governments incurred substantial fiscal deficit^.^ Successive fiscal deficits were covered by rapidly increasing external borrowing. External debt more than doubled between 1970 and 1975, then increased ninefold between 1975 and 1980, grew another 74 percent between 1980 and 1985, and still another 25 percent between 1985 and 1988. 1.6 The eighties were a decade of severe external shocks and slow growth. In 1982, the terms of trade declined sharply, petroleum receipts stagnated and there was a sudden cessation of credit from external sources. The collapse of oil prices in 1986 precipitated a prolonged period of fiscal crisis and macroeconomic instability that peaked in 1988. In January 1987, the country stopped servicing its external debt with commercial banks, and interest arrears climbed to almost US$1.O billion by the end of the Febres Corder0 administrationinAugust 1988. Inthe meantime, inflation was on the rise and reached 75.8 percent in 1989. 1.7 The period 1988-1991 was marked by initial, but unsuccessful,efforts at stabilization: The government also undertook significant structural adjustment initiatives, including a partial tax reform, trade liberalization, and progress toward financial sector liberalization. However, inthese years, inflation persisted at high levels, and, in the run up to the 1992 elections, government expenditure rose sharply. This was, in sum, a transition period in which some advances toward stabilization and structural reform were made. 1.8 Only in 1992 did the Ecuadorian government launch a relatively successful macroeconomic stabilization program. On the structural reform front, financial liberalization was basically completed, the foreign exchange market was fully unified, trade liberalization was deepened, most of the smaller public enterprises were privatized, and some progress was made in State modernization through the 1992 Public Budgets Law and the 1993 Modernization of the State Law. The government also made substantial progress in controlling public finances. It reduced non-financial public expenditure, in part through a program that reduced public sector staffing by nearly 10 percent. Inflation was reduced to the 20 to 30 percent range from the 40 to 50 percent range that prevailed in the early 1990s (de la Torre and others 2001). As Beckerman (2002) points out, these were hard-won reforms, secured inthe face of broad political opposition. 2. Real GDPgrowth in 1972 was 14.4 percent, and in 1973 was a staggering 25.3 percent! 3. There was a deficit in the accounts o f the non-financial public sector every year from 1975 to 1982. After the fiscal adjustment episode of 1983-85, deficits resumed with a vengeance, and the average deficit during 1986- 88 was 6.6 percent of GDP, untila new adjustment effort reduced it to 0.3 percent of GDP in 1989. 4. Real public spending was reduced 5.0 percent in 1988, and was kept relatively constant through 1991. In 1990, the government resisted pressures to expand expenditures when the Gulf crisis generated a large windfall in petroleum prices. Most cuts took place in current expenditures (wages, goods and services, interest, and transfers), and not in public investment. As a result of the stabilization effort, real growth recovered to between 3.0 and 5.0 percent during 1990-92. 2 1.9 In the early nineties, the wave of foreign private capital flows engulfed the Ecuadorian economy? In late 1992, Ecuador joined other emerging markets as a receiver of important flows of capital, which were largely intermediated through the domestic financial system. Private capital flows surged from an average of US$200 million a year (about 2.0 percent of GDP) during 1989-1992 to US$730 million a year (about 5.0 percent of GDP) in 1994. The inflows led to a lending boom that peaked inthe second half of 1994, when bank credit to the private sector was expanding at annual rates of around 60 percent in real terms. The euphoria of the credit boom intensified information asymmetry problems, resulting in a generalized underestimation of the nature and extent of the risks taken by financial intermediaries. With asset prices soaring, the fast expansion of firms' balance sheets during 1993-94 implied much more risk-taking than what was then perceivedby bankers. The inflows included a significant component of short-term speculative capital: non-FDIprivate flows to Ecuador swelled from a cumulative US$12 million (0.1 percent of GDP) in 1991-1992 to US$506 million (3 percent of GDP) in 1993-94. And the stock of registered private-sector debt (much of which was short-term in nature) jumped form US$258 million at end-1992 to US$832 million at end-1994. Because of their volatility, the flows of short-termcapital introduced a major factor of vulnerability for the economy. 1.10 The lack of an effective regulatory and supervisory framework deepened the vulnerability to volatile capital flows. Three Ecuador-specific shocks,6 compounded by the general slump in confidence brought about by the Mexican crisis of 1995-96, triggered a gigantic reversal in short-term capital flows in the mid-1990s. The accumulated outflow of non-FDI capital in 1995-96 was US$1.5 billion (8 percent of GDP), three times the accumulated non-FDI inflow during 1993-94. Real credit stagnated. Interest rates went up. Highinterest rates sharply raised firms' debt-service obligations, which caused a large drop in the value of f m s and their collateral. Nonperforming loans increased from less than 4.0 percent at end-1994 to 9.0 percent by the end-1996. A medium-intensity banking crisis emerged. Real GDP growth fell from about 6.0 percent in the second half of 1994 to 0.8 percent during July 1995- June 1996. 1.11 The post-1995 crisis period was characterized by worsening macroeconomic imbalances. This was also a period of policy inaction vis-&-vis the reforms needed in the financial system, high political instability, and a slowdown economic activity. According to Jicome (2004), in spite of a favorable external environment, Ecuadorian governments muddled through in 1996 and 1997, keeping the economy and the financial system inan unstable equilibrium. 1.12 Starting in 1998, a new string of exogenous shocks occurred that led to the end-1990s crises. El NiAo had struck in late 1997, destroying agricultural crops, which impaired the assets of several banks of the coastal region. Oil prices sank below US$10 dollars per barrel. But by far the decisive shock was the sudden stop in capital flows brought about by Russia's partial debt repudiation in August 1998. Ecuador's capital-flow reversal was impressive by any standard. Net inflows of about US$2.2 billion in 1998 changed into net outflows of US$1.3 billion, a reversal of US$3.5 billion, equivalent to 20 percent of 1998 GDP, or 56 percent of that year's credit to the private sector. The adjustment in the capital account also affected the current account, which swung from a deficit of 11percent of GDP in 1998, to a surplus of 6.9 percent of GDP in 1999, as the country was forced to undergo a substantial real exchange rate depreciation. 5. The narrative about the 1990s draws on de la Torre and others (2001); Beckerman (2003); JQcome (2004); IMF (2000); andIzquierdo (2002). 6. The three Ecuador-specific shocks were: (a) the war with Peru in early 1995; (b) a prolonged drought that caused a major energy crisis; and (c) the forced resignation of Vice President Dahik, who was considered by market participants at home and abroad as the economic reform leader pur excellence in the Durin BallCn administration. 3 1.13 The Ecuadorian downturn combined three crises. The banking crisis started in April 1998 with the traumatic closure of a small bank. It gained momentum in August 1998 when a medium-size bank was closed after having been unable to honor its obligations and, subsequently, when the largest bank of the system requested assistancefrom the Central Bank. The debt crisis exploded inAugust 1999, when Ecuador defaulted on its Brady bonds as the logical result of rising public sector indebtedness. The currency crisis started in the aftermath of the decision, in February 1999, to float the exchange rate in order to limit international reserve losses. Over the following weeks, the exchange rate lost 30 percent of its value. Eventually, the different dimensions of the crisis fed on each other. The initial runs against the banks prompted the Central Bank to provide liquidity. Fresh liquidity fueled the attack against the currency, which depreciated in the context of the free float. The nominal devaluation adversely affected balance sheets, eroding the solvency of debtors and, ultimately, that of the banks. This prompted more liquidity assistance and reinitiated the devaluation-insolvency spiral. Inthe end, a vicious cycle set in of currency depreciation, corporate insolvency, growing bank insolvency, runs against the banks, banks' declining liquidity, and Central Bank assistance that further fueled the run on the currency. The process didnot stop endogenously. Rather, it was haltedby an exogenous (political and administrative decision), namely, the abandonment of the sucre as the national currency and the formal adoption of the dollar. 1.14 This brief summary of macroeconomic developments in the last three decades before dollarization is useful to identify the key structural patterns that drive Ecuador's macroeconomic performance. First, the dependence on oil is a major theme. The positive aspect of this dependence i s that oil booms are a source of short-term economic expansion, potential fiscal savings and foreign exchange earnings. The downside of this dependence i s twofold. The most salient effect i s to make the economy vulnerable to oil-price volatility. An additional effect is the resultant Dutch disease that causes damage to the competitiveness of the non-oil tradable sectors. A second maior theme i s that the oil- driven boom of the seventies was unwisely managed and this had long-run repercussions that are still felt today. Instead of saving the transitory excess revenues, they were used to finance consumption, the reduction of non-oil taxation and the expansion of the public administration and state enterprises. The collateral of oil was used to contract a huge amount of debt, which today has turned into an inherited burden that is handicapping the conduct of economic policy. A third theme is that, with financial liberalization in the early nineties and with the advent of greater degree of capital mobility in the world economy, the nineties brought on new sources of vulnerability for the Ecuadoran economy. Inparticular, private capital flows came to be capable of driving investment and growth, as in 1993-1994, but also of sinking the economy into crisis and disarray, as in 1998-1999. A fourth theme concerns the exposure to natural disasters as an additional factor of vulnerability and a threat to economic growth and macroeconomic stability. In addition, all these patterns have developed in the context of a rentist, conflict-prone and fragmented political economy, adverse to macroeconomic adjustment and monetary and fiscal discipline, whose characteristics will be presently discussed (see paragraph 1.21below). B. The Role of FiscalPolicy ina Dollarization Framework 1.15 Dollarization created a new framework for both private sector economic activity and public sector policymaking. There is a whole body of literature on the economic costs and benefits of dollarization, which has produced a partial consensus as to the virtues and limitations of this exchange- rate regime, but important differences of opinion remain, and some of the issues raised in the debate warrant further research. Box 1.1 summarizes them.' Unlike most discussions of the issue, the box examines the particular adjustment mechanism of a dollarized economy in the presence of negative external shocks. 7. This discussion draws heavily upon Calvo (1999); IMF(2000); Berg and Borenzstein (2001); and Antinolfi and Keister (2001). 4 Box 1.1. The CostsandBenefitsof Dollarization A definite advantage of dollarization is that it eliminates the risk of a nominal devaluation of the country's exchange rate. Inother words, the risk o f a currency crisis disappears. This is tantamount to saying that one important component of the country risk premium is eliminated, with the consequence that interest rates on foreign borrowing will be lower. However, while the interest premium attributable to devaluation risk will disappear, sovereign risk will not and, therefore, total country risk and hence interest rate risk premia will, in all likelihood, continue to be different from zero. The absence of currency risk does not insulate a country from swings in market sentiment toward its economic policy and/or prospects. Moreover, foreign lenders will normally perceive that there is a risk that the particular country may default on its obligations. The elimination of currency risk makes it possible for both the government and the private sector to borrow at lower cost with positive effects on the private agents' balance sheets and the governments' fiscal accounts. Dollarization also rules out one kind o f sudden stop in capital flows, namely, that particular type that i s motivated by fear o f devaluation o f the national currency. However, dollarization cannot eliminate the risk o f all external crises, as investors may flee because o f fears stemming from lack o f sustainability o f the fiscal position, or about the soundness of the financial system. The elimination of currency risk also removes an important source o f vulnerability inthe financial systems o f emerging market countries. This can contribute to building a stronger financial system, one more capable o f actively participating in international financial markets. Financial integration is then facilitated by dollarization. B y the same token, there is a cost to this improved integration: if the authorities believed that some insulation of the domestic financial system i s needed for the purposes of, say, improving its stability, through, for instance, the imposition o f capital controls, they would find that insulating measures are difficult to accomplish, because it would always be possible for the private agents to convert their assets to dollar cash. Another advantage of dollarization is that the likelihood i s high that the inflation regime that would set in will be characterized by lower inflation rates than under alternative exchange rate schemes. The reason i s simple. Under dollarization there is no domestic monetary authority that could expand the quantity o f money, and thus the greatest potential inflationary factor i s eliminated. As a matter of fact, this is a promise on which Ecuador's dollarization i s finally delivering, as the rate of inflation dropped from 91.O percent in2000 to 6.1 percent in2003. Dollarization also makes commercial integration with the rest of the world easier because it eliminates the transaction costs associated with currency exchange. In particular, in the case o f Ecuador, dollarization is an obvious advantage for pursuing commercial integration through a trade agreement (bilateral or otherwise) with the United States. Finally, dollarization has a well-known disciplining effect on fiscal policy inthat, inthe new system the Central Bank cannot monetize fiscal deficits. However, this does not prevent fiscal indiscipline. This i s because it does not prevent excessive borrowing to finance large fiscal deficits. Thus, dollarization increases the incentive to borrow vis-i- vis what would be the fiscal authority's behavior in the alternative scenario o f an emerging market country that maintains its own currency. In the same vein, Chang (2000) also argues that, under exchange rate arrangements other than dollarization, changes in exchange rates or interest rates make the costs of a lack o f fiscal discipline immediately visible. Dollarization takes those incentives away by allowing the costs o f present fiscal profligacy to be shifted to the future (inthe form of, say, higher future taxes). Among the disadvantages o f dollarization, the most cited are: (a) the sacrifice o f an independent monetary policy; (b) the loss of the lender o f last resort; (c) the sacrifice of the inflation tax; and (d) the loss o f seigniorage Advocates and opponents of dollarization have debated at length on these issues. A fundamental concern is that the adjustment mechanism to a shock under dollarization has a particularly worrisome feature: After a negative external shock (say, a sudden stop incapital flows), a real exchange rate depreciation is necessary for the economy to return to ar equilibrium situation. In a dollarized economy this can only be attained through price deflation. However, if domestic prices and wages are sufficiently inflexible in a downward direction, a prolonged period o f deflation may be necessarj to carry out the required adjustment. Adding to this, there is the connected issue of the possibility o f debt deflation: Ar unanticipated collapse inprices may lead to bankruptcies, even if the borrowing firms are efficient. Thus, if bankruptcj i s costly, debt deflation carries a deadweight loss. Calvo (1999) asserts that debt deflation i s perhaps the most serious threat for a dollarized economy. Another related vexing question is that dollarization does not isolate the non-tradable sectors from changes in the real exchange rate. Although the dynamics i s quite different from the case o f non-dollarizec economies (where the nominal exchange rate shoots up), there will be a relative-price pressure on non-tradables, once domestic prices start their adjustment process after a negative shock. 5 1.15 There should be no doubt that having given up monetary and exchange rate policies, fiscal policy is the most important tool available for macroeconomic management, and that a strong fiscal stance i s therefore a key prerequisite for the sustainability of dollarization, and the best shield against negative shocks. On the other hand, a policy agenda must address key policy challenges of a dollarized economy such as (i) the risk that adjustment to external shocks should take the form of prolonged deflation and (ii) the lack of a lender of last resort for the financial system. On these issues, two general recommendations follow: 9 There isa highprobabilitythatadjustment to externalshockswilltake the formof a prolonged deflation. To prevent this, it i s necessary to introduce structural reforms. Reforms should aim at introducing flexibility in goods and factors markets, but especially in the labor market. This will require a social consensus on how to change prices and wages inperiods of recession. 9 The lack of a lender of last resort requires strengthening the management of liquidity in the financial system. Powell's (2003) proposals are a good starting point. This author suggests that there are definite advantages to centralizing the liquidity that the Ecuadorian financial system has available. He also proposes that the rules and operations of the existing liquidity support mechanisms, including the Liquidity Fund, be improved and that a new contingent liquidity facility be negotiated with a foreign institution. C. Structural ConstraintsonFiscalPolicy 1.16 Ecuador entered the 21st century suffering the deleterious effects of three Figure 1.1 Total Public Sector Net Worth decades of predominantly wrong fiscal (index basedon 2001US$) policies. Throughout 1970-99, there was a 100 sizeable decline in the public sector's net worth, indicating that oil revenues have been largely 80 consumed rather than invested.* Estimates of the decline in real net worth calculated in two 5) 4 60 separate studiesg range between 32 percent and & 48 percent. Figure 1.1 shows the estimated decline trajectories found inthe two studies. This 40 erosion of net worth i s equivalent to having =3 Ecuador financing i t s excess consumption 20 through the depletion of a single resource. Moreover, the problem i s not only that oil wealth 0 was not invested. It was even worse than that: oil 0 b 0 a3 0 m 0 Y c m o o o ou c u c u o o wealth was also used as collateral for a large debt In In 7 7 7 c u buildup. External indebtedness was the road to source: World Bank (2004a). financing a long succession of fiscal deficits since 1975. Thus external debt doubled between 1970 and 1975, then increased ninefold from 1975 through 1980. The public external debt kept steadily rising throughout the 1980s" and early 1990s and reached an unprecedented level of 87.3 percent of GDP in 1995. During 1996-98, the external-debt-to- GDP ratio dropped to an average level of 66.9 percent, as a result of the 1994 debt-and-debt-service- reduction deal with commercial bank creditors, but, with the advent of the 1999 crisis and the precipitous fall in output that year, the external debt ratio climbed to an all-time high of 98.3 percent, the heaviest 8. The discussionof public sector net worth drawson Traa (2003), andWorld Bank (2004). 9. The studiesare Traa (2003), andFierro-RenoyandNaranjo (2003). 10. Dataonthe growthof externaldebt inthe 1980scanbefound inWorld Bank (1991). 6 burden by far among Latin America's 10 largest economies at that time (Beckerman 2002). During 1970-2000, the budget received an estimated cumulative nominal US$23 billion in direct oil revenues. Nonetheless, the public debt never stopped increasing (Traa 2003). 1.17 The available estimates indicate that there was a pause in the declining trend of public sector net worth during 2000-02 (Traa 2003; UNICEF2003). However, with the rate of oil extraction bound to increase with the new heavy crude oil (OCP) pipeline, public sector net worth may resume its declining trend. The decline may even proceed at a faster pace than before, leading eventually to public sector insolvency, unless disciplined fiscal performance prevented this from happening. The fact i s that, with the new pipeline, the remaining oil reserves may last less than 30 years. 1.18 Three other structural factors should be added to have a complete picture of Ecuador's current fiscal fragility. They are: first, the highrevenue volatility causedby external shocks; second, the pro-cyclical character, extreme rigidity, low quality, and limited transparency of spending; and, m,the serious liquidity problems of the public sector that prevent it from keeping current inthe payment of the public debt. 1.19 The relevant exogenous shocks are well known. First, the Ecuadorian economy i s exposed to the volatility of oil export revenues. Second, it i s also exposed to other sources of economic external shocks, prominent among them being those associatedwith private capital flows. Third, various forces of nature have been another standing source of contingency for Ecuador's economy. As pointed out by Beckerman (2002), the country i s prone to earthquakes, volcanic eruptions, landslides, El NiAo phenomenon, and extendedperiods of either drought or excessive rain." 1.20 The ways in which the different exogenous shocks tighten (or ease) the constraints on fiscal policy are also well known. The volatility of oil prices in the international market translates into volatility of oil fiscal revenue in Ecuador. External financial shocks, through their effects on aggregate demand, aggregate income, and the health of the financial system, have a bearing on fiscal performance. For instance, capital inflows stimulate aggregate demand and output, and may thereby result in increased non-oil tax revenue. Capital outflows, by contrast, may throw the economy into recession and thus lead to falling non-oil tax revenues. Furthermore, they can-and this effect may be even worse-augment the government's contingent liabilities associated with the fiscal costs of financial crises. Just to mention another example, international interest rate shocks can increase the cost of the external debt service, as was the case in the early 1980s. Last, random natural disasters have the double effect of lowering tax receipts (via their effect on output) and prompting increased spending (to cope with the emergency). 1.21 Ecuador's fiscal fragility to shocks does not lie in the external shocks themselves, but inthe political economy regime that prevents the country from adopting proper adjustment measures to offset them. Box 1.2 highlights the conflictive character of this political economy, the towering difficulties faced by the national actors to reach consensus, the acute competition for oil rents, the role of the state ina rent-seeking society, and the ensuing weak governance. 11. Beckerman (2002) points out that while many countries face standing risks from natural phenomena, if one were to list the world's economies according to the frequency and variety o f their natural disasters, "Ecuador would surely rank relatively high" (p. 25). 7 The political economy regime was shaped in the early 1970s under the powerful formative influences stemming from the oil boom. It was born as a response to the key questions of what to do with, and how to distribute, the (then) newly found oil rents. The social makeup. The main social actors in the post-oil-discovery political economy have been (a) the elite class, the backbone o f which is a small number of powerful economic groups, some of them family dominated, which own fairly diversified business conglomerates operating in both the financial and non-financial sectors of the economy (de la Torre and others 2001); (b) the powerful military that, as an institution, has participation inoil revenues and i s the owner of a number o f state enterprises; (c) the unions, particularly the public sector unions, which have traditionally been able to mobilize their membership to exert pressure on governments and politicians; (d) the middle class of professionals, small entrepreneurs, shopkeepers, and bureaucrats; and (e) the indigenous movement that has emerged as a social and political force in the 1990s. The poor have mostly been integrated into the political system as passive recipients of services (and, sometimes, o f goods as well) provided through clientelistic networks and a patronage system. The conflictive nature of social relationships. The political system has been appropriately characterized by Eifert and others (2002) as a factional democracy. In the view of those authors, oil-exporting factional democracies have several features that distinguish them from mature democracies. First, income distribution is unequal and social consensus is elusive. Second, politically powerful interests attached directly to state spending, such as bureaucratic and political elites, public sector unions, and the military, tend to capture the State. The driving incentive is the competition for oil rents. Each social group wants to maximize its share o f those rents. The diverse, predatory interest groups can be stronger and more continuous than political parties and governments, and try to lock in their claims on the oil rents through devices such as the earmarking o f government revenues. The conflicts around the distribution o f oil rents got interwoven with the previously existing regional, ethnic, and social cleavages in Ecuadorian society. Prominent among the preexisting conflicts is the traditional deep-seated division between the dominant classes of the trade-oriented Costa and the agrarian-oriented Sierra (Gelb and Marshall- Silva 1988). I t i s well known that the socioeconomic differences and rivalries between the dominant two regions go a long way in explaining Ecuador's history of bitter infighting and political instability. Not surprisingly, as the oil boom of the 1970s got under way, the interregional rivalry quickly became woven together, in complex ways, with the disputes around the use o f the oil rents, and must be seen, therefore, as an integral element o f the prevailing political economy regime. The other major cleavage that plays a role in the Ecuadorian political economy stems from the social exclusion to which the indigenous and Afro-Ecuadorian populations have been subjected (World Bank 2004b). This social exclusion-a lingering legacy of the colonial period-has been a secular feature of Ecuadorian society and shows up inthe considerable gaps in income and social indicators between the indigenous and Afro populations, on one hand. and the mestizo and white populations, on the other. The economic role of the State. During the oil boom of the 1970s, the decision was made to assign propertj rights on petroleum resources to the Ecuadoran state. B y virtue of this decision the State became both the administratox of the oil rents and the crucial arena where the struggles for oil-rent distribution are waged and the institutional and policy decisions on this crucial matter are made. In consonance with this new role, the government became a produce] of goods and services that were sold at subsidized levels, a rescuer of private firms in distress, a subdsidizer o f credit tc the private sector, an employer o f last (and sometimes of first) resort, and a builder o f infrastructure. (World Bank 1993). This political economy breeds weak governance and weak institutions. Ecuador governments are unstable*; the relationship between the Executive and Congress is very conflictive (Jones 1995; Mainwaring anc Shugart 1997; Mejia-Acosta 1999); the political party system is highly fragmented (Coppedge 1998); there i s a chronic electoral divorce between parties and voters (Conaghan 1994, 1995); a large number o f players are able to wield vetc power over governmental decisions (Pachano and others 2004); and "last-ditch" veto players (in the form of, say popular protest or judicial decisions) have the potential to block the implementation o f policies (Pachano and others 2004). Furthermore, political party fragmentation has brought about a crisis of representation. A number of socia groups do not feel they are adequately represented by the existing political parties. As a result, social and economic demands are channeled intensively via the social organizations (Jkcome 2004) instead of through the political parties, a: is the case in mature democracies. 8 This political economy cannot but breed weak institutions. While the country's presidency has considerable legislative powers, including the possibility o f introducing fast-track legislation in economic matters (with bills defined as urgent by the President becoming law if Congress fails to act within 30 days) and significant veto powers,12 presidents have seen these advantages offset by the fact that, as a result o f political party fragmentation, they have enjoyed little support in the legislature. On average, since 1979, the president's party has controlled only 26 percent of the seats inCongress and, in that period, no president has ever commanded a single-party majority. The efJicacy of public administration is severely damaged in such a context. In addition to the endemic shortcomings o f the civil service, the central-government ministries are further weakened by the short tenure of the ministers, a common occurrence in governments whose political base o f support is often quite narrow, and are hence vulnerable to external pressures on the ministers to resign, a situation compounded by the use of congressional censure as a political weapon on the part o f opposition parties in the legislature. The turnover of key officials has been quite high. Between 1979 and 1998, the average Minister o f Finance lasted 336 days inoffice. On the other hand, inthe same period, 22 percent of all cabinet members were subjected to a confidence vote in Congress, and 10percent were actually censured (Pachano and others 2004). Things have not improved after dollarization though. Between February 2000 and March 2003, the country had six Finance Ministers who stayed in office for an average o f just seven months. High cabinet and key-agency instability has a deleterious impact on the Executive's policymaking ability. As for the Ecuadorian Congress, besides being known for the weakness of the alliances and coalitions, it often lacks the technical skills and staff support needed to deal responsibly with the technical issues involved inlegislating on economic and social matters. * Forinstance,five presidentsgovemedEcuadorduring 1995-98 (Jkcome 2004). 1.22 There exists an internal relationship between the prevailing political-economy regime and some deeply engrainedpractices and institutions that frame the conduct of fiscal policy. Throughout the last three decades, the internal connectionbetween the political-economy regime and fiscal policy has been visible infour systemic features of fiscal policy and public finance, namely: 0 The size of Government and its role as a producer and provider of non-public goods and services, frequently at subsidized prices. 0 The existence of a set of tax expenditures, whose function i s to channel (potential) government revenue to the private sector. The main mechanisms for this channeling of rents have been the reduction of the tax burden on non-oil economic activities and the enactment and maintenance of consumptionsubsidies. 0 The fact that policymakers' incentives tend to be biased toward a short-term horizon. In particular, savings-oriented behavior i s frequently shunted aside and incentives are biased toward increased expenditure. This is the obvious way of accommodating interest-group pressures, maintaining political clienteles, and "greasing" patronage networks. The lack of a long-term horizon makes public policy in general, and fiscal policy in particular, very sensitive to the varying political shocks occurring in the public sphere. This explains why it is so difficult for Ecuadorian governments to persevere in efforts at fiscal adjustment, as expenditure-reduction measures prompt social and political pressuresthat may even take the form of rioting and general strikes. 0 The deeply flawed institution of earmarking. Lacking a consensus on a set of rules for budget allocation, a number of particular constituencies have attempted to "lock in" their entitlement to public resources. There are more than 50 legal provisions mandating that preestablished percentages 12. A partial presidentialveto can be overridden by Congress only with a two-thirds majority and within 30 days. If the president appeals to a package veto, Congress is prevented from addressing the bill in question for a year (Pachano and others 2004). 9 of particular taxes, or of oil income, must be given to particular levels of government, agencies, public institutions, or programs. As the 1993 Ecuador Public Expenditure Review pointed out, earmarking of large shares of government revenue has been a trademark of the country's fiscal policy since the 1970s (World Bank 1993). The rationale for this approach was that it would safeguard allocations to particular programs, shielding them from discretionary policy measures. But the abusive expansion of this practice only helped establish entitlement claims that eventually hindered government efforts to adjust to the changing macroeconomic and fiscal realities (ibid.). This is so because earmarking ends up making budgetary allocations arbitrary and automatic, and deters the required review and reassessment of priorities when situations change or when a budget crisis arises. Earmarking i s carefully assessedthroughout this report. 1.23 A good example of the internal relationship between the political economy and fiscal policy i s how oil booms have made possible for Ecuadorian governments to reduce non-oil taxes, mainly through tax exemptions. As oil revenue increased from 3.9 percent of GDP in 1973 to 9.3 percent of GDP in 1982, non-oil tax revenue dropped from 17.0 percent of GDP to 10.5 percent (WorldBank 1984). The decline proceeded during the decade of the 1980%and by 1988 non-oil tax revenue had reached a low value of 9.0 percent of GDP (World Bank 1991). The creation of the Internal Revenue Service in 1997 has brought about a notable recovery of non-oil tax revenue. However, advances made in the non- oil-tax arena are incomplete and fragile. A number of tax exemptions still exist. There i s resistance by some interest groups inthe political economy to the idea of increased non-oil taxation. The confrontation between reformers and interest groups resisting tax reforms i s permanent in Congress. The corporate sector continues to press for tax concessions (IMF 2003b). 10 ChaDter 2 FiscalTrendsand Challenges 2.1 Ecuador's fiscal position at mid-2004 seemed paradoxical. On one hand, oil prices and oil fiscal revenues were flying high. On the other, the fiscal authorities were advocating (and practicing) austerity as they reiterated their commitment to a primary surplus of 5.0 percent of GDP, and had embarked on negotiations with the International Monetary Fund (IMF) about a possible Arrangement. What lied behind this apparent paradox was, a somewhat improved, but still fragile, fiscal position that resulted from a long accumulation of deep, structural vulnerabilities and weaknesses. This chapter analyzes the fiscal situation in seven sections. Section A deals with the volatility of fiscal variables; section B compares Ecuador's fiscal performance in the pre- and post-dollarization periods; section C explores revenue management; section D analyzes expenditure trends; as a way of paving the road for policy recommendations, section E analyzes the Fiscal Transparency, Stabilization and Responsibility Law; section Ftackles fiscal sustainability issues; and section G offers fiscal policy recommendations. A. The Volatility of Fiscal Variables 2.2 While output has been less volatile under dollarization, most fiscal variables--e.g., petroleum and non-petroleum revenue and current and capital expenditure-have been more volatile (Table 2.1). The standard deviation of real GDP growth has declined from 3.98 during 1994-99 to 1.16 during 2000-03. Petroleum revenues have been more volatile than current expenditures. The volatility of expenditure variables has also increased, especially wages and interest payments and, to a lesser extent, capital expenditures. Table 2.1 Volatility of NFPS Variables, 1993-2003 StandardDeviations 1993-99 2000-03 Total Revenue 1.94 1.16 Petro1eum 1.32 1.56 Non-Petroleum 0.63 1.74 otw VAT 0.26 0.61 Total Expenditures 1.38 0.87 Current Expenditures 1.10 1.34 Interest 1.19 1.58 Salaries 0.43 1.68 Goods and Services 0.41 0.56 Capital Expenditures 0.60 0.81 Fixed capital formation 0.44 0.38 Central Government 0.67 0.28 Municipal governments 0.25 0.17 Overall Balance 1.99 0.63 Primary Balance 1.54 1.71 GDP 3.98 1.16 Source: MEF. 11 B. Fiscal Performance Beforeand After Dollarization 2.3 More disciplined policies in the post-dollarization period have brought a definite improvement in fiscal performance, but the fiscal situation continues to be fragile and requires further consolidation. Both the primary balance and the overall balance clearly improved during2000- 03 compared to 1995-99 (Figure 2.1). The overall balancejumped from an average deficit of 3.2 percent of GDP in the pre-dollarization period to an average surplus of 1.0 percent of GDP in the post- dollarization period, a turnaround of 4.2 percentage points of GDP. The increase inthe primary balance was about equivalent: from an average of 1.4 percent of GDP during 1995-99 to an average of 5.5 percent of GDP during2000-03. Never-theless, an important caveat i s in order. The fiscal improvementreflects in part higher oil revenues, rather than an Figure 2.1 Non-Financial Public Sector Balances increased effort at expenditure reduction. Non-oi1 revenue increased from an average of 20.6 percent of GDP during 1995-99 to an average of 25.2 percent during 2000-03. Duringthe same period, average total non- financial public sector (NFPS) expenditures -- increased by 0.8 percent 1995 1996 1997 1998 1999 2000 2001 2002 2003 of GDP, partly offsetting revenue gains +Overall Balance 1)- Primary Balance (Table Al). Source: BCE. 2.4 Estimates of the structural fiscal balance confirm the improvement in the fiscal stance during 2000-03.''2 Table 2.2 shows how the nominal primary balance, the structural primary balance, the fiscal stance, and the fiscal impulse evolved during 1990-2003 (a detailed analysis i s in Annex A). The average fiscal stance during2000-03 is a surplus of 0.8 percent of GDP, in contrast to the average fiscal stance during 1990- 99, which was a deficit of 0.2 percent of GDP. The estimates of the fiscal impulse also show an improved fiscal performance: In 2000 there was a strong contractionary fiscal impulse of 3.1 percent of GDP, followed, however, by an expansionary impulseof 1.5 percent of GDP in 2001. In2002 and 2003, fiscal impulses were again positive (contractionary). 1. This paragraph is basedon Artana and Moskovits (2004). 2. The structural balance is a measure of the fiscal balance that is independent of the particular position of the economy inthe business cycle at a given point intime. On the basis o f the structural balance, the fiscal stance at any given year is defined as the difference between the actual and structural budget balances. When the structural balance approach i s chosen, another key concept to bear in mind is the fiscal impulse, defined as the change in the fiscal stance from one year to the next. Ifthe stance remains unchanged, the government i s said to have kept fiscal policy neutral (the fiscal impulse i s zero). A positive fiscal stance means that the authorities are imparting a contractionary impulse on aggregate demand. A negative fiscal stance is tantamount to the authorities imparting an expansionary impulse on aggregate demand. 12 Table 2.2 Ecuador: Actual and Structural Fiscal Balance, FiscalStance, and FiscalImpulse, 1990- 2003 (percentages of GDP) NFPSPrimary Surplus Actual Structural FiscalStance FiscalImpulse 1990 6.8 5.8 +1.0 1991 5.2 5.6 -0.4 -1.4 1992 3.5 3.5 0.0 +0.4 1993 4.1 5.O -0.9 -0.9 1994 4.1 4.5 -0.4 +OS 1995 2.9 3.1 -0.2 +0.2 1996 1.3 0.7 +OS +0.7 1997 2.1 1.7 +0.4 -0.1 1998 -0.6 -0.2 -0.3 -0.7 1999 3.2 4.8 -1.6 -1.3 2000 8.1 6.6 +1.5 +3.1 2001 5.1 5.1 0.0 -1.5 2002 4.2 3.7 +OS +OS 2003 4.7 3.4 +1.3 +0.8 Source: Annex A. 2.5 Several elements reinforce prospects for continued fiscal adjustment and reform but political-economy and institutionalfactors are still major obstacles to fiscal stability. Three elements point towards a structurally strengthened fiscal situation in the medium-to-long term. First and foremost, the new institutional framework provided by dollarization creates a structural bias toward fiscal discipline because it prevents monetization of fiscal deficits3 Second, both passage of the Organic Law on Fiscal Transparency, Stabilization and Responsibility (FTSRL) and the establishment of the Fund for Stabilization, Investment, and Public Debt Reduction (FEIREP) considerably strengthened the institutional framework for fiscal discipline. Third, a structural change towards a permanent increase in non-oil tax revenue i s inthe making, as will be presently shown. These advances are the clear expression of the existence of pro-reform forces in the political economy engaged in a major, medium-to-long-term attempt both at redressing the cumulative effects from past, chronic lack of discipline and at laying the ground for the fiscal institutions and policies that dollarization demands if it i s to be sustainable The depth of the 1998-1999 crisis opened a window of opportunity for reform. The arrival of a new Administration in 2003 has, until now, given an upper hand to the advocates of reform. However, medium-term structural adjustment continues to be resisted by an array of forces that, at best are suspicious of fiscal discipline, and, at worst, are adamantly opposed to it. This i s why progress made so far i s fragile and the timingof reformvulnerable to the contingency of uncertainpolitical alliances. C. The Challenges of Revenue Management 2.6 Non-oil tax revenue has recently improved: One major feature of the Ecuadorian oil-driven political economy over the last three decades has been the reduced burden of taxation on non-oil economic activities. The sizeable increases in public-sector oil revenues in the seventies made it possible for the government to reduce non-oil taxes, mainly through tax rebates and tax exemptions. As shown in 3. However, to the extent that it eliminates the foreign exchange risk of external debt, it also, unfortunately, creates incentives toward that particular form of fiscal indiscipline the content o f which i s excessive indebtedness of the public sector (see Melo 2003). 4. These sections draw heavily from Schenone, Osvaldo (2003), Tax Policy and Administration, in Vicente Fretes- Cibils, Marcel0 M. Giugale, and JosC-Roberto L6pez-CAlix, (eds.), Ecuador: An Economic and Social Agenda in the New Millennium, World Bank, Washington, D.C., pp. 43-63. 13 Figure 2.2,5 non-oil tax revenues declined between 1974 and 1984, partially recovered between 1986 and 1989 (due, among other things to president Borja's fiscal reform) and declined again between 1990 and 1996. The creation of the InternalRevenue Service (SRI inits Spanish acronym) as an independent entity and its strengthening inthe late nineties brought about a significant recovery innon-oil tax revenue. SRI- collected tax revenue increasedfrom 8.0 percent of GDP in 1997 to 11.3 percent in 2002 (see Table 2.3). Following dollarization, policymakers found a real incentive to make tax collection efficient. The SRI was given the managerial autonomy to simplify procedures, impose and enforce sanctions, and work with taxpayers. As a result, the government has improved its fiscal position. The SRI i s also supporting modernization of Customs administration, less with the objective of raising revenue, than of facilitating international trade and competitiveness. Figure 2.2 Oil Fiscal Revenue and Non-Oil Tax Revenue of the Central Government (% GDP)and Important Tax ReformEfforts, 1964-2004 18 1 16 14 12 n a 10 (3 8 8 6 4 2 Source: World Bank staff calculations, based on data from Central Bank of Ecuador (2003). 2004 values are estimates. 2.7 Ecuador does not have a problem of low fiscal revenues but there are obstacles to higher revenues. The government should deal with three main shortcomings of the tax system, which are constraining the possibility of increasing revenue. These shortcomings are: (a) the proliferation of taxes, with the resulting fragmentation of the tax system; (b) the number and scope of tax exemptions, which narrow the tax base; and (c) the pervasiveness of earmarking. 5. In Figure 2.2 "Tax Revenue" stands for non-oil tax revenues and "Non-Tax Revenue" stands for oil fiscal revenues. 6. InFigure 2.2 "Tax Revenue" stands for non-oil tax revenues and "Non-Tax Revenue" stands for oil fiscal revenues. 14 Table 2.3 Total Revenue of the Non-Financial Public Sector (NFPS) and Central Government (in % GDP), 1995-2003 Central GovernmentRevenue Tax Revenue Within SRI Year NFPS Total Oil Outside Total Income VAT SpecialConsumption Other revenue SRI Tax Tax (ICE) Taxes 1995 22.8 15.5 5.9 1.8 7.1 1.7 3.1 0.4 0.1 1996 21.9 15.1 7.0 1.3 6.6 1.6 3.0 0.5 0.3 1997 19.9 14.6 5.1 2.0 8.0 1.7 3.3 0.6 0.4 1998 17.3 13.9 3.8 2.7 8.6 1.5 3.6 0.5 0.2 1999 21.1 16.1 6.0 2.0 8.9 0.7 3.3 0.4 2.5 2000 25.9 20.4 8.8 1.4 10.2 1.8 5.2 0.5 1.2 2001 23.5 18.3 6.1 1.7 11.3 2.2 6.4 0.7 0.3 2002 25.8 18.8 5.6 1.7 11.3 2.2 6.3 0.9 0.2 2003 25.7 17.8 5.8 1.4 10.4 2.2 5.9 0.7 0.2 Source: SRI (2004). A Fragmented Tax Structure 2.8 A central shortcoming of Ecuador's tax structure lies inits fragmentation. Currently, there are more than 80 taxes (Table 2.4), but, from the standpoint of their revenue-collection potential, most are irrelevant. Five taxes-namely, the value-added,tax (VAT), the personal income tax (ISR), the Special Consumption Tax (ICE), customs tariffs, and the vehicle tax-generate more than 75 percent of tax revenues. The Central Government would not lose much revenue by eliminating insignificant taxes. Proliferation of taxes remains a major obstacle for collection and administration. For both the taxpayers and the tax authority it implies unnecessarily high administrative costs. It discourages payment by taxpayers and undermines supervision by the tax authority. 2.9 The SRI has introduced a number of significant improvements to tax administration. A centralized tax payers registry (Registro Unico de Contn'buyentes, RUC) has been created, with 1.1 million registrants (80 percent individuals and 20 percent legal entities). This allows SRI to impose sanctions on delinquent or tardy taxpayers who, in the past, were able simply to move from one regional office to another to avoid punishment. Under the revamped Registro they can no longer do that. The SRI i s also using third-party information to increase compliance. Crosschecks during 2001 and 2002 detected approximately 100,000 people not registered in the RUC.' As a result of these measures, tax administrationindicators show satisfactory levels for 2001-03, as indicated inTable 2.5. 7. The fact that companies and public entities are subject to the VAT offers excellent possibilities for crosschecking information, because when these entities request a reimbursement of VAT credits, they automatically reveal their identities to their suppliers. Because of the large volume o f purchases made by the State, the volume of resultant crosschecked information is also very large. 15 Table 2.4 The Structure of the Tax System Beneficiaries Type of Tax Number Central Munici- Provincial Other of Taxes Gov. palities Councils Income and Capital Earnings Tax 5 3 1 1 Payroll Tax 1 1 Property and Net Worth Tax 26 1 11 2 12 RealEstate Transfer Tax 24 8 2 14 Financial Asset Tax 6 1 5 Sales Tax on Goods and Services 17 7 3 7 VAT 1 1 ICE 7 6 1 Telecommunications 1 1 Electricity 4 4 Public Entertainment 3 1 Betting 1 2 ForeignTrade Tax 1 1 Various Taxes 4 2 1 1 Totals 84 15 24 4 41 Note: The "Other" category is composed o f an enormous variety of institutions, including the Guayaquil Beneficence Council, the State University o f Guayaquil, the Guayas Transit Commission, the Ecuadorian Social Security Institute, the Osvaldo Loor Foundation, the Potable Water Company, the National Promotion and Development of Sports, the Superintendency o f Companies, the Superintendency of Banks, the Ecuadorian Tourism Corporation, and the National Children's Fund. Source: SRI. Table 2.5 Selected Tax Administration Indicators, 2001-03 2001 2002 2003 Debt collected versus outstanding debt 23% 41% n.d Forced collections versus collections under management 1.7% 2.8% 5.7% Taxpayers who declare nothing versus taxpayers who declare a 47.2% 41.4% 43.5% gain Taxpayers who declare versus taxpayers who should declare 43.4% 33.5% 53.7% Notifications to actual non-filers versus (special) non-filers 94% 89% n.d. Notified taxpayers with crosschecked differences versus 47% 91% n.d. planned notifications Source: SRI. Tax Exemptions and the Erosionof the Tax Base 2.10 In Ecuador, the tax base is constantly eroded by exemptions. Efforts of tax administration authorities to reduce them are permanently offset by waves of new exemptions. Authorities reduced exemptions between 1999 and the first half of 2001, in particular with regard to the VAT.* Since then, and within less than one year's time, new exemptions were created or attempted, related not only to the VAT, but also to the personal income tax. Exemptions imply a significant revenue loss for the Central 8. The ISR exemption was eliminated for the financial sector on the income from securities and shares issued by the government, for cooperatives and provident societies (except for the ones established by farmers or officially recognized indigenous people), and for promoting development (directed primarily at tourism and industrial endeavors). Inaddition, the list o f items subject to the VAT was replaced with a tax list o f VAT-exempt services (fundamentally, housingrentals, and financial services)-which, therefore, leaves all other services subject to the VAT. 16 Government and reduce the fiscal space. Calculations of tax exemptions of the internal VAT (that is, excluding the VAT on imports) indicate that in 2001 the tax cost was approximately US$237 million, or 1.1percent of GDP (Chapter 4). Earmarking: The Undermining of BudgetFlexibility 2.11 Historically, earmarking has been a feature of Ecuador's tax system. Earmarking can be appropriate under certain circumstances. It can establish stable financing for favored programs or services. It may also heighten taxpayer compliance by linking taxes to benefits. Back in the seventies, when this practice was adopted, its rationale was that it would safeguard allocations to particular programs, shielding them from arbitrary, discretionary changes. But the expansion of this practice helped establish entitlement claims that eventually came to hinder the government's efforts to adjust to changing macroeconomic and fiscal realities. When earmarking becomes a widespread phenomenon, it limits the flexibility of government to formulate and execute the budget inresponse to shifting public priorities. It undermines the possibility of channeling funds into areas where their productivity i s higher, promotes an inefficient, and possibly also inequitable, use of tax revenues and reduces the fiscal space. InEcuador, it has served as a means of appeasing strong local elites and balancing the political interests of different regions. In 2001, there were 32 earmarkings from tax revenues and 25 earmarkings from oil income, the latter being 100 percent earmarked. The FTSFU abolishes earmarking (Article 22), but it remains a standard practice inthe country's fiscal system. 2.12 Earmarking undermines the efficiency and flexibility of budget management. In 2004, earmarked resources represented 4 percent of GDP, of which 2.4 corresponded to pre-budget earmarked tax revenues and 1.4 percent to off-budget earmarked oil resources (Table 4.5). This represents an increase over earmarked revenues of 3 percent of GDP in 2001, of which about 0.7 percent of GDP corresponded to the VAT and 0.8 percent of GDP corresponded to the income tax. D. ExpenditureTrends Figure 2.3 Total and Primary Expenditures 2.14 This section analyzes trends in ( % of GDP) public expenditures. First, it looks at the overall andprimary expenditure of the non- financial public sector and the Central Government. Second, expenditures are 25.0 broken down by their economic and sectoral classification focusing particularly on wages 20.0 and public investment. Detailed data can be found inthe StatisticalAnnex (Al-10). 15.0 2.15 An expansionary trend inprimary I 10.04 I expenditures contrasts with the constant 95 96 97 98 99 00 01 Between '02 03 behavior of total expenditure. *NFPSTotal Total +NFPSPrimarj 1997 and 2002, total and primary CG +%- CG Primary expenditures trended upward for both the NFPS and the Central Government (Figures I Source: BCE. 2.3 and 2.4), though they dipped in 2003.' Primary expenditure rose by about 4 percent 9. The Central Government accounts for about three-quarters of the NFPS, which explains why the two series track each other fairly closely. 17 - of GDP during 1997-2003. However, while there has been an upward trend inprimary expenditure, total expenditure has remained constant around a close 24-25 percent of GDP range, a fact mainly explained by the drop in interest payments as a share of GDP. What i s of concern i s that the expansionary trend in primary expenditures is more pronounced since dollarization than before. Figure 2.4 Trends inNFPS and Central Government Expenditures I Y I 4.0 - 0.04 1 I , , I , , I , I I0.0 0.0 0.0 ~ 93 94 95 96 97 98 99 00 01 02 03 95 96 97 98 99 00 01 02 03 -w- Interest ++Salaries tInterest +Salaries tGwdsandServices -Capital Expenditures tGodsand&vices -Capital Expenditures &Total Exmditures F W +Onrent ExaendituresRHS +Total ExpendituresRHS +Cumem ExpendituresRHS Source: Statistical Annex. 2.16 The upward momentum on primary expenditures is mainly propelled by current spending and especially the wage and salary component. While NFPS current expenditures increased from an average of 17.7 percent before dollarization to an average of 18.4 percent since, average capital expenditures increased by just 0.1 percent of GDP The expansionary bias in current spending i s partly due to budgetary inertia resulting from public sector wage indexation and to earmarking." Inertia i s also reflected in the low volatility of total expenditure (see Table 2.1), which contrasts with the much higher volatility of revenue variables (on this, see also L6pez-Cilix 2003). Political-economy factors play a major role in explaining both that inertia and the overall trend towards expansion of primary spending. The civil service and state-enterprise unions, the pensioners, the teachers and the indigenous movements have all continuously pressedfor increased spending (IMF2003b). For their part, past governments have often raised public sector wages andor expand public employment as a way of strengthen their political base and maintainthe loyalty of political clienteles. 2.17 The functional composition of Central Government expenditure is relatively constant, but skewed towards defense and public order (Figure 2.5). Figure 2.5 indicates that average defense and public order spending accounts for about 22 percent of the budget, which is roughly equivalent to the share of health and education expenditures. In a sample of 12 countries (the United States and eleven Latin American countries) Ecuador ranks as the country with the lowest ratios of spending on education and health with respect to GDP, about 3 percent and 1.5 percent of GDP, respectively. Furthermore, education has a relatively large share of spending only because of its large wage bill (Shepherd 2004). 2.18 The rates of growth of some components of current expenditure have been high in recent years, and are unlikely to be sustainable in the medium term. Inreal terms, the goods-and-services component grew 12.5 percent in 2001 and 39.0 percent in 2002. The wage and salaries component grew 21.7 percent in 2001, 35.4 percent in 2002, and 19.5 percent in 2003. The most unionized segments have the greatest ability to win concessions, as attestedby the fact that real wages in the education sector grew ~~ 10. Since earmarking implies the ex ante pre-allocation of projected tax revenues in the proforma budgets, this implies that any increase inearmarked revenue automatically translates into increased spending. 18 an annual average 18 percent during 2000-02, and the share of wages within education expenditure jumped from 67 percent during 1995-99 to 80 percent during 2000-02. This has had a major impact on both intra-sectorand inter-sector budget allocation (see on this Chapter 4). Figure 2.5 Public Expenditure by SelectedFunctions(Percent of total expenditure) Note: In connection with the data on functional composition of expenditure a caveat is in order: these data are inaccurate as they only take into account spendingby the institutional sector responsible of the particular function and exclude spending by other institutional sectors in that particular function. For instance, the expenditure in educationdoes not take into account possibleeducationspendingby the Defensesector. Source: Statistical Annex Table A6. 2.19 Rigidity makesthe Central Government's budgeta bluntpolicyinstrument becauseit limits the ability to reduce or reallocate expenditures. Political-economy constraints shape Ecuadorian public finances and make the budget highly inflexible. There are some expenditure categories in connection with which, once a given level of spending has been reached, it i s extremely difficult to cut back to lower levels. This is the case of the wage bill in a highly unionized public sector. It i s also the case with most of the earmarked revenues, as its beneficiaries, both social and regional, vehemently defend their share inthe public resources. The high degree of budget rigidity is harmful for the design and fulfillment of any comprehensivepoverty reduction plan with strategic priorities. Detailed estimates of inertial andrigid spendingare developedinChapter 3. Calculations show that (a) as an average during 2001-04, the rigid components of the Central Government budget added up to 81 percent of total expenditures (including debt amortization); and(b)rigidity rosefrom77.6 percentin2001to an estimated 82.7 percent in 2004. In addition, (c) the non-rigid components of expenditure include some categories that could appropriately be characterizedas inertial expenditures; and these components accounted for 11.9 percent of expenditures in 2001-04. Given these rigidities, the flexible components (namely, a part of investment spending, the whole of the expenditures on goods and services, and certain non-wage, non- interest, andnon-transfer components of current expenditures) amounted, on average, to only 7.1 percent of expenditures. Moreover, as the rigid componentsincreased, the flexible components shrank from 13.6 percent of total expenditure in 2001 to 4.0 percent in 2004 (Table 3.5). The budget should be the main instrument to achieve the government's objectives, but these rigidities leave the Ecuadorian government with little scope to act. But above and beyond these considerations the fact stands that flexibility i s crucially needed ina small open economy vulnerable to external shocks. Moreover, it should go without saying that the call for greater flexibility must not be seen as equivalent to the return to the old practices of discretionary, unaccountable spendingthat pre-datedthe heavy reliance on earmarking. The flexibility required for macroeconomic purposes i s premised on a transparent, accountable management of public finances. 2.20 Wages and salaries account for the largest and fastest growing share of current expenditures, and this trend is unsustainable. In 2003, 45 percent of current expenditures went to 19 wages and salaries, up from 25 percent in 2000. While the 50 percent salary increase for public employees in 2002 redressed the severe predollarization decline in real wages (about 70 percent), its impact as a permanent expenditure is troublesome, and its expansionarytrend i s not sustainable.Among middle-income countries, average GeneralGovernmentemployment is 4.3 percent of the population." In Ecuador, the equivalent figure is 3.1 percent. When contrasting public employment levels in Ecuador with other countries inLatin America the comparisons are less favorable (see the Statistical Appendix). Nonetheless, by these rough indicators, Ecuador's government is not grossly overstaffed. Public employment has been growing, however. From 2002 to 2004, Central Government employment increased by 7.3 percent (see the Statistical Appendix). Much of the increase occurred during the last year, most notably for police (8 percent) and armed forces personnel (18 percent). Naturally, these employment increasesplace additional fiscal strain on government. The projected wage bill for 2004 is nearly 27 percent higher innominal terms than in2002. Of course, partof this increaseis the result of the aforementionedemployment increases. However, employees' wages and salaries have grown even faster than employment. Monetary allowances, ratherthanbasepay, account for most of the increase. In2003, prior to salary unification, base pay averagedonly 20 percent of public employees' total remuneration.12 From 2000 to 2001, while overall employment levels were largely unchanged, the wage bill growth of individualministries ranged from 21 percent to 114per~ent.'~Increasesin pay have slowed since then, buthavecontinuedtheir upwardtrend. Averagenominal remunerationfor central governmentemployees i s projected to he 18percenthigher in 2004 than in 2002. Other sources of wage-bill expansionand other shortcomings of the currentpublic-employment institutional framework include: 0 The proliferation of salary components, which undermineseffective wage policy and administration. On top of base monthly salary, a public employee i s paid up to 22 additional components. The net result is salaries that are four to six times bigger than the base salary, making it difficult for the governmentto control pay policies. The lack of reliable data on public employment. There is no government office collecting the information to provide acomprehensivepicture of public employment. There i s not eventhe capacity to ensure that employees who have been releasedfrom public service (and duly indemnified) do not return to public employment. 0 Differences in salary levels are frequently ineq~itable,'~hiring i s not subject to rules but is frequently left to discretion of the hiring officials, andthere are cases in which professionalsworking in similar types ofjobs receivewidely divergent remuneration. 2.21 The Central Government accounts for about 81 percent of the NFPS wage bill and, within this, a few sectors are dominant. In2002, the education sector accounted for 43 percent of the wage bill, a significant increasefromthe 32 percentin 1995-99. Thishasbeen offset by decreasesinthe shares allocated to defense (21 percent), public order (12 percent), and health and social welfare (11 percent) (Figure2.6) (Shepherd2004a). 11. See the World Bank dataset on public employment and wages. These data are for various years from 1996- 2000(httD://www 1.worldbank.org/publicsector/civilservice/cross.htm). 12. Calculated from salary composition data gatheredby SIGEF. 13. Based upon detailed budget figures for the following ministries: Presidencia y Secretaria General de la Administracidn, Vicepresidencia de la Repliblica, Agricultiira y Ganaderia, Bienestar Social, MEF, and Tourism. 14. According to Shepherd (2004), who quotes the Ecuadorian Observatorio de Poli'tica Fiscal, there are substantialdifferencesbetweenthe current salaries for employees inthe same gradeaccording to whether they work in institutions of the Central Government or in certain decentralized agencies. Whereas the average monthly salaries for employees in the 5th and 11th grades are US$385 and US$959, respectively, for employees at the ministries, employees in the same grades in the Port Authority of Guayaquil earn basic monthly salariesof US$1,300 and US$4,320, respectively. 20 Figure 2.6 Compositionof Wages by SelectedSectors (inpercentof total wage bill) Infrdsmraue mastructur Health 3 8 e Eakh 4% Educahlm 32% 14% 199599 2002 Source: MEF. 2.22 The performance of the last two administrationsindealingwith the wage billhasdiffered in some respects. The Noboaadministration approvedthe 2002 salary raise, which boostedthe wage billby 1.7 percent of GDP. The Gutierrez administration has taken some measuresto contain the growth of the wage bill and, more generally, current expenditures. A 2003 executive decree contained the "Nonns on Patriotic Incentives to Savings in the Public Sector." The decree froze wages for regular civil servants, suspended overtime allowances, reduced the President's salary by 20 percent and the salary of political appointees by 10 percent, reduced the number of new positions, and restrained spending on goods and services. The decree was successful in freezing the share of the wage bill in a subset of the Central Government. The decree does not affect, however, some key sectors of the Central Government (see the discussion in Box 2.1). It does not apply either to employees in public enterprises, sectional and provincial governments, social security institutions, and other autonomous agencies, because all these entities havetheir own laws and statuteson matters of employment andlabor relations. Box 2.1 The Civil Service and the Wage Unification Law: A Stepinthe Right Direction The Ley Orgdnica de Servicio Civil y Carrera Administrativa y de Unificacidn y Homologacibn de las Remuneracionesdel Sector Pliblico was approvedon September 28,2003, and amendedonJanuary 28,2004. It is essentiallegislation for the governmentto controlpublic employment and wage expenditures. Though the changes inthelaw are significant, itscoverageis incomplete. The first key accomplishmentof the law is the elimination of a myriad of monetary allowances, except for df?cimo tercer sueldo, df?cimo cuarto sueldo, and vidticos, subsistencias, dietas, horas suplementarias extraordinarias, encargos, and subrogaciones. The proliferation of allowances had undermined the coherence of government pay scales and made it difficult to calculate, let alone control, annual increasesin the wage bill. The number of permissible allowanceshas beenreducedto a minimumfor those covered by the law. However, the law exempts teachers, military, police, and civil servants of subnational governments from the salary unification. These groups are 55 percent of all public sector employees. Total Public Employees (less workers instate enterprises): 360,000 o/w Armed Forces: 56,000 o/w Subnational Governments: 34,000 Total Civilian Central Government: 270,000 o/w Teachers: 113,000 o/w Police: 26,000 21 o/w Judiciary andLegislature: 5,200 TOTAL COVERED BY THE LAW: about 45 percentof public employees. To oversee the number of public employees, their classification, and remuneration, the law establisheda Secretaria Nacional Tkcnica de Desarrollo de Recursos Humanos y Remuneraciones del Sector Pliblico ( S E W S ) . This Secretaria was granted broader powers than its predecessor (OSCIDI). Because the law left doubts about the duplicating role of CONAREM, the amendments eliminated it so S E W S will approve the new UnifiedMonthly Pay Scale. Salary unification will strengthen the government's capacity to monitor and control the wage bill. The amendmentsto the law correctedchanges that the Legislature introducedinthe originalbillsubmitted by the Executive. Congress had frozen employees' social security contributions at the amount paid inSeptember 2003. Therefore, the real value of contributions to IESS would have declined over time. In addition, the Legislature raised payments for dismissed workers (raising the cap from US$lO,OOO to US$30,000), and allowed many public employees who received indemnity payments between 1993 and 1998 to demand a reliquiducibn so that their payment would equal that permitted by law in January 1998. The fiscal impact of this measure was sizable. Amendments to the law not only eliminated such deficiencies, but extended untilJanuary 2005 the time for starting to apply the "uniform" salary scale, and increasedthe transition time for full application to five years, instead of three. In addition, it set a 20 percent gradual increase in the base pay subject to an additional contribution to IESS, to be applied from 2006-10. Since public salaries have been frozen in 2004 and 2005, contribution to the IESS remains the same. Finally, the amended law set the annual ceiling for dismissed workers at 1percentof the economically active population. 2.23 A new Civil Service Law was passed in 2003. The Civil Service and Public Administration Career and Public Sector Remuneration Unification Law represents a significant step towards regaining control over the wage bill. Among other provisions, it reduces salary categories (from2lgrades structure to 14grades), eliminates or rationalizes monetary allowances, makes wage policy more transparent, and facilitates single registration of public employees (Box 2.1). The drawback is the fact that the law covers only about 45 percent of public servants. As a complementary action, the government has completed reviews of the ministerial staffs and these reviews have identified redundant personnel in a number of ministries. The five largest ministries personnel-wise (agriculture, the central administrative core of the Ministry of Education, social protection, health, and public works) initially remained under the 21-grade scale. However, their staffing reviews were completed inDecember 2003, so they could move to the 14- grade pay scale in 2004. In addition, the Central Bank has established a consolidated, electronic database for registering the payments and benefits of all public employees. This system will be instrumental in identifying ghost employees. 2.24 Spending on goods and services is decentralized and volatile, leading to limited accountability and poor medium-termmanagement of resources. The discretion available to fiscal authorities with respect to goods-and-services expenditures is limited by the fact that 65 percent i s spent by the local governments, municipal- and province-level government-owned enterprises, and the autonomous institutions, which, together, make up the rest of the General Government. The central budget authorities are not able to track these expenditures, nor control them in the context of fiscal adjustment. A consequence i s that in recent years, budgeted operation and maintenance expenditures have been utterly insufficient: while the Ministry of Public Works has estimated US$1.5 billion will be needed over the next five years for the operation and maintenance of roads, the 2004 budget allocated only US$52 million for that purpose. This is very deleterious and costly for the government's net worth, because reconstruction i s more expensive than maintenance, and higher levels of future investment will be required simply to maintain the current level of public sector capital stock. 2.25 Deep structural problemsmake the pensionsystem a major social and fiscal issue. In2000- 2003, transfers accounted, on average, for 5.6 percent of Central govemment expenditures. The main 22 types of transfer payments are the social security transfers and the Bono S~lidario'~ transfers16. The discussion here will center on the pension system and will emphasize its fiscal implications. 2.26 Ecuador's pension system" has serious problems regarding: (i) low coverage as provider of its adequate retirement income for the elderly; (ii) its institutional arrangements, which limit efficiency and transparency; (iii) its redistributive effects, which are not consistent with equitable social development; and (iv) its medium- and long-term sustainability and the burden it impliesfor the fiscal accounts. 2.27 As to the coverage issue, only one in seven individuals receives benefits from the system. This low access results from the combination of three factors: the high prevalence of informality at younger ages, which implies that workers end up reaching retirement age without having made enough contributions to qualify for benefits; the contributive character of the system, which i s designed to protect only those in the formal sector; and the low percentage of contributors: Ecuador i s the LAC country with the second-lowest percentage of contributors to formal pension schemes. When it comes to the institutional issues, the main shortcomings are: Ecuador still i s one of the few countries in the region where the health service and the pensions fund have not been separated;I8 the absence of a government agency in charge of policy definition and planning in the area of social security, which means that the Ecuadorian Social Security Institute (IESS) has no counterpart in the Central Government regarding policy decisions; and the lack of transparency, as the IESS does not publish any periodical report providing information on the evolving status of the programs it managesor their expected evolution inthe medium and long term. The issue of the re-distributive implications of the current social security scheme revolves around the 40-percent mandatory contribution that the Government must pay every year to the IESS (on which, more below). The fact of the matter i s that the contribution in question represents a subsidy from the society as a whole to the relatively small, predominantly middle-income segment that receives pensionbenefits. 2.28 Fromthe pub1ic:finance standpoint, the pension systemposes the followingproblems: 0 By law the Central Government i s responsible for 40 percent of pension payments, and must transfer those resources to the EcuadorianSocial Security Institute (IESS).20 0 The 40-percent payment was maintained by the new Social Security Law, with the added problem that it also applies to increases in pension benefits, which are unilaterally determined by the IESS, in what i s tantamount to a flawed, moral-hazard-ridden scheme of incentives. The 40-percent transfer creates in fact the wrong incentive as it reduces the cost of raising pensions to social security agencies, inducing them to grant higher pensions than would otherwise be the case. 0 The government has accumulated a stock of arrears in its obligations to the IESS on account of the contributions it has to pay as an employer. The 1998 Constitution requires that the government and the IESS reach an agreement on the value of the total debt owed by the government to the IESS, and 15. Bono Solidario transfers consist of continuous cash stipends given to unemployed mothers and disabled and elderly people living mostly inpoor areas. Recently it was rechristened Bono de Desarrollo Humano. 16. Other transfers include those destined to the Defense Board, the Armed Forces Social Security Institute, and the Police Social Security Institute. 17. The discussion on the pensions system draws heavily upon Rofman (2004). 18. However, the 2001Social Security Law was a significant step ahead as it separated the accounting systems for the four insurance schemes the IESS manages, namely, the pensions, labor accidents, health, and Social Security for Peasantscomponents. 19. Bono Solidario transfers consist o f continuous cash stipends given to unemployed mothers and disabled and elderly people living mostly inpoor areas. Recently it was rechristenedBono de Desarrollo Humano. 20. The Central Government became legally required to cover 40 percent of the cost of pensions following the emergence of an actuarial deficit inthe public pension fund in 1941. During 1959-64, the Central Government also became obligated to pay military and police pensions, as well as special additional pension schemes for other groups (see World Bank 1991). 23 that it be completely paid by 2009. The JESS estimates this debt i s US$2.3 billion, but MEF estimates it i s US$500million. 0 High dependency ratios raise concerns about the social security system's actuarial positionz1. Although a young population currently helps the system as a whole to maintain a sound cash flow position, the system shows a large actuarial deficit (more than 30 percent of 2002 GDP) if the government subsidies are excluded from the calculations. Without these subsidies, the IESS would exhaust its reserves in2038. However, the pace at which pensions have been increased inthe last several years raises concern that the pension expenditure dynamics may be turning explosive. In 2002 pensions were raised by 30 percent in January, then by 60 percent in July 2002, followed by another 60 percent in October. In 2003, pensions were raised by 27 percent inJanuary, followed by another 27 percent inJuly. Further increaseshave been approved in2004.22 Expenditure on pensions as a share of GDPincreased almost six-fold as itjumped from 0.3 percent in2000 to aprojected 1.8 percent in 2004. 2.29 Taking into account the large weight of public enterprises, capital expenditures ought to be higher. With respect to total NFPS expenditure, capital expenditures were, on average, 24.7 percent during 1995-99 and 24.3 percent during 2000-03. When compared with the Latin American average (14.3 percent), those ratios seem high. But the Ecuadorian public-enterprise sector i s large and includes key capital-intensive sectors like oil, telecommunications, and electricity and, from that standpoint, a more relevant benchmark for comparison would be the level of public investment in economies with a strong public-enterprise sector. Such comparisons would show that public investment in Ecuador i s pretty much on the low side.23 2.30 The role of local governments in implementing public investment has increased significantly, certainly exceeding their capacity, and with little accountability. As a result of the Distribution Law of 2002, which increased from 10 percent to 15 percent the share of Central Government current revenue transferred to the local governments, local governments have become important executers for public investment. In2003, local governments accounted for 29 percent of public fixed investment, up from an average of 18 percent during 1995-99 (Figure 2.7). Public investment by the municipal governments increased from an average of 1.0 percent of GDP during 1995-99 to 1.7 percent during 2000-03. 2.31 Resourcestransferred to the local governments for investment expenditures are sizable and represent about half of total Central Government investment. In 2003, for instance, out of US$1.3 billion in Central Government capital expenditures, US$642 million (that is, 48.7 percent) consisted of transfers to the local governments. There i s some evidence, however, that some transfers are spent on current expenditure items, including wages and goods and services. A number of local governments have limited management capabilities, making it questionable whether they can successfully execute investment projects (Banco Central delEcuador 2003). 21. Bear in mindthat this statement refers to the complete social security system, including the IESS as well as the Police and the Armed Forces social security agencies. 22. In June 2004, Congress approved an increase of the minimumpension to US$135.62, a measure that would have an estimated fiscal impact according to MEF of US$210 million. A pension increase of 15 percent was also approved, and its fiscal impact estimated at US$65 million. Similarly, resources from the Reserve Fund could be devolved to pensioners (amount equivalent to US$230 million). Finally, a proposal for retiring pension deposits every three years has been advanced. 23. Venezuela, which has a similarly large public-enterprise sector, boasts a public-investment-to-GDPratio of about 8 percent, whereas Ecuador's i sjust about 3.5 percent. 24 Figure2.7 Institutional Compositionof Public Fixed Investment(inpercentof total) Others Others Local 6% government 1(1 _In S 185% Central 6o vernment 43% Centra! government 50% ._. ... enterprises 33% 1 9 9 5 - 9 9 enterpriseS 4- 10m m 2 0 0 3 Source: MEF. 2.32 Public enterprises play a major, albeit declining, role in public investment. The share of public enterprises within total public investment dropped by almost half (from 33 percent to 18 percent) between the predollarization period and 2003. This i s the result of a downward trend in fixed-capital- formation on the part of the non-financial public enterprises. By 2003 investment spending by these fm accounted for 0.8 percent of GDP compared to an average of 1.8 percent during 1995-99 (Shepherd 2004). Public enterprises are handicapped by pricing policies setting tariffs below the true economic costs, and by uneven managerial quality. 2.33 The residual character of capital expenditures leads to their volatility. NFPS capital expenditure has decreased from an average 5.8 percent of GDP during 1995-99 to 5.4 percent of GDP in 2003 (Table Al), while current expenditures went inthe opposite direction. The differences between the behavior of current expenditures and capital expenditures do not bode well for the future o f public investment in Ecuador, unless serious institutional changes are implemented that would protect capital expenditures from the effects of fiscal volatility. Due to the inflexibility of the budget, investment spending i s the adjustment variable par excellence when expenditure needs to be cut. Public investment thus bears the brunt of fiscal adjustment, giving it considerable variability vis-&vis current expenditures. This pattern also reflects institutional weaknesses like the absence of savings inthe goodyears to finance expenditure in the bad years, and insufficient medium-to-long-term planning capabilities in the public sector. E. The FiscalTransparency,Stabilization andResponsibilityLaw 2.34 Ecuador's adoption of fiscal rules and stabilizationfunds is a major step ahead towards an institutional framework favorable to fiscal discipline. The most important steps Ecuador has taken to offset the vulnerabilities stemming from its volatile revenue base and rigid spending structure are: (i)the fiscal rules contained inthe Fiscal Transparency, Stabilization and Responsibility Law and (ii) the setting up of two stabilization funds, namely, the Oil Stabilization Fund (FEP inits Spanish acronym) and the Fund for Stabilization, Social and Productive Investment, and Reduction of Public Debt (FEIREP). Details on these funds are provided inBox 2.2. 25 Box 2.2 Ecuador's Oil Funds The Oil Stabilization Fund (FEP). Article 58a of the Public Finance Reform Act created the FEP, principally using oil revenues that were not foreseen or that were higher than initially budgeted. The FEP i s now allocated as follows: 45 percent to the FEIREP; 35 percent for the construction of the Amazon Highway; 10 percent for comprehensive development projects in selectedprovinces; and 10percent for the National Police for a period o f 5 years, of which half must be spent inthe Amazon region. The Fundfor Stabilization, Social and ProductiveInvestment, and Reductionof Public Debt (FEIREP). In June 2002, Article 13 of the Organic Law on Fiscal Accountability, Stabilization, and Transparency created the FEIREP. 0 Funding sources o f the FEIREP are: transportation revenue from state-owned oil rated lower than 23" on the API scale (that is, heavy crude oil) goes to the fund; budget surpluses of the Central Government; and 45 percent o f FEPresources. FEIREPfunds are allocated as follows: 70percent for the repurchase of public debt and the service of the debt with the Social Security Institute o f Ecuador (IESS); 10percent for health and education expenditures; and 20 percent are destined to a stabilization account of up to 2.5 percent o f GDP, to be used to respond to legally declared emergencies and to compensate for the decline inrevenue when oil prices are low. ~~ Stabilization Funds 2.35 The main objective of a fiscal stabilization fund inan oil-exporting country is to reduce the impact of volatility of oil prices on government expenditures. The FEP was created in 1999. It i s designed to offset fluctuations in world oil prices by saving when prices are high and enabling the government to draw on the fund when prices are low. Historically, however, FEP funds were earmarked, and little remained as a contingency fund. For its part, FEIREP was created in 2002, with two main sources of funds: (i) some revenues from FEP and (ii) additional revenues from the opening of a new the oil pipeline. FEIREP has three objectives: (a) to repurchase public debt, (b) to function as a macroeconomic stabilization fund, and (c) to contribute to social expenditures. 2.36 An analysis of FEIREPmechanisms finds weaknesses that should be addressed. The rules for resource accumulation inthe fund may force the government to save too muchat certain times and not enough at others. The formulae also entail a conservative estimate for how long oil reserves will last, forcing the government into a more restrictive fiscal policy than necessary at times, thus increasing the temptation to skirt the rules, as happenedinMay 2004. 2.37 The goal of accumulating up to 2.5 percent of GDP to fund FEIREP's stabilization component provides insufficient insurance for countercyclical fiscal policy. This stabilization component aims to provide resources when the price of oil i s low or when the country suffers from a natural catastrophe or any other national emergency. Both events reduce fiscal revenues and the latter also raises the demand for public outlays. The amount estimated necessary to provide full insurance for an eventual period of low oil prices, based on average oil revenues for 2000-03, would be 6.9 percent of GDP, compared to 3.9 percent in 1998 (Artana and Moskovits 2004). This is two times higher than the funds available in the FEIREP stabilization component. In addition, the fund is assumed to provide insurance against losses from natural catastrophes. In the last big natural disaster, the El Niiio of 1998, the government made an emergency budget allocation of 1.4 percent of GDP, but this was insufficient: the total cost (both public and private) of the damage was estimated at US$2 billion dollars (almost 10 percent of GDP). Although, a priori, the probability distribution of natural catastrophes and low international oil prices are independent, the 1998 period i s an example of bad luck when both negative external shocks were present at the same time. Though there i s no economic imperative to provide full insurance, authorities should be aware that with FEIREP, they have only partial coverage in a worst-case scenario. 26 2.38 FEIREP does not accumulate resources in an efficient manner to protect the budget. A well-designed Figure 2.8 "Optimal" and Actual Stabilization stabilization fund would have funds entering when the Accumulation Scenarios actual oil price i s higher than the target price, but Ecuador's funds have a more complicated formula. The FEIREPreceives money from two sources: 45 percent of 400 the FEP resources (which capture oil windfalls above a 200 certain price), and fiscal revenues from increases in heavy oil transported through both pipelines beyond a set 0 production level. Now, even if the actual price i s lower -Ideal Fund than the target price set for FEP, FEIREP would still be - ` " " +EcuadorFu ~ d earning resources from the second source. This would not -400 1 be the case in an "Ideal" Fund. Adding this result to the -600 1 previous one shows that the FEIREP may save more than 0.6 0.8 0.9 1 1.1 1.2 1.5 optimally desired when oil prices are lower than the Actual vs Target Price Ratio reference price, and less when there i s a boom in prices 1 Source: World Bank staff calculations. (Figure 2.8). 2.39 Debt repurchasing should be implemented to conserve the country's oil wealth. Because a reduction in government debt i s similar to the creation of a government asset, using oil resources to repurchase debt can be seen as a way of replacing a dwindling asset. Ideally, the amount of the non-oil deficit should equal the return on the size of the oil wealth. Assuming a 3.5 percent discount rate and a conservative oil outlook, the net present value of Ecuador's oil wealth i s US$18.9 billion.24Therefore, if Ecuador were able to have a non-oil deficit of US$600 million a year and to save the difference in assets, yielding a 3.5 percent annual retum in real terms by the time that crude oil reserves are completely depleted (by 2020), it would have assets equivalent to 35 percent of GDP. Ecuador's rules do not follow this "optimal" path. Table 2.6 compares how much FEIREP would accumulate for debt reduction with the amount that would be saved (or would be available for buybacks of public debt) if Ecuador had an "optimal" rule that would maintain a constant oil wealth. Estimates show that Ecuador saves "too little," or in other words, the country i s consuming part of the oil wealth inthe current generation (and implicitly transferringto the future more debt or lower assets than under the optimal rule). Table 2.6 "Optimal Fundversus FEIREPFlows for Debt Reduction (MilnsUS$) 2004 2005 2006 2007 2008 203 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 EcuadcaRule(70TiofFElREP) 191 249 290 331 373 414 420 427 433 440 447 4.54 460 467 474 481 489 "Ideal"LMxRedxion 329 411 506 591 676 761 796 828 861 895 930 966 1 W 1013 1083 1124 1167 - . Note: Prices are forecasted as equal to the long-term trend. Source: World Bank staff estimates. Fine-TuningFiscalRules 2.40 To stabilize the fiscal accounts, the FTSRL introduced quantitative rules for fiscal performance: (a) primary Central Government expenditures should not increase more than 3.5 percent per year in real terms, or the real growth rate, whichever i s lower-this growth rate for expenditures i s reduced to 2.5 percent for the operating expenses of the Financial Public Sector; (b) the non-oil deficit of the Central Government should be reducedby 0.2 percent of GDP per year-for this purpose the non-oil deficit i s calculated by subtracting oil revenues derived from exports (which i s about two-thirds of what ~~ 24. Technical details on this calculation are available from Artana and Moskovits (See Annex A). 21 the government includes as petroleumrevenues) from total revenues; (c) public debt shouldbe reduced by 16 percentage points of GDP during the period January 2003-December 2006 and debt reduction must proceed until a debt-ratio of 40 percent of GDP is reached; thereafter, this 40-percent ratio will remain as the ceiling for public debt; and (d) there are also constraints on sub-national debt: any provincial or municipal government's debt cannot be higher than its annual revenues, and the debt service cannot be higher than 40 percent of the given sub-national government's annual revenue. In addition, Central Government guarantees on private debt are forbidden, and the Executive can only guarantee state debt issued for investment projects. 2.41 The government haslargely met its targets. Although the FTSRL became mandatory in 2003, data for 2002 are also shown to provide some perspective (Table 2.7). In2004 the rule governingplanned real primary spending and the non-oil deficit were met. The scheduled reduction of public debt-to-GDP ratio was also on track in 2003. The executed real primary expenditure, instead of increasing, fell by almost 1percent of GDP in 2003. Only the non-oil deficit reduction was not achieved in 2003, but after two consecutive years of reduction. This i s discussedbelow as a rather restrictive target. Table 2.7 Evaluation of Compliance with the FiscalRules 2004 Measurement Legal 2002 2003 Planned Application Unit Clause Executed Executed (Proforma) Primary central govern- real growth 3.5% 15.1% -0.9% 3.4% expenditure ment budget Non-oil deficit central annual change -0.2% -0.3% 0.0% -.02% government (% of GDP) Public total public debt reduction each 16.0% -12.0% -5.0% debt/GDP 4 years' 3.8% annually Notes: reported figures are for the central government. up to 40 percent of GDP. Source: World Bank staff estimates. 2.42 The 3.5 percent target for growth in real primary expenditures could be ineffective. Rules that constrain the growth in public expenditures are easier to control and more understandable by the public than rules that restrict the overall balance or the debt level. However, the design of this rule raises two concerns. 0 The 3.5 percent target growth is calculated on the budgeted broforma) public expenditures of the previous year, and not on the executed expenditures. Alternatively, it could be the extrapolated state of the expenditures. In the case of the 2004 budget, the rule might become very lenient, since projected real growth inprimary spending could end up being about 14 percent in2004. 0 Continuous budget amendments affect compliance of the 3.5 percent rule. The latter should be applied to the proforma budget approved by Congress. 2.43 The condition to reduce the non-oil deficit by on average 0.2 percent of GDP per year is too restrictive on public finances. Starting from a non-oil deficit of 2.4 percent of GDP in 2003, a 0.2 percent annual reduction would bring the non-oil balance to zero in 12 years, at which point the country would, in theory, no longer depend on oil revenues. Even the most pessimistic estimate assumes that oil reserves will last more than 17 years; therefore, the annual rate of non-oil deficit reduction could reasonably be set at 0.15 percent of GDP per year. 2.44 This condition could be, moreover, unnecessary.To maintain oil wealth for future generations, it may be sufficient that the reduced non-oil deficit become equal to the future real interest yield on 28 returns from real oil wealth preserved in FEREP. Should public debt become no longer a significant problem, a well-designed oil fund would generate savings through a well-diversified portfolio that would finance a non-oil deficit equal to the real return on this asset. This would make part of the oil wealth available for future generations. 2.45 To allow flexibility inresponding to shocks, Ecuador should reduce its public debt below 40 percent of GDP. Ecuador's public deb$ at 53 percent of GDP, is below the ceiling of 60 percent by the European countries under the Maastricht Treaty. However, for developing countries a sensible ceiling would be about half of what i s appropriate for European countries,25when due account i s taken of the narrowness of domestic capital markets, the higher interest rates, and the lower public revenue inrelation to GDP in Latin America.26 This reinforces the need to increase public savings and lessen the vulnerability associated with a fragile fiscal position. Several factors, including low liquidity, low credibility inthe legal framework, and the need to "buy" some confidence after episodes of default on the sovereign debt, suggest that for highly-indebted, oil-rich, emerging countries a lower ratio would be desirable. Thus, it may be better to begin by reducing a government liability, and bringit to a low ratio to GDP to prepare for contingencies, and once an appropriate lower ceiling is attained, to start saving into a government asset. Therefore, while the 40 percent cap on public debt to GDP i s acceptable, it would be advisable to reduce it in the medium term. Alternatively, a contingent clause could be introduced to further reduce debt ratios in years when the oil price i s not excessively low, and there are no natural disasters. 2.46 The FTSRL introduces positive reforms to reduce contingent liabilities. The Central Government cannot guaranteeprivate debt, and guarantees to provincial and sectional governments' debt are limited to investment projects. Although money i s fungible and therefore the loan can end up financing current expenditures, this "golden rule" i s a reasonable objective. Together with quantitative restrictions on states' debt, it puts limits on how much debt subnational governments may contract. By doing so it reduces the size of a contingent increase inCentral Government debt becauseof the bailouts of subnational governments that are so frequent inthe region. F. Issuesindebt sustainability 2.47 A major weakness of Ecuador's fiscal situation is the high level of public debt, the inherited result of decades of undisciplined fiscal management. The country experienced a full-fledged debt crisis in 1999 and defaulted on its debt obligations. After a turbulent period of involuntary debt restructuring, including a 40-percent "haircut" in the restructuring agreement, and thanks also to the ongoing economic recovery, Ecuador was able to reduce its public external debt from 102.0 percent in 2002 to 37.2 percent of GDP as of the first half of 2004. Despite this decline, both the debt level and the debt service are still excessive and the risk of default i s not negligible. In addition, the Treasury has repeatedly faced liquidity problems in serving its debt, which has frequently forced the country to run sizeable levels of arrears with its creditors. Despite progress in reducing arrears27,liquidity shortfalls are still a problem as of the writing of this report. 2.48 A Debt Sustainability Analysis (DSA) indicates that primary surpluses between 4.5 and 5.0 percent of GDP are required gradually to reduce the debt-to-GDP ratio to 40 percent, the 25. For more details see Artana, L6pez Murphy, and Navajas (2003); or Reinhart, Rogoff, and Savastano (2003). 26. The 60-percent-of-GDP ceiling for European countries, expressed as a fraction of government revenues, turns out to be 150 percent. This i s similar to a widely used ratio for private companies, which i s the debt-to-sales ratio. Considering that Ecuador Public Sector Revenues are about 25 percent of GDP, a debt ceiling o f 150 percent o f revenues would yield a "reasonable" debt-to-GDP ratio of 38 percent. The 30 percent "rule-of- thumb" ratio takes into account a higher liquidity risk. 27. According to Minister of Finance YBpez, President GutiCnez's government was able to reduce arrears from $780 million inJanuary 2003 to nearly $280 million inearly June 2004 (see Reuters, 2004). 29 benchmark defined in the FTSRL.28 Most DSA exercises aim at calculating the primary surpluses needed to stabilize the ratio of debt to GDP, that i s to say, the fiscal effort needed to avoid an explosive future evolution of the debt ratios. The exercise carried out in this report takes as a parametric point of departure the policy and legal decision to reduce the debt level to 40 percent of GDP, an objective enshrined inthe FTSRL, and determines what are the primary balances requiredto reach this benchmark by 2006. The base case scenario i s premised on the assumption that the economic program will be successful in sustaining dollarization (Box 2.3). Focusing on the medium-term, no oil windfall i s assumed under the base case scenario. Hence, it i s a conservative one (a high case scenario based on higher oil prices i s presented in Box 2.3). It assumes that the country would have steady growth and low inflation, favored by slow expansionary spending, an increase in non-oil tax revenues, and sizable primary surpluses. Stability will induce a reduction in interest rates to foster private-foreign and domestic-investment and exports as the main sources of growth. Restructuring of the banking system will continue. Funds from the FEIREP will be usedto repurchase debt as mandated by the fiscal rule. All inherited public debt arrears are cleared in 2004, and any default over the next three years i s ruled out. Partial grant financing of the priority social expenditure will be given to the GOE to bring adequate protection to the poorest, while minimizing the potentially negative short-term impact of adjustment measures, thus preserving minimum social and political consensus over the difficult measures to be adopted. Under the base case, obtaining primary surpluses between 4.5-5 percent of GDP would allow the government to cover their amortization payments with little additional financing and continue the gradual decline of the public debt-to-GDP ratio to the sustainable ratio of 40 percent of GDP in 2006, (Figure 2.9). The active repayment of expensive domestic and external (global) debt with the primary surpluses i s a major factor explaining this result. Box 2.3 The Debt Scenarios Base Case: During 2004-06, key assumptions are: growth averaging 4.0 percent, with single-digit inflation rates, converging to international levels (2 percent) by end-period; decline in oil prices in 2005-06, with falling prices partly offset by new oil exports. o A WTI oil price assumption of $32 per barrel is used for 2004, and $20 per barrel for 2005-06. o Steady private investment in the oil sector occurs. Oil production would increase by about 40 percent, or 64 million barrels (an increase from 400,000 bpd to 570,000 bpd), during 2003-06. PetroEcuador's production i s expected to remain relatively flat, increasing only 9.5 percent during this period, whereas private companies should see their production levels increased 81percent. o The increase inoil production will support growth rates o f 4.0 percent inthe forecast period. o As a result, the NFPS primary surplus should average 4.5-5 percent of GDP(for anNFPS overall surplus of 2.2 percent o f GDP) arising from temporarily increased oil and non-oil tax revenues, austerity measures and declining public spending in real terms. o Compliance with all fiscal rules contained in the Fiscal Transparency, Stabilization and Responsibility Law i s also assumed. High Case: This scenario assumes a higher growth rate of 6 percent in2004, led by higher oil prices, with similar inflation rates as inthe Base Case. The results of the high case show: o A WTI oil price assumption of $40 per barrel for 2004 and $30 per barrel in2005 and $26 in2006. o Higher primary surpluses are generated, averaging 5.1%, reducing the fiscal financing needs. o Higher oil revenues also lead to a greater accrual of FEIREP resources and higher levels o f debt repurchasing. o Public debt to GDPratios will achieve the 40 percent target ayear earlier in2005. 28. The baseline DSA is taken from World Bank 2004a. 30 2.49 Additional debt sustainability exercises are carried out taking into account shocks to the financing of the current account. Like all emerging markets, Ecuador faces high volatility of capital inflows. Sudden outflows of capital can be interpreted as a shock to credit or a shock to the capital account and could lead to solvency problems if vulnerabilities are not addressedinadvance. A fall inthe financing of the current account deficit implies that the country must follow a forced adjustment in its absorption of tradable goods. As Izquierdo and Panizza (2003) argue: "To the extent that consumption of non-tradable goods i s a complement inconsumption of tradable goods, a fall in the latter will imply a fall in the former, leading to a decrease in non-tradable prices. Since, for a small open economy, tradable prices are taken as a given, this implies that the real exchange rate will have to adjust. Adjustment inthe real exchange rate will generate valuation effects on the debt-to-GDP ratio, which, in turn, affect fiscal sustainability." Figure 2.9 Base and High Case for Debt Sustainability 60% , PrimarySurplus (%GLP) PublicDebt ("hGGDP) 1 60% 1 1 4.0% 3.0% 2.0% 1.w/o 0.0% - 2003 2004 2005 2006 - 2003 2004 2005 2006 -Base Hgh -Base Hgh Source: World Bank staff calculations. 2.50 Ecuador is exposed to direct shocks to the current account that could work just like a sudden stop. Ecuador's economy i s highly vulnerable to a fall in oil prices (or a fall in remittances). This is becausethe government does not have access to international credit markets and therefore has no tool to smooth out oil price shocks. 2.51 The usefulness of fiscal sustainability exercises lies in their helping policy makers to assess the likely impacts of shocks. Two exercises aimed at assessing the impact of potential shocks on the economy are presented in what follows. Their theoretical rationale and methodology are developed in Volume I1(see Chapter 1). Additional discussion on the theoretical model underlying these exercises i s found inDiaz, Izquierdo, and Panizza (2004). All exercises are carried out with reference to the state of the Ecuadorian economy in late 2003, that is, before the recent sharp increase in oil prices, and supplement the conventional fiscal sustainability analysis aimed at identifying the primary surplus needed to achieve a constant debt-to-GDP ratio under given interest-rate and GDP-growth assumptions. The & exercise identifies the primary surplus needed for fiscal sustainability, taking into account the change in relative prices required to adjust to the shock and its impact on the debt level. In this exercise two cases are analyzed, a 50 percent drop in the price of oil and a sudden drop in capital inflows. The latter includes an assessment of the additional impact of an increase in contingent liabilities that may surface as a consequence of a depreciation of the real exchange rate needed to adjust to a drop in capital inflows. The second exercise gives an indication of how close or far the economy i s from being in a crisis situation given its current debt level and the volatility of government revenue. Exercise No. 1. Fiscal Sustainability under a Sudden Drop in the Price of Oil, Surge in Interest Rate, or Capital Ouflows 31 2.52 A sudden drop inthe external resourcesneeded to sustain the level of aggregate demand of the economy (at a given income) has proven very costly in terms of output in developing countries (Calvo and Reinhart 2001). Ecuador is no exception. I t s crisis of the late 1990s was partly the consequence of a sudden stop in capital inflows that led to fiscal solvency and sustainability problems (Izquierdo 2002). A fall in the financing of the current account deficit implies that the country must follow a forced adjustment in its absorption of tradable goods. This implies an adjustment in the real exchange rate that may be accompanied by a materialization of contingent liabilities that would deteriorate fiscal sustainability. 2.53 Impact of an oil shock and an interest rate increase. A 50 percent drop inthe price of oil from US$24 to US$12 per barrel would lead to a drop innet export proceeds equivalent to 13.3 percent of 2003 imports (a proxy for absorption of tradable goods). For this, a 33.3 percent depreciation of the real exchange rate would be required, because the drop in imports would be accompanied by a similar percentage fall in complementary non-tradable goods. This means that tradable goods would become more expensive relative to nontradable goods (e.g, a real exchange rate depreciation). This will not change the value of Ecuador's dollar debt in terms of tradable goods, but it will lower the value of output-that has a large nontradable component-in terms of tradable goods, thus leading to an increase inthe debt-to-GDP ratio. At dollar interest rates of 11percent, Ecuador would require a primary surplus of around 4.4 percent of GDP to sustain 2003 levels of debt. At this interest rate, a drop of 50 percent in the price of oil would require a primary balance of 5.2 percent of GDP, that is, almost 1additional point of GDP. But, if the shock is accompanied by an increase in 200 basis points ininterest rates and a fall of 1percentage point in GDP growth, the required primary surplus would rise to 7 percent of GDP. Debt sustainability results prove very sensitive to interest rate changes. 2.54 Impact of a sudden stop in capital inflows. Under the same assumptions, closing Ecuador's 2003 current account deficit would require a primary fiscal surplus of 4.9 percent of GDP. Adding an increase in interest rates of 200 basis points, and a fall in GDP growth of 1percent would require 2.2 points of GDP in adjustment, bringing the total needed primary surplus to 6.7 percent. If resulting contingent liabilities from the dollarized banking system surface,29the public debt would increase by 2.5 points and the economy will require an additional adjustment inthe primary surplus of about 0.3 points of GDP, bringing the total to 7 percent of GDP. As a result, when occurringjointly, each of these different elements, that is, contingent liabilities, interest rate increase, and GDP growth deceleration, they would add combined pressure to the less stringent adjustment required if each of those factors were to operate individually and inisolation. Exercise No. 2. Fiscal Sustainability under Tax Revenue Volatility 2.55 This exercise accountsfor tax revenue volatility ina context of expenditure inflexibility and constraints to access international credit markets. The implicit assumption in this exercise i s that the government cannot accumulate more debt than the level it could service if it were to enter a fiscal crisis (defined as the case in which the primary balance remains forever at its lowest possible value3'). Thus, even under the assumption that the government would always be willing to pay its debt, there would still be a limit above which no additional lending would take place, because the government may not have sufficient resourcesto pay back under a sustained crisis scenario. While inthe standard approach to fiscal sustainability the primary balance (to GDP) that stabilizes the current debt-to-GDP ratio i s assumedto be its steady state level, in this exercise only a threshold level of debt to GDP i s defined, but this level i s neither the equilibrium that will necessarily be observed nor the optimal level of debt. With this information, the government should strengthen fundamentals so that the probability of hitting the upper 29. Assumes that 20 percent of loans from the non-tradable sector go bankrupt (see Izquierdo 2004). 30. This lowest possible value is dictated by the lowest possible realization of public revenues and the maximum adjustment inexpenditure that would be feasible for the government. 32 bound of government debt remains low. Lower values of government revenues per GDP (which would result from taking into account revenue volatility) would require that sustainable debt levels be adjusted downwards, especially if the government cannot commit to large expenditure cuts while facing a shock. Taking into account these considerations, this exercise shows that, with 50-percent probability, it would take six years for Ecuador to reach the debt threshold, and with 35-percent probability it would take three years. This information should help the authorities to correct fiscal imbalances well before hitting debt thresholds. 2.56 Both exercises show additional reasons why Ecuador needs to prepare for potential oil, interest-rate, or capital-outflow shocks by strengthening its fiscal accountsto achieve fiscal primary surpluses at least between 4.4 and 5.2 percent of GDP. Were authorities willing to protect the economy from the additional potential impact of resultant output slowdown, interest-spread increases, and contingent liabilities, the required primary fiscal surplus should be around 7.0 percent of GDP. As these exercises imply, negative shocks could quickly add substantial pressure to debt sustainability in the future. G. PolicyRecommendations 2.57 Inadopting dollarization, Ecuador chose a very demandinginstitutionalframework for the conduct of economic policy. The country faces significant challenges stemming from dependence on volatile oil revenues, unpredictable capital flows, and the devastating effects of nature-induced supply shocks. These disturbances are bound to operate regardless of the monetary regime, and would seriously test the most skillful and inspired team of policymakers. But in the context of dollarization, the tests are even more stringent. While, from the standpoint of stabilization objectives, dollarization gives policymakers a head start by the sheer impossibility of a currency crisis and hyperinflation it entails, demand management is restricted to fiscal policy, and the buffer of the nominal exchange rate i s no longer available. Moreover, the economy's adjustment mechanism after a negative shock i s likely to be through recession and deflation, thereby magnifying the challenges. In addition, for all its advantages for financial stability, dollarization in the context of an open capital account needs to be supported by a robust, well-regulated financial system where information-asymmetry problems can be kept under control. Finally, there is the challenge posed by the need to enhance competitiveness in the tradable sectors, where real exchange rate depreciations are costly and painful when, as in the Ecuadorian case, wages and prices are inflexible in the downward direction. To address these challenges, the next recommendations focus on a proposal for a fiscal agenda that could help strengthen the dollarization framework. 2.58 The fiscal agenda, however, face formidable political-economy and institutional obstacles in Ecuador. As it was pointed out inthe first chapter, the ultimate root of Ecuador's fiscal stance lies inthe political-economy regime that took shape in the wake of the oil discoveries of the late sixties and i s still prevailing. This i s the political-economy regime that begets-and keeps reproducing-the incentives that bias fiscal policy towards inertial expenditure expansion and against flexibility of resource allocation. The current oil price boom intensifies the resistance to fiscal adjustment as it creates a public opinion atmosphere where, in the eyes of some social groups, fiscal discipline seems both unjustified and unnecessary. However, in the last several years, large segments of public opinion have come to support the reform camp as the citizenry increasingly understands the inextricable connection between the need for fiscal adjustment and the creation of the fiscal space needed for a poverty reduction strategy. 2.59 Thus, the medium-term objective of fiscal policy is to strengthen and preserve the sustainability of the fiscal accounts. This could be achieved with the support of an implicit or explicit Fiscal Pact, whose main general goals would be to: (a) deal with the public sector's insolvency risk; (b) carry out a comprehensive tax reform; (c) eliminate the inertial bias toward payroll expansion and investment cutback in the management of public finances; (d) reform the civil service and pensions 33 system; (e) use the oil wealth for debt reduction; (f) while duly complying with the fiscal rule requirements inthe short tem (and thereby building credibility), amend in a few years the FTSRL to make it a sharper, more coherent, and more powerful instrument for fiscal consolidation; and (g) address the issue of the Treasury's short-termliquidity problems. To lower insolvency risk, the authorities should obtain the large primary surpluses required to reduce the debt to a sustainable level. Under a base case scenario, the PER'S debt sustainability analysis indicates the annual NFPS primary surplus needed to meet interest payments and to reduce the debt ratio to at least 40 percent of GDP during 2004-07. The only way to achieve primary surpluses of this magnitude i s to generate new revenues while controlling expenditures. Whereas the 40 percent ratio is the one contemplated by the FTSRL, an additional 5 percent would be desirable to allow for contingencies and shocks. The government should persevere in the pursuit of a comprehensive tax reform. Following a return of oil prices to their historic average level, the objective of tax reform should be to raise the non-oil tax revenues by at least 2.5 percent of GDP. This i s equivalent to the amount that must be accumulated in the FEIREP to comply with the FTSRL's rules. Tax reform should be directed toward: (a) eliminating or rationalizing all non-constitutionally mandated earmarked expenditures; (b) eliminating or rationalizing exemptions to VAT, income-tax, and customs duties, with the exception of the VAT exemptions for unprocessedbasic foods; (c) simplifying the tax structure, and eliminating taxes that generate minimal revenue; (d) completing Customs reform; and (e) strengthening tax administration. Political-economy obstacles to tax reformare substantial as it affects powerful vested interests. This explains why multiple proposals for new tax legislation have been repeatedly rejected by Congress. However, this only means that the reform camp will have to persevere in the task of creating the political conditions that favor reaching a Fiscal Pact. Eliminate the inertial bias toward payroll expansion and investment cutback. This can be done through: (a) containing the growth of the wage bill (more details on this below); (b) a review of programs in every spending ministry to find those that can be eliminated because they either do not adequately represent current social priorities or are duplicative or wasteful; (c) better targeting of social programs and elimination of those that cannot be directed sufficiently toward the poor; and (d) using FEIREP funds to repurchase the most expensive debt first (particularly in Global Bonds) to reduce interest payments). Complete civil service reform. Three measures that the government could consider are: (a) extending the wage unification law to employee categories that were not covered by the FTSRL; (b) a program of public employment retirement, complemented, when necessary, by hiring bans, elimination of vacant positions, and early retirement programs; and (c) permanent review of staffing needs. The pensions system must be reformed so as to render it more equitable; improve its institutional design, make it more efficient and transparent; reduce its current fiscal cost; and preempt future Central Government 1iabilities:l The Government should develop and implement several policies aiming to solve or reduce the problems of low coverage, inadequate institutional arrangements, costly and inequitable subsidies, and long term sustainability risks. This report suggests: (1) adopting an inclusive approach inthe design of the system, providing income security to groups with little or no coverage but great needs, in a fiscally sustainable manner; (2) changing the sector's institutional scheme, by completing the administrative separation of the social security component and the health component and by creating an instance within the central administration in charge of designing and implementing social security policies; (3) reviewing the autonomous status of 3 1. Policy recommendations on how to deal with the social security system's health component are presented in Chapter 4 of Volume 11. 34 IESS to avoid political interference in its daily operations; (4) increasing transparency, by adopting a permanent program to prepare, disseminate and evaluate information on the status of IESS and its different programs; (5) eliminating the government contribution to IESS benefits, adjusting the parameters of the system to achieve actuarial sustainability in the medium and long term; and (6) reviewing the system's legal framework to create a simpler, more transparent model. A structural reform (including the creation of individual accounts and private management) may be an option, but only if it can be implemented with a reasonable social and political consensus and in a fiscally sustainable way. >Using the oil wealth for debt reduction is an optimal strategy. Given the large country risk premium on public debt, the best alternative i s to use oil revenues to reduce public debt well below 40 percent of GDP in a gradual manner until 2006, and once this ceiling i s attained, start accumulating into financial assets that will allow the country to have a reasonable non-oil deficit, even after oil reserves are depleted. This strategy has several advantages. First, it will reduce interest payments for the government. A quantitative exercise shown in the Statistical Appendix demonstrates that, under suitable assumptions as to amount of additional debt repurchases, annual interest payment reductions can be as highas 0.2 percent of GDP in2005 and 2006 (for details, see Table A76). Secondly, ceten's paribus, the fiscal deficit will be lower. Thirdly, real interest rates for the private sector will also be lower. And this through two channels: (i) government borrowing in domestic capital markets will decrease and (ii) the interest rate for the sovereign, which tends to be a minimumfloor for borrowing costs for the private sector, will go down. >Buying back the external debt would improve government credibility and strengthen the economy's growth prospects. This is the debt strategy that will moderate the change in relative prices that i s created by the oil windfall and help strengthen the markets' perception that Ecuador's debt sustainability i s improving. However, it must be borne in mind that a policy of repurchasing external debt may be politically costly as it i s rejected by some segments of public opinion. Indeed, the legal guarantee involved inGlobal Bonds may raise the debt's repurchasing price inthe secondary market, thus reducing initially anticipated savings. Moreover, using available FEIREPresources to buy back domestic debt could add pressure to aggregate demand, unless the holders of domestic debt take their payments out of the country. This i s surely not the case with the pension system debt, and i s probably not the case with other debt held by residents, either. So, to prevent demand pressure, it must be borne in mind that debt buyback is no substitute for expenditure discipline. If the government cannot cutback inefficient expenditure, the rollover of domestic debt mightjust be higher domestic debt in net terms. Hence more fiscal vulnerability and lower reduction of the overall debt ratio than projected. Even repurchasing external debt might not prevent this outcome. Hence, assuming expenditure cuts take place, it would then be advisable to keep assessingthe advantages and disadvantages of repurchasing domestic or external debt with FEIREP resources. The repurchasing debt policy should weigh the benefits of reducing the country risk and gaining access to financial markets with the increased cost of the renegotiated repurchased price of bonds. In general, expensive debt, such as the Global Bonds, should be exchanged for less expensive one ina transparent process. 9 In the short term, strictly complying with FEIREP rules to prevent diversion of its debt buyback funds would strengthen its credibility. The pressure to dip into these resources i s high, and windfall oil revenues have always been used to finance poorly selected public investments. High oil prices such as currently prevail, make changing FEIREPrules more tempting. 9 FEIREPshould startdeveloping its debt repurchases to gain early credibility, butinthe future it should clearly separate the stabilization function from the savings and debt-reduction functions. This will require allocating a higher amount of FEP revenues to the stabilization fund, to deal properly with the volatility in the international price of crude oil. Now, by contrast, the rules force the government to save too much when prices are depressed and too little when prices are high. 35 With regard to the second function, a more precise definition is neededon how muchto save for debt reduction and how much for accumulation infinancial assets. The current rules force the government to save only about half of what i s desirable. All production of crude oil should be involved in this Savings Fund instead of targeting only the increases in production of heavy oil, and how much to save should depend on an estimate of the country's oil wealth and a target for a sustainable non-oil deficit. Ifthe design of the oil fundis improved, the mandatory reduction of the non-oil deficit at an annual rate of 0.2 percent of GDPmightbe too restrictive inthe future. >FEIREPresources should be partly invested in highly liquid assets abroad. Because FElREP funds will be used when crude oil prices are low, at which time Ecuadorian assets are more likely to be illiquid and depressed in price, it i s prudent to invest these funds in liquid financial assets independent of Ecuador's risk. A portion of the stabilization fund should be invested in the same way as the Central Bank's foreign reserves-abroad and in relatively safe assets. However, the downside i s the low rate of return on highly liquid assets. Because the size of shocks varies, it i s more likely that partial amounts of the fund would be used than all the funds at any given moment. Then, to increase returns, a part of the funds should be invested in a diversified portfolio of longer- term assets with a higher return, but exposed to short-term fluctuations in price. Moreover, a well- designed stabilization fund will also try to offset the macroeconomic consequences of oil price variations (the Dutch disease), by investing abroad. 9 With respect to other reforms of the FTSRL, itis recommended that discretion be reduced in some additional areas. One such area i s the definition of the reference price for crude oil that i s included in the budget. The reference price should be best estimated on the basis of a technical criterion that could be included inthe FTSRL. Another improvement would be to apply the criterion of the 3.5 percent growth rate for primary expenditures with reference to the executed budget of the previous year, and not to the approved budget. 9 The Treasury's short-term liquidity difficulties badly affect the authorities' credibility and the country's reputation. Inearly 2004, the public sector has total net financial requirements of US$1.8 billion. Once the identified sources and amounts of financing are taken into account, a financial gap of US$548 still remained. A similar gap is projected for 2005. Given the composition of the public debt, closing the liquidity gap in 2005 i s critically dependent on the design of programs of substantial structural reformas a basis for agreements between the government, the IMF,and other multilaterals. Structural reform, supported by sound macrofiscal management will enable the country to obtain cheaper, rapidly disbursing and freely disposable funds from the multilateral agencies. This i s also the road to enhancedcredibility of the debt reductionplan. 36 ChaDter 3 Pro-Poor Expenditures and the Fiscal Space 3.1 InMay 2004 PresidentGutidrrez signeda decree committinghis Government to developing a Poverty Reduction Strategy (PRS). Findingresources for implement the strategy i s a a significant challenge. The task seems enormous, given the structural and fiscal constraints highlighted in other chapters. On one hand, fiscal performance remains mixed. Chapter 2 depicts a fragile consolidation of fiscal accounts during the 2000s. NFPS fiscal revenues have increased inpart due to tax receipts, albeit these declined in2003. Total NFPS expenditure has remained constant, but expenditure composition has worsened in favor of current spending, especially the payroll, and against investment spending. Debt ratios have decreased, but these are still high and their further reduction requires sustained efforts in building sizable primary surpluses in the next several years. On the other hand, expenditure flexibility has worsened with rigid expenditure-wages, earmarkings, and interest payments-accounting for an increasingly large portion of budget revenues and leaving almost no funding for development activities. In this context, serious questions remain about the level and efficiency of public (mainly social) expenditure and the ability to finance an effective poverty reduction strategy. This chapter examines those issues. 3.2 Authorities recognize that the present fiscal stance is inconsistentwithin the framework of a poverty reductionstrategy. Their goal i s to take advantage of the process leading to the PRS to improve the amount and quality of public expenditures: level, composition, and execution capacity while augmenting budgetary flexibility, and diversifying the sources of financing. To do this, they intend to (a) increase the level of education and health spending, to at least a combined 5 percent of GDP; (b) correct the composition of public expenditure between capital and current outlays from a 20:80 ratio to a 25:75 ratio; (c) de-concentrate public investment beyond the three provinces where it i s currently concentrated; (d) reduce budget rigidity; and (e) modify the de facto separation of financing of current spending by non- oil tax revenue, and investment by oil revenue (Salazar 2004). Those goals identify important aspects of public expenditure size and "quality" and move in the right direction, but officials will need to further detail their implementation. The answers to four more fundamental questions would be helpful: (a) I s the social spending level enough to support poverty reduction? (b) Are the efficiency and effectiveness (targeting) indicators-two other key aspects of expenditure quality-of resources assigned to the PRS satisfactory? (c) Given needed fiscal retrenchment in the next .two to three years, how can the country gain additional fiscal space and improve efficiency and targeting of public expenditures to finance a PRS? and (d) How can the government link expenditure to desired intermediate outcomes and final outputs, especially those related to the Millennium Development Goals (MDGs)? This chapter answers all four questions, and provides policy alternatives. However, it highlights complex trade-offs which require strong political will and the creation of a national consensus. A. I s Social Expenditure Enough for Poverty Reduction? 3.3 Budgets for the social sectors are low by regional standards and declining (Table 3.1). Following dollarization, social sectors expenditures peaked up to 5.1 percent of GDP in 2002, thus reversing a long term decline and almost reaching their level of the early 1980s. However, the levels of education and health expenditures remains low-about half the combined LAC average of 7.5 percent of GDP-and are declining since 2002.' There are two main reasons for the increasing trend in social 1. In 2001, according to World Development Indicators, LAC averages for education and health (no pensions included) were 4.1 and 3.4 percent o f GDP respectively. An important caveat for Ecuador ratios as suggested 37 outlays in the post-dollarization context: On the one hand, the need to increase wages and salaries in the social ministries (see Chapter 2); on the other hand, the importance to strengthen the safety net under social welfare programs like the Bono. As a percentage of GDP, social welfare programs increased from 0.3 percent of GDP in 1995 to a peak 1.3 percent of GDP in 2002. These programs were especially critical to protect vulnerable groups in a dollarized economy vulnerable to shocks (see Chapter 4 for a detailed review of the performance of the most important programs). Payroll increases have been achieved at certain costs: in education, a significant deterioration of school infrastructure; and in health, severeproblems indrug supplies (see below). Table 3.1 Ecuador: Social Expenditure' of Central Government as a Percentage of GDP2 1973 1975 1980 1985 1990 1995 2000 2001 2002 2003 20043 Total Social Exvenditure 3.5 3.3 5.3 4.7 4.4 3.6 4.0 4.3 5.1 5.0 4.9 Education 2.9 2.5 4.3 3.5 2.8 2.4 1.9 2.2 2.7 2.6 2.6 Health 0.5 0.7 0.9 1.1 1.3 0.9 0.8 0.8 1.3 1.3 1.2 Social Welfare 0.1 0.1 0.1 0.1 0.3 0.3 1.3 1.3 1.1 1.1 1.1 Cash transfer programs 0.8 0.7 0.5 0.6 0.6 Other 0.1 0.1 0.1 0.1 0.3 0.3 0.5 0.6 0.5 0.5 0.5 Memo: Education & Health 3.4 3.2 5.2 4.6 4.1 3.3 2.7 3.0 4.0 3.9 3.8 Notes: 1. Social expenditures refer to the Central Govemment budget only. Social expenditures include education, health, social welfare, and labor, and cash transfer programs. Cash transfer programs refer to the Bono Solidario for 1999-2002, and includeBecaEscolar and Bono de DesarrolloHumano thereafter. 2. The social expenditure share of GDP is calculatedon the basis of constant price series in dollars of 2000. The share at current prices i s slightly higher on average (0.3 percentagepoints for the 1990s, and 0.1percent for the whole series), but the trends are the same. The difference between the constant and current price shares i s explained by the difference in deflatorsfor govemmentspending andGDP, the former being-on average-slightly higher. 3. Numbersfor 2004 refer to the provisionalbudgetfor social expenditure data and CentralBank projectionof GDP growth. Sources: CentralBank and MEF data. Series from Vos andothers (2003), updatedandadjustedby Vos andPonce(2004). 3.4 There has been a dramatic decline in real public spending on education and health over the last two decades. Despite the recovery in recent years, by 2003 the real level of education spending per beneficiary was 40 percent below that of 1980.2 Because much of education spending i s on teacher salaries (about 80 percent of the total in 2000), and recent budget increases have been mainly driven by salary adjustments, by applying the wage deflator, there i s no actual recovery inreal education spending. This is consistent with the hypothesis that most, if not all, budget increases after 2000 went into raising teachers' nominal salaries. Indeed(a) the shares by educational level have remained fairly constant, such that this factor does not seem to explain the trends in education; (b) private school enrolment has increased somewhat, but not substantially, implying that this i s not a relevant reason either; (c) student- teacher ratios have fallen, while enrolment has increased such that declining trends in spending cannot be attributed to a lack of supply of teachers, inprimary education the student ratio fell from 30 to 23 students per teacher, and in secondary education it fell from 13 to 11; and (d) rising cost-sharing to cover school maintenance and other operational costs of schools suggests that this i s the area where public spending has been faltering. This is, however, difficult to corroborate because of data problems. For its part, public health spending per capita remains among the lowest in Latin America (only Haiti spends less), and recent budget increases have also mainly gone into salary increases for health workers as well. Two additional common factors inboth sectors-as well as inthe basic infrastructure sector-are the persistent by Ecuadorian officials, i s that they do not fully comply with the functional budgetary standard classification, but arejust proxies. 2. The measurement of the recovery in real education spending from 2000 onward i s sensitive to the choice of deflator, not so much for the years prior to that. Here, education spending is deflated using, alternatively, a weighted price for public consumption and investment (weighted for their respective importance in education spending), and a nominal wage index for public employees (Vos and Ponce 2004). 38 poor targeting of resources and inequalities in access to service delivery. These findings point to the need to assess the pro-poor focus and efficiency of the social and basic infrastructure programs. B. I s Social and Basic Infrastructure Expenditure Pro-Poor? Main Social Programs 3.5 About half of total social spending is not pro-poor. Incidence analysis allows examining to what extent expenditure i s pro-poor (that is, a negative differential between the share of spending received by the richest and poorest quintiles, thus favoring the consumptionlevels of the poor relatively more than those of the rich, and hence reduces consumption inequality), mixed or non-pro-p~or.~ In2003, andin absolute terms, the poorest quintile received 12 percent of social expenditure, compared to 27 percent to the richest quintile (Vos and others 2003). Table 3.2 combines findings from Vos (2003) with the actual amount of subsidies in 2003, which gives a broadly disaggregated picture of pro-poor and non-pro-poor social expenditure. InEcuador: 0 About 45 percent of total social expenditure can be considered non pro-poor, since a significant amount of spending benefits households inthe top of income distribution. Table 3.2 Pro-Poor and NonPro-Poor Social Expenditures, 2003 Fiscal Direct Distributive Effect Direct Distributive Effect cost (in US$M, 2003) (with respectto the equity line) Typeof US$M, Pro- Non Mixed Pro- NonPro- Type of Expense Program 2003 Poor Pro-Poor Poor Poor Mixed UNIVERSAL 1,706.6 Education 1,123.8 Primary Universal 453.4 453.4 X Secondary Universal 304.0 304.0 X University (public)' Universal 366.4 366.4 X Health 582.8 Per service provider 381.5 IESS ssc2 Affiliates 191.4 191.4 X Universal 35.9 35.9 X MSP Universal 160.2 160.2 X Per care service 195.3 Hospitals Universal 114.4 114.4 X Centros de salud3 Universal 80.9 80.9 X FOCALIZED 254.6 Schoolbreakfast Focalized 11.0 17.0 X Infantcare Focalized 11.7 11.7 X Bono Solidario4 Focalized 159.9 159.9 X SUBSIDIES 221.0 Cookinggas Universal 221.0 221.0 X TOTAL 2,182.2 743.9 974.1 464.2 Notes: 1. Refers to earmarked transfers to universities. 2. This program is not financed by the Central Government, but by the Social Security Institute, and by employee and employer payroll contributions. 3. Includes sub-centros de salud y puestos de salud. 4. Excludes the Beca Escolar. Source: Vos and others (2003),MEF, and Bank staff's estimates. 3. A mixed ranking represents a close to nil differential, and a non-pro-poor a positive differential (see Figure 3.1 for a graphic representation). Notice that incidence analysis also allows examining progressivity or regressivity o f social spending. Progressivity is defined as the degree to which spending improves the consumption levels of the poor with the rich under a non-proportional distribution. Conversely, a social program can be non pro-poor ifproportionally higher resources are allocated to the richest quintiles. For concrete examples on Ecuador, see World Bank (2004b). 39 Primary education and, to a mixed extent, secondary education, are pro-poor, and combined they account for about two-thirds of education spending. The Health Ministry's spending (MSP) can be considered pro-poor, but combined with the Seguro Social Cumpesinofinanced by the Social Security Institute, account for barely one-third of the health sector. Tertiary education (universities) and IESS spending are most relevant among non-pro-poor spending, and represent about one-fifth and one-third of education and health sector resources. All social welfare-targeted-programs (school breakfast, infant care, and the cash transfer Bono Solidario [recently converted to BDH]) are pro-poor, but together represent only about 10 percent of social expenditure. Notice that these programs still have, however, significant improvementsto do bothinterms of targeting and nutritional impact. The subsidy to the cooking gas is definitely not pro-poor, and represents a meaningful 10percent of all social expenditure (equivalent to resources allocated to all focalized social welfare programs). i 3.6 Incidence analysis allows ranking Figure3.1 Incidenceof SocialExpenditureand Energy social expenditure and energy subsidies from Subsidier .99! - less to more pro-poor. Taking the difference Schmlbreakisst between the richest and poorest quintiles, it appears that the school breakfast, primary school Primaryschml spending, and the Bono cash transfer, in that &F richest Ban0 order, are the most pro-poor programs (Figure Heal&care SSC 3.1).4 At the bottom of the classification, university education, JESS health care and the Healthcare MSP cooking gas appear to be the most non-pro-poor secwdary school outlays. Before its disappearance in 2003, the Cookmggar gasoline subsidy, in particular, did not have any impact on the two poorest quintiles, and 85 percent of this subsidy was concentrated on the richest quintile. If public expenditure needs to UZIOllnr 1YD1IYI l l l 1 l l I be reoriented toward financing of a PRS, these 0 10 20 30 40 50 60 70 SO 90 findings point to key areas of reform. Source: Vos, Ponce, Cuesta, and Borobrich (2003). 3.7 The pro-poor share of total social expenditure is even lower. Social expenditure i s defined in a wider sense, since it includes off-budget financed programs-like IESS health care payments or the Seguro Social Campesino outlays, and excludes self-financed ones-like IESS pensions? Combining the distribution per quintile of household income found by Vos and others (2003) for 1999 with the actual amount of social expenditure and subsidies in 2003 allows disaggregating cumulated outlays by quintile. Table 3.3 extracts these conclusions: e Social subsidies are progressive. They represent a higher share of poor income households, inasmuch as they representtwo-thirds of their household budget (67.2 percent). 0 The cooking gas (41 percent) and university tuition (24 percent) subsidies, and IESS health care (16 percent) together represent 80 percent of total subsidies to the richest households quintile. Notice 4. Two important caveats to these findings are that (a) even though the School Breakfast and the Bono appear as more pro-poor, that does not mean that they do not require important re-targeting efforts (both programs are rather fine-tuning their database of beneficiaries); and (b) the Seguro Social Campesino do not depend on the Central Government, but on IESS, which makes more difficult to channel increased resources to it directly. 5. Table 3.3 does not include all social outlays. Important missing ones are IESS, ISSFA and ISSPOL pensions, which are all very sizable and regressive (e.g. IESS spends about US$500 million in pensions per year). 40 9 9 9 9 9 E E Z E E ? ? ? ? ? m o o 0 0 ??h)?9 9 9 9 9 9 9 0 0 0 0 0 %41?;;'" E 9 9 T Y . r q 9 9 ? ? N 0 0 0 0 0 o o o o * that IESS health care payments are off-budget and financed through an earmarked payroll tax, not through the Central Government budget. In aggregate terms, the poorest quintile receives 17.4 percent of total outlays, compared to 21.4 percent for the richest quintile. The most pro-poor program, school breakfast, receives an annual amount (US$17 million) that represents less than 1percent of total social expenditure, and is 20 times less than the most regressive subsidy, the one devoted to the university tuition. Notice that the gasoline subsidy was rightly eliminated in 2003 with the increase in gasoline prices early adopted by the Gutierrez's administration. Notice that despite this finding, the school breakfast-served by PAE-still has important shortcomings regarding its targeting and nutritional impact, and i s under review (see Annex C). The pro-poor SSC health care receives 5 times less resources (US$35.9 million) than the non-pro- poor IESS health care spending. The Bono receives a little less resources than IESS (US$160 million compared to US$191 million), but the number of beneficiaries is four times larger than the number of retirees inthe pension system (Rofman 2004). 3.8 Incidence analysis has several caveats but they do not affect the main conclusionspresented. First, not all social programs are designed exclusively for poor households. Second, even if a programis targeted to the poor, its level of progressivenessdepends on how effective the targeting mechanism is. As a result, the higher the leakages in public spending, the larger the losses in progressiveness. Third, the cost of providing a service may vary across groups or areas, and often service delivery to poor areas, especially rural areas, i s more expensive. Fourth, even pro-poor programs, like the Bono, have difficulty maintaining adequate pro-poor targeting, but the welfare index-known as SELBEN (the Sistema de Identificacidn y Seleccidn de Beneficiarios de Programas Socia1es)-which ranks Ecuadorianhouseholds according to demographic and structural characteristics-remains a solid targeting instrument for social programs locatedinthe first and second quintiles (World Bank 2004b).6 3.9 Ecuador has had a mixed performance in education indicators? Despite the significant amount of non-pro-poor expenditure in tertiary levels, Ecuador has good coverage of primary education. Poor quality and low transition ratios to secondary education remain the major issues in the sector. Inequality in access to education has also increased between urban and rural populations, and between the rich and the poor. Only the gender gap in education appears to be by-and-large closed. Positive educational outcomes show the advantage of making education expenditures more efficient. 3.10 Educational outcomes. Indicators continued to improve during the 1990s and into the new millennium: 0 There has been continuous growth in the average level of schooling since the 1970s: in 2001, the average adult had completed 7.3 years of schooling, up from 6.7 years in 1990. This level is above the Latin American mean, and i s about the same as that of East Asia's population. 0 By 2001, the gender gap had practically been closed: 7.5 years for males compared to 7.1 years for females. Thus, educational levels of the female population have risen much faster than that of males, suchthat, interms of net enrollmentrates, girls already outperformboys at all educational levels. 0 Net enrollment in primary education increased from 88.9 to 90.1 percent between 1990 and 2001, approximating the MDGof primary education for all. 6. In2003, the estimated targeting errors o f the BDH are as follows: (a) using income poverty, the exclusion error i s 23.4 percent and the inclusion errors 9.7 percent; and (b) using SELBEN, these percentages are respectively 22.2 and 7.3 percent (Table A54). 7. This section draws from Vos and Ponce (2004). 42 3.11 Important concerns, however, remain in the education sector. First, the transition rates from primary to secondary education and from secondary to tertiary did not improve in the 1990s. Second, important disparities remain, affecting rural, indigenous, and black populations. The average level of schooling of the rural population i s less than half the one of the urban population, and the gap is even larger for the indigenous and black populations. This i s also the case for illiteracy rates. Third, the overall quality o f education i s poor, with math and language test scores worsening between 1996 and 2000 and starting from an extremely low baseline (Ecuador scores lowest for the Latin America region). Fourth, internal efficiency indicators, measured by desertion and repetition rates, have also worsened, with the number of years pupils need to complete primary education increasing from 6.7 years to 6.9 years between 1995 and 2001, a decline concentrated in urban schools; and higher dropout by girls in secondary rural schools and by boys in urban schools, seemingly for economic reasons. Retention rates and education quality also appear affected by the highrate (15 to 18 percent) of teacher absenteeism and frequent teacher strikes in Ecuador (Box 3.1).* Fourth, budget allocations to primary education schools are not approved according to standards in terms of coverage, vulnerability and poverty ratios, studentdteacher ratios by school, and type of school establishment. Box 3.1 Ecuador: Teacher Absence inPrimary Schools As part of the PER work, andincollaborationwith amulticountry study initiated by the World Bank as abackground paper for the World DevelopmentReport 2004, a pioneer national teacher tracking survey was carried out in Ecuador. (See Rogers et al. 2004 in Volume 11.) It aimed to determine teacher absence rates and their main correlates in Ecuador. During 2002-03, survey teams interviewed a random sample of 720 teachers in 102 primary schools in 51 randomly selected parroqufas were surveyed. Each school was visited twice, in December 2002 and again in January-February 2003, to allow two observations of teacher attendance. Of the 670 fulltimeteachers who would normally have beenteaching at the time of the survey visit, 86.5 percent. On average, teachers were found in the classroom (or accompanying the enumerator) 79 percent of the time, although in a fifth of those occasions the teacher was not teaching at that moment. The overall teacher absence rate, measured as the fraction of teachers who could not be found anywhere in the school, was about 14 percent, with a relatively small number of teachers apparently accounting for a larger share of the absence than in other countries. About half of the overall absences were not accountedfor by the schooldirectors. Percentage of time the teacher was found ... December January- Total 2002 February (both 2003 rounds) Inthe classroomteaching 59.5 64.8 62.1 Inthe classroomnot teaching 16.0 14.9 15.5 Out of class on scheduledbreak 0.0 0.0 0.0 Out of class but in schoolDremises 6.0 5.1 5.6 Source: 3gers et al. (2004). The study also analyzed the correlates of (and hence possible reasons for) absence, with some notable results. Overall, Bangladesh, India, Indonesia, PapuaNew Guinea, Uganda, and Zambia all recordedhigher teacher absence rates than Ecuador, and only Peru's absence rate was lower (Chaudhuryet a1 2004). Surprisingly, in Ecuadorabsence i s not more likely in remote, rural areas-as i s often found in other countries-but rather in urban areas. In fact, teachers in urban schools (outsideQuito) are absent at twice the rate of teachers in the most remoteruralschools,perhapsbecause of greater employment opportunities and distractions in the urban areas. Also, one-teacher schools (escuelas unidocentes) are not associated with higher teacher absenteeism, althoughmulti-teacherpolidocente schools apparently are; thus as a general matter, the effects of having to teach various grades at the same time are ambiguous. In general, a better community (lower poverty rate) and institutional environment are associated with reduced teacher absence, whereas individual characteristics of teachers (age, tenure, education, unionmembership) do not seem to make muchof adifference. Some factors that might be expectedto raise performance through formal monitoring and enforcement of attendance-notably, proximity to a Ministry of Educationoffice, 8. When compared to other countries, though, the Ecuadorian absence rate i s in the low range (Rogers and others 2004). 43 and the past use of discipline by the school director-are indeed associated with lower absence. On the other hand, active parent committees, per se, do not appear to reduce teacher absence. Finally, and perhaps more significant, contract teachers (especially those not hired at the school level) have a higher probability of absence than regular teachers, despite the greater leverage that schools and communities might be thought to have over contract teachers. Hence, special contracts may not be the right "quick fix" for the problem. Inthe analysis of access to education, no doubt absenteeism will affect the quality of education, but the survey is less clear to what extent it will influence school enrolment. As indicated, rural school enrolment rates are much lower than enrolment rates inurbanareas, and are thus seemingly unrelated to teacher absence. Also, teacher quality does appear to have a positive influence on school enrolment, but does not affect absence. 3.12 Ecuador has also improved health conditions and this is another reason why an increased social budget would help improve living conditions. Since 1970, infant mortality rates have been cut by 70 percent to 34 per 1,000 live births. Health provisioning has shifted toward greater emphasis on primary health care inpublic provisioning, and inpatient hospital care by private health providers. Health policy reforms have includedincreasesinuser fees for public services, decentralization of public services, special programs providing free health care for the poor (including the free maternity program), and introduction of demand subsidies for health through the conditional cash transfer program Bono de Desarrollo Humano. However, reforms as implemented so far have yet to address some fundamental problems in health service delivery, including access to services, the quality of services, and inequalities across social groups and geographic areas. Important inequalities remain, showing muchhigher mortality rates and limited access to health care for the indigenous population, the poor, and those living in rural areas. Inequalities in access to health facilities have also increased, partly because of expensive user fees and partly because health inputs (particularly drug supplies) have fallen well behind requirements. Reforms were implemented under tight budget constraints, because the health budget (in real per capita terms) declined for most of the 1990s. However, continued expansion of the immunization program and the introduction of the Free Maternity Program have compensated for this effect, at least for young children and pregnant women. This could explain the continued decline in infant mortality during the 1990s. 3.13 Health outcomes. Health indicators have also improved during the 1990s and into the new millennium: 0 Life expectancy at birth increased from 48 years to 72 years between 1950 and 2000. This upward trend was sustainedduringthe 199Os, adding another 5 years to life expectancy. Parallel declining trends are found in child and infant mortality rates. The overall mortality rate dropped from 13.8 per 100,000 inhabitants in 1960 to 4.5 per 100,000 inhabitants in 2001. This rate did not change much duringthe 1990s. Incontrast, since 1970, the infant mortality rate has fallen by 70 percent inEcuador, which i s as impressive as the achievements inthe rest of the Americas, where the rate has fallen on average at a similar rate.g The infant mortality rate has followed an almost linear trend since 1950, reaching 33 per 1,000 live births during 1995-2000, down from 140 during 1950-55. Child mortality rates follow similar trends because they mainly reflect infant mortality since most child deaths are concentrated inthe first year of life (WHO 2003). 0 The drop in infant mortality coincides with a long-term decline in fertility rates. Fertility dropped from almost 7 per woman in the 1950s and 1960s to 2.8 during 2000-05. Duringthe 199Os, fertility dropped faster in rural than in urban areas, but the rate i s still 1.5 times higher for rural women. Fertility and infant mortality may be mutually dependent, because higher fertility raises the risk of early childbirth, whereas infant deaths may induce higher fertility due to a replacement effect. 9. Only achievements in Chile, Costa Rica, and Cuba have been more impressive, with reductions of over 80 percent. 44 3.14 Important concerns remain. First, the decline infertility and the increase inlife expectancy are changing Ecuador's demographic profile. Second, the epidemiological profile (causes of mortality) i s moving away from traditional child diseases (malnutrition, respiratory and infectious diseases) toward diseases associated with higher levels of economic well-being and urban lifestyle (cardiovascular and cancer health risks). Third, preventable diseases remain the main causes of child (1to 5) and infant (0 to 1) mortality. Fourth, the prevalence of AIDS has increased, with 10 times more cases reported in 2002 than in 1990. Fifth,malaria trends remain closely associatedwith the occurrence of the El Niiio effect. SubsidiesinBasic Infrastructure Services 3.15 The water and sanitation, electricity, and telecommunications sectors face similar problems. These are poor coverage (especially in rural areas), low efficiency and quality of services, irregular allocation of resources for new investment, and-with the exception of the electricity sector-incomplete regulatory and institutional frameworks. In general, the Government should encourage greater local, national, and international private sector participation, consolidating institutional and legal agreements, and-especially in the water supplyhanitation sector-transferring resources from the center to improve services and coverage. In addition, the individual sectors face some specific problems (Giugale, Fretes- Cibils, and L6pez-CAlix 2003): The water and sanitation sector is characterized by poor cost recovery through rates, and high dependence on transfers from the Central Government to cover deficits. There i s no integrated national system for managing water resources. Since all water and sanitation services are decentralized, providers depend on municipal governments, and the Central Government has limited i t s role to improving the quality and efficiency of services, and guaranteeing coverage to urban and rural populations not yet served. The electricity sector problems include heavy power losses and operational inefficiencies in companies, an incomplete sector reform process, institutional vulnerability, the questionable sustainability of the Electric Wholesale Market (MEM), the need for rate adjustments to ensure the sector's financial sustainability, incomplete implementation of the sectoral environmental policy, and the lack of a rural energy strategy. The telecommunications sector faces challenges including artificially low and unsustainable rates for fixed and residential telephone service; limited competition in the cellular market, resulting in user costs that are among the highest in the region; very limited Internet access; and an outdated institutional and legal framework not suitable to attract greater participationfrom the private sector in the capital of ANDINATEL and PACIFICTEL. 3.16 From the point of view of public expenditure, subsidies to basic infrastructure have little impact on the poor. Subsidies to electricity, water and sanitation, and telephone service represented about US$341 million in 2003, equivalent to 1.3 percent of GDP. They divert funds from other priorities without effectively reachingthe poor. They can hardly bejustified on economic grounds. 0 There are high losses in the power sector and a large portion i s due to theft and inappropriate billing-these can be considered implicit subsidies (Lecaros 2004). 0 The water sector has hadmassive shortfalls of resources for non-wage and maintenance expenditures. Subsidies to the sector, both from the Central Government and municipalities, have more than halved from US$146 million in 2001 to US$67 million in 2003. The Ministry of Urban Development and Housing's (MIDUVI's) transfers for municipality investments have been drastically reduced to less than one-tenth their amount in 2002-from US$52 million to US$5 million in 2002. Seventy percent of sector resources are concentrated in Quito and Guayaquil (Sotomayor 2004). The telecom sector has poorly targeted subsidies in a context of low operational surpluses for the telephone companies. First, by the nature of the service, poor and rural users are less likely to have a 45 fixed-line telephone connection: in 1999, telephone penetration was almost 10 times lower in the poorest quintile (Figure 3.2), and only one out of twenty rural households had telephone service in 1999. Second, targeted subsidies require substantial financial capacity, but the financial burden of the current subsidy scheme is very high. Third, many users that can afford to pay unsubsidized tariffs are being benefited because they reside in eligible areas. Thus, subsidies benefit all residential consumers, regardless of their income level, which has no rationale whatsoever (G6mez and Vidal 2004). 3.17 The next important question is, to Figure3.2 Telephone Penetrationby House IncomeQuintile what extent do the poor benefit from the subsidies? Without 701 reliable data, and with the 61.1 1 decentralized system of water, 60 which makes it virtually impossible s% o 5 50 to estimate the water subsidies, a ,$S $ Z $ 40 30 feasible approximation can be 20 based on household income 10 distribution from the 1990s (Table 0 3.4). 1 2 3 4 5 Incomequlntlles Source: ECV, 1998-99. Table 3.4 Basic ServicesSubsidiesby Expenditure Quintile, 2003 Consumption Expenditure Quintile Subsidy lSt2nd 3rd 4th sth Overall Source (Inmillions of U S dollars) Electricity' 12.8 18.2 24.9 35.6 60.5 152.1 Central Bank Water 5.3 8.1 10.2 15.9 28.0 67.5 WB staff estimate' Telephone 6.1 10.9 15.7 30.3 58.1 121.0 G6mez and Vidal (2004) TOTAL 24.2 37.2 50.8 81.8 146.6 340.6 (% distribution) Electricity 8.4 12.0 16.4 23.4 39.8 100.0 LSMS (1994) Water 7.9 12.0 15.2 23.6 41.3 100.0 LSMS (1994) Telephone 5.0 9.0 13.0 25.0 48.0 100.0 ECV (1998-99). TOTAL 7.3 11.2 15.0 23.9 42.6 100.0 Notes: 1. Dataobtainedby MEF andCONELEC for electricity (deficit turifario). 2. Excludessubsidiesfrom municipalities. 0 Overall, subsidies to basic services are not pro-poor: 7.2 percent accrue to the poorest households located inthe lowest income quintile, while 43 percent accrue to the richest households. This means that the poorest households are receiving about one sixth as much subsidy as the richest households. 0 Comparing these subsidies to those of the social sectors, their non-pro-poor level i s similar to the most unequally distributedone: those devoted to university tuition. 0 Among the three main public services, subsidies to telephone rates are the most regressive. About half of them benefit households in the richest income quintile. Again, in nominal terms, this means that the poorest households are receiving only about one-tenth as much subsidy than the richest households. 46 0 Electricity and water have similarly regressive distributions: the two richest quintiles "capture" about two-thirds of subsidies to both sectors. 3.18 Inequality is higher when one considers that water and sanitation services are highly subsidized for those householdsthat have a connection (the more affluent segments), while they are not subsidized for those with no connection. As a consequence, service to the poorest segments of the population end up being of higher cost and/or lower quality (Box 3.2). Box 3.2 Household Expenditureson Water: The Case of Machala, ElOro Machala is a city with 90 percent water coverage. Those not covered are the poorest segments o f the population. I t is possible to estimate the surprising differences in price and consumption for the poor, and the impact in terms o f his or her monthly expenditures on water and sanitation as a percentage o f family income. The table below shows that an average family served by a water connection consumes three times more than an average family without a water connection. Those households without a connection spend about 9 percent of their monthly income for a service o f lower quality, while those with access to the network spend less than 1 percent of their monthly income on water. Machala:MonthlyExpenditureonWater of Households Withand Without a Connectionto the Water Network User Concept WithWater Servedby Connection Tankers Estimatedaverage monthly consumptionof apoor family 15 4 to 5" (inm3) Estimatedaverage monthly payment (inUS$) 1.2Ob 29.00 Mbnthly expenditure inwater as apercentageof monthly 0.4 9.0 family income Notes: a. Estimatedby Yepes andGomez based on averageconsumptionof approximately 30 liters per person per day. b. Tariff depends on house characteristics. The indicatedvalue is basedonthe lowest tariff. Source: Yepes, Gomez, and Carvajal(2002); and Sotomayor(2004). C. How MuchFiscalSpace is Available for a Poverty Reduction Strategy? 3.19 Ecuador's fiscal space is declining. "Fiscal space" i s the sum of resources available to finance the current-both non-wage and non-pension-primary expenditure required by individual programs and investment projects associated with the PRS." Between 2001 and 2003, and as a percentage of GDP, public investment declined from 6.6 percent to 5.4 percent of GDP. This cut was almost equivalent to the one-third of the increase in the public payroll in the same period from 5.5 to 8.5 percent of GDP, which shows the fiscal space lost by accommodating wage increases. An important finding i s that interest payments on public debt opened new fiscal space, as they declined by almost 2 percent of GDP in the 10. This is an intermediate definition of fiscal space between two extremes. A stricter definition would restrict it to investment spending, but this would exclude goods and services required to implement it, e.g. on basic infrastructure maintenance (roads). Alternatively, an extended definition would include wage and benefits, especially outlays associated to human capital formation (like teachers' salaries and pensions). However, this would associate pension increases to pro-poor spending (which is not the case, even when restricted to the social sectors). 47 same period. These "savings" were mostly used to finance Central Government (CG) transfers to the Ecuadorian Social Security Institute (JESS)." 3.20 Identification of financing mechanisms is a major issue in the design of a PRS. In heavily indebted poor countries (HIPCs), for example, PRS financing has been provided by the debt relief obtained from debt reduction. In some middle-income (non-poor) countries developing a PRS, the little progress of their strategy has been associated with poor targeting and insufficient internal resources to meet its goals. Failure was due not to donor delays in financing, but to confusing priorities between the PRS and other "national" agendas, and low local implementation capacity. Hence, creating fiscal space, externally or domestically financed, i s not enough to finance a PRS. Consensus about and clarity of agreed PRS objectives, and the effectiveness and efficiency in the use of dedicated resources, are also extremely important. 3.21 At present, Ecuador has a small fiscal space to finance a PRS. For a country that has a budget envelope of 24 percent of GDP-in the top Latin America and the Caribbean (LAC) range-and public spending that can reach 30 percent when off-budget activities are included, finding an extremely small fiscal space may sound odd or suggest severe inefficiencies. Indeed, between 2001 and 2004, CG public investment declined from 4.2 percent to 2.2 percent of GDP, a cut of 2 percentage points in four years. As a result, the amount of resources that, in strict terms, could be reallocated to a PRS under the 2004 budget framework i s around 0.5 percent of GDP (US$145 million), that is, the flexible part of the budget devoted to investment (Table 4.1). This amount is not adequate for an effective strategy and the government needs to find additional resources. Hence, implementing a PRSrequires Ecuador to solve the paradox of having a relatively bigbudget coupled with a small fiscal space. Reversing Current Expenditure 3.22 Budget inertia has raised rigid expenditure to unsustainable levels. Within the 2004 Central Government budget-the Presupuesto General del Estado (PGE)-rigidity in the use of funds (besides earmarking) i s also substantial, and an already-tight fiscal situation i s becoming an unsustainable trend (Table 3.5). In the budget, about 83 percent i s rigid, composedof 32 percent for wages and salaries, 30 percent for debt service, and 21 percent for transfers (mainly to sectional governments, but also to the IESS) and the BDH. The "rigid share" of the budget i s driven by payroll, the share of which has increased by about 50 percent since 2001, and now represents almost one-third of total CG expenditure. CG transfers, especially to IESS, also show a similar increase in their "rigid share" during this period. Both trends have offset the savings obtained from debt service reduction inthe last four years. 3.23 Reversing the rigid share of the budget for current expenditure should bring the structural investment rate to 5.7 percent of GDP. In the last five years, the difference between the actual and structural primary surpluses has been an annual average of 0.3 percent of GDP (Annex A). In the same period, the average investment rate has been 5.9 percent of GDP. Subtracting the latter from the former gives a structural investment ratio of 5.6 percent of GDP. This coefficient would also be consistent with an estimated structural fiscal surplus of 4.8 percent of GDP. As a result, and taking into account that the investment ratio is projected at 5.4 percent of GDP in 2004, structural reformefforts concentrated on the CG budget-aimed at reducing the wage payroll and/or the burden from pensions-should target potential savings of about 0.2 percent of GDP, in order to bring the investment ratio back to its structural level. The austerity decree of 2003 and the recent passage of the new civil service law are encouraging steps inthe right direction (Chapter 2) that should prove Ecuador being able to curb payroll expenses this time, despite unsuccessful attempts inthe past. 11. Transfers to municipalities from the Central Government, which could partly be considered as fiscal space- for most transfers should be invested-also declined inthe same period. 48 Figure3.3 RigidComparedto FlexiblePublicExpenditure,2004 3.24 Barely 2 percent of the total budget is actually non-rigid and 100 flexible (Table 3.5 and Figure 3.3). In the non-rigid component, equivalent to 80 17.3 percent of the budget, three aggravating factors are that: (a) an 60 additional 13 percent of the budget is inertial due to investment pre- 40 commitments (counterpart funds and others) andtemporary salariesbeing paid 20 through the "services" component'*; (b) the flexible share has rapidly decreased to minimum levels (4 percent of the U ' fi budget) inthe last four years; and (c) the Total Revenues Budget Non-rigid flexible investment share i s about half of Source: BCE,IMF, andWB. such amount (2.4 percent of the budget). This residual investment i s equivalent to about US$145 million (0.5 percent of GDP). Should this pattern continue, the flexible part of the budget devoted to public expenditures would disappear in 2007. Potential fiscal space in inertial investment is capped at 1.7 percent of GDP in 2004, but its materialization depends on the authorities' ability to shift resourcesto apro-poor PRS. Table 3.5 Rigidity of the Central Government Budget, 2001-04 Percentof Budget Percentof GDP 2001 2002 2003(p) 2004(e) 2001 2002 2003(p) 2004(e) RigidComponent 77.6 81.8 81.9 82.7 19.0 21.1 18.4 18.3 Debt Service 39.9 30.9 31.9 30.2 9.8 8.0 7.2 6.7 Amortization 21.7 17.8 18.2 16.2 5.3 4.6 4.1 3.6 Interest 18.2 13.1 13.7 14.0 4.5 3.4 3.1 3.1 Wages and Salaries 21.1 26.7 30.9 31.6 5.2 6.9 6.9 7.0 Transfers' 13.8 22.1 16.3 17.8 3.4 5.7 3.7 3.9 olw Municipalities 10.9 9.9 10.6 10.4 2.7 2.5 2.4 2.3 olw Social Security 0.7 2.5 2.8 3.6 0.2 0.6 0.6 0.8 BonoSolidario 2.9 2.2 2.8 3.1 0.7 0.6 0.6 0.7 Non-RigidComponent 22.4 18.2 18.1 17.3 5.5 4.7 4.1 3.8 Inertiul Expenditures 8.7 12.2 13.2 13.3 2.1 3.1 3.0 2.9 Services 2.3 4.9 5.2 5.8 0.6 1.3 1.2 1.3 Investment 6.5 7.3 8.0 7.5 1.6 1.9 1.8 1.7 External credit (predestinado) 4.7 4.4 4.3 3.8 1.2 1.1 1.o 0.8 Domestic credit (BancoEstado) 1.2 1.3 1.o 0.7 0.3 0.3 0.2 0.2 Autogestion 0.5 1.1 1.1 1.8 0.1 0.3 0.2 0.4 Grantdcounterpart na 0.4 1.6 1.2 na 0.1 0.4 0.3 Flexible Expenditures 13.6 6.0 4.9 4.0 3.3 1.6 1.1 0.9 Goods 0.1 0.2 0.2 0.2 0.0 0.0 0.0 0.1 Other Current Spending 3.0 3.4 1.5 1.4 0.7 0.9 0.3 0.3 Investment* 10.5 2.4 3.2 2.4 2.6 0.6 0.7 0.5 Total Expenditures Amortizations 100.0 + 100.0 100.0 100.0 24.5 25.8 22.4 22.2 Notes: p =Preliminary;e =Estimate. 1. In2002, there was a US$467 million transfer to water, sewer, andelectriccompanies. 2. In2004, allocatedto other (urrastre) projects andother commitmentswith extemalprojects. Sources: BCE, IMF, andWB. 12. If we assume that all investment financed by external credit and grants (including counterpart funds) is inertial, butaddressing PRSneeds, their combined amount still remains small: 1.1percent of GDP in2004. 49 Making an OptimalSelection of Public Investment 3.25 Institutional constraints that affect public investment planning further restrain the small fiscal space remaining. Under the new administration, the Executive has made noteworthy improvements in public investment planning: projects are now validated at the Ministry of Economy and Finance (MEF), a public investment central database (Banco de Proyectos) has been created, and limited monitoring of selected projects has been introduced. However, despite these improvements, authorities recognize that there are still misallocations of the fiscal space. Duringinterviews, authorities conjectured that about half (1.1percent of GDP) of the non-rigid investment budget goes to non-priority projects, and that at least one-fifth of it (0.2 percent of GDP) couldbe reoriented. Inaddition: MEF lacks tools to plan and monitor effectively all public investment and to overcome several institutional shortcomings (WB/IADB 2004). There i s no national system of public investment. There are, instead, two public entities with duplicative and uncoordinated roles at the planning stage: ODEPLAN inthe Office of Planning inthe Presidency, and the Secretariat of Public Investment at MEF. No coordination mechanisms exist between ODEPLANand the Secretariat of Public Investment and INECI, at the Ministryof ForeignAffairs, which prevents effective coordination with external grants. N o information system, or physical and financial indicators, are available for past or current monitoring of public investment projects, and the time series are not consolidated. There i s no unifiedor consistent methodology for formulating public investment projects. Congress approves the investment envelope by sectors and expenditure item, but does not assign the allocations project per project, which limits transparency and social accountability about approved projects. Projects executed by public enterprises and autonomous institutions are not included in the investment envelope of the annual budget. These are simply off budget. Assessing Defense Spending 3.26 Another potential source of intra-budget fiscal space may be defense spending. It is necessary to examine Ecuador's own military expenditures and see how these fit within budget priorities, but this analysis is not developed here. Indeed, whether a defense budget is high or low can only be determined relative to the perceived threats to national security and the government's goals and intentions, which i s not the purpose of this report. However, some international comparisons are relevant. Can Ecuador afford to continue increasing military expenditures, especially now, during a time of severe fiscal adjustment, or do they contain a potential source of fiscal space inthe near term? This i s what a thorough analysis could further develop. 3.27 The number of members of the Ecuadorian armed forces is increasing. According to 2003 budget numbers, Ecuador has roughly 57,000 armed forces personnel. This is equivalent to 0.5 percent of the population in 2000. This percentage is high compared to the LAC regional average of 0.3 percent, albeit average when compared to the 0.5 percent found in middle-income countries. However, from 2003 to 2004, the number of army members registered in the budget increased by 18 percent, or 10,000 new effectifs (Table 3.6). Police members also increased by about 3,600 during the same period. This represents a costly payroll increase for a country making important efforts to streamline its wage payroll. Collier and Hoeffler (2002) show that foreign aid helps countries to raise their military spending outside their income level. InEcuador's case, foreign assistance to combat drug traffickers from neighboring countries and to facilitate coca eradication efforts may have temporarily increased personnelnumbersand military expenditures. The problem i s that aid i s temporary, but current expenditure created associatedto it tends to persist over time. 50 Table 3.6. Ecuador: Personnel Involved inDefenseand Security (2002-04) 2002 2003 2004 National Police 26,008 33,303 36,907 Armed Forces 56,193 56,193 66,193 Source: CentralGovemment Budget, 2004; MEF. 3.28 The defense budget is also high by international standards. When comparing Ecuador to L A C and developed economies, inGDP terms, the country's military expenditures (above 3 percent) were almost double the L A C mean (1.6 percent), and higher than the developed economies mean (2.4 percent). The same result is found when military expenditures are assessed as a share of total government expenditure. Again, Ecuador significantly outspends its LAC neighbors and developed economies. Only Chile, inLAC, devotes a higher share of its budget to defense (Figure 3.4). Figure 3.4 Ecuador: Military Expenditure, 1998-2001 (inpercent of GDP) (inpercentof Government Expenditure) 10- Note:LAC mean not available for 2002 as a percent of GDP. Source: International Institute for Strategic Studies (IISS) Yearbook 1997-2003. 3.29 Bringing Ecuador's defense budget back to its historic level of spending would generate a potential fiscal space of about 1 percent of GDP, but this should be preceded by a thorough assessment. The rising level of military expenditure in a given country appears "strongly influenced by the levels chosen by its neighboring governments" (Collier and Hoeffler 2002). Ecuador confirms that pattern, closely following Colombia's recently increased defense spending. Colombia, Ecuador's northern neighbor, has faced internal conflict for over 50 years, but since 2000, it has deepened its campaign against drug traffickers and guerillas with foreign assistance. Ecuador has followed such a trend, raising its defense budget from an averageratio of close to 2 percent of GDP by the end-1990s to 3 percent in the early 2000s. Among its immediate neighbors, however, Peru shows a very contrasting picture. To finance its pro-poor agenda, President Toledo deliberately shifted its budgetary priorities away from defense spending, by reducing it by more than half a percentage point of GDP between 2000 and 2003 (World Bank and Interamerican Development Bank, 2003). Thus, Peru made a conscious decision to shift spending from defense to the social sectors and, not less important, to make military expenses fully transparent and information about them available to the public through its financial management system on the Internet. Peru's reform of military spending demands Ecuador's close attention. 51 D. How MuchOff-Budget FiscalSpace CanBeFoundfor a Poverty Reduction Strategy? Reducing Off-Budget Activities 3.30 Off-budget activities, especially budget earmarking are, in general, bad for budget flexibility. Although some tax or expenditure earmarking can benefit specific goals, provided they are adequately targeted and monitored, in general they have caused several shortcomings in Ecuador's budgetary policy (Reis 2003): 0 Fiscal policy becomes constrained, because the government's ability to shift resources or modify the budget i s significantly reduced. 0 In a process of a budget adjustment mandated by a fiscal rule, such as the one for Ecuador, rigid earmarking provoke the compression of public investment, the flexible variable inthe budget. 0 Perverse incentives develop, as the pressure for cutting expenditure brings additional requests for additional earmarking. 0 Prospects for a countercyclical fiscal policy become dim, because inertial cyclical spending leaves little room (and resources) for countering a recession. 3.31 There is also potential fiscal space arising from the elimination of off-budget activities (Table 4.5). There are two main types: pre-budget taxes, off-budget oil revenues, and off-budget subsidies also financed from oil revenues. Figure 3.5 Decomposition of Budget and Off-Budge Most pre-budget Financing Sources (percent) earmarking of taxes i s constitutionally man- External Domestic dated and mainly affect allocations to Pre-budget earmarked oil provincial and =.-.-.-.-.-.-.-.-revenue 59i sectional governments, and universities. There i s very little that Oil subsidies 5 % can be done about them, short of amending the be-budget earmarked tax revenue Constitution. In 2004, Revenues 8% they account for about 52% 2.4 percent of GDP (8.2 percent of the Source: Budget Office, MEF, and Table A25. total budget) (Figure 3.5). 0 Off-budget earmarked oil revenues amount to 1.6 percent of GDP (5.5 percent of the budget) in 2004. These include resources to the Fund for Stabilization, Investment, and Public Debt Reduction (FEIREP) (1 percent of GDP), which is well targeted, has a well-defined purpose, and is adequately monitored. DiscountingFEIREP, their elimination would open a potential fiscal space of 0.6 percent of GDP. Regressive subsidies financed by PetroEcuador, e.g. (gas, diesel, and electricity) are an expenditure "quasi-earmarking." In2004, these subsidies represent 1.3 percent of GDP and could generate some fiscal space by reform, but this would depend on the proposals adopted (see Table 4.5). 52 Reducing Tax Expenditure 3.32 The fiscal space is also constrained by tax exemptions (also known as tax expenditures). Exemptions narrow the tax base, make tax administration more complex, and deviate resources needed to finance priority public expenditure. Authorities reduced the value-added tax (VAT) exemptions between 1999 and the first half of 2OOl,I3 but since then new exemptions were created or proposed on the VAT and the personal income tax (Schenone 2003). 3.33 Tax exemptions increase income inequity. They are often regressive and benefit the poor least. In1999, out of every 100sucresthe lRS couldnotcollect becauseof the exemptions oneducation, books, health, transportation, water, and electricity (in other words, except basic food items, house rental, and financial services), 43 sucres benefited the richest 25 percent of the population and only 14 sucres benefited the poorest 25 percent (Kopits and others 1999). Except for VAT exemptions-which could most likely be eliminated without changing incidence significantly-tax expenditures re-enforce Ecuador's income inequality. Two factors favor such distortions: the proliferation of taxes, and the concentration of high-income taxpayers in a small but well organized group that i s able to shift the tax burden to a large number of "small" taxpayers. In Ecuador, about 3,000 large companies generate 80 percent of revenue collected by the Sewicio de Rentas Zntemas (SRI). Table 3.7 Estimated Tax Expenditure of the InternalV A T Exemptions, 2001(millions of U.S. dollars) Sector Total Export of Taxable Income Exempted Tax Credit Fiscal Income Goods and Income Before Actually Estimated on cost Services Exempt Taxed Income Purchases - Agriculture 1,339 363 976 86.6 889.4 533.6 42.7 Commerce 7,162 507 6,655 5,148.6 1,506.4 903.8 72.3 Construction 523 1 522 467.8 54.2 32.5 2.6 Energyand gas 572 0 572 89.9 482.1 289.3 23.1 Energy& mining 571 431 140 65.3 74.7 44.8 3.6 Finance andinsur. 1,499 63 1,436 1,190.0 246.0 147.6 11.8 Industry 5,468 899 4,569 3,469.9 1,099.1 659.5 52.8 Others 25 1 24 13.0 11.0 6.7 0.5 Comm. services 3,089 909 2,180 1,982.4 197.6 118.6 9.5 Transp. &comm. 1,405 30 1,375 998.7 376.3 225.8 18.1 Total 21,654 3,204 18,450 13,512.4 4,936.6 2,962.2 237.0 Source: Schenone (2003) based on MEFdata. 3.34 Tax expenditures entail a significant revenue loss of about 3.8 percent of GDP for the Central Government. Although there are no reliable estimates of overall tax expenditure in Ecuador, broad calculations of tax expenditure on the internal VAT (that is, excluding the VAT on imports) indicate that in 2001, its cost was approximately US$237 million, or 1.1 percent of GDP (Table 3.7).14 Approximately half of this was related to the retail and manufacturing sectors. Then, tax exemptions on customs tariffs and the VAT on imports amounted to 1.4 percent of GDP. Finally, according to SRIdata, tax expenditure on the income tax i s estimated to have a fiscal impact of 1.3 percent of GDP (1percent 13. The personal income tax (ISR) exemption was eliminated for the financial sector on the income from securities and shares issuedby the government, for cooperatives, and provident societies (except for the ones established by farmers or officially recognized indigenous people), and for promoting development (directed primarily at tourism and industrial endeavors). Inaddition, the list o f items subject to the VAT was replaced with a tax list of VAT-exempt services (fundamentally, housing rentals and financial services)-which, therefore, leaves all other services subject to the VAT. 14. Based on a sample of 3,055 large taxpayers. An alternative estimate, using national accounts data, gives a combined result of V A T tax exemptions and evasions o f 1.9 percent of GDP (SALTO/AID 2003). This result i s not inconsistent with the figures presented above. 53 for firms and 0.3 percent of GDP for individuals). Hence, a 25 percent reduction of VAT and SRI tax exemptions, such as authorities indicated in their initial multiyear reform program (World Bank 2003b), would bringfiscal space for about 0.9 to 1.O percent of GDP. Making Transparent and Integrating Off-Budget PublicInvestment 3.35 There are other potential sources of fiscal space in the budget. Other off-budget or quasi- fiscal activities are those developed by (a) public enterprises grouped under the Solidarity Fund; (b) regional development agencies (Organismos Regionales de Desarrollo, ORDs); and (c) provincial and sectional governments.l5 0 The Solidarity Fund (Fond0 de Solidaridad) i s an autonomous body, created in 1995, to address human development through education, health and public services. It i s the sole holder of the 5 power generation companies, 1power transmission company, 19power distribution companies, and 3 telephone companies (ANDINATEL,PACIFICTEL, and TELECSA). Its pricingpolicy and financial balances and activities are subject to strong political interference, which leads to complex and non- transparent cross-subsidies, and important quasi-fiscal activities. Their audited financial balances are unknown, and this prevents making any estimates about the potential fiscal space that could be created through reform of their pricing policies. Worst of all, the Solidarity Fundmight be a source of contingent liabilities. Indeed, the IMFestimated that the stock of debt of the power distribution companies was US$571 million up to December 2002 (IMF2003). The minimumfiscal space that could be generated is 0.1percent of GDP. 0 Regional development agencies (ORDs) depend on the Executive, but receive off-budget earmarked revenues and manage their budget independently. Some of their projects are known by the Investment Office at MEF, but there i s no accountability and transparency about their pro-poor impact. In2003, RDBs managed a budget of US$246.8 million (equivalent to 0.9 percent of GDP) (Table A26). Making these investments fully transparent and/or evaluating their use prior to integrating them into the budget are reasonable options to consider while looking for additional fiscal space. 0 Provincial and sectional governments are autonomous, but receive earmarked transfers and also have their own sources of revenue. There is no estimate of their contingent liabilities. A significant step, however, has been the database of municipal finance that MEFhas recently developed. The system will be fully running in November, cover subnational fiscal accounts, and be opened to the public. Subnational transfers represented 2.4 percent of GDP in 2003 and, in theory, 70 percent should have been devoted to investments. Inpractice, MEF authorities believe that the actual investment share of transfers i s around 50 percent, and that the ensuing investment by subnational ("sectional") governments i s mostly disconnected from the priorities of the Central Government's expenditure agenda.16 Had municipalities agreed devoting at least one-third of their receipts to the priorities of a PRS agenda would generate about 0.4percent of GDP of additional fiscal space. 15. Other public entities, like the Central Bank, the Deposit Guarantee Agency, or the group of 22 enterprises associated to the army that are registered as SociedadesAndnimas, also develop quasi-fiscal activities, which are not always duly registered and transparent. 16. According to Pablo Lucio Paredes, a well-known Ecuadorian economist, "sectional" governments devote less than 10percent of their total transfers to education and health needs. 54 E. Toward a Results-Oriented Budget: Attaining the MillenniumDevelopment Goals 3.36 Ecuador has to accelerate progress to meet the Millennium Development Goals (MDGs). MDGs are integral to the Government's social strategy and goals (Table 3.8). Progress toward meeting them i s mixed. According to official statistics, the government has already achieved the gender ratio gap and, provided progress in primary education i s sustained, it i s in line to achieve universal enrollment in primary education. Similarly, on health indicators, the target for reducingby two-thirds mortality for girls aged 1to 5 has already been achieved, and a simple linear extrapolation suggests that the goal for infant mortality (aged 0 to 1)could be reached by 2010 for boys and 2008 for girls (Vos 2004). Table 3.8 Progress by Ecuador inMeeting the M lenniumDevelopmentGoals MillenniumDevelopmentGoalsTarget Ecuador'sPerformance 1. Reduce by half the share of population earning less than There are no reports on the share of population earning less than US$1a day (extreme poor), from 17.7 percent in 1998 to 8.9 US$l a day and on prevalent child malnutrition in recent years. percent in2015, andreduceby half the share of childrenunder The World Bank projectsareductionin the share of the population age 5 with prevalent malnutrition, from 38 percent in 1990 to living below the extreme poverty line from 28 percent in 1990to 19percent in2015. 22 percent in 2007 (World Bank 2004a). However, more recent estimates found that the share of the population living below the national poverty line increasedduring 1990-2001. There are no morerecent data available(World Bank 2004b). 2. Enroll all children in primary school by 2015 and increase Net primary enrolment was 90 percent in 2003. The share of male the share of male and female pupils starting grade 1 and and female pupils starting grade 1 and reachinggrade 5 increased reachinggrade 5 from, respectively,40 percent and 41 percent to, respectively,77 percent and79 percent in2001/02. in 1990/91to full completionin 2015. 3. Make progress toward gender equality and empowering Thefemale/male net primary enrolment has been met. The ratio of women by increasing the female-to-male enrolment ratio in female-to-male enrolments in primary and secondary school primary and secondary school of 97 percent in 1990/91. By increasedto 100percent in 2001/02. 2015, gender disparities at all levels of education should be eliminated. 4. Reduce infant (aged 0 to 1) mortality rate by two-thirds, Infant mortality rate per 1,000 live births declined to 25 in 2002, from 61.3 per 1,000 live births in 1990 to 20.4 in 2015, and and the childmortality rate per 1,000 live births declined to 29 in child mortality (aged 1 to 5) rate by two-thirds, from 57 per 2002. Alternative figures releasedby the Government of Ecuador 1,000 live births in 1990to 19in 2015. Increaseimmunization show infant and child mortality rates of, respectively, 11.5 and 34 against measles and DPT in children aged 12 to 23 months in 2003. Immunization against measles and DPT increased in from, respectively,46 percent and 51 percent of children aged 2002 to, respectively, 80 percent and 89 percent of children aged 12to 23 monthsin 1990to full coverage. 12 to 23 months. 5. Reduce matemal mortality rate per 100,000 live births by Data on matemal mortality are not very reliable and what can be three-quarters, from 150 in 1990 to 50 in 2015, and increase safely said is that this rate has fallen considerably in recent the rate of births attended by skilled health staff, from 66 decades. For instance, accordingto INEC, this rate dropped from percent in 1990to 100percent. 203 per 100,000 live births in 1971to 46 in 2002, but according tc ENDEMAIN, matemal mortality had an average of 302 pel 100,000 live births in 1981-87 and 159 in 1988-94. Births attendedby skilled health staff increasedto an average 69 percenl during 1995-2000. 6. Halt and reverse the spread of HIVIAIDS. Eliminate Prevalence of HIV increasedfrom 0.8 to 6.0 per 100,000 between incidence of malaria, tuberculosis, and measles from, 1990 and 2002. In 2003 the number of reported cases were respectively, 15 cases, 160 cases, and 15 cases in 1990to zero respectively797 (HIV: 497, AIDS: 300). Incidenceof tuberculosis by 2015. per 100,000 inhabitants was 52 in 2002. In 2002, there were 145 cases of measles, down from 779 in 2000. In 2002, there were 87,547 reported cases of malaria (that is 692 per 100,OOC inhabitants,down from 775 per 100,000 inhabitantsin2000. Sources: World Development Indicators (2003); PAHO (1994, 02); MOH. 55 3.37 Despite progress, unfavorable poverty trends and budget constraints raise questions about whether the goals will be reached. Social expenditure is declining, and its level i s low. Unless dedicated efforts are made, special population groups-like indigenous, Afro, and rural poor-could be left further behind. Indeed, the most recent data showed that the national poverty rate increased from 40 to 45 percent during 1990-2001 (Box 3.3), especially inthe urban areas. Box 3.3 Main Conclusionsof the World Bank Ecuador Poverty Assessment Poverty is increasing in Ecuador. Over the decade 1990-2001, the national poverty rate deteriorated from 40 to 45 percent, and the number of poor increased from 3.5 million to 5.2 million. According to the most recent survey and census, headcount poverty rates inurban areas, both inthe Sierra and the Costa, experienced a sharp increase of 100 percent and 80 percent, respectively, whereas rates inrural areas appeared stable over the decade. In contrast to 1990, in2001 the poor population appeared concentrated inurbanareas-the Sierra (20 percent) and the Coast (26 percent). As a result, the absolute number of poor people increased by 500 percent inthe urban Sierra and by 300percent inthe urbanCoast. Most surprisingly, growth inthe number of poor in urban areas far surpasses that of the total population in those areas, and this poses formidable challenges both interms of employment andinprovisionofbasic social services Patterns of domestic migration suggest that over 30 percent of the population live in a place different from where they were born. About one-third o f all internal migration occurs within provinces, and two- thirds occurs across provinces. Most migratory movements (60 percent) are urban-to-urban, rather than rural- to-urban, and Quito and Guayaquil alone are the destination of 20 percent of all internal migrants, with 13 percent and 16 percent, respectively. Poverty increased significantly in 44 out of 220 cantons over the decade. The largest increases (15 to 25 percentage points) occurred in cantons located inthe provinces of Azuay, Bolivar, Cotopaxi, Guayas, Manabi, and Pichincha. Individuals heading poor householdsare more likely to be self-employed and less likely to be public salaried workers than heads of non-poor households. Inparticular, 40 percent of poor heads of households are self-employed, compared to 30percent of non-poor ones. Similarly, only 4.5 percent of poor heads o f households are public sector servants. 3.38 The implications for achieving the MDGs are consistent with the general principles that should guide expenditure in the social sectors and the budget as a whole. The standardjustification for public spending depends on how well the expenditure line item compares with other sectors and programs on the following three grounds: efficiency, ability to implement, and equity. While equity issues have already been addressed, this section examines efficiency and ability to implement. To do so, . an input-output model i s used in a two-step method innovated by Vos and Ponce (2004). The model allows estimating financing requirements for meeting three MDGs-primary and secondary enrolment and infant mortality-and comparing them to financing requirements for other selected goals estimated under altemative approaches (UNDPRJNICEF 2003). Only models for primary enrolment and infant mortality are explained below. For a full explanation of the model applied to secondary enrolment, see Vos and Ponce (2004). 3.39 An innovative input-output method is introduced to estimate the financial inputs required to achieve specific MDGs. It aims to (i) define a production function for specific outputs, (ii) theisolate main determinants of such output; (iii) estimate cost per unit of producing such output, and, based on estimated elasticities and unit costs ensuing from above, (iv) calculate the financing needs of reaching a key MDG. This is similar to the World Bank model known as SimSip (see World Bank, 2004b), but its main difference i s that its estimated coefficients are Ecuador-based, while those assumed in SimSip result from L A C averages. 56 3.40 Meeting universal primary education. Estimation proceeds in two steps. First a model of the determinants of schooling i s estimated. Second, unit costs are assessed. This allows the identification of the various determinants of schooling, and their unit cost, making it possible to establish input-output relationships betweenpolicy interventionsand expected educational outcomes measuredby net enrolment rates. First step. The schooling model establishes the relative importance of each possible determinant of enrolment in primary education. It finds an elasticity expressing the impact of a 1percent change of a given determinant on school enrolment. Then, these determinants are linkedto unit costs to obtain a basis for making budget projections for alternative resource allocations. Complete details about elasticity results of the determinants of net enrolment and coefficients of cost-effectiveness in primary education are found inTable A42. Secondstep. Three static cost-effectiveness simulations are performed for the main determinants chosen, that is, inputs change to approach a given target (say, 100 percent net enrolment in primary education) usingmodel elasticities and unit costs to estimate the costs of three selected interventions: (a) an increase in the shares of trained teachers to 100 percent, (b) an expanded Beca Escolar (now tranformed into BDH), and (c) a combination of the two previous policies and an increase of rural school infrastructure." However, there are two basic assumptions: (a) the pupil-teacher ratio i s kept fixed during the simulation period; (b) the baseline (no policy change) scenario keeps all inputs constant (although the number of teachers may rise as a result of the fixed pupil-teacher ratio assumption), except unit costs which adjust for projected inflation only. As a result, the primary education budget remains at 1.3 percent of GDP, and the budget for secondary education remains at 1.2 percent of GDP, while net enrolment rates do not improve from their 2003 levels inthe baseline scenario. 3.41 Following are results from the three budget scenarios (Table 3.9): 0 The share of trained teachers increases to 100 percent (up from 90 percent in 2003) and the share of teachers with central appointments decreases to 84 percent (down from 94 percent) by 2007. This scenario works to get all urban non-poor children in school by 2007, and induces the urban poor to a net enrolment rate of 97 percent (up from 89 percent in 2003). This policy should also help increase enrolment among the rural poor, but without additional investment in rural school infrastructure the effect will be more than offset by the ensuing increase in the number of pupils per classroom. The budget implications of this policy are minimal. All other things being equal, it would require an increase of about 7 percent in the (nominal) education budget by 2007, or about US$30 million per year compared to a baseline of no policy change (0.1percent of GDP)." 0 The Beca Escolar (Bono)program i s expanded to cover all urban poor. School subsidies (or reduced schooling costs) appear to significantly influence enrolment for this population group only. This involves additional budgetary costs, not just from the increase in coverage of the cash transfer program, but also due to rising teacher costs as enrolment increases and the pupil-teacher ratio i s kept fixed. Having all urban poor in school by 2007 would involve a slightly smaller additional budget effort of 6 percent over the baseline projection by 2007, or about US$28 million (0.1percent of GDP) per year. 0 Combining the two policies and allowing for an increase inrural school infrastructure by a sufficient amount such that the rise in the number of pupils per classroom does not have a negative effect on 17. Interaction effects are included. For instance, an expansion inthe conditional cash transfer program would lead to higher school enrolment, but also to a higher number of students per classroom, which has a negative effect on enrolment for some groups o f the school-age population unless more i s invested in classrooms. 18. This policy can be as effective, and even cost saving, ifthe assumption of a fixed pupil-teacher ratio o f 23 i s dropped and the ratio is allowed to increase to 25, implying a 3 percent reduction in the number o f teachers (or about 3,700 teachers). 57 enrolment of the rural poor. This combination of policies leads to universal access to education for the urban population, but leaves the rural population without any visible benefit. The cost would be only marginally higher than for the second budget scenario (reaching 0.2 percent of GDP by 2006), mainly due to the extra investment in schooling infrastructure and the rise in demand for teachers as enrollment increase^.'^ Table 3.9 EstimatedFiscal Cost of Attaining Key MDGsand Social Targets (inpercent of GDP)20 Primary Secondary Basic Infant Child Child Total Year Education Education Health Mortality Nutrition* Care 2004 0.1 0.2 0.3 0.02 0.1 0.1 0.82 2005 0.1 0.3 0.3 0.02 0.1 0.2 1.02 2006 0.2 0.4 0.3 0.02 0.1 0.3 1.32 2007 0.2 0.6 0.5 0.02 0.1 0.5 1.92 Notes: * includes only the free school meals (colacidn escolar and others by the Programa deAlimentacidn Escolar). Sources:Vos and others (2004) for primary and secondary education, and infant mortality; and UNDP (2003) for others. 3.42 Meeting the Infant Mortality Target. The same methodology is applied: an input-output relationship in health is established, that is, between policy interventions and expected health outcomes measured by infant mortality (Vos et al. 2004). First step. The relative importance may be expressed as the elasticity measuring the impact of a 1 percent change of a given determinant on, respectively, the probability of professionally assisted child delivery and child survival. This i s linked to unit costs, to make budget projections for alternative resource allocations. Elasticity results of the determinants of infant mortality and cost-effectiveness coefficients for infant mortality are found inTable A54a. Second step. Three static simulations are considered: First, modifying the impact on access to health services. This specification is for the probability of professionally assisted childbirth and prenatal controls. The key policy variables are access to health insurance and availability of health services and medical personnel. In addition, we assume that expansion of the free maternity program will increase professionally assisted childbirthcommensurately.21 Second, results of policy simulations on the demand for maternal care (at child delivery and prenatal controls) are subsequently usedinthe health "production function" for child survival. A survival model further suggests that breastfeeding has a strong positive effect on avoiding early child death. According to the model findings, this leaves the coverage of the immunization program as the policy variable with the most important expected effect on reduction of infant mortality. Ethnicity and female education variables are considered the relevant household variables in determining infant mortality. Third, unit cost estimates are added for relevant public health input 19. This policy package could be financed by allowing for a gradual increase in the pupil-teacher ratio to 27.5 by 2007, but requiring a reduction of 9 percent inthe number of teachers (that is, affecting about 11,000 teachers). 20. Estimates for secondary education follow the same two-step procedure indicated above for primary education. For a complete description o f the exercise, see Vos and Ponce (2004). The input variables that require financing in this case are the number of teachers with a university degree (for urban enrolment), the coverage of students with Beca Escolar (Bono) in secondary education (the one with expected major impact on rural poor students) and, to a minor extent, some increase ininfrastructure (especially urban). 21.. Household determinants include the educational level of the mother and per capita household consumption. The averages of these two variables are assumed to change at fixed rates of 1.5 percent per year, thus imposing a trend inrising access to health services due to improving socioeconomic conditions. We assume further that use o f prenatal controls has similar determinants and impact as the probability o f medically assisted child delivery. 58 variables, such as salaries of health workers, construction and maintenance of hospitals and health centers, and of the special programs, specifically the Immunization and Free Maternity Programs. Key assumptions are that the nominal health budget (for all items) i s adjusted for a given inflation rate (3 percent per year) to account for changes in input costs, and nominal salaries of health workers are adjusted this way, and thus real wages for doctors andnurses are kept constant. 3.43 Results from the three budget scenarios target an overall reduction of infant mortality from 34 to 19 per 1,000 live birthsbetween 2004 and 2015. For the poor, the reduction should be from 42 to 22 per 1,000 live births, and for the indigenous population reduction should be from 66 to 33. Baseline simulation projects improvements in education and per capita consumption forward to 2015 under the assumptions indicated above, but assumes that health programs show no further expansion from their coverage reached in 2003. The baseline then projects that without health input improvements, infant mortality would reach only 30.5 per 1,000 live birthsby 2015. e Neither reaching full coverage of the Immunization Programnor the expansion of the Free Maternity Programtargeted at the poor by themselves i s sufficient to reach the MDGtargets for infant mortality. The expansion of the Immunization Program would reduce infant mortality to 20.1 per 1,000 live births by 2015. Because this program is universal and initial coverage does not differ much across population groups, this policy would reduce infant mortality for all, but would not narrow differences between poor and non-poor or indigenousand non-indigenous. e A targeted expansion of the Free Maternity Program would narrow such gaps. Generating unrestricted access to medically assisted child delivery and maternal care for the poor and indigenous population would reduce the overall infant mortality rate to 28 per 1,000 live births. The higher probability of early child deaths among the poor compared to the non-poor would fall from 45 percent around 2000 to 21 percent in 2015, and the gap between the indigenous and non-indigenous population would be almost halved from 118percent to 67 percent. 0 The additional budgetary cost of the expansion of these uromams as simulated would be about the same for each: between US$3 million and US$4 million per year (or 0.01 percent of GDP). Combining the two would sum to an annual cost of US$7.2 million over the baseline budget (or 0.02 percent of GDP). The combination of these two policies would be sufficient to reach the MDG targets for the poor and indigenous population groups, but would still fall slightly short for the non- poor (Table 3.9). F. Policy Recommendations 3.44 In the context of fiscal adjustment, an efficient and effective allocation of increased resourcescould make substantial inroads into poverty. Findingadditional fiscal space i s not the only concern for Ecuadorian authorities. Current spending on the education and health sectors inEcuador have already led to positive educational and health outcomes and there i s room for significant improvements in equity. The value of targeted spending in selected programs of the education, health and social welfare sectors i s underscored by the opportunities foregone when budgetary resources are used for less-priority and non-pro-poor items. Thus, ensuing policy recommendations open additional room for redirecting social spending and subsidies to basic infrastructure. However, the emphasis on targeting and a more equitable distribution of resources should not preclude addressing the institutional aspects of poor performance. This i s why the next chapter deals with institutional shortcomings in social programs, very relevant in a context of resource constraints. Inthe meantime, here i s a set of proposed actions. 59 To Shift Public Expenditure Toward a Pro-Poor Focus 3.45 Developing a poverty reduction strategy is something that just cannot be done with a non- pro-poor budget. Therefore, reviewing the non-rigid part of the budget to increase its pro-poor focus i s also critical for a PRS. This implies: 9 Making sure that the additionalfiscal space created is allocated entirely to pro-poor spending, especially inthe education, health, and social protection sectors. 9 Revising the current allocations of social expenditure and projects. Figure 3.6 Chile Fondo Concursable The development of a Competition- Started in 2000 (combined with policy rule) Based Fund (Fondo Concursable) could be tantamount to a cooperative Eliminates traditional incrementalist practices "game" among uncooperative New and reformulated projects must apply every year "players" in a cash-rationed budget. Old projects compete with new projects Indeed, there are not many choices for Ecuador except to optimize the use of Current spending associated to projects i s also considered existing resources. In that context, a /- modified version of the Fondo I --. 04- Non-Rigid Concursable could deal, with at least, --. 6:. Component the non-rigid share of the budget whose size is about 1percent of GDP Budget (Figure 3.6 and Box 3.4).22 In this Rigid Component case, the Ecuadorian Fondo would deal with assigning- residual monies I I I I - according to some pre-established Source: 22. Some LAC countries, like Guatemala, also integrate investment by subnational governments into the Listado Nacional de Proyectos that accompany the budget proforma, but Ecuador is barely starting to simply register such projects. 23. Some caution should accompany this recommendation, since the Chilean Fund works because it has a technically competent budget office, within a Cabinet-level Expenditure Review Committee that initially prepares a medium-term financial programming (although not a detailed multi-year budget). Ecuador does not have a similar body. 60 The World Bank (1996) estimated that without proper compensation, the pure elimination of the subsidy would lead to a welfare loss (reduction in consumption) of more than 5.3 percent for the very poor. A detailed analysis of the alternatives for retargeting the cokking gas subsidy i s included in Annex E. k Implementing a new Household and Living Conditions Survey is essential. The last one was performed in 1999, before dollarization. This would allow an updated poverty map (crossed with the census) and benefit incidence analysis of public subsidies. It would also help analyze most recent poverty trends. Box 3.4 The Chilean Competition-Based (Concursable)Fundfor Public Programs Goals and Requirements. The Fund's goals are to provide better information on the budget-making process and improve the allocation of public resources to new, restructured, or substantially enlarged programs, thereby reducing the inertial nature o f the budget. Since 2001, the budget has been prepared using a procedure that differs in significant ways from the system applied inprevious years. This procedure has two main steps: 1. Ministries develop their budget proposal based on information from a budgetary framework linked to their inertial spending (determined by laws, medium- and long-term commitments, and so forth). This enables them to assign all residual resources to a Central Fund for government priorities (Fondo Concursable). This Fund examines all new, restructured, or substantially enlarged projects. 2. The initiatives submitted to the Fund must be presented in a standard format. This includes background to allow study on the need for and relevance of the initiative. Most o f the headings included in this form correspond to those used in the standard methodology of evaluation of government programs. This methodology is the basis for the analysis and selection of initiatives for financing. Focus and Methodology. The public institution prepares a Program (Project) Presentation file in a required format of the draft budget. The main features o f this format are: justification, target population and beneficiaries, budgetary cost and budget request, goal and purpose, description o f components, monitoring indicators, and means of verification. Institutional Framework and Actors Involved. Projects are presented as part o f the overall framework of annual budget preparation. The Finance Ministry, originally through its Budgets Office, i s the responsible authority for assessing them. Projects to be financed by the Fundare selected in meetings with the President of the Republic at which the overall budget for public spending is presented. The selected ones become part of the draft annual Budget, which must be debated and approved by the National Congress, which scrutinizes it. Beginning with the 2002 budget, the initiatives are presented to the Ministry of Planning (MIDEPLAN) for review and assessment of the basic prerequisites, thus becomingjointly involved with the Budget Office in the decision-making process of the draft budget at the Ministry of Finance. Monitoring. The selected projects included inthe budget are monitored annually, based on previously established performance indicators and a well-establishedproject evaluation methodology. To Re-Target PublicSubsidies 3.46 None of the three subsidies to basic infrastructure services caters to the poor, and they are highly distortive from an efficiency perspective. The implicit subsidy for telephone service i s the most unequally distributed, followed by the largest subsidy, for water. While tackling the telecom and electricity subsidy i s a priority for the Central Government, dealing with the water subsidy requires collaboration from subnational governments. k Electricity would be, intheory, the easier sector to tackle. Restructuring the tariff system to reduce the total subsidy to consumers below a maximum amount of electricity consumption should not have a major impact on the electricity companies, or on non-poor consumers. An alternative fo further lower the 15 percent actual subsidy to residential consumers below 300 Kilowatt-hours would be to modify the ceiling itself to target the truly poorest households. 61 The telecom sector should start by reducing the tariffs for public telephones, which are 10 times higher than tariffs for residential users; and by eliminating cross-subsidies through completion of the tariff rebalancing between domestic andinternationalrates approvedby CONATEL in2003. The water sector i s the more complex to handle, due to its decentralized management. Cross- subsidies should be linked to operational performance, and the amount of subsidy allocated per connection should be defined considering the size and income level of the population, to avoid existing disparities intransfers that actually benefit the largest and richer water companies. 3.47 While the bias in social expenditure is much less pronounced for pro-poor expenditure than insubsidiesfor basicservices,itis neverthelessnotdesirable. 9 Reductioninoverall university tuition subsidiescouldfinance access for poorer groups, for instance, to secondary education. 9 Retargeting of the subsidy of the cooking gas makes sense inthe context of an increaseto a well- targeted and fully financed Bono de Desarrollo Humano, or a revamped SELBEN. In the past, the BDH was in fact introduced to compensate the poor for the elimination of the subsidy on gas. Therefore, retargeting i s an appropriate policy substitute to the cooking gas subsidy, since welfare losses would be compensated through an increased BDH and its revised database of beneficiaries or revamped SELBEN, and fiscal savings would be generated from lower administrative costs and reducedleakages. 9 Pro-poor programs, likeprimaryeducationandschool breakfast, shouldbeapriority, buttheseefforts should be accompanied by rapid completion of the on-going re-targeting process in the school breakfast. Once this i s achieved, increasing its very small budget i s alsojustified. 9 As raisingthe budget of the Seguro Social Campesinofaces legal and institutionalconstraints inside IESS, other low-budget pro-poor programs providing similar services, like the Ley de Mutemidud > Grutuita should see an increased budget allocation. All pro-poor programs would highly benefit from a results-oriented, performance-based budgeting. This would imply defining standardmonitoring indicators and evaluation mechanisms, and promoting civil society participationthrough social accountability mechanisms. Table 3.10 Potential Sourcesof FiscalSpace and Estimated Annual Impact Measures Percent of GDP Intra-budget Curb capital spending ratio toward the "structural" level (paragraph 3.23) 0.2 Interest savings from debt repurchase (Table A76) 0.2 Reduce defense spending to end-1990s level (paragraph 3.29) 1.o Make optimal use of public investment (paragraph 3.25) 0.2 Off-Budget Reduce selected off-budget earmarking o f oil revenues (paragraph 3.31) 0.6 Incorporate oil-subsidies to budget (gas, diesel, and electricity) (paragraph 3.3 1) 2.3 Eliminate 25 percent of overall tax exemptions (paragraph 3.34) 1.o VAT 0.7 Internal 0.3 External 0.4 Income 0.3 Firms 0.2 Individuals 0.1 Rationalize spending of ORDs (paragraph 3.35) NA Integrate 10% of subnational spending with national priorities (paragraph 3.35) 0.4 Allocate resources from Solidarity Fundto the PRS (paragraph 3.35) 0.1 TOTAL 6.0 Source: World Bank staff calculations. 62 3.48 Ecuador has multiple medium-term sources of a potential fiscal space for pro-poor spending adding up close to 6 percent of GDP. Most possible sources of fiscal space for Ecuador are identified inTable 3.10. Identified sourcestax policy and tax administrationchanges (other than removal of selected tax exemptions, i.e. not tax reform i s considered. This provides authorities with several options to consider, both intra- and off-budget. Intra-budget sources represent about one-third and off- budget two-thirds. Due to the lack of information, contingent liabilities are not included, but the table does include most off-budget and some quasi-fiscal activities. It i s important to remember that the materialization of these sources is constrained by the annual quantitative ceilings set by the fiscal rule on primary spending, non-oil deficit, and public debt reduction. In other words, whatever solutions are adopted, they will have to be accommodated within the annual fiscal ceiling mandatedby the rule. Trade-offs betweenIn- and Off-Budget FiscalSpace 3.49 Intheprocessofassessingalternatives, itis importanttoconsider severalrulesofthumb: Curbing the increase in wage payroll-in GDP terms-is absolutely essential to prevent full elimination of the very small fiscal space remaining in the budget. There i s simply no choice. This could also imply considering expanding the coverage of the civil service law to the rest of uncovered public servants. Ditto for the growth intransfers to IESS. Off-budget activities, by definition, should gradually go into the budget. Perhaps easier to handle are those that depend on the Executive alone. A natural candidate i s oil-financed (and non-pro-poor) subsidies (cooking gas, diesel, and fuel oil). It i s important to realize that their introduction to the budget does not mean their full elimination, but rather their reorientation, and increased monitoring and transparency. The level of tax and expenditure earmarkings should not only go into the budget, but be reduced in real terms. Their elimination has been attempted several times in the past decades with little success. It has also been suggested to start by eliminating those that are non-constitutional. In reality, there might be two feasible alternatives: (a) freeze the level of the earmarking in real terms, which preserves its existence, but prepares beneficiaries for its gradual phase out; or (b) adopt a temporary reduction of a group of them, which allows obtaining fiscal space faster. Reis (2003) suggests that a 20 percent across-the-board suspension of all earmarkings, (including transfers to subnational governments), would generate annual savings of about US$400 million. Resources from such a temporary suspension could go to a trust fund (fideicomiso), following the best-practice experience of Brazil, and be closely scrutinized intheir useby social accounting mechanism^.^^ Defense expenditure may be evaluated, so as to identify possiblesourcesof fiscal savings. To Reach SelectedMDGs 3.50 Selected MDGs are within reach by Ecuador provided social expenditure is well targeted, effective, and financed by low-cost programs explicitly linked to specific outcomes. In the past decade, net primary education enrolment and infant mortality rates have come down steadily, despite low education and health expenditures, apparently poorly functioning education systems, a significant amount of non-pro-poor spending in both sectors, and a high incidence of malnutrition. Continued overall improvements in education, urbanization, fertility rates, and sanitary conditions explain these paradoxical trends. However, no linear extrapolationguarantees that these trends will continue to improve. Usingan input-output model, however, that identifies the main factors (inputs) that determine achieving key goals, 24. This mechanism has been extended three times and will last 14 years inBrazil, until2007. However, the initial earmarkings considered eligible for the Fund have been reduced by the exclusion o f those for subnational governments following the first extension. 63 there i s a way to target cost-effective interventions that may guarantee reaching at least three key MDGs: universal primary and secondary enrolment, and infant mortality. These four cost-effective factors are teacher training and the expansion of the Bono de Desarrolo Humano for primary and secondary education, and expansion of the coverage of the Immunization and Free Maternity Program for infant mortality. Notice, however, that significant institutional shortcomings can impede progress: highrotation of personnel, deviation of resources toward non-priority items, corruption and lack of transparency, and political clientelism in the allocation of spending remain. Some of these shortcomings have been identified inChapter 4. Linkingbudget support to performance requires: > Providing the additional budget needs, which are reasonably small, but focused on specific key programs. Financing the four cost-effective factors amounts to a combined additional budget representing about 0.3-0.4 percent of GDPper year for 2004-05, and 0.6-0.8 percent of GDP per year for 2006-07 (Table 3.9). Taking advantage of complementary estimates done by UNDPLJNICEF (2003) and usinga very different methodology, it i s also possible to approach the low cost of meeting a wider number of MDGs, also including those of child malnutrition, basic health, and child care. Their inclusion would raise total additional budget needs to about 0.8 percent of GDP in 2004, 1 > percent of GDP in 2005, and almost 1.9 percent of GDP in 2007. Defining a set of performance indicators that would allow monitoring progress. Performance indicators ought to be an integral part of the PRS and result from a consensus-building exercise. In the case of the main two goals examined, most frequently used indicators are: planned and actual expenditures in the associated programs, enrolment rates for boys and girls, number of teachers trained, cost of primary education, differences inunit costs, and immunization rates. Obviously, high data quality, participation by civil society, and transparency and accountability of executing agencies favor a good tracking of progress. An interesting conclusion of this approach i s that not all sector budgets need to be linked to performance indicators, but only those that are critical for achieving the goals of the PRS (including the MDGs).~~ 25. In the EuropeanUnion, perhaps the most advanced region in developing results-oriented budgeting, so far only 30 percentof theEuropeanCommission's budget i s linkedto performance indicators (World Bank 2004~). 64 Chapter 4 Performanceof PublicExpenditureManagement 4.1 This chapter examines public expenditure management (PEM) in Ecuador, introducing an innovative methodology for reviewing budgeting by social programs and subnational governments. It assesses the country's capacity to maintain discipline in overall fiscal balances, allocate resources to policy priorities, and channel resources to expenditure programs efficiently, effectively, and with transparency. Thus, this assessment completes the standard three-level PEM analysis (Table 4.1). As discussed in Chapter 2, Ecuador has made progress in regaining aggregate fiscal discipline in the early 2000s and in reverting unsustainable fiscal trends. As shown in Chapter 3, it has also reached positive outcomes in the social sectors, despite declining outlays, a sizable share of non pro-poor social spending and a very small and reducing fiscal space. In this Chapter, it i s showed that such progress is fragile, but for an additional reason: the multiple and severe institutional shortcomings surrounding the budget process. Under any scenario, no effective poverty reduction i s possible without institutional and structural reforms in the budget system. Hence, there are legitimate questions as to whether current PEM practices support fiscal discipline or makes it difficult to sustain, facilitates an optimal pro-poor allocation of resources or distributes them with little prioritization, and implements an efficient mix of resources in the social sectors or wastes them. Table 4.1 Basic Elements of Public Expenditure Management: The "Three-level Analysis" Aggregated FiscalDiscipline Budget totals should be the result o f explicit, enforced decisions, not merely accommodate inertial trends and spending demands. Aggregate ceilings on totals should be set before individual budget decisions are made, and these should be sustainable over the medium term. Allocationto Strategic Budget allocations should be based on government priorities and on effectiveness Priorities o f public programs. The budget system should shift resources from lesser to higher priorities and from less to more effective programs. Operational Efficiency Line agencies should produce goods and services at a cost that achieves ongoing efficiency gains and is competitive with market prices. Source: Schick (1998). 4.2 Poor performance of social expenditure in Ecuador is closely linked to PEM shortcomings. Recent reviews of international experience with poverty reduction strategies have concluded that inmany countries, the practice of PEM i s an obstacle to the achievement of poverty reduction objectives (Judge and Klugman 2003). Ecuador i s no exception. Failures in the budget process and institutional bottlenecks systematically lead to underexecution of social programs. These shortcomings result in underbudgeting or in long interruptions and delays in the channeling of budgeted resources. Perhaps the most important failures are unrealistic budget planning, deviations between budgets approved and executed-with a bias in favor of defense and security forces and against spending in the social sectors- a lack of effective interventions resulting from budget fragmentation through a myriad of overlapping social programs, the presence of significant off-budget funds, and delays in the actual transfer of resources, arising from cash rationing and poor execution capacity at the level of line agencies. This chapter explores all those issues indetail. 4.3 Developing an effective poverty reduction strategy for Ecuador requires, as a precondition, an overall reform of the budget processand, more broadly, of all levelsof PEM. A sound PEMi s the key policy instrumentthat articulates the country's fiscal ceilings and rules with, on one hand, priorities reflected in the budget and, on the other hand, improvements in public sector performance and service delivery. Hence, PEM reform requires an enhanced performance of the budgeting system, rapid 65 upgrading of its budget and financial management procedures, a complete overhaul of budgeting procedures by social agencies in charge of priority social programs and of provincial and sectional governments receiving transfers, transparent information access at all levels of government to allow results-orientedbudgeting inthe future, and, only when previous reforms have gainedground, a medium- term expenditureframework that would allow aligning expenditureinputswith expectedsocial outcomes. A. The PEMProcessand itsRecentPerformance 4.4 The PEM process in Ecuador takes place in a framework where the role of the Central Government is being diminished. In 2003, the Central Government directly accounted for about 60 percent of total spending, down from 64 percent in 1998 (Figure 4.1 and Table 4.2). Fiscal decentralization has meant that the relative share of the rest of the public sector (mainly the Social Security Institute of Ecuador (IESS), and municipal and provincial governments) has increased by 7 percentduringthis period. Public enterprises, on the other hand, have also had a declining trend: overall, their share has declined by 3 percent since 1998. Figure4.1 Changes inShareofNFPSExpenditures,by Government Tier 1998 2003 Riblii sector Central Government Central 60% Government 64% Source: BCE. Table 4.2 Allocation of Public Spendingby Tiers 2001 2002 2003 US$ Mln Percent US$ Mh Percent US$Mln Percent Non-FinancialPublic Sector 4,853 100.0 6,117 100.0 6,585 100.0 Central Government (less transfers) 3,206 66.1 3,773 61.7 3,952 60.0 Rest of Public Sector 1,28 1 26.4 1,992 32.6 2,289 34.8 olw SubnationalGovernment 706 14.5 982 16.1 1,102 16.7 olw IESS 252 5.2 546 8.9 656 10.0 olw Universities andTechnical Schools 171 3.5 222 3.6 305 4.6 Public Enterprises* 252 5.2 356 5.8 295 4.5 olw PetroEcuador 164 3.4 282 4.6 210 3.2 Note: * Onlyincludes nonoperatingand capital expenditures. Source: BCE, MEF,andIMF. 4.5 Since 2003, Ecuador has taken significant steps to improve its overall PEM. The passage of the Fiscal Law (FTSRL), in2002, has provided a significant impetus to this effect. Efforts include setting 66 quantitative rules for budget formulation,' initiating the groundwork for multiyear budgeting, and requiring subnational entities to submit monthly revenue and expenditure reports. Besides the Law, in recent years Ecuador has also been prudent in its budget formulation: despite high oil prices, it has adhered to using a conservative oil price estimate of US$18 per barrel, though average oil prices were above US$24 per barrel in the last two years. Excess oil revenues are feeding the Fundfor Stabilization, Investment, and Public Debt Reduction (FEIREP). In addition, it has also reinitiated the expansion of coverage of its integrated financial management system (SIGEF), with the goal of covering 90 percent of the Central Government by end-2004; built a single database for Central Government-financed public investment; developed an interbank payment system (IPS) at the Central Bank that records salary payments to public employees; started to develop a central registry database for all government employees at SIGEF; and developed CONTRATANET, an electronic public procurement system, on a pilot basis and initially as an informational system. 4.6 Despite these improvements, a standard overall assessment of the Ecuadorian PEMreveals significant shortcomings inall areas. Since the late 1990s, the InternationalMonetary Fund(IMF)and the World Bank have jointly developed a survey tool called the Country Assessment and Action Plan (AAP)questionnaire to evaluate budgetary management practices worldwide (World Bank 2003~).This survey involves a series of 16 indicators covering the various stages of budget management, including 7 on formulation, 4 on execution, 2 on reporting, 2 on auditing, and 1on public procurement. The survey was originally designed to help heavily indebted poor countries (HIPC) identify key budget management areas for improved execution and tracking of increased poverty reducing expenditures. Indeed, this assessment was a precondition for preparing HIPCs for the upcoming debt relief, which would increase fundingto be made available for poverty reduction. Likewise, Ecuador is also experiencing an increase in revenues associatedwith the new oil pipeline. Although 70 percent of these revenues are devoted to debt reduction, 20 percent are assigned to countercyclical policy and 10 percent to social investment, especially in education and health. Besides setting a baseline, applying this survey to Ecuador highlights areas for improving budget management. 4.7 The survey results show that Ecuador barely meets 1of the 16 global benchmarks (Table 4.3). According to IMFNB standards, if less than half (7 or fewer) of the benchmarks are met, the country's PEM system requires substantial upgrading. Ecuador's performance i s dismal across the different budget stages. On the formulation side, an outdated classification system and significant amounts of extrabudgetary resources limits the government's ability to plan, control, and direct expenditures. Inaddition, coverage of general government activities i s insufficient, and there i s a lack of a medium-term expenditure framework. An important caveat for the only satisfactory item-the overall ratio between the executed and approved budget--is that the execution performance by sector widely varies from 0.31 in housing, and 0.35 in social welfare to 1.45 in energy and 2.05 in administration, which reduces the budget reliability as a guide to resource allocation. On the execution side, the key reason for Ecuador's low scoring i s the persistence of significant domestic arrears and cash-rationing mechanisms arising from the lack of timely and centralized information managed by Treasury. This cash rationing does not really reflect the Government's true cash situation: While Treasury knows how much budgetary resources are transferred from the Cuentu Unicu to private accounts from public institutions, these funds are either kept in deposit at some 2,400 accounts at the Central Bank, or in the private bank accounts, becausethey are not necessarily immediately disbursed by these institutions. Hence, whereas Treasury i s short of funds and must raise money, in reality there are undisbursed funds lying in the 1. Inthis respect, three important measures were taken. First, a limit is set on the real growth of the non-financial portion of the budget at the lower o f either the previous year's growth rate or a growth cap o f 3.5 percent on real noninterest expenditure. Second, the debt-to-GDP ratio is to be brought down to below 40 percent of GDP, and henceforth capped at that amount, which means that fiscal surpluses are needed, as well as savings from FEIREP mostly devoted to repurchase public debt, which also prevents diversion from its resources to current expenditure. Third, the non-oil deficit i s to decrease by 0.2 percent of GDP per year. 67 institutional accounts. On the reporting side, the lack of prompt and virtual consolidation of the budget executed (the so-called devengado) by SIGEF, and shortfalls in auditing, mainly explain low scores. Despiterecent efforts to expand its coverage, SIGEF's decentralized system i s based on separate software databases, which produces a virtual disconnect among themselves, and problems in collecting reliable data from one institutional database to SIGEF central, a shortcoming that i s expected to be corrected by year's end. No legal requirement to audit the fiscal accounts exists, and there i s high-level political interference at the General Comptroller's Office (CGE) by political parties, which prevents substantive cases from being presented or sanctions being applied. Table 4.3 Review of Ecuador's Public Expenditure Management and Benchmarks InternationalBenchmark Ecuador (inparentheses) Rating Rationalefor Ecuador's Rating Formulation 1. (A) Fiscal reporting matches the IMF B Coverage of the general govemment budget is 78 percent of the definition of general govemment sector NFPS (and 90 percent of non-financial public enterprises). with coverage (by value) of at least 95 Consolidated reporting of the general government i s weak and percent, whether it is funded through the untimely, especially ex post. Ex ante reporting of activities by budget or not. subnational government data is available, but ex post reporting is available only since 2004, and does not include quasi-fiscal activities. 2. (A) Government activities are funded C There i s no comprehensive and reliable information estimating through extrabudgetary resources, but extrabudgetary resources, but they are significant (above 10 these funds represent less than 3 percent percent). Earmarkedrevenuesnot includedinthe budget represent of total spending. 23 percent of total budget (correctedby off-budget financing) in 2004 alone. The budgets from the social security agencies for the army and the police, public enterprises, and autonomous institutions are not reported and, since only recently, their financial accounts are expectedto be known. 3. (B) The level and composition of the B During 1995-03, deviations represented an average 6 percent budget outturn is quite close to the of under-budgeting. Although there was no dominant pattern original budget appropriations of under- or overexecution in those years, systematic (deviations between 5 and 15 percent) underexecution of public investment appears (and for at least two years. overexecutionof wages-see Table A57). 4. (A) Budget reports, full ex ante and B Ex post reports on budget execution are neither timely nor timely ex post-include data on external complete. Grants, especially in-kind, are partly incorporatedinto financing-extemalloans or grants. the budget, and with delays. 5. (B) Classification of budget expenditure C Budget classificationdoes not conform to intemational standards i s done on an administrative, economic, (ROSC 2003). There are no programmatic classifications. and detailedfunctional andprogrammatic Ecuador's economic and functional classifications do not fully basis. conformto the acceptedintemational standards. 6. (A) Identification of poverty reducing C Poverty reducing spending i s not tagged. The lack of a virtual expenditure is clear, through a virtual or poverty fund, or the establishment of a special tracking an actualpoverty fund. mechanism for pro-poor expenditures adds to the lack of a programmatic classification as a major obstacle to identify pro- poor spending. 7. (A) Multiyear sectoral expenditure C The FTRSL approved a multiyear program, but its proposed projections exist, and are integrated into implementationi s delayed to 2005. The lack of functional and the budget formulationcycle as indicative programclassificationcouldpreventthe adoptionof detailedinter- ceilings. andintrasectoralbenchmarks. Execution 8. (A) No stock of payment arrears (or very C The level of arrears has declined, but still i s about 5 percent. It few), with little accumulation of arrears manages a cash-rationing system, with priority social spending over the previousyear. protectedup to acertain extent. 9. (A) Intemalcontrolsystemis effective. C Intemal financial audits are weak in their breadth, depth, and frequency. 68 InternationalBenchmark Ecuador (inparentheses) Rating Rationale for Ecuador's Rating 10. (B) Public expenditure tracking surveys No PETS have been piloted or implemented. CGE does financial (PETS) o f -funds are piloted to but not performance audits, sometimes verifies whether resources supplement weak intemal control as a reach the final users or service providers, and does not publish its secondbest. audit findings andrecommendations. 11. (A) Satisfactory reconciliation of fiscal The Cuenta Unica is reconciled daily and through electronic and banking accounts is undertaken means, but institutional accounts can take longer than one month. monthly. Statistical discrepancies in monthly reporting are still significant. Reporting 12. (B) Intemal budget execution reports are Delays in receiving institutional reports o f budget execution from received between two and four weeks of spending units may last much longer than a month. Delays are the end o f the relevant period. particularly significant indecentralized execution units. 13. (A) Good-quality classification of Because there i s no virtual fund or an altemative permanent poverty reducing spending is reflected in tracking mechanism of poverty reducing expenditure, reporting i s regular in-year budget reports. made on an ad hoc basis, and its quality is poor, in terms o f both a function@ or programmatic classification basis. 14. (A) Routine transactions are entered into Transactions are entered within a six-month period. Full extension the main accounting system within two o f SIGEF is reducing these delays. months after the end o f the fiscal year. 15. (B) An audited record of the financial Audits are made ad hoc, and there is no legal requirement to audit outtum should be presented to the overall fiscal accounts, which significantly reduces accountability legislature between 6 and 12 months of for financial results. the end of the fiscal year. Procurement 16. (A) The public procurement system The procurement system has unclear rules and weak enforcement. promotes efficiency and effectiveness in There have been numerous cases of suspected corruption inpublic the use o f public resources, through clear procurement contracts. An interinstitutional commission i s rules that promote competition, preparing a new legal framework. CONTRATANET has been - . I I transparency, and value for money. used as an informational tool, but not yet as a transactional tool. Benchmarks met 1 Notes: Bold= Benchmark met. A = Good, B = Fair, and C =Poor. Source: ROSC and CFAA, and WB/IADB staff survey. 4.8 Inthe same vein, Ecuador PEMglobal rankingis low comparedto those of HIPCs.Ecuador fares worst on about half of all indicators, thus placing it in the group of less-advanced countries (Table 4.4). Alternatively, when compared to Peru, a strong Latin America and the Caribbean (LAC) performer in PEM, Ecuador fares worse in all but three indicators. This is in part due to Peru's usage of a well- functioning integratedfinancial management system supported by a sound regulatory framework. The gap has widened since the survey was taken, because Peru has made strides in participatory multiyear budgetingand PETS piloting. B. Budget Management Review inthe Central Government 4.9 Whereas the previous section provides valuable insights and benchmarking, this section highlightsbudget formulation and execution features that are relevant to understand key shortcomings in line agencies and subnational governments. More comprehensive and detailed analysis of budget management has been developed in parallel with the IMFReport of Observance of Standards and Codes (ROSC) (IMF 2003), and the WB/IADB Country Financial and Accountability Assessment (CFAA) (WB/IADB 2004). Annex B also presentsan overview of the budget cycle inEcuador. 69 Budget Formulation 4.10 There are several technical and institutional weaknesses that inhibit Ecuador from having an effective budget formulation. The main ones include (a) extensive off-budget activities, especially through earmarking of revenues and expenditures (see Chapter 3), which constraints the scope of the budget as a policy instrument; (b) incrementalism, whereby allocations are decided mostly as semiautomatic adjustments to the previous year's allocation and turnbudgeting into a very rigid exercise; (c) bad incentives to overestimate revenues in the initial budget, so as to "armor" it before facing Congress's proposed reallocations that often are far from best budgetary practices; (d) absence of a results-oriented budget, shown in the lack of physical and financial indicators and, more generally; (e) absence of a multiyear budgeting framework (MYBF). Table 4.4 Rankingof Ecuador's PEM inRelationto Peru, Bolivia,and HIPC Indicators Indicator Ecuador Peru Bolivia HIPCs DetailedHIPCs Distribution 2004 2001 2001 Average (24 countries, 2000/01) 2000/01 A B C Formulation 1. Good coverage of general government 8% 67% 25% 2. Fullreporting of extrabudgetary sources 54% 38% 8% 3. Reliable budget as programming tool 0% 67% 33% 4. Registered data on external financing 42% 54% 4% 5. Sound classification of transactions 17% 38% 46% 6. Tagging o f poverty reducing spending 46% 38% 17% 7. Integration multiyear & annual budget 17% 46% 38% Execution 8. Timely reporting of payment arrears B A B A 46% 33% 21% 9. Good quality o f internal control system C B C B 13% 71% 17% 10. Regular spending tracking surveys done C C C C 8% 46% 46% 11. Proper reconciliation of accounts B A B B 38% 38% 25% Reporting 12. Timely budget reporting C B C B 8% 71% 21% 13. Regular reporting of pro-poor spending C A C B 23% 54% 23% 14. Timely accounts recording and closure B A A B 25% 42% 33% 15. Timely audited accounts C A C C 0% 17% 83% Procurement 16. Efficient and effective procurement C na na na na na na Note: A = Good, B =Fair, and C =Poor. Source: World Bank/IADB (2003a); and response to WB/IMF survey. 4.11 A central problem in budget formulation is the sizable extent of off-budget resources, mainly earmarked revenues. When off-budget resources are significant like in Ecuador, they make the size and financial situation of the government opaque, complicate Treasury management, and reduce budget transparency (Table 4.5). Budget fragmentation also weakens the government's capacity to prioritize policies and expenditures. Froma total of US$3.9 billion projected to be collected intaxes in2004, about US$705 million, or 18 percent) are "preassigned" to different entities including local governments, the (Fondo de Desarrollo Seccional) (FODESEC), universities, the Fondo Salvamento Cultural, the Comisidn de Transit0 de Guayas, the Centro de Rehabilitacidn de Manabi, water companies, the Colporacidn Aduanera Ecuatoriana (CAE), and the Sewicio de Rentas Znternas (SRI). 70 Off-budget resources from oil revenues are also sizable. From a total of US$2.1 billion projected to be collected in oil revenues, about US$855.5 million (or 40 percent) are "preassigned" to different entities-FEIREP, sectional governments, the army, and others. As part of oil revenues, PetroEcuador's off-budget resources are remarkable as well. While about 57 percent of Petroecuador's net oil surplus (revenues-expenditures) of US$2.7 billion is assigned to budgetary operations, the remaining 43 percent (US$1.2 billion) is assigned to off-budget activities, including about one-third (US$387 million) going to subsidies (cooking gas, power companies, and diesel), and the remainder to FEIREPand other earmarkings to institutions and regions. An overall total of about one-fifth of the total public resources (including off-budget activities) beyond the government's control-and earmarked (preasignado) resources-represents a major limitation on the government's ability to shift budgetary resources to pro-poor policy priorities, especially if Ecuador i s committing to a poverty reduction strategy.2 Table 4.5 Sources of Government Financing, 2004 (In US Mlns,Percent of Total, Percent of GDP) 2004 Percent Percent of Proforma Financing GDP Budgetary Sources 7022.6 81.8 24.2 Fiscal revenues 4501.3 52.4 15.5 Tax 3243.6 37.8 11.2 Petrol 1257.7 14.7 4.3 Autogestion" 391.3 4.6 1.4 External credit 584.0 6.8 2.0 Domestic credit 1401.5 16.3 4.8 Other 144.4 1.7 0.5 Local counterpart 99.6 1.2 0.3 Nonreimbursable (grants and technical assistance) 44.8 0.5 0.2 Off-Budget Sources 1560.9 18.2 5.4 Pre-budget earmarked non-oil tax revenues 705.4 8.2 2.4 Off-budget oil revenues 855.5 10.0 3.O Pre-budget earmarked 468.5 5.5 1.6 o/w FEIREP 292.1 3.4 1.o Subsidies (gas, diesel, electricity) 387.0 4.5 1.3 Total 8583.5 100.0 29.6 Total tax revenues 3949.0 46.0 13.6 Total oil surplus revenues 2113.3 24.6 7.3 Note: *Self-generated revenues from fees, and other sources. Source: Budget Office, MEFand Table A25. 4.12 Budget incrementalism has raised inertial expenditure to unsustainable levels. Within the Central Government's budget-the Presupuesto General del Estado (PGE)-rigidity inthe use of funds (besides earmarking) i s also substantial, making an already tight fiscal situation run an unsustainable trend that i s exhausting an already small fiscal space. Overall, in the total budget, about 83 percent i s rigid, comprised of 32 percent for wages, 30 percent for debt service, and 18 percent for transfers (mainly for municipalities), and 3 percent for the Bono de Desarrollo Humano (see Chapter 111). 4.13 Some overestimated budget assumptions prevail, although deviations have been declining. These overestimations appear as contingencies to offset Congress's proposed modifications inthe budget 2. An aggravating factor i s that for earmarked taxes, increasingrevenue collections automatically triggers additional spending resources to the earmarked entity, thereby diminishing added revenue to the Central Government from improved tax collection. 71 approval process. Congress has no capacity to modify the ceiling on total expenditure set by the proforma proposedby the Executive. However, it can modify its sector allocations arbitrarily, often with few technical criteria. To offset this, the Executive i s tempted to overestimate its budgetary assumptions (Table 4.6), although deviations have been declining. This has not prevented Congress from approving a few budgetary malpractices. For instance, inthe debate on the approval of the 2004 proforma, Congress arbitrarily reduced the amount of interest payments to make room for additional increases in other categories of current spending. This decision placedEcuador in potential debt default, a very dangerous move for a country slowly rebuilding credibility with domestic economic agents, and its reputation in international markets. This move forced the Executive to readjust current spending in other areas, but ex post, only after Congress approved it. Table 4.6 Budgetary Assumptions and Actual Values 2002 2003 2004 Assumption Actual Assumption Actual Assumption GDP (US$ mln) 19702 24311 27092 26913 29707 GDP growth 5.0% 3.3% 3.5-4% 2.6% 55.5% Inflation year end 8-10% 9.4% 6-8% 6.1% 4-5% Current revenues (US$ mln) 4239 4572 4503 4771 4429 Crude oil price 20.0 21.8 18.0 25.7 18.0 Petroleumproduction (Mlnbarrels) 157 143 152 152 192 Source: MEF, BCE, and IMF. Execution 4.14 The practice of using Emergency Decrees to modify the approved budget has also steadily increased. Between 1998 and 2003, their annual number rose from 3 to 9 (Figure 4.2). Emergency Decrees have two important implications: (a) they overcome cumbersome procurement procedures, while opening room for eventual corruption in Figure 4.2 Emergency Decrees direct contract allocations; and (b) they modify budget priorities while shifting 101 I resources from the original budget, further reducing the already small fiscal space for social investment. Certainly, among the recent list of emergency decrees, in most cases there i s a reasonable justification linked to a natural disaster, but this i s not always the case: in recent years, a few budget modifications not linked to natural disasters have also o~curred.~ Inthe absence of a contingency budget reserve, as i s typical in other LAC countries, these emergency situations should rather be considered as 1998 1999 2000 2001 2002 2003 eligible for the use of FEIREP resources, as Source: MEF. i s now mandated by the FTRSL. 4.15 Absence of a Results-Oriented Budget Framework. The budget has an extremely simple format, and does not contain any fiscal ratio (IMF/ROSC 2003), or any physical or financial indicator. Lack of a medium-term expenditure 3. For example, in 2002, an emergency decree was issued during a period of public unrest that led to road closures and work stoppages in the construction of the new oil pipeline. 72 framework that ties expenditure priorities to government policies also prevents the government from allocating resources in line with its long-term goals. This makes virtually impossible to shift resources between and within sectors in accordance with medium term goals, and plan capital investments that require multiyear commitments. Thus, decisions on budget trade-offs are not based on systematic analysis grounded in clearly articulated policy, and expenditures tend to be biased towards short duration projects or towards current spending. or under-spendingabove 15 percent. 2/ Includesthe Electoral Tribunal and the ConstitutionalCourt. 3/ Includestransfers to subnational governments. Source: Ministry of Finance. 4.16 Whereas the reliability of the overall budget has improved in recent years, its decomposition shows big winners and losers. In examining Ecuador's actual versus budgeted expenditures over the past seven years, several noteworthy features emerge (Table 4.7). First, the reliability of the budget has generally increased after dollarization, particularly in 2002, when actual expenditures accounted for 98 percent of budgeted expenditures. An average underspending of 6 percent during 2000-03 means a marked improvement compared to overspending of 12 percent of the budget during 1995-99. Second, although overspending dominates underspending over the whole period-5 out of 9 years-there i s no persistent trend. Third, there i s a considerable variance inthe undedoverspending by individual sectors, the overall trend showing fewer sectors with a greater than 15 percent deviation in 2003 compared to 1995 (10 compared to 16). Fourth, very disturbing, however, i s the constant underspending in the social sectors-particularly in social welfare, which during 1995-2002, on average 73 spent only half its budget. Other continuously underspending sectors include environment, labor, and tourism. Fifth, on the flip side, sectors with a general history of heavy overspending-some with an average of over 200 percent during 1995-2003-include administration, finance, energy, other general services and, particularly, defense. 4.17 There is heterogeneity among sectors as to the relationship between their weight in the overall allocation of budget resourcesand the degree to which they exhibit over- or under-spending (Table 4.8). For example, whereas executed spending in "Other General Services" averaged 300 percent higher than budgeted during 1995-2003, it accounted for only 0.2 percent of the executed budget, and therefore was not critical to the overall budget allocations. By contrast, overspending greatly impacted the allocation of budget resources in Defense and Interior, which accounted for 2.0 percent of the budget; inFinance, which accounted for 2.4 percent of the budget; and on Public Debt, which accounted for 8.1 percent. The latter case reflects a weak technical capacity to properly budget for what are foreseeable expenditures. Overspending in these sectors was mainly compensated by underexecution in social welfare (2.4 percent of the budget), Transport and Communications (1.1 percent), and Agriculture (1.1 percent), that is, a significant underspending in critical social and infrastructure sectors, which appear more vulnerable to fiscal cuts. Table 4.8 Changes inthe Budget Share between Executed and Approved 1995 1996 1997 1998 1999 2000 2001 2002 Average 95-02 Legislative 0.1 0.4 -0.1 -0.2 0.0 0.1 -0.1 0.0 0.0 Judiciary 0.0 0.9 0.0 0.0 -0.3 -0.1 0.1 0.2 0.1 Administration 1.4 3.4 0.8 0.2 -0.5 0.5 0.1 0.3 0.8 Planning -0.2 -0.1 -0.2 -0.1 .. -0.1 Environment 0.0 0.0 0.0 0.0 -0.2 -0.1 0.1 -0.1 0.0 Interior 1.8 6.8 1.7 -0.4 -1.4 0.4 -0.9 0.1 1.o Defense 0.2 5.7 1.5 1.4 -2.4 2.3 0.0 -0.8 1.o External Affairs -0.1 0.4 -0.1 0.0 0.5 0.3 .0.1 -1.2 0.0 Finance 0.6 2.5 0.3 3.9 -1.o 12.2 0.4 0.5 2.4 Education -3.1 8.9 -2.1 -1.1 -1.o 0.4 -0.3 1.9 0.5 Social Welfare -2.2 0.1 -1.0 -2.7 -5.0 -4.0 -1.5 -2.5 -2.4 Labor -0.2 -0.6 -0.1 0.0 -0.1 0.0 0.0 0.0 -0.1 Health -1.o 2.1 0.1 -0.8 -1.1 0.2 0.4 -0.8 -0.1 Agriculture -3.2 1.8 -1.7 -2.5 -1.5 -1.7 1.2 -1.3 -1.1 Power and Natural Resources 0.0 -0.5 1.3 3.7 0.4 0.0 0.1 0.1 0.7 Industry -0.1 0.1 0.0 0.0 -0.1 -0.1 0.0 0.0 0.0 Tourism -0.1 -0.1 0.0 0.0 -0.1 0.0 -0.1 0.0 0.0 Transport a n d Communications -4.0 2.4 0.9 -1.o 0.1 -1.3 -2.9 -2.5 -1.1 Housing and Urban Development -0.3 -0.6 -1.4 0.1 -1.3 0.1 1.4 -2.2 -0.5 Others General Services 1/ 0.5 0.8 0.3 0.0 0.0 0.1 0.0 0.0 0.2 P u b l i c Debt 11.0 37.3 1.3 -1.6 14.5 -7.7 2.2 8.1 8.1 Others 2/ -1.4 0.1 -0.9 1.2 -3.3 -1.6 .. -1.o Memo: Social Sectors -8.3 11.1 -3.0 -4.6 -7.9 -3.4 -1.4 -1.4 -2.1 Note: Shares are estimated as a percentage of the total budget. Changes are the difference between the executed and approved shares per sectors. A negative amount reflects executed share that i s lower than the approved share. Source: Ministry of Finance 4.18 Budget cuts were made with inadequate information. A shortcoming of the current SIGEF i s the delayed recording of expenditures made by public institutions, making it very difficult for the Ministry of Economy and Finance (MEF) to consolidate executed expenditures with accuracy. In this way, while MEF i s aware of the amount it transfers to institutions, reconciling its transferred amounts with actual spending can be delayed by several months, which in turn impacts the decision on how much Treasury should over- or under-assign to the executing line agency. 74 4.19 Cash rationing is declining, but arrears still represent a recurrent problem in budget Figure 4.3 Stock of Arrears management. Another direct consequence of poor (US$ Mn) budget execution management has been the 35% occurrence of arrears. In Ecuador, these are '*O0 30% particularly problematic, and the large size of 1000 25% arrears carried into the next fiscal calendar year 800 20% weakens the execution of that year's budget allocation. Figure 4.3 illustrates the stock and size 6oo 15% of the arrears relative to expenditures in the last 400 10% four years. Less than 5 percent is the international 200 5% benchmark; however, Ecuador substantially 0% exceeded that amount in the previous three years. Problems also arise in defining arrears, which the Arrears, LHS +% of NFPS expenditures government classifies as the difference between budgetary transfer requests and actual budgetary Source: MEF. transfer payments, though the former may not reflect actual expenditures, thereby leading to a potential overestimation of arrears. Nor does the budgetary accounting system used provide the Central Government with timely information on the size and composition of the arrears. Accumulating arrears impacts the government's credibility and lead to disruptions inthe delivery of goods and services. C. BudgetManagementReview inSocial Programs and Subnational Governments 4.20 This section introduces an innovative approach to review budgeting-especially at the execution stage-in two key segments of Ecuador public sector: social programs and subnational governments. The Central Government has persistently under-executed both social programs and transfers to provincial and municipal governments. For instance, the combined cutbacks on approved expenditures for education, health, and social welfare were 16 percent during 2000-02, from an average of only 5 percent during 1995-1999. This section reviews the main reasons for this deteriorating budget underperformance within social sectors, particularly within those priority social programs (PSPs) considered of utmost importance by the Government of Ecuador (GOE). The next section deals with subnational governments. Following an overview of the institutional arrangements underpinning the provision of social service, the budget process within social programs i s reviewed in detail, using a spending tracking methodology focused on time execution delays, rather than on resource flows. Taking into account the multiplicity of social programs, out of a sample of the 45 existing programs, a sample of five PSPs was selected according to size, targeting, and sources of financing. The sample includes: the Bono de Desarrollo Humano (BDH), Plan Ampliado de Znmunizaciones (PAI), Ley de Matemidad Gratuitay Atencidn a la Znfancia (LMG), Programa de Asistenciay Desarrollo Comunitario (PRADEC), and Programa deAlimentacidn Escolar (PAE). Results are revealing. 4.21 Work on reviewing budgeting procedures inline agenciesis inthe pioneer stage worldwide. It builds upon seminal work developed in LAC in the nineties, particularly on institutional reviews and staff tracking surveys for Honduras (The World Bank, 1997 and 2001) and, more recently, on public expenditure tracking surveys for Peru (The World Bank and IDB, 2003). The surveys inHonduras were able to identify the main constraints to personnel management and the relationship between trends in budgetary assignments and unobserved in health and education outcomes, especially because either resources never reached the programs in whose names they were consignedor inefficient service delivery arrangements prevailed. Those from Peru were able to identify shortcomings in the transfer of resources, and quantify leakages at each ladder of the chain of the delivery of resources until the final beneficiary. Instead, the Ecuador budgeting review focus on quantifying perceived delays in the transfer of resources 75 and institutional shortcomings at the level of social programs and subnational governments that explain, to a significant extent, observed budget underexecution. ' 4.22 The tracking methodology employed has some limitations and its limited focus has some pros and cons. On the pros, it identifies major bottlenecks in expenditure execution, while showing systemic problems (either at the level of Treasury or the line ministries) that are in need of reform. On the cons, the limited scope-five social programs-does not negate its contribution, but also raises further few questions that would require additional work. For instance: how did the multiplicity of programs come into being and did the donors have to do with it? what are the social ministries' capacities for budget management-especially at the Ministry of Social Welfare that shows a significant under- execution capacity? and do these programs have performance indicators, if any? So, it must be beard in mindthat to fully understand some of the key issuesthat are identified inthis review would have required a complementary study at the sectoral level that was beyondthe scope of this PER. Social Programs 4.23 The institutional organization of social programs is very simple, and involves increasing local participation. It consists of (a) the management structure-within the social ministry andor local agency administering it; (b) the service intermediary, generally the ministerial Direccidn Nacional or Provincial; and (c) the service provider. The management in each social program consists of a technical administrator, a planning committee (the Comite' de Gestidn), and a supervisory committee (the Comite' de Vigilancia). The technical administrator i s the social program's executing entity, and has main responsibility for its direction. The planning committee i s composed of representatives of the social ministry and local organizations who design the scope of the program and make arrangements about the technical role of all players participating in the program, including setting up activities. The suDervisorv committee i s also composed of representatives of the government and local constituencies to ensure that the program i s adequately executed and inconformity with the technical and budget planning. The treasury function is part of the management structure, and i s carried out by the ministry hosting the social program. It executes any financing arrangement agreed between the program and MEF, autonomous agencies, or other entities. In general, budgetary resources of the social program for either the current operational expenses or capital investment are transferred by MEF to the social ministry, which acts through its intermediary agencies to finally reach the service provider. Diverse forms of local participation have recently developed, but these are rather exceptions (Table 4.9). Bono de Desarrollo Humano (BDH) d a Programa Ampliado de Inmunizaciones(PAI) 76 comites cantonales promoting vaccination programs, and 176 municipios participating in the provision of services Programa de Maternidad Gratuita (LMG) 40 out of 220 municipios participating inthe provision and supervision of services Programa de Asistencia y Desarrollo Comunitario 443 out of 786 juntas parroquiales participating in the (PRADEC) provision and supervision of services Programa de Alimentacidn Escolar (PAE) d a Service providers are institutions that serve local communities. They are either a school or a health unit in a local community, paid either through the intermediary social ministry or directly by MEF. PAI, for instance, has hospitals and other health centers devoted to the vaccination of infants and 76 children to prevent tropical diseases. For its part, PAEhas a significant degree of local communities' participation. It delivers food to the education units, and mothers' committees prepare breakfast and lunch for children. Monetary contributions are also frequent. Others, like PRADEC, manage food supplementary activities through comedores comuniturios, meeting the needs of children and elderly populations inthe poorest areas. 4.24 Priority social programs have grown increasingly fragmented and disorganized. The number and volume of PSPs has almost tripled to 45 in 2004, from 17 in 2001, and their approved budgetary resources have climbed to US$398 million (1.3 percent of GDP) in 2004, from US$62 million (0.3 percent of GDP) in 2001. Inparticular, the budget for PSPs inthe social welfare and health sectors skyrocketed to 10 times its 2001 level. Their creation reflects the GOE's response to political and social pressures and reverse poor household income losses that resulted from the economic crises and the transition to full dollarization. They address the needs of the poorest and most vulnerable groups, including women, school children, disabled and elderly people, and indigenous groups. However, this effort has not been supported by sound budgetary practices.. Consequently, the existence of too many social programs translates into too many overlapping administrations to control, less-efficient service delivery, diversity of financing arrangements, and discretionary government intervention (Table 4. Table 4.10 Budget of Priority Social Programs Grouped by Social Sector, 2001and 2004 2001 2004 Number of Budget Number of Budget Sector Programs US$Mn % Programs US$Mn % Education 6 27.6 44.7 8 49.9 12.5 Social welfare 5 23.0 37.3 14 279.5 70.3 Health 2 4.4 7.2 10 37.5 9.4 Housing and urban development 4 6.7 10.8 8 30.0 7.5 Labor 0 0.0 0.0 5 1.o 0.2 Total 17 61.7 100.0 45 397.8 100.0 Source: STFS. 4.25 Overall, budget formulation in the PSPs suffers from the same discretionary and unreportedamendments made to budget approval ex post. The major problems, however, are located at the stage of budget execution, and budget evaluation and control, due to inadequate cash management, budget reporting, and limitedtechnologicalresources and trained personnel inline agencies. 4.26 Discretionary government intervention reduces the budget initially approved for social programs, especially for those depending on the Ministry of Social Welfare. Their budgets are formulated in conformity with a preestablished timeline, accounting and classification procedures, and within overall expenditure ceilings. Once approved, there i s no transparent timely reporting of changes in the amended budget (codifkudo) throughout the year. Thus, cash constraints originating from reduced tax collection drive MEF to amend its budget without discussing it at a technical or local level with the program managers. It turns out that not only programs, but ministries, frequently do not have the most 4. The cluster of programs grouped as "Aid to Rural and Indigenous Households" is the one with the largest number of social programs, totaling twenty-one. This cluster provides 6 different services to mainly rural and indigenous women, including a conditional cash transfer for consumption (BDH), a credit for productive activities (Credit0 Productivo Solidario), eight low-cost housing programs including the Vivienda Bono Solidario, four labor development and training programs, and three primary health programs including a mobile station service, all managed through four different ministries. The next-largest cluster is the one for "Child and Infant Support," with ten programs or activities embodying four food security and nutrition programs (PAE, P A " , PRADEC, and Micronutrientes), five shelter and development programs (Direccion de Proteccion de Menores, FONDEJU, Nuestros NiAos, ORI, and Erradicacion del Trabajo Infantil), and 1 primary care program (Cuidado Materno Infantil), all dispersed infour ministries as well. 77 updated version of their budget. Therefore, changes in the budget are unknown, which prevents decisionmakers and stakeholders from knowing ex ante and ex post how public funds are actually reassigned in agreement with government priorities. As a result, in 2003, all PSPs had an average 11 percent reduction in their budget, but those grouped under the Ministries of Education or Social Welfare-particularly PAE, BDH, and PRADEC-suffered major cutbacks-close or above 20 percent- of their initial budget (Table 4.11). Table 4.11 Budget Approved and Executed for Priority Social Programs, 2003 (inmillionsof US dollars, exceptwhere otherwise noted) Approved Executed Initial Budget Revised % Change % Change Sector/Program (Pro Forma) (Codifcado) (over initial) Devengado (over initial) Ministry of Economy and Finance Bono de Desarrollo Humano' 203.1 164.6 -19 161.8 80 Ministry of Social Welfare PRADEC 12.2 10.0 -18 7.0 57 Ministry of Health P A " 2000 5.7 5.7 -1 5.7 99 PAI 10.0 9.0 -10 9.1 91 Ley de MaternidadGratuita 19.9 25.5 28 23.9 120 Ministry of Education PAE 30.7 17.0 -45 14.2 46 Investmentexpenditure2 90.8 98.4 8 26.0 29 Total 372.4 330.2 -11 247.7 66 Notes: 1. IncludesBeca Escolar. It will be transferred to MSW in2005. 2. IncludesSecretaria Nacional de Deportes. Source: Shepherd(2004). 4.27 Other relevant problems are found in other aspects of budget execution. Indeed, bureaucratic hurdles and delays inthe transfer of resources contributeto program execution falling behind schedule (Figure 4.4). 4.28 Bureaucratic hurdles include an excessive and duplicative number of required documents. Three documents are required for budget execution: (a) a monthly Programacidn Periddica de Caja (PPC) form to MEF requesting Treasury to establish a cash commitment up to the amount indicated therein; (b) an electronic entry of progress indicators updating the database system (SIGOB); and (c) a Liquidacidn de Gastos or Resultados Econdmicos spreadsheet of the social programs submitted to the Budget, Public Investment, and Accounting Departments in MEF. There i s room for elimination of eventual duplications among these requirements, to counteract the strong complaint that these hurdles are deliberately set to delay transfers of resources. 4.29 Cash rationing induces multiple budgetary malpractices. These include: The lack of knowledge about any technical criteria that the Budget or Treasury Offices apply to approve-partially or in full-or disapprove a PPC. The rationale for cutting the planned budget follows directives of fiscal discipline, as MEF adjusts expenditures to the amount of actual revenues collected during the year. But apart from such a general directive, there i s little knowledge about what specific criteria are actually considered to reallocate resources. The lack of a Cash Committee, made up at least of representatives of Treasury, Budget, Public Credit, and Economic Policy only makes this matter worse. The proliferation of informal financing arrangements to avoid major disruptions in program operations. This includes the use of off-budget funds-through autonomous agencies or donor resources-that often are not registered or are registered late in the initially approved budget. For example, in 2003, the PA1 raised about one-third of its additional budget through resources not 78 reported in the pro forma, which were donated by the Government of Luxembourg and multinational private firms (Shepherd 2004). A n ineffective intermediation of ministries inchannelling resources, resultingindelays or leaks. Once MEFtransferred resourcesto ministries, e.g. MEDor MSW, often these ministries held funds in their accounts before actually delivering them to the decentralized executing units, programs or service providers (Annex C). Not surprisingly, the one PSP-the Free Maternity Program (LMG)- that has an appointed Executing Unit (EUs) was the most successful in terms of budget execution in 2003. This is due to the fact that it i s ruled with autonomy, has its own special account and received non budgeted exceptional financing from the Solidarity Fund. Missclassification of PSP spending as between current and capital spending is common. No clear-cut criteria are defined. For instance, the classification of expenditure assigned to the BDHas a current transfer i s questionable, because this operation could also be conceived to be a temporary capital investment that produces a return of an increased number of children attending schools, particularly in the poorest areas. Bear in mind, however, that these expenditure classification problems are endemic and appear inmost social (and cashtransfer) programs worldwide. Figure 4.4 Budget Executionof Social Programs i submits PPC DIRECCIdN PROVINCIAL Y consolidates PPC Actual < ejecutora? projected revenues? k+a!F Cut down the budget FINANCE I I I I [PPC < up to date? -I N+ Cashavailable? CRITICAL POINT Source: MEF. 79 4.30 Institutional hurdles and budgetary malpractices are the main reasons for budget underexecution of priority social programs. Budget protection had a mixed performance in 2003. All PSPs, except the LMG, underexecuted their approved budget, but the level of underexecution, however, was significant-below 80 percent-only inthree out of six selectedprograms (Table 4.12). 4.3 1 A simple tracking survey of transactional steps inthe channelling of resourcesmanaged by the PSPs reveals that delays are significant, and that programs with the worst budget performance are, not surprisingly, those with most significant disbursement delays. This innovative survey, based essentially on interviews and records collection, tracks not only processing delays in payments- especially by MEF-to intermediaries and service providers, but also processing and reimbursement of purchasing orders by social ministries and all five PSPs in 2003 (Table 4.12). Even though this analysis has a limitation, for it i s very difficult to discriminate exactly on who is responsible for each type of delay-MEF or the executing ministqdagency-it provides insights and sets a baseline performance benchmark for each procedure analyzed. The main survey results for 2003 are worth highlighting: On average, an intermediary-ministry or agency-got fully reimbursed by MEF after more than two months. There were cases, such as the BDH, whose reimbursements took more than three months. On the other extreme, the LMGtakes only one month. In a similar vein, the payment of overhead fees for program administration-frequently incurred by donors-took on average more than three months. It could go to four months inthe case of the BDH. Moreover, purchase orders took an average of more than two months just to be processed, and more than four months to be fully reimbursed. The more effective PSP was the LMG,in part due to the existence of an Executing Unit, and to support from international cooperation agencies in the latter. Other programs, like PRADEC, signed a technical cooperation with the UNWorld Food Program (WFP) that has provided not only advanced financing, but also a transfer of technology and training in an entrepreneurial-like management environment. Similar arrangements with the WFP and UNDP are inplace regarding the PAE, and also with PAHOinregardto the PAI, but transactional delays remainedvery highfor both paying overhead administrative services and reimbursing purchase orders. Table 4.12 Budget Executionand Transactional Delays within PSPs, 2003 BDH PA1 LMG PRADEC PAE Average 2003 InitialBudget (inUS$millions) 203.1 10.0 19.9 12.2 30.7 Revised budget (inpercent)' 81% 90% 128% 82% 55% 87% Budget execution, dtvengado (inpercent)' 80% 91% 120% 57% 46% 79% Transaction delay (days in average) Intermediary being reimbursed at least 50%' 78 63 32 43 47 53 Intermediary being reimbursed in full2 93 74 65 62 62 71 Service provider being paidin full3 30 92 73 73 85 71 Program administration services (overhead) paid ... ... 65 102 135 107 Purchasing order processed4 ... 62 ... 47 112 74 Purchases being reimbursed5 ... 134 ... 113 172 140 Notes: 1. Inpercent of the initialbudget approved by Congress. 2. Funds actually transferred by MEFto the intermediary (PPC Aprobado y Recibido),inpercent of the initial. 3. Funds actually transferred to the service provider (PPC Aprobado y Recibido). 4. Includes the time since the order was processedup to Customs clearance that takes 35 days. Purchases ordered twice a year. 5. Refers to intemational cooperation agencies contracting with suppliers and then getting reimbursedby Treasury. Sources:Ministry of Finance; and social programs estimates. 4.32 The same tracking survey found a strong seasonality in budget execution for the five PSPs. Inaggregateterms, data show the following: 80 In 2003, the amount of cutbacks in the approved budget were constant until the third quarter, but deepened in the last quarter. Data for all five programs show that the modified (codificado) budget for PSPs was reduceddrastically inthe last quarter (Figure 4.5). Figure 4.5 Seasonal Patterns inBudget for Social Programs, 2003 (Flows, US$mn) BDH 1 PAE 30 - -Planned budget 20 - 43-Coddicado 10 - +Transfers (paid) 0, 7 1 6 1 12 1 6 - PA1 5 - 4 - Q1 02 Q3 a 4 QI Q2 9, Q1 a 1 @2 a 3 a4 Source: Secretaria Te'cnicadel Frente Social . Evidence suggests that MEF held back its transfers to the PSPs in the first half of the year-an average reduction of US$30million (43 percent) per quarter, but then went on to comply with the committed average level of US$70 million. This suggests that cash rationing was the real budget constraint. In individual terms, it is clear that BDH, PAE, and PA1show a common trend in their modified budget, whereas wide diversity is found inactualtransfers. 0 Among all five programs, PAEwas clearly the program most severely affectedbecauseit received no budget resources in the first half of the year. However, it recovered its level of budget execution in the third quarter and peaked inthe last quarter, reaching about 100percent of committed resources for the year. 4.33 Finally, it would have been interesting to explore whether these trends-individual or aggregate-reflect a seasonal pattern intax collection, but no complete data were obtained from Treasury to explore this question. 81 Provincial and Municipal Governments Figure46 PlannedcomparedtoExecutedTransfersof the 15-RrcentLaw, 1997-2oa3 4.34 Perhaps the most outstanding shortcoming of 16 budget formulation by sub- national governments is the 14 unrealistic projection of the 12 level and timeliness of Central Government transfers. This is 10 largely due to the particularities 8 of the management of inter- governmental transfers in 6 Ecuador. Although the 15- 4 Percent Law i s formally a revenue-sharing arrangement, 2 which mandates that resources are transferred to local levels based on revenue availability, actual budgetary practices differ from this norm. In theory, in Source: Frank(2004). January of each fiscal year, the MEFcommits itself to a certain level of transfers, formalized inAcuerdos Ministeriales, which lists both the revenue sources and the amounts to be transferred. Local governments then take these Acuerdos as a reference for their own budget formulation. However, in practice, despite a significant improvement in 2003, the amount of total revenue transferred deviated substantially from the amount planned for many years back (Figure 4.6).5 Under the current system, intrayear budget adjustments are frequent, and local budgeting, especially for subnational governments heavily dependent on transfers, becomes a cash- strapping, year-long, continuous exercise, perversely affecting their budget execution. 4.35 Budgetformulation at the local level replicates similar deficienciesat the central level. Most municipalities and provincial councils do not align their spending priorities with those from the national level, have no strategic plan, and have no incentives for allocative and productive efficiency. Emphasis i s placed more on budget recording than on reporting and planning. Most of them consider budget formulation a routine exercise that provides little value addedfor effective government. 0 Since 2000, most municipalities and provincial councils formulate their budgets in isolation from national strategic directives, either because there was no legal requirement to do so6;because of lack of knowledge of national objectives; or because of the inability of central institutions to formulate national objectives, communicatethem, and buildconsensus. 0 As far as sub-national governments are concerned, the Fiscal Law (FTSRL) provides a weak framework for fiscal discipline. For instance, the Law establishes limits for yearly budget growth: but subnational governments are excluded from it. Therefore, the only formal budget constraint on municipalities and provincial councils i s the maximum allowed level of indebtedness, which itself is 5. This i s a general disadvantage of revenue-sharing arrangements, not attributable to the particular circumstances in Ecuador. It is therefore highly unrealistic to introduce multiyear subnational budgets as mandated by the Fiscal Law (Art. 20). Eventhe mandatory quarterly budgetary projections are difficult to achieve (Art. 34). 6. The Fiscal Law, adopted in 2002, expressly exempts municipalities and provincial councils from sending their operational plans to the MEF (Art. 2, LRF). 7. 3.5 percent inreal terms (Art. 3). 82 problematic.8 Because their maximum level of debt is measured against current revenue, and this includes intergovernmental grants, its implications are obvious: subnational governments achieve higher debt ceilings if they obtain a higher level of transfers. LAC experience has shown that this rule canresult in an unsustainable debt path.g Budget formulation is not used as a means of raising allocative efficiency and matching public services with the priorities of citizens. Moreover, participatory budgeting i s rarely performed, and only a few public entities, mostly municipalities, engage communities regularly. An outstanding exception i s the Cuenca municipality, where the rural Juntas Parroquiales-electednationwide since 1998-are formally involved inthe yearly budget exercise. Inaddition, a minor share of its municipal budget i s now directly executed by this social committee." Future expenditures are projected based on past levels, which replicates inefficiencies and rigidities of central management. No incentives are given to local politicians to find creative solutions interms of costing exercises for improving their budget efficiency." Figure4.7 TransfersandPaymentsof the 15-PercentLaw to Municipalities(inthousand USD),2003 45,000 I 1 40,000 35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 P aym ents Source: Frank (2004), basedon informationprovidedby MEF. 4.36 Budget execution is affected by arrears intransfers, not only on a yearly, but on a monthly basis. In contrast to other LAC countries-Bolivia and Colombia, for example-that apply revenue- sharing arrangements, payments to subnational government inEcuador are not made on a daily basis, but on an irregular basis: each month there can be several payments, only a single payment, or none at all. Consequently, the amount of transfer payments i s very volatile. In2003, for instance, transfers remained relatively low until the middle of the year, but in the second half, MEF scaled up the amounts in order to meet the agreed yearly target (Figure 4.7). Similar seasonal trends can be observed in 2001 and 2002 (Frank 2004). Oil revenue volatility accounts for most of transfer volatility. When oil is excluded from 8. A debt stock relativeto current revenue 100percent; debt service lower than40 percentof current revenue (Art. 7,LRF). 9. Bolivia i s probably the clearest example. After 1999, nine, mostly large municipalities-among them La Paz, Cochabamba, and Santa Cruz-went bankrupt. A rescheduling of debt showed its positive effect in local government returningto solvency a few years later. 10. Other examples of participatorybudgetingincludethe Guamote municipality inChimborazoprovince, and the Cotacachi municipality in Imbabura province, both of which have a high share of indigenous population residinginthe localjurisdiction. 11. Some advances for costing have been made in the education sector and the Ministry of Education (MEC). With Cooperacidn Te'cnica Alemana (GTZ) assistance, a costing methodology has been developed where transfers would take into account the number of students attending a particularschool. 83 the revenue base that is used to distribute the 15Percent Law, the volatility of funds appears lower, and the flow of funds much smoother (Frank 2004). The volatility of transfers undermines the ability of municipalities and provincial councils to cover the cost of services that are usedregularly. It also impedes efficient delivery of local services in the water, sewerage, and road sectors, and often leads to interruptions in service provision. 4.37 Since 2003, however, MEF has strictly complied in keeping monthly transfers stable, but this has increased rationing in a cash-strapped Treasury. Transfer volatility does not necessarily affect the monthly amount of total transfers made to subnational governments. However, instead of letting transfers fluctuate according to actual revenue obtained each month (or each day), MEF has predetermined (and complied with) its own schedule of transfers, but at the cost of sacrificing other financing priority needs of the Central Government (Figure 4.8)12. Therefore, anew subnational spending "cycle" has emerged, featuring "savings" by the National Treasury when actual monthly revenue has exceeded the level of transfers; or "borrowing," when the reverse occurs, that is, when there i s a shortfall of revenue. In the former, MEF has used the uncommitted part of funds to finance expenditures in sectors, other than the subnational governments. Unfortunately, these "savings" have become contingent liabilities of the Central Government that are later claimedby mayors and prefects. Figure 4.8 Transfers of the 15-Percent Law and the New SubnationalBudget Cycle, 2003 70 I I 60 E -I- Planned h .l- 50 transfer L according to 3 40 15% -law 0 0 2 30 1" Executed .-c transfers = 20 0 2 10 I I I 0 1 Monthly Transfers Note: Monthly transfers that are displayed may consist o f several payments. Source: Frank (2004) ,based on informationprovided by MEF(2004). 4.38 Budget control and monitoring insubnational governments remain weak and cumbersome. Both municipalities and provincial councils conduct their own internal auditing. In addition, there i s external monitoring from the Comptroller General, which has the formal powers to initiate administrative sanctions. Annually, roughly 150 municipalities are found in conflict with administrative rules and regulations. The majority of subnational governments are able to justify legal behavior ex post, but a minority of administrations-roughly 20 a year, or 10 percent of the total of 2 2 0 4 s subject to judicial procedures. Much weaker i s budgetary oversight and monitoring by citizens and users. 4.39 Reporting on local budgets remains poor. Only since 2003, has the Central Government made an effort to improve the information flow among levels of government. Reports on executed municipal and provincial budgets are now required by MEF before the end of March of every year, but a 12. The fact that the Ministry o f Economy and Finance i s now, since January 2004, planning the level o f transfers based on the quarterly average will not substantially improve this situation. 84 considerable number of local administrations still send incomplete information, or simply disregard this rule entirely. Although there are no official data on compliance, roughly 20 percent of municipalities and provincial councils still do not send their executed budget back to MEF: as of April 2004, about 177 of the 220 municipalities hadreported their 2003 executed budget. D. BudgetTransparency, Accountability, and Participation 4.40 Ecuador has the lowest L A C regional ranking in budget transparency. In a 2003 regional survey of budget transparency, Ecuador ranked last among 10Latin American countries, with only about a third of respondents saying budget policy was transparent inEcuador (Figure 4.9 and Table 4.13). Figure 4.9 L A C Index of Budget Transparency, 2003 (Percent of Respondents) 46 45 44 44 Source: Latin American Index of Budgetary Transparency (2003). Inthe survey, positiveresponsesfluctuatedbetween 0and 36percent. The variables with the highest percentages of positive responses were: information on macroeconomic criteria (36 percent), time given to the Legislature for analysis of the national budget (42 percent), and quality of information and statistics (24 percent). However, these values were below the average of the selected group of LACcountries. Areas with dramatically low positive responses were the evaluation of the internal comptroller (0 percent), budgets developed on a National Development Plan (0 percent), reported budget execution information that allows its monitoring (0 percent), information on public servants' salaries and benefits (0 percent), and timeliness of budget information being made public duringthe execution and auditing phases (0 percent). Inbetween these extreme ratings, low ratings were also received on proper verification of non- financial objectives of the budget by the Comptroller (2 percent), proper supervision of resources managed by public enterprises (3 percent), citizen participation inthe budget process (3 percent), and budget oversight by civil society (4 percent). This reflects that despite recent efforts developed by the Fiscal Policy Observatory and other civil society organizations, there i s still a scarcity of effective institutional mechanisms (and tools) to incorporate social auditing and accountability norms into the budget process. 85 I m m s >4 d n m m n 9 . 9 - u -d 4 29e Y 3Ya 3 b 0 E 8o 4e, 5 c) e, Q bl U 0 13 9 1 + CO 3 5 e, 0 c 3 c 0 C ; JZ Y 3 B Ycd 5 e E 0 z3 El s i E Y P 0 E m tC 3 8 j;c c 3e, zi 8i a !t L fe, S ;i v1 i 4za e, f 0 zb C 4 3i e, 2 fc Y 5 c s I 4 r e, J 3 - E. Are There SufficientConditionsfor a MultiyearBudgeting FrameworkinEcuador? 4.41 A well designed Multiyear Budgeting Framework (MYBF) should lie at the core of government efforts to builda poverty reduction strategy and Ecuador's commitment to meetingthe Millennium Development Goals. The MYBFi s the core instrument available to articulate the country's poverty reduction strategic spending priorities within the projected and sustainable fiscal resources envelope. It has three objectives: improved macroeconomic performance, especially fiscal discipline; better inter- and intra-sectoral resource allocation; and more efficient use of public resources. Complementary objectives are greater budgetary predictability for line ministries, increased political accountability for public expenditure outcomes through more legitimate decision making processes, and greater credibility of budgetary decision making by enforcing political constraints conveyed ex ante. Finally, budget flexibility can also be enhanced by continuously reviewing budget priorities and managingthe budget inthe context of hardbudget constraints. 4.42 The MYBF provides the framework that allows expenditures to be driven by policy priorities and disciplined by budget realities. It combines a top-down fiscal resources envelope with a bottom-up estimation of the current and multiyear costs of existing national policies and, ultimately, the matching of those costs with available resources in the context of a multiyear budget process. The "top- down" resource envelope-often known as the "ceiling" expenditure-is determined by a macroeconomic model that projects fiscal ceilings and estimates target revenues and expenditure for the next years. As such, it also requires a pre-definition of the priorities of the national agenda. The "bottom- up" approach-often called the "floor" minimum spending-summarizes the sectors' review of their main program and project priorities, with an eye to optimizing the minimum allocations. Whereas the top-down approach i s often determined by the central fiscal authorities, some fully developed MYBFs- like those of Brazil or Uganda-have developed the bottom-up approach in a participatory way with subnational governments and civil society. Once the country becomes ready for a MYBF satisfying a series of pre-conditions (see below), it will essentially have to cover six stages (Box 4.1). A variant of this approach could consider starting a pilot MYBF focused in the social sectors. If such pilot process becomes successful, it would give credibility to medium-termprogramming. =Six Stagesof a ComprehensiveMultiyear BudgetingFramework I 1. Development of a Macroeconomic Fiscal Framework: A macroeconomic model that incorporates projections of revenue and expenditures inthe mediumterm (multiyear). 2. Development of Sector Programs: Agreement on sector objectives, outputs, and activities review, development of programs and subprograms, and preliminary cost estimation. 3. Development of Sector Expenditure Frameworks (SEFs): Analysis of intra- and inter-sector trade-offs. 4. Definition o f Sector Resource Allocations: Setting current and medium-termbudget ceilings. 5. Preparation o f Sector Budgets: Medium-term sector programs based on the matching o f sector priorities with budget ceilings. 6. Final Political Approval: Presentation of budget estimates to cabinet and Parliament for approval. I Source:World Bank (1998). I 4.43 Ecuador is not ready for a MYBF. International experience collected inthe implementation of poverty reduction strategies (PRSs) has, however, concluded that malpractice in public expenditure management i s an obstacle to achieving the full benefits of an MYBF. In addition, the fiscal space is extremely reduced for a multiyear investment plan. Improving budget formulation in a multiyear fashion should not be seen as a panacea to solving the multitude of PEM problems that exist, and it even risks distracting attention away from the need to improve the basic framework. To its credit, Ecuador budget practices have several positive aspects: (a) macro-fiscal projections were realistic and within an reasonable range until 2003; (b) deviations between budget formulation and budget execution were not 87 significant until 200213;(c) budget ceilings are introduced in line agencies to ensure their plan i s consistent with overall macro aggregates; and (d) financial information is slowly becoming more disaggregated, timely, reliable, and transparent. However, key issues that Ecuador needs to resolve in order to have a complete MYBFare: Negotiations between line agencies and MEF typically do not involve technical discussions on trade- offs, but are focused on the extent of inertial allocations. 0 Line ministries rarely consult with spending unitsto get informed costing exercises of selected goals, so there is no alignment of expenditures (inputs)with strategic outcomes. 0 Few agencies collect data systematically on activities or programresults, so there i s little scopeto link budgetallocations with agency performance. 0 While in a well designed MYBF, capital and current spending budgets should be integrated, in Ecuador they are currently determined by separate entities, and these should be integrated in an MYBF. 0 Municipalbudgets have no clear connection with defined national priorities and lack adequate budget formulation, reporting, and control. 4.44 The MYBF should be seen as a complement to-not a substitute for-basic budgetary reform. A piecemeal approach i s likely to fail, since improvements in one area will not translate into gains through the budget cycle. By precedingthe formulation of annual budgets, the MYBFmay have at least a partial, but significant, impact especially at the budget formulation stage. Henceforth, the approach to the introduction of an MYBF should necessarily be gradual. It can start by defining ceilings on the major aggregate fiscal variables and on key sectoral spending. Then, a multiyear sector budget could be piloted in those ministries where predictability of funding and transparent outtums would warrant good monitoring. Those building blocks should be in place before moving forward to more advanced stages of an MYBFinthe foreseeable future. F. Policy Recommendations 4.45 The performance of PEM needs to improve in all levels. Priorities might be to first consolidate annual budget procedures that support aggregate monitoring of fiscal discipline, while simultaneously introducing measures to improve the strategic allocation and operational effectiveness not only at the central level, but at the level of subnational governments. In the short term, developing austere and realistic budgets (Chapter 2) and sound budget management of a still cash-rationed economy are priorities for fiscal consolidation. In the medium term, continued efforts to refine the design of the poverty reduction strategy, and to develop adequate sector strategies will guide intra- and inter-sector allocation and programprioritization (Chapter 3). 4.46 Key recommendations to sound budget management are well summarized by a recent ROSC completed by the IMF (Box 4.2). Besides those general recommendations, a few essential ones derivedfrom this chapter analysis are to: P Create a Cash Committee at MEF, composed of representatives of the Budget, Treasury, Public Credit, Public Investment, and Economic Policy Offices. Its main function would be to design and implement a cash management policy, based on clear and simple prioritization criteria. In the short term, and while arrears still exist, define temporary rules for applying cash rationing through in-year budget modifications. All resolutions from the Cash Committee should be made public, to allow social accountability. 13. The lack o f closure of the 2003 budget accounts has prevented authorities from updating budget execution at the sectoral level for 2003. 88 9 Revertinertialbudgeting throughstructuralchange. Curbingthe positive growth ofthe wagebill i s critical. There i s no substitute for this and it cannot be emphasizedenough. 9 Gradually integrate off-budget activities into the Treasury Cuentu Unica gradually. This would bring Treasury control of resources now sleeping in commercial bank accounts. For instance, an obvious candidate i s PetroEcuador subsidies. Box 4.2 Main Recommendations of the IMF-ROSC InDecember 2003, an IMFReport on the Observance of Standards and Codes (ROSC) evaluated Ecuador's fiscal transparency practices compared to the IMF Code of Good Practices on Transparency in Monetary and Financial Policies. The ROSC recommended the following measures: 9 Improvethe legalframework for fiscal management: 0 Approve a new organic budget law regulating the main aspects of budget management. Simplify tax legislation and eliminate unproductive taxes and exemptions. Periodically review private sector regulations and procedures to maintain transparency. 9 Increasethe comprehensivenessofthe budget: 0 Apply gross budget accounting (that is, not netting out expenditures from revenues). Bring all fiscal policy activities carried out inentities into the Central Government budget. Report inthe budget all quasi-fiscal activities, tax expenditures, and contingent liabilities. 9 Bolster operationalcontrol: SIGEFshould beredesigned to collect timely, relevantfiscal information. 9 Increaseaudittransparency andeffectiveness: 0 Make public the audited government accounts and public enterprise financial statements. Publishand enforce the conclusions andrecommendations of audit reports. 0 Increase the number and training o f audit personnel. 9 Improvethe presentationofthe budget: 0 Present clearly the revenue, expenditure, overall balance, and financing components inthe budget, and in reports o f budget execution. 0 Present the functional, economic, and program classification of revenues and expenditures. Include inthe budget a discussion of the main objectives o f the different budget programs. 4.47 Regarding budget management in social moa-ams, it i s clear that that the budgets in many social programs have no relationship to their goals and expected outputs. The fact that many social programs have a lower than 80 percent level of execution i s a problem. Critical recommendations are to: > Review budgeting policy in social programs so as to establish a close association between budgets and goaldresults. Setting monitoring-financial and physical-indicators i s a f i r s t step. These should include coverage, targeting, and cost efficiency indicators. 9 Review the overall budget protection policy. Budget protectionpolicy hastwo sides: onone hand, it increasesbudget rigidity; on the other hand, in a context of fiscal retrenchments, it guarantees that cash-rationed mechanisms do not affect essential pro-poor expenditure. A first step would be to assess existing protected programs. For example, budgetary protection i s actually given to seven priority social programs (PSPs), five of which are examined in this Chapter: LMG, BDH, PRADEC, PAE, P A " , PA1 and investment by the ministry of education. Then, some criteria should be defined and applied for reconfirming programs to receive budgetary protection annually, so as to make sure they effectively represent key priorities of the national agenda. International experience indicates that the number of selected programs should be small. Finally, the amount of resources to be protected should be seen as a function of past and present execution capacity, as well as of the outcomes to be produced. In any case, programs should receive the minimum amount of resources neededto achieve their goals. 89 9 Consider the possibility of converting the de facto budget protection policy into a "Virtual" Poverty Fund. This enhanced mechanism would consist of tagging selected budget programs that are poverty-reducing, i.e. the PSPs (the so-called "virtual" fund). Drawing from the experience of Uganda's Poverty Action Fund (Box 4.3), virtual funds rapidly expanded through Highly Indebted Poor Countries (HIPC).14 In some countries, its setting up has also involved making changes to the existing budget and accounting systems (Zambia); or introducing new reporting templates to track spending lines (Mauritania, Mozambique and Tanzania); or producing additional information on all poverty-reducing spending (Honduras) (The World Bank and IDB 2003). A Virtual Poverty Fund, tailor-made to Ecuadorian budgetary protection policy, would essentially require adding two new features: (i) the development of its institutional framework by creating a coordination body and facilitating open access and reporting; as well as (ii) the definition of monitoring indicators and evaluation mechanisms to be followed in a participatory way with civil society (an Observatory of Social Policy, for instance). Such a proposal would definitely strengthen a medium-term social policy with one of the finest tools for a solid poverty reduction strategy. Box 4.3 Ecuador: Institutionalizinga Virtual PovertyFund: A Look at Best-PracticeUganda The virtual Poverty Action Fund (PAF) i s focused on implementing the Government o f Uganda's highest priorities within the Poverty Eradication Action Plan (PEAP). The PAFis funded by a combination o f HIPC debt relief, donors-general or sector-earmarked-support, and Government's own resources. The operational framework of the virtual PAF has the following components: (i) budget priorities are defined by a rolling 3-year MTEF; (ii) it is not a separateFund,but a subset of the overall budget; (iii) Government the commits not to cut its funds; (iv) all expenditures fall under full Congress and auditor general oversight; (v) it is managed by sector and ministerial level Working Groups; and (vi) PAF programs have a structured and participatory institutional framework that ensures that they are properly planned, budgeted and implemented. This includes: Eligible Criteria: A program qualifies if it meets 4 criteria: it is part of the PEAP; it directly reduces poverty (with involvement by the poor); it delivers a service to the poorest 20" o f the population; and has a well-developed implementation plan (with costs, outcome, and output targets clearly identified). Programs are reviewed once a year. Administration: 5 percent o f PAFresources are set aside for improving program effectiveness and transparency; Size: at least a constant proportiono f the original budget; Reporting: An overall quarterly report by central or local governments should be produced by the Ministry o f Finance and distributed countrywide. Biannual sector performance evaluations are also required. Audits should cover all Central Government-led and at least 60 percent of district-led programs. Sector expenditure tracking studies are required when there is inadequate audit information. Civil society independently monitors reviews. Safeguards: (i) case of budgetary shortfalls, under-allocation to PAF should be lower than cuts in innon-PAFprograms; (ii) caseofunder-performance, fundscanbereallocatedto other PEAPprogramsor in to PAFprograms inthe following fiscal year. 14. In 1996, the IMF and the World Bank launched the HIPC initiative to reduce the debt stock of debt-stressed countries. In 1998, they launched the Enhanced HIPC initiative to provide faster, deeper and broader debt relief, and strengthen the links between debt relief and poverty reduction through a poverty reduction strategy produced by the country. Uganda i s committed to channel cash-flow savings on interest payments resulting from the reduction of the stock of debt under HIPC toward selected poverty programs. 90 9 Merge or eliminate duplicative socialprograms. The current number of 45 social programs is too high and suggests waste of resources (Table A58).I5 Obvious candidates for elimination are those programs that show a significant degree of low budget and/or significant budget underexecution and/or poor targeting. A first step is to assess the performance of existing programs, and also find out whether these are donor-financed, which would improve chances of their coordination. Impact evaluations would help deciding which programs are more effective and deserve to take the lead in a particular area. For instance, the creation of the Sistemu Zntegrudo de Alimentucidn y Nutricibn- S I A N i s a very positive step in the right direction and should be replicated. The on-going retargeting of the BDH, PAE and PRADEC i s also a welcome development. 9 Eliminate cash constraints inthe first half of the year. Arrears reappearedinthe first quarter of 2004. Preventing them in future years would require proper planning of cash management by the proposed Cash Committee. 9 Rationalize, simplify, and if possible automate, budget procedures and forms for requesting reimbursement of payments. Bureaucratic controls over individual commitments and payments could be replaced by an advanced quota system covering an overall envelope of resourcesplanned in the monthly PPCs. This process, however, would depend on the extent to which SIGEF records proper and timely information about budget execution. A system of management contracts should be designed combining incentives for timely reporting of financial and physical indicators and sanctions for noncompliers in selectedPSPs. 9 Resist the temptationto convert an additional number of PSPsto ExecutingUnits,even though it might produce short-term improvementsinbudget execution. Another variant of this approach that should also be prevented i s the creation of further "trust funds" as a financing alternative that would allow those programs to have a special regime of operational and procurement procedures. 4.48 Regarding budget management in provincial and municipal governments, it is essential to revert to a sound processof budget decentralization. Problems have arisen inL A C from arrangements that involved automatic unconditional transfers from the center through rigid tax-revenue-sharing formulas such as those applied inEcuador. Critical recommendations are to: 9 Design a strong regulatory and institutional framework that clearly assigns expenditure responsibilities, accompanying the devolution of revenue, and in line with subnational government administrative capacity. This framework can consider establishing standard incentives for the transfer of resources in exchange for devolved expenditure responsibilities (for example, standardized per student criteria in assignment to education). 9 Isolate oil revenue from intergovernmental transfers, which would smooth fluctuations in the planned flow of funds toward subnational levels. When oil is excluded from the revenue base that i s used to distribute the 15-Percent Law, the flow of transfers appears smoother on a monthly basis (Figure 4.10). This i s necessarily so because oil revenue amounts for roughly one thirdof the volume of transfers relatedto the 15-PercentLaw. 15. The generation o f fiscal space should not be the guiding criteria in merging or phasing-out programs, but the elimination of inefficiencies and prevention o f overlapping areas o f work. InEcuador, 3 social programs take about 75 percent o f the overall budget. Besides, there are important political economy considerations to consider, especially when programs are administered by different sector ministries, or particular constituencies have been created around programs, e.g. indigenous groups. 91 Figure4.10 RevenueBasefor 15-PercentLaw with/without OilRevenue, 2003 500 I 4 5 0 400 3 350 300 g 250 200 ' .- =8 150 100 50 o ! , I , , I I --- Months _- __ ___ __ _ --- Revenue Base Revenue Base without petroleum '2 per. M ov. A v g . (Revenue Base) '2 per. M o v . A v g . (Revenue Base without petroleum) Note: Dotted line represents trend line (moving average). Source: Frank (2004) basedon informationprovided by MEF. 9 Adopt daily automatic transfers basedon revenue available on daily accruedincome as is often an international practice. This would prevent accumulation of local "savings" in transfers. 9 Condition actual delivery of transfers on timely and reliable budget reporting by subnational governmentsas mandated by the Fiscal Law. This would constitute the minimum conditions that are requiredto achieve more transparency. 9 Promoteresponsiblesubnational borrowingby establishingfurther legislation underwhich the Central Government might intervene in those local governments that violate the fiscal rules, specifying the main elements of subnational fiscal adjustment programs, ' reprogramming debts, and excluding the possibility of any bailout. 4.49 The budgetshould be mademore accountable,transparent, andparticipatory. In2003, the government prepared a Transparency Plan basedon three actionsthat should be taken: (a) complete the upgrade of the SIGEF financial management system to consolidate NFPS financial statements and produce monthly, timely, and reliable reports on consolidated executed expenditure to be published on the government's website (Portal) beginning in 2004; (b) support the e-government system-CONTRATANET-to perform transactional public procurement; and (c) buildanationalpublic investment system, startingwith a central databaseand project bank. In the medium term, budgetary transparency should be institutionalized at all levels of government, including subnational governments, while civil society participation should be encouraged.16 This may include a national dialogue on the multiyear program, possibly within the 16. The Civic Anti-Corruption Commission was established in 1999 to combat corruption in Ecuador. It was established as an independent entity not controlled by the Executive. Its president is elected by civil society 92 context of the ongoing preparation of the poverty reduction strategy. This would facilitate civil society organizations, such as the Fiscal Observatory, playing a better-informed and more effective watchdog and advocacy role in the implementation of key fiscal reforms, like those of Customs and of the civil service. 9 The recent approval of the Law of Access to Public Information represents an important milestone, but its Reglamento (norms) i s pending approval. Norms should guarantee gradual access to information on all public budget sectors, including defense, to ensure detailed oversight and monitoring. 4.50 Developing a MYBF requires fulfilling a number of prerequisites and carrying out several preparatory actions. A credible MYBF requires strong political support from the President and the Cabinet. Politicians must be aware that a MYBF will tie their hands and reduce their discretionary bargaining powers on budget issues. Therefore, developing a consultative process at the outset i s both legitimate and desirable, since this would prevent Legislative and public pressure from dissuading the Executive to change its priorities infurther years of implementation of the MYBF.Initial measures are to: 9 Define national priorities and a reduced set of key target outcome indicators, as part of the PRS. Ecuador already has identified a few social indicators, which should serve as a good starting point. 9 Continue improvingthe realism of growthandrevenuebudget projections, as well as of sector budget ceilings. 9 Prepare the first MYBF, including an overall macro framework and propose sector expenditure ceilings. 9 Consider the implementation of a competition-basedFundfor investment. The ChileanFundis a positive reference the main feature of which i s the simplicity of its focus to allocate the flexible component of budget expenditure (see Chapter 3). 9 Reclassify the budget per programs, which is an extremely helpful prerequisite to match strategic priorities to the PRS, prepare groundwork for integrated costing (capital and recurrent), and develop monitoring indicators. 9 Select pilot ministries for preparing sector expenditure frameworks (strategy, objectives, key programs and costs, and performance indicators), especially focused on the sector priorities of the PRS. Inadvancedstagesof the MYBF, this would leadto defining budgetaryfloors for selected PSPS. 9 Appoint a highly qualified staff to an interministerialexecutive body of the MYBF. It should develop a roadmap encompassing stepped-up phases for the preparation of a MYBF (in some best practice countries-like South Africa-this team has even standardized its procedures ina Manual on Financial Planning and Budgeting). 9 Convey national priorities to municipalities, to induce the alignment of their budgets. This can be achieved through expenditure ear-marking (provided there i s the possibility of monitoring and sanctioning) or through co-sharing arrangements such as matching grants. These and other incentives must beestablishedinsuch a way that subnational budgets are aligned to nationalpriorities. 9 Set a clear schedule between the multiannual and annual budgeting processes during each year's budget formulationprocess. Successful MYBFcountries conclude a first draft by April each year, just in time to use it as a key input in the guidelines for the preparation of the annual budget to line ministries. organizations every four years under the supervision of the Superior Electoral Tribunal. It is 95 percent financed by public funds, and the remaining 5 percent comes from international donors. 93 Volume I,Annex A An Estimationof the PotentialOutput and the Structural FiscalBalanceinEcuador' Why Estimate a Structural FiscalBalance? A l . l It has long been recognized that the nominal budget balance is an imperfect indicator of the government's true fiscal stance.' The problem i s that, in general, the fiscal balance depends not only on authorities' tax and expenditure decisions, but also on the extent to which autonomous spending decisions by the private sector are reflected in the country's output and national income levels, and thereby on tax collection. Thus, if the economy enters a recession, becauseof, say, a reduction inprivate investment, tax revenue automatically declines, irrespective of the authorities' fiscal stance, and the fiscal balance worsens. Conversely, ifan autonomous increase in private spending leads to an increase in national income, tax revenue increases and the fiscal balance improves. These changes in the fiscal balance take place automatically for a given tax structure. This implies that the nominal fiscal balance i s a distorted measure of whether government fiscal policy i s expansionary or contractionary. A given fiscal stance is consistent with a worsening or an improvement in the nominal balance, depending on whether private spending decreases or increases. A1.2 To address the issue of the need to have a measure of fiscal policy that i s independent of the particular position of the economy in the business cycle, the concept of cyclically adjusted or structural budget balance has been developed in the macroeconomics literat~re.~The basic idea i s to carry out a simple decomposition of the nominal budget balance (B) into two unobservable components-the structural component (SB) and the cyclical component (CB)-in such a way that the following equation holds for period t: B(t) =SB(t) + CB(t) (1) A1.3 Under the admittedly inaccurate, simplifying assumption that government spending i s not a function of the state of the businesscycle, the structural component i s defined by4: where SR(t), the structural revenue, i s the fiscal revenue that would be collected if actual output in t were the level of output determined by the long-term trend, which will be called trend output. G(t) i s total government spending. SR is estimated by means of the following equation: SR(t) = R(t)[7Y(t)/Y(t)]o (3) where R(t) i s the observed value for the government revenue, TY(t) is the economy's trend output, Y(t) the actual level of output, and 13 i s the income elasticity of total government revenue. Completion of the 1. This Annex was prepared by Daniel Artana and Cynthia Moskovits. 2. Most o f this section is based on Annex B o f the Public Expenditure Review for Peru (World Bank 2003). 3. See, for instance, Hagemann (1999) and Marcel and others (2001). 4. In the European Community expenditures are also adjusted by excluding expenses related to the business cycle (for example, payments of unemployment insurance). As in most Latin American countries, private firms compensate fired employees directly through severance payments, and it has been customary to adjust only government revenues for the cyclical component (see Marcel and others 2001). Another issue i s the impact of inflation on interest payments. This correction is not necessary when the focus o f the analysis i s on the primary balance of the public sector. Focusing on the primary balance is a good approximation when nonresidents hold most public debt. 95 Volume I,Annex A calculations implied by equations (2) and (3) requires that estimates for both trend output and the income elasticity of revenue be generated.5 A1.4 To estimate the potential output, two alternatives are used: a production function and the estimation method proposed by Hodrick and Prescott (1997). The second alternative i s used in most cases because it i s easier to calculate. Interms of this method, if the logarithm of Y(t) is denoted y(t), the time series for y(t) can be decomposed inits trend component ty(t) and its cyclical component cy(t): y(t) = ty(t) + cy(t) ;fort = 1, ..,T . (4) then the Hodrick-Prescott (HP)trend output, hpy (t), is the series that minimizes the expression: where the sums are carried out from t = 1to t = T and h is a parameter that determines how smooth the trend line will be. The method's idea i s to minimize the sum of two terms where the first term i s the sum of squares of the cyclical component and the secondterm i s the sum of squares of the trend component's second differences. A1.5 Inthe case of oil-producing countries, the adjustment to the nominalbudgetbalance shouldtake into account the impact of volatility of the price of crude oil. Therefore, a first adjustment i s to recalculate the nominal balance net of the effect of fluctuations inthe price of oil (these should better be saved in an oil stabilization fund). A second step i s to correct for the cyclical component of tax revenues to get a structural budget balance. Estimating the Output Gap and the Public Sector Structural Balance A1.6 This section applies GDP at 2000 constant prices during 1960-2003. Data came from the IMFandthe CentralBankofEcuador. As is customary inthis exercise, someadditionalyears are added to moderate the instability that the estimate has at the tails of the sample period. For 2004 and 2005, most conservative estimates provided by Consensus Forecasts as of April 2004 and from 2006 onward were assumed: a real annual growth rate of 3.5 percent. Several alternatives for adding years (3, 5, or 10) were applied, but they didno major changes. Hence, the series produced forecasts for only three years. The Output Gap A1.7 To determine the output gap, the potential GDP must first be estimated. Two alternative methods are available: (a) assessing a production function; or (b) filtering actual real GDP (the most common filter i s that built by Hodrick and Prescott). Though the first alternative i s probably more accurate, it i s not available for Ecuador. The results obtained by making use of the second method, with the usual h = 100 parameter for annual time series, cover 1960-2006 (Figure A.A.1). 5. Equation (3) considers only the effect on tax revenues of the economic cycle. But in a more sophisticated estimate it would be better to allow for other changes in revenues produced by the cycle (some countries adjust for the changes in the tax base o f capital gains or take into account different lags in the perception o f taxes (see for example, Donders and Kollau 2002 and Braconier and Forsfalt 2004). These topics are not very relevant for Latin American countries because taxation o f capital gains is usually low and fiscal lags are relatively short in the most relevant taxes. 96 Volume I,Annex A ACTUAL ANDFiP~TENTIALREAL GDP ure A.A.l 2 0 , m 16,wO 16,wO e m 3 14,oM) 0 0 0 cu 1 2 , m c -20 0 - x.- 10,WO 8,OW 6.000 4.000 2 , m 4 ....................... 1960 1962 19M 1568 1968 1970 1972 1974 I970 1976 1980 1982 1984 1986 1968 1990 1992 1994 1996 1998 2wO 2002 2004 2006 Source: Author's calculations on BancoCentral del Ecuador (BCE). A1.8 Natural disasters proved to injure the economy during 1970, 1982/83, and 1987, although it cannot be detected from the data inthe last El Niiio phenomenon that took place during 1997-98. Barely minor differences appear when extending the actual GDP estimates up to 2010, 2015, and 2020, respectively, in order to avoid the "tails problem" of the Hodrick-Prescott filter (Artana 2004). The Structural FiscalBalance A1.9 Ina second step, the structuralfiscal balance is estimated. The informationavailable is on the Non-Financial Public Sector for 1983-2003 of the Central Bank of Ecuador and the World Bank. Expenditures were not adjusted by cyclical components because there are no payments directly related to the cycle (like unemployment insurance). In a preliminary analysis, this assumption appears to be supported by the data, which suggests that expenditures are independent of the business cycle inEcuador. For simplicity, oil revenues are considered only those originated in exports of crude oil. Although a fraction of the revenues obtained from sales for derivatives could also be included, as collected at the upstream level, its adjustment i s small. To remove the impact of the volatility of the crude oil price, oil revenues are recalculated every year according to the average export price for the Ecuador mix during 1991-2003 (equal to US$17.2 per barrel). The adjusted series i s named Structural Oil Revenues. To estimate them, equation (3) has to be estimated with different values for the income elasticity of tax revenues. Results are virtually the same as using an income elasticity of 1.1. Following this, the Structural Primary Balance (surplus in the case of Ecuador) can be obtained (Table A.A.1).6 Figure A.A.2 compares Actual and Structural Primary surpluses for every year since 1983. The line shows the impact of departures of the crude oil price from its long-termreference. 6. A comparison for a longer period o f time i s available inFigure A.A.2, and data are available from the authors. 97 Volume I,Annex A In most years, the volatility in oil prices and its effect on oil revenues accounts for most of the difference between Actual and Structural Primary Surpluses. This underlines the importance of oil prices inthe economy and of improving the design of the Oil StabilizationFund.' Table A.A.l StructuralFiscalBalance (percent of GDP) 1591 19,w 18,Ph 18,Ph 7,1% 7,5% 5,Ph 5,6% 1992 21,5% 17#h 17>Ph 77% 7,Ph 3,5% 3,5% 1993 18,Wo 17,6% 17,5% 5,Ph 63% 41% 5,Ph 1934 17,Ph 183% 17,Ph 3,5% 44% 4,1% 45% 3,4% 35% 44% 4,Ph 2,6??0 29% Figure A.A.2 Primary Balance (percent of GDP) -3%. Y - 4% - =Actual El!!!!! Structural Actual vs. Structural Oil Rev. Source: Author's estimates on BCE, Ministerio de Econom'a y Finanzas (MEF) and IMF. Source: Based on Table A.A.l 7. The Output Gap measured for every year of the period 1991-2003 is relatively small. There is a puzzle in the data, because 1998 appears with a Potential GDP below Actual GDP in spite of the impact o f El Niiio. This puzzle is also present in the estimates done inIMF(2003 b). 98 Volume I,Annex A It i s also apparent than in periods of high export prices of crude oil, the government relaxes the collection of non-oil tax revenues (for example, in 1990, 2000 to 2003), and the opposite holds in periods of low prices (1993 to 1995, 1997, and 1998).8 This i s clearly the case with the revenues obtained from sales of derivatives in the domestic market, which are negatively correlated with the price of oil. Previously, during the 21-year period for which we had statistics, the cumulated actual primary surplus was slightly higher than the cumulated structural one (1 percent of GDP, for an annual average of less than 0.05 percent of the GDP). If the sample i s restricted to the last 10 years, this figure grows to 1.4 percent of GDP (a similar figure i s obtained for the last five years), equivalent to an annual average of about 0.3 percent of GDP. The Structural Fiscal Balance i s in surplus after dollarization. Inparticular, the level of the Structural Primary Surplus was 3.9 percent of GDP in 2003. Recent excess actual surplus over the Structural Primary Surplus i s explained by high oil prices and the need to devote additional fiscal resources to debt reduction. 8. This has beendone mostly through tax exemptions and incentives, introducingdistortions that should be avoided. 99 Volume I,Annex B Summary of the BudgetProcess Budget Formulation Bl.l The national budget formulation process in Ecuador begins in July, when line ministries and institutions must develop their annual proforma budgets.' These are then passed and reviewed by the Budget Office (Subsecretaria de Presupuesto) in the Ministry of Economy and Finance (MEF), and incorporated into the pro forma annual budget, which the Development Policy office also reviews. The pro forma annual budget is then sent in September to Congress, which can change allocations, but cannot raise the budget ceiling. Congress has until November 3 0 to ~ approve the budget, and on December lst, the new budget (PresupuestoGeneral del Estado,PGE) becomeseffective. B1.2 The national budget does not include provincial and municipals expenditures. Formulation of these budgets i s not covered by the PGE, except for the assignment of transfers to the subnational governments. Budget Execution B1.3 Budget execution i s the responsibility of the Treasury and Budget Offices inthe MEF.Movement of most of the public funds i s done through the Treasury's Cuenta Unica (Single Account). Tax and petrol revenues, and external and domestic financing, fund this account and, inturn, the account channels budgetary transfers to the institutions, makes subnational transfers, and pays the debt. This account i s reconciled on a daily basis. Figure A.B.1illustrates the flow of funds inthe budget process. 1 FigureA.B.l. Flow of BudgetaryFunds B1.4 Each public institution has TI three accounts: two in the private banks lnstaionalAccount 1 transferred and one in the Central Bank. Of the in Private Bank alter4days I lnstitionalAccount #3 two accounts in the private bank, one is I Autogestion revenues in Central Bank I 1 I I I for deposit of autogestion (self- I lnstitional Account # 2 I t generated revenues from fees, and so , . . forth). These deposits remain there for four days before being automatically transferred to the institution's account in the Central Bank. This holding period was introduced as a means to Fundingfrom: Petrol, SINGLE ACCOUNT help support private sector banks after externalfinancing Central Bank the banking crisis in 1999. The second I private bank account i s a payment account on demand. Checks are written Source: World Bank against the institution's private bank, not the one from the Central Bank. Payments withdraw the money transferred to the private bank account from the institution's account in the Central Bank. Thus, in essence the second private bank accounts have zero balance. Although withdrawal of the funds i s codified using a two-digit classification, a recurrent problem is the lack of control over whether withdrawals are consistent with their assigned budget classification. The only control i s that the amount withdrawn cannot exceed the institution's account balance inits Central Bank account. B1.5 The institution's account in the Central Bank i s funded from autogestion revenues and from budgetary transfers from the Cuenta Unica. Each institutional budget has a payment plan-the Programacidn Periddica de Caja (PPC)-stating the monthly amount of budgetary transfers to be made to the institution. The PPC i s initiated by the Budget Office inMEF. However, this i s not necessarily the 1. This Annex was preparedby ElaineTinsley. 101 Volume I,Annex B monthly amount that i s transferred from the Cuenta Unica to the institution's account. Instead, it provides a "promise of payment" that i s merely indicative of the intention to pay, but i s not enforceable. The actual amount transferred depends onTreasury cashavailability and MEF's decision to transfer. B1.6 Decentralization of expenditure managementto virtually all public institutions creates severe cash management problems for Treasury. While money i s flowing to decentralized spending units that do not report whether they have actually spent the money in deposit in their private bank accounts, Treasury finds itself short of funds. This generates artificial arrears inTreasury's Cuenta Unica, even though there i s cash sitting in the institutions' private accounts. While Treasury i s empowered to and does take back some of the sitting funds, institutions strongly resist this. In 2003, cash management shortfalls forced Treasury to raise short-term money on the markets, and hence increase budget-financingcosts. However, with the recent build-upof the old oil stabilization fund (FEP), Treasury now borrows short term against it. Still, in 2004, it has become apparent that this action has not resolved severe cash management problems. If the government had cash management under control, it would allow it to implement the budget more efficiently, minimize the cost of borrowing, and maximize return on government deposits. B1.7 Financial management i s also weak. Although Ecuador has an integrated financial management system (SIGEF) in place, this system i s primarily used for budgetary formulation and accounting, rather than execution. With only 70 percent of Central Government institutions covered by it, SIGEF does not permit timely and complete consolidation of fiscal information yet. As a result, the Government of Ecuador does not have access to complete and accurate budget execution information, which could be used as an indispensable tool for decisionmaking and management control. There are also problems in assessing the impact of expenditures and targeting future expenditures on poverty reduction programs. Expenditure monitoring i s partial and done through two mechanisms. First, the Central Bank uses the fiscal database to detect expenditure withdrawals using a two-digit classification from the institution's account in the Central Bank. However, as explained above, there i s no way to verify that this institution used resources as approved. Second, SIGEF provides some expenditure reporting that is later used by Treasury to verify payments. Budget Control B1.8 In the public sector, there are two levels of controls-internal control units and the external Controlaria General de Estado (CGE). There are about 300 internal control units employing some 800 auditors. The CGE itself has about 1,740 personnel, of which 55 percent are auditors (957). Internal control units report to and are hired by the CGE, though their wages come from the respective line ministries. The CGE also approves their work program, which i s based on a series of ad hoc tasks. Once internal units have issued their reports to the CGE, their task i s done, because only the CGE has the authority to further investigate, or issue fines and penalties for infractions. Each institution i s to be audited at least once every five years, though the overall fiscal accounts are not audited. According to the CGE, public institutions have the capacity to conduct an annual internal audit of the fiscal accounts; however, becausethey have not been required, they have not done so. B1.9 Internal auditing. The internal control environment i s weak, due to ineffective accounting and information systems, inadequate control over fixed assets, and inadequate monitoring of controls. Internal audit capacity relies excessively on external auditing. The internal audit function in line ministries i s not oriented to providing assurance on the control environment, but rather to performance of ad hoc special reviews. The internal financial audits are weak in their breadth, depth, and frequency, with procedures mainly done manually. Norms governing internal audit have not been adequately promulgated. Internal control units do not do routine analysis, or auditing of accounts; this i s mainly the responsibility of the CGE. Instead, internal control units investigate processes, and report on noncompliance of civil and administrative problems to the CGE. It also examines whether the budgets match the development goals. There i s no full auditing of the budgets of public institutions, only parts. Reports about internal audits 102 Volume I,Annex B (and external audits) are not published-though names of people committing infractions are published in the papers. B1.10 External auditing. The CGE covers all public sector entities and private entities that receive public funding. In general, the norms and procedures governing external audit are adequate and in agreement with international standards. CGE's principal activitjl i s to conduct financial audits, although recently it has started a pilot planto conduct performance audits. It does not audit the public debt. Audit functions are not directed toward efficiency gains, but rather toward ex ante and ex post compliance. The CGE also takes an active role inmonitoring public works and preapprovalof public contracts. Despite its role of monitoring the use of public resources, the CGE does not publish the conclusions and recommendations of its audits, though it does monitor the fulfillment of the recommendations it makes. Political interference is seen as the main reason why the CGE has not tackled larger cases of corruption. Though it enjoys a reputation for its thoroughness, its work has targeted minor infractions rather than greater allegations. The process of the Comptroller's selection i s subject to political interference. The President selects the Controller General from three candidates presented by Congress. This i s often problematic in a multiparty Congress having difficulty agreeing on the initial three candidates, who often represent different political parties. As a result, the Controller General's position remained vacant for over a year. Budget Reporting B1.11 With expenditure management decentralized to the institutions, reporting is another particularly weak area of public expenditure management throughout the whole spending cycle. The MEF receives information from each institution on the execution of the PGE. In some cases, this information i s transmitted by internet via SIGEF, and i s timely. However, about 40 percent of public institutions do not send their information in a timely manner. This leads to delays inprocessing by the Budget Office, which in turn authorizes monthly budget transfers to the institutions through Treasury, which materializes the transfers and delivery of funds to the institutions. Because the spending information i s not delivered in a timely way, Treasury has a hard time finding the information to make the appropriate decisions on budgetary allocations. Published budget information i s also too broad and does not reflect all fiscal activities. Periodic reports mainly contain information about the central administration. The budgetary information i s reported on a cashbasis, but authorities have indicated their intention to move to an accrual basis. The strong decentralization of the budget does not facilitate the conciliation of data over and under the line. B1.12 Reporting by public enterprises. The government's Accounting Office receives financial statements from the subnational governments. This includes information on an accrual basis, and on cash transactions. Inaddition, public enterprises are legally required to publishtheir financial audits, but this i s not being done. Part of the problem i s the lack of a legal definition of a public enterprise. For instance, companies of the Solidarity Fund are considered autonomous and subject to the Law of Companies, and therefore are classified as private companies rather than public. Lack of a legal definition of public enterprises also results in overlapping jurisdictions. For example, the electricity companies of the Solidarity Fund, due to their quasi-public/private status, are subject to five different levels of controls: internal audit controls, the Comptroller's Office, the Superintendency of Companies, private external auditors (international), and a so-called Comisario, the functions of which are not well known. 103 Volume I,Annex C M a j o r Budgetary Issues inPriority Social Programs The Institutional Framework and Budgetary Constraints C1.l Following i s a description of the organization of five priority social programs.' It helps explain the institutional weaknesses of the budgetprocessdetailed inAnnex D. C1.2 The Bono de Desarrollo Humano (BDH). The Bono de Desarrollo Humano was originally created in late 1997 on a smaller scale, as the Bono Solidario. While the program started in 1998 with a budget of US$57 million, its budget was increased from US$151 million in 2002 to US$202 million in 2004. The BDH i s now Ecuador's biggest social program. The Central Government has primary responsibility for it, and the number of beneficiaries will rise by 40 percent in 2004, as 12 new provinces gradually become part of the program. (21.3 The BDH has a dual organization: a treasury authority and a technical administrator. The treasury authority i s the Ministry of Finance's responsibility for transferring cash stipends to beneficiaries through the private bank system, BANRED.The technical administrator i s the Ministry of Social Welfare (MSW), which has responsibility for technical administration, development, and supervision and control activities. (21.4 This duality of mainstream responsibilities is setting an undesirable, embittering precedent and creating serious cleavages within the BDH institutional structure. The existing law designated the MSW to be the only entity responsible for the functioning of this and other relatedsubsidy programs. The BDH's Decreto 486-A, of June 7, 2000, was issued to regulate the operation of subsidy programs, while replacing other legal provisions that mandated the Ministry of Finance (MEF) to administer subsidies at first, then passed the responsibility on to the President of the Republic's National Modernization Commission (CONAM), and finally to the MSW. The law instituted the creation of the Social Protection Program (PPS), and charged it with the responsibility to, among other things, "administer and transfer focalized subsidies for development of projects of social compensation directed to most vulnerable population groups." In practice, however, it was the MEF that ended up taking over the treasury and financial management functions, and transgressing the legal mandate, while subordinating the technical management functions to the PPS' 39 staff members. At present, MEF transfers the monies to the beneficiaries and has the power to stop their flowing without prior consent or discussion with PPS. The BDHdoes not have a planning committee that gathers rural women and other key beneficiaries. Rather, that task i s performed by PSS in coordination with the Beneficiaries Selection Service's (SELBEN) specialists in order to verify the actual compared to projected number of beneficiaries. C1.5 A major organizational flaw is that the founding law does not recognize the BDH as an executing unit2and, consequently, funding is channeled by MEFeither through MSW or directly to the service provider (Banred). Usually, MSW receives the funds transferred from Treasury for administrative purposes, and then withholds them. There were serious delays in the functioning of the program. These became critical duringits implementation in 12 Provinces, when Treasury transferred the planned budgetary resources in the third quarter of 2003, but MSW withheld the funds for 124 days, and then only partially released them. InFigure A.C. 1, the current organization shows the dual organization inadministeringandproviding the subsidy through the private banks network. 1. This Annex was prepared by Jorge Shepherd. 2. An executing unit i s any official entity lawfully endowed with an administrative capacity and autonomy, which reports use of public budget funds directly to the Ministry of Finance. Even though the budget of an executing unit is part of the hosting ministry, the ministry will neither be able to control nor intervene in any way in the funds that were approved by Congress and conveyed with the financing sources at the beginning of the fiscal year. 105 Volume I,Annex C Figure A.C.l Organizationof the BDHoperating in14 Provinces C1.6 The BDH has suffered cutbacks. I t s budget in 2003 was originally approved in the MINISTRY OFSOCIAL MINISTRYOF WELFARE FINANCE amount of US$203 million, or 0.8 percent of gross - - domestic product. BDH has reportedly benefited TECHNICAL - 1,043,826 mothers, as well PROGRAM ADMINISTRATOR SUPERVISOR as 226,848 elderly people, and 8,840 handicapped people living inthe first and second quintiles of the population, mainly in rural 3 DIRECCION I SERVICEPROVIDER PROVINCIAL ( B W W areas. The goal in 2003 was I to help 1,025,882 mothers, 244,234 elderly people, and 9,884 handicapped people. BENEFICIARIES BDH faced major financing problems in the second half of 2003. I t s budget was Source: Programa de Proteccion Social (PPS) reduced by 20 percent, to $165 million. Cutbacks were attributed to the changes of authorities in MSW. When-measired by the approved initial budget, BDH executed 80 percent of the social program in 2003 (Table A.C.l). The program i s entirely financed by the Treasury. Table A.C.l Budget Execution at BDH, by FinancingSource (inmillions of US dollars) Total Treasury Fondode Solidaridad WFP Others Initial budget 203.1 203.1 0.0 0.0 0.0 Revised budget 164.6 164.6 0.0 0.0 0.0 Executed (devengadobasis) 161.8 161.8 0.0 0.0 0.0 Executed (cash basis) 161.8 161.8 0.0 1.o 0.0 Source: STFS. C1.7 There is a major problem with the supervision and administration of the BDH. Its expenses, as stated earlier, were repaid late by Treasury and then withheld by MSW. That led the technical administrator (PPS) and universities to go unpaid for a longer time, and paralyzed the regular supervision tasks for the Second Phase. Evidence shows that the Treasury took two months to transfer about US$1 million to the program and, after an additional five months, MSW released only one-third of it, in mid- March 2004. Today, it i s still uncertain when and how much of the remaining two-thirds will be released by MSW.Thisjeopardizes the planning of the program. C1.8 The Plan Ampliado de Inmunizaciones (PAIL Since 1989, the Pan-American Health Organization (PAHO) has made it possible for the PAIto arrange purchases of vaccines and other inputs through a revolving fund PAHO created to help countries in the region eradicate, in a more efficient manner, polio, measles, tuberculosis, and other major diseases affecting children. PAHO helps facilitate good planning of the program by paying in advance for the purchase of vaccines, and getting reimbursed by the government later. Effective and instrumental as PAHO has been, however, it i s clear that the MEF does not reciprocate by reimbursing PAHO in a timely manner for the purchases and, distribution expenses. Table 3.12 shows that MEFtook more than three months on average to repay PAHO for such purchases. 106 Volume I,Annex C C1.9 PA1 is one the finest and most efficient social programs, but is not yet protected by law. Program operations are regulated by the Immunizations Law at the Ministry of Public Health (MPH), and aimto ensure adequate provision and programmingof primary immunizations (immunization campaigns). Institutionally, the program continues to run the risk of government intervention because it is funded entirely by the Treasury. Official funding i s enabled by the Constitution, which states that Ecuadorians are entitled to free access to primary vaccination services. Yet, despite this constitutional mandate, PA1 was not instituted as an executing unit that could otherwise permit protecting its approved budget from future reductions. For the first time, in 2004, the approved budget was cut-and reduced inhalf under the commitment of FEiIREPto provide the other half of financing. This arrangement has still not been agreed inwritingby FEIREPandthe CentralGovernment. C1.10 Another example of good formulation, planning, and supervision when local communities are involved. Formulation and programming of immunization campaigns (nationwide information campaigns on regular or extraordinary vaccination programs supported principally by the government) are prepared in accordance with the basic needs of the population detected in the prevalence and incidence statistics collected through various local surveys. Vaccinations are planned ahead of time and executed by the Ministry's Provincial Branches (Direcciones Provinciales) through the District Branches (Areasde Salud) to the 1,743 hospitals throughout the country. Hospitals include the Health Centers, Health Sub- Centers, Health Posts and providers belonging to the Hospital Matemo Znfantil network. There are also Immunization Brigades, groups of health staff who provide the service in the Amazon region and other remote areas of the country. MPH created a technical administration office to direct the formulation, planning, and supervision activities in coordination with local health providers and communal organizations. There i s active participation by local constituencies in the planning and implementation efforts through the Canton Committees, the Watch Committees, and the Immunization Brigades, which are acknowledged as being key players inthe improvement and vigilance of the services providedby PAI. . - . Infact, PA1isacknowledgedto beone of the most successful social programs, Figure A.C.2 Organizationof PA1Operating Nationwide 1 helping to reduce extreme poverty by eradicating polio 12 years ago and measles 6 years ago. MINISTRY OF HEALTH C1.ll There are no major operating problems identified within service intermediaries and providers. Problems related to unpaid wages in recent months have not led to disruptions in the service. TECHNICAL ADMINISTRATOR At present, 176 of 220 municipalities participate in the provision of vaccination services, and it i s expected that all COMITE DE PROGRAM municipalities will participate in next GESTI6N DJRECCI6N PROVINCIAL DE SUPERVISOR year's nationwide campaign. There are SALUD numerous individual bilateral agencies and non-governmental organizations (NGOs) participating in the verification of the service delivery. Among those are the SERVICE PROVIDER U.S. Agency for International Development (USAID) and Physicians Source: Ministry of Health Without Borders, each working in different regions. In other areas such as technological development and training, there has also been enormous support provided by international cooperation programs and private donors. Among those are the Government of Luxembourg, PAHO, the United Nation's Children's Fund(UNICEF), Petroecuador, and multinational firms such as Glaxo SmithKline Laboratories. 107 Volume I,Annex C C1.12 The PA1 operates well, but is supported with extra-budgetary resources. The initial budget for PA1approved by Congress for 2003 was US$lO.O million, but it was cut to US$9 million (Table A.C.2). According to PAI, however, the actual budget was US$13 million, the difference funded by extra-budgetary resources other than Treasury. Like most social programs, PA1i s funded entirely by the Treasury, which makes its operations vulnerable to changes in projected tax revenues and other urgent cash payments by Treasury. However, to avoid delays in purchases of vaccines, PA1 entered into an agreement with PAHO. This intemational organization set a revolving fund available for all countries in the region to make advance payments to the supplier and thus ensure timely delivery of the vaccines to areas at risk. This has helped significantly in the vaccination planning, and has saved valuable time, considering that it takes 50 to 73 days on average for the purchase order to be processed through the PAHO bureaucracy, and then shipped and cleared incustoms. Table A.C.2 Budget Executionat PAI, by Financing Source(in millions of US dollars) Total Treasury Fondode Solidaridad WFP Others Initial budget 10.0 10.0 0.0 0.0 0.0 Revised budget 9.0 9.0 0.0 0.0 0.0 Executed (devennadobasis) 9.1 9.1 0.0 0.0 0.0 Executed (cash basis) 9.1 9.1 0.0 0.0 0.0 Note: *Devengado i s the executed budget. I t is the sum of budget commitments recognizedby Treasury and supported by documentation certifying the service delivery. Source: STFS. C1.13 The major problem with PA1lies in the funding of the operational expenses to the service providers. Inthat respect, Treasury has transferred the monies to the program, but often the MPHholds resources for unknown reasons. Interviews suggest that MPH owes about US$2 million of the US$11.1 million, inclusive of overdue payments from the previous year, transferred by the Treasury to the program. Yet, despite this problem, budget execution generally has been high, reaching 91 percent of the approved initial budget and 101percent of the revisedbudget. C1.14 Another problem detected with budget execution at PA1 i s the circumvention of external and other financial and technical assistance off the budgetary process. In addition to the US9.1 million that the Treasury transferred through the Ministry of Health, there were extra-budgetary resources inthe amount of US$2.9 millions, that is, 32 percent more than was originally allocated to the program's budget. These are funds donated with the official cooperation of Luxembourg (US$2.7 million), consisting of cold storage rooms and other equipment necessary to improve the local cold storage chain to ensure that adequate vaccinations are available, especially in the remotest areas of the country. Other donations included those from PAHO, UNICEF, Petroecuador, and the vaccine supplier Glaxo Smith KlineLaboratories. C1.15 PAI, nevertheless, was successful in reaching almost 90 percent of the target population nationwide during 2003. A total of 5.1 million vaccines were provided to infants, school-age children, and females aged 10 to 49. The following vaccines were administered: 355,000 doses against Bacillus Calmette-GuCrin (BCG); 552,349 triple doses against diphtheria, tetanus, and polio; 657,110 against pentavalentes (a vaccine that immunizes against diphtheria, tetanus, pertussis, type-B influenza, and hepatitis B); 1,382,750 doses against polio alone; 90,558 doses against hepatitis B alone; 295,201 doses against measles; 518,332 doses against diphtheria and tetanus for children, and 1,085,912 doses against diphtheria and tetanus for women; and 157,882 doses against yellow fever. The goal for 2003 was to administer 5,264,99 1 vaccines to beneficiaries. An evaluation of beneficiaries has been successfully conducted by the Watch Committees and NGOs. C1.16 The Free Maternitv and Infant Care Program (LMG]. The LMG provides health services to pregnant women and to children under age 5, and preventive health services for women, such as screening for cancer, HIV, and other infectious diseases related to women and infants. This i s the only priority 108 VolumeI,Annex C social service administered under executing units and i s administered in an autonomous manner. That is, it has its own executive organization, and its funding i s deposited directly into a special account at the Central Bank of Ecuador, without any interference from the Ministry of Health. C1.17 The program has an organizationwith ample grassroots participationinvolving key official institutions and special-interest groups. There is a top executive board (the Support and Monitoring Committee) chaired by the First Lady, as head of the National Institute for Children and the Family, and includes the MPH, the Executive Director of the National Women's Committee (CONAMU), and the President of the National Association of Municipalities (AME),among others. The Board appoints the Technical Administrator, the Executive Committee, and the Local Management and Monitoring Committees. Technical Figure A.C.3. Organizationof LMGOperating in40 Municipios administration resides in I I the Executing Unit of the Free Maternity and Infant MINISTRY OF HEALTH EXECUTIVE Care Program (UEPMGAI), which was created in mid-2002 to provide technical, financial, supervisory, and (UEPMGAI) training services to social COMITEDE COMITE providers nationwide. The - GESTION EJECUTIVO Executive Committee comprises representatives ~ of other local and DIRECCI6N international agencies. PROVINCIAL DE ALCALDIA SALUD Finally, the Local Management and Monitoring Committee comprises representatives of health intermediaries SERVICEPROVIDER and providers, women, and indigenous organizations, and local governments. I Service providers are all hospitals belonging to the BENEFICIARIES MPHand the LMG system (Figure A.C.3). Source: Programa de Maternidad Gratuita y Atencion a la Infancia (LMG) C1.18 In 2003, LMG reached a women and infant population coverage ratio of 16 percent within 40 of the 220 municipalities participatingin a pilot maternal and infantcare program. Incontrast with other social programs, LMG i s enabled by its charter to get the municipalities and other local organizations involved inthe development and provision of the social service. The Solidarity Fundprovides most of the funding for the program, but financial arrangements with this agency are subject to annual revision and renewal. According to LMG data, maternal care in rural areas expanded its coverage significantly in 2003: 90 percent of pregnant women received first and subsequent pre-delivery tests, and three of five pregnant women even received dental care services. However, only very low numbers of pregnant women received HIV tests and sexually transmitted disease exams. C1.19 The LMG was created by a law, and is supported by a broad-based structure of powerful women and indigenous organizations, which gives it the capacity not only to protect and execute the budget as initially planned, but also to retrieve large overdue funds not paid in the previous year. 109 Volume I,Annex C As a result, the program was able to raise 28 percent more of budget resources than initially planned in 2003, as Treasury fundingwas doubled to US$10.4 million, from US$4.8 million. In2003, LMGreported a budget execution of 120 percent, when measured by the approved initial budget of US$19.9 million. The program's budget execution declined to 94 percent when measured by the revised increased budget of US$25.5 million (Table A.C.3). This i s the only social programthe initial budget of which grew, and considerably, whereas the revisedbudgets (codificado)of other programs mostly declined. TableA.C.3 BudgetExecutionat LMG, by FinancingSource (inmillionsof US dollars) Total Treasury Fondode Solidaridad WFP Others Initialbudget 19.9 4.8 15.1 0.0 0.0 Revisedbudget 25.5 10.4 15.1 0.0 0.0 Executed(devengadobasis) 23.9 8.9 15.0 0.0 0.0 Executed(cash basis) 22.1 7.1 15.0 0.0 0.0 Source: STFS. C1.20 LMG i s funded mostly by the SolidarityFund (59 percent) and by earmarkedtax revenues earmarked fund represents the 3 percent of the Special Consumption Tax . As mentioned, the program (41 percent). The Solidarity Fund provides a fixed amount every year (US$15.1 million), whereas the operates as an executing unit within the MPH. This provides the advantage of getting all budgetary resources deposited directly by Treasury and the Fondo de Solidaridad into a special account at the Central Bank of Ecuador. In principle, this prohibits the MPH from using these resources for other purposes. Once the funds are deposited in the program's main account they are transferred by the program, according to its own evaluation guidelines and controls, to the service providers' individual bank accounts within the Provinces, but through the program's intermediaries (the Management Committee). C1.21 The Food and CommunitvDevelovment Prom~m(PRADECL A reformed social program with an entrepreneurial vision, PRADEC provides food security services to children aged 2 to 6, and to handicapped and elderly people. In addition, it provides technical assistance in the development of ' productive activities that can improve the income and consumption of rural households. PRADEC has undergone a major organizational change, starting to reshape its managerial and monitoring strategies through more vigorous community participation and more efficient use of financial resources. Compared to previous years, supervision and evaluation of the service are now monitored more closely. 110 Volume I,Annex C C1.22 Like the PAI, PRADEC conducts its financial management Figure A.C.4. Organizationof PRADEC Operating in and food purchases through an 443 Juntas Parroquiales arrangement with an international cooperation agency. Inthis case, the UNWorld Food Program (WFP) has MINISTRY OF SOCIALWELFARE valuable expertise in the financial FINANCIAL management and provision of - MANAGER PRADEC's food and its connection with multinational food suppliers. TECHNICAL The financial manager agrees to ADMINISTRATOR execute the funding arrangement with Treasury, and the purchaser agrees to COMITE DE advance the purchases of food, and PROGRAM then get reimbursed by the GESTIdN COMITE JUNTAS SUPERVISOR Government of Ecuador. This PARROQUIALES facilitates better planning and helps avoid major disruptions in times of SERVICEPROVIDER liquidity stress at Treasury. Another (COMEDOR reason for PRADEC's performance i s COMUNITARIO) its revamped internal organization and the active participation of local communities. Service providers are the community cafeterias (comedores BENEFICIARIES comunitarios), facilitated by the Ministry of Social Welfare in Source: Programade Alimentacion y Desarrollo Comunitario (PRADEC) coordination with the parish committees and local communities. There are practically no intermediaries because food deliveries are made directly by the financial manager through a chain of distributors that i s regulated by internal rules, and are executed by the program's technical administrator (Figure A.C.4). C1.23 PRADEC suffered a heavy budget cut in2003. Treasury, its sole financial source, reduced its funding due to emerging conflicts over the reporting of expenses. As a result, budget execution was 57 percent when measured by the approved initial budget, and 70 percent when measured by the revised codificado budget (Table A.C.4). The program, however, progressed rapidly as the role of the parish committees evolved positively by enhancing the supervision and control functions through a system of inspectors. In spite of the budget reduction, the program was able to meet the service goals within the first and second quintiles of the population, and the food supplies were purchased and delivered in a timely fashion. C1.24 The School Nutrition Program (PAEL PAE, like PRADEC and P A " are part of the Integrated System of Children Care-known as SIAN). It is the second biggest social program in Ecuador. It provided lunch services in the Sierra Region to nearly 520,OO children aged 5 to 14 (120 days) during 2003. In the Coastal Region, lunch services were provided to 751,710 students and breakfast to 371,821 students, for 40 days each. The goal for 2003 was to provide lunch services to 1,450,000 during a minimumof 160 days. Over the years, PAE has turned out to be the most ambitious of all food programs inEcuador, aimed at providing both breakfast and lunch supplements to school-age children for a period of 160 days in the Coast and the Sierra. Compared to other countries, this program i s much larger, because only breakfast i s provided by most countries (such as the Glass of Milk Program in Peru). As a result, the budget increased from US$17.5 million in 2001 to US$30.7 million in 2003. Yet, the budget is not protected against possible reductions by the Central Government, the Solidarity Fund, or other parties. 111 Volume I,Annex C allows steady execution of MINISTRY OF MINISTRY OF the budget process and EDUCATION proper financial arrange- -- FINANCIAL MANAGER ments, particularly with the Solidarity Fund, is a major TECHNICAL. problem. In 2003, Congress ADMINISTRATOR approved an initial budget of GESTI6N DIRECCIbN SUPERVISOR PROVINCIAL Later, however, it was cut by EDUCATNA 45 percent because the 112 Volume I,Annex C Table A.C.5 Budget Executionat PAE, by FinancingSource (inmillions of US dollars) Total Treasury Fondo de Solidaridad WFP Others ~~ Initial budget 30.7 16.0 12.0 2.7 0.0 Revised budget 17.0 16.0 0.0 1.o 0.0 Executed (devengado basis) 14.2 13.2 0.0 1.o 0.0 Executed (cash basis) 15.2 14.2 0.0 1.o 0.0 Source: STFS. C1.28 PAE also suffers from a severe pro-poor targeting problem. In 2003, among all programs belonging to SIAN, PAE had the poorest targeting ratio in quintiles 1 and 2. Whereas P A " and PRADEC reached 67 and 100 percent of their intended beneficiaries, PAE barely reached 60 percent. Hence, close to half of their intended beneficiaries were not reached by this program. C1.29 Poor targeting is also compounded by an inadequate selection of its main beneficiaries. According to studies developed by SIAN, the nutritional impact of PAE on its intended beneficiaries, children from 6 to 15 years, i s almost nil, and this explains why SIAN i s gradually shifting food-aid resources toward children inthe 0-5 years range, where malnutrition rates are highest andneeds strongest. C1.30 Other issues are the slow reporting ofjustiJicaciones3to MEF, and the bureaucratic delays in payments. As stated, the financial management of the program is conducted by UNDP and WFP, which also subcontract and purchase the food supplies. These international agencies pay in advance for the food items to the suppliers, and then are reimbursed by the MSW. Given delays in MEF transfers, this methodhas provenvery useful also enabling the programto operate with transparency. 3. Justi'caciones are the documented explanations on the use given to public funds transferred by the Treasury to the social programs that are required to justify the next payment. 113 Volume I,Annex D Budgetary Framework of Decentralization inEcuador D1.1 Since 1997 Ecuador has developed into an increasingly decentralized unitary state.' After repeated failures in the preceding years, subnational government successfully gained access to additional national tax and oil revenue. The centerpiece of this fiscal bargain i s the "15-Percent Law," which transfers a portion of Central Government revenue to municipalities and provincial councils in a revenue- sharing arrangement. As a result of the implementation of this law, the amount of all 17 intergovernmental grants has increased significantly: while in 1996 transfers represented 1.2 percent of gross domestic product (GDP), this share rose to 2.7 percent in 2003,2 and in terms of current Central Government revenue, the share increased from 8.4 percent in 1996 to 16.5 percent in 2003 (Figure A.D.1). From an international perspective, this represents a fairly advanced degree of fiscal decentralization. Similar degrees of transfers are employed, for instance, in Bolivia and Vene~uela.~ Both in depth and speed, in recent years the decentralization process in Ecuador has advanced significantly-at least on the revenue side. Figure A.D.l Intergovernmental Transfers, 1996-2003 2.7 3.00 / 2.50 2.00 0TotalTransfers a in%of Central 1.50 n Government 8 Revenue 1.00 0.50 +Total Transfers + in%of GDP axis) 0.00 2, q+ +P +iO ++ +s" $$+\ %$+? >$P Years Source: Author's calculations basedon informationprovided by MEF(2003) and Central Bank (2003). D1.2 Much less pronounced was decentralization on the expenditure side. With a few exceptions, transfers of responsibilities inthe health, education, and roads sectors do not occur on a large scale or ina consistent manner, and only a few municipalities have adopted new responsibilities in those sectors. To break this inertia, during 2001 and 2002 a large-scale bargaining process took place for the transfer of expenditure responsibilities in the environment, agriculture, roads, and tourism sectors. Yet this attempt also ended indeadlock. Although several agreements were signed among Central Government ministries, 1. ThisAnnex was preparedby JonasFrank. 2. This is still much less than the 6 percent GDP that federal Argentina makes available to its provinces; however, they are responsible for health and education services, and the fiscal problems Argentina faced due to a deficient intergovernmental setup speak for themselves. 3. Both Bolivia and Venezuela also transfer roughly 20 percent o f Central Government revenue to subnational governments. This, however, is not a degree o f fiscal decentralization comparable to Colombia, which makes roughly 45 percent of Central Government revenue available to departmental and municipal governments. However, Ecuador i s far more advanced than many Central American countries, where the share o f transfers usually does not exceed 5 percent. During the 1980s and early 199Os, Ecuador transferred roughly 9 percent o f current revenue to subnational levels. 115 Volume I,Annex D provincial councils, and municipalities? they were not applied in practice. Expenditures today remain as centralized as they were inthe 1970s, when the oil boom led to unparalleled public sector growth. So far, decentralization of expenditures has fallen short of expectations. D1.3 This particular type of decentralization came at a cost, however-a cost that is only beginning to emerge and become fully apparent. The sudden increase intransfers of revenue was not followed by a commensurate transfer of expenditure responsibilities. Municipalities and provincial councils-as well as their associations, AME and the Consorcio de Consejos Provinciales del Ecuador (CONCOPE), respectively-have repeatedly and expressly refused to link transferred revenue to the explicit adoption of new expenditure responsibilities, and the center has so far been unable to convince local politicians of the need to overhaul the fiscal framework in order to regain the much-needed budget flexibility-a condition to sustain dollarization. D1.4 This has putadditionalfiscal stress on the nationalgovernment. As shown above, never were transfers higher in terms of GDP, and never was a higher share of Central Government revenue transferred to subnational government than in 2003. Inthe medium to long term, this implies a severe sustainability problem for the national level: while the real average GDP growth during 1996-2003 was 1.8 percent, and Central Government revenue grew an average 3.4 percent, the increase in transfers during the same period was 14 percent without any further accountability in return. In other words, intergovernmental grants have grown over-proportionately relative to the growth of the economy and the revenue the nation was able to raise. To comply with the amount of intergovernmentalgrants, the center has sacrificed budget flexibility on one hand, and fiscal responsibility on the other. D1.5 Some of the dilemmas originate inthe way transfers are administered. A particular feature i s earmarking of revenue. This type of management comes in two forms: (a) for transfers in general revenue-sharing arrangements, which are applied, among others, for the 15-Percent Law: and (b) for transfers that tap resources before entering the national budget ("off-budget assignments"). While the first form of earmarking provides subnational government with a transparent form of financing, the second form i s worrisome since it establishes discretionary property over some sources of revenue. Thus, subnational governments receive earmarked revenue from, among other sources, the income tax, oil revenue, the vehicle tax, surpluses of public enterprises, import taxes, and a tax on domestic credit operations.6 Off-budget earmarkingi s a threat to budget flexibility and undermines transparency. D1.6 Although negative from an overall perspective, earmarking is appealing for subnational governments. Not only does it protect financing in an environment of competing claims on scarce resources, but, more important, earmarking acts like a countercyclical revenue policy tool for municipalities and provincial councils. Recipients of earmarked funds usually receive funds in a timely, stable, predictable way. In contrast, recipients of allocated non-earmarked resources from the national budget suffer inparticular duringtimes of economic crisis. This is evidenced by the crisis years 1999and 2000, when transfers from earmarked resources remained stable, and hence their share in the amount of total transfers increased significantly (Figure A.D.2). In a volatile economic environment, as often prevails in Ecuador, earmarking fulfills an important stabilizing function for subnational govemments- despite its negative impact on national public finances. 4. In total 140 of 220 municipalities, and all 22 provincial councils, formally participated in this bargaining process. 5. The establishment of the Fondo de Desarrollo Seccional (FODESEC) in 1990 represents the first regular revenue-sharing arrangement in Ecuador, and hence a break from the past and the practice of transferring extrabudgetaryfunds. 6. This is channeledthrough the FODESECFund. 116 Volume I,Annex D Figure A.D.2 Sourcesof Revenuefor Transfers to SubnationalGovernments(inpercent), 1991- 2002 100% 90% 00% OOther Revenue 70% 60% 0IncomeTax 50% 40% Oil 30% General Budget (Tax) 20% Revenue 10% 0% Years Note: "Other Revenue" includes transfers from Central Govemment agencies (Contraloria, among others) that are transferred separately anddirectly from different agencies. The beneficiaries of earmarkedfunds obtain a share of the vehicle tax, a share of the surplus o f public enterprises, a share o f import taxes, and a share of taxes of domestic credit operations. GeneralBudget(Tax) Revenueincludesa share of oil pertainingto the 15-Percent Law (1997onward). Source: Author's calculationbasedon DevelopmentState Bank (BEDE) (Sistem de Informacibn Municipal [2001]) and (unpublished)information providedby the Ministry of Economy andFinance (the yearlyAcuerdos,which define the sourcesand amountsfor transfers to subnational govemments). D1.7 Consequently, earmarking has historicallybeenhighduringeconomic crises. This has ledto an inefficiently highlevel of earmarking. The rationale behind this i s that social groups inEcuador were trapped into a social dilemma: all of them could have been better off if they reneged unilaterally on earmarking. But a unilateral restraint would have benefited all others, a perverse incentive mechanism that prevented them from achieving collectively superior outcomes. Local politicians and the different provinceswere trappedintothis very dilemma. D1.8 While the recent trendinfiscal decentralizationcame at a cost for the CentralGovernment, the level of executed spending and service delivery efficiency at the local levels has been underminedfrom at least four perspectives. First, the intergovernmental transfer systemestablishes a fragmented and contradictory incentive framework that sends confusing signals to subnational governments. Transfers are based on 17 different grants. Each of these grants establishes different criteria for resource distribution (see Table A.3 in the Annex), thereby creating a complex web of rules and regulations for the use of funds. Although there is no earmarking for specific expenditures-for instance health and education-there are multiple rules limiting their use for current and investment spending. The 15-PercentLaw, for instance, mandates that at least 90 percent of these resourcesmust be used for investment purposes, and 10 percent for current spending. However, in practice, current expenditures take a much higher share of municipal and provincial government budgets: in 2003 some 56.4 percent of expenditures were destined to wages and salaries, and goods and services. Only 43.6 117 Volume I,Annex D percent were usedfor proper investmentpurposes(Table A.D. 1). Yet monitoring of and sanctions7on the use of resourcesremain cumbersome, andare not appliedregularly.' Table A.D.l Current and CapitalSpendingof Municipalities and ProvincialCouncils(inthousand USD),2003 Total Expenditure 1,102 Current Expenditure 621 Wages and salaries 169 Purchasesof goods and services 352 Interestpayments 3 External 3 Current transfers 97 To IESS 10 To other public sector 27 To private sector 60 Capital Expenditure 480 Fixedcapital expenditure 480 Other 0 Source:MEF. D1.9 Municipalities and provincial councils have taken advantage of an underregulated subnationalsector. Inan environment where the national level has failed to signal spendingpriorities to local government, they have quite naturally developedtheir own expenditurepatterns and priorities. The lion's share of subnational expenditures is destined to providing two services: urban streets and roads (32.9 percent) as well as anumberof non-specified public works (Table A.D.2). The importanceof these two sectors is contrasting with expendituresthat are funded only by a minor share of municipal budgets. Ecuadormustdecide whether these localspendingpurposesarejustified when lookingat the highlevelof transfers-particularly low are investmentsinthe education (3.6 percent) and health(0.5 percent) sectors, yet it is these services that traditionally require a high share of Central Government grants. This i s particularly so because of the positive externalities associated with these services. In other words, transfers compensate subnational government for producing benefits that spill over into other local jurisdictions. A highlevelof transfers would precisely constitute an appropriateincentive framework not only to deal with positive externalities, but also to give the Central Government the leverage to redirect subnational expenditures into sectors it feels provide most value added to communities and socioeconomic development. Transfers inthe formof earmarking, matching grants, and other means, are specific models of fiscal managementthat could be usedto achieve this goal. None of these, however, has beenapplied inEcuadorina consistentmanner. D1.10 Second, transfers have weakened accountability relationships with local citizens and users of services. Because of the highlevel of transfers, mayors and prefects do not necessarily needto tax their citizens, and hence become accountableto the center. Local own revenue did not grow proportionate to the level of transfers. In fact, between 1997 and 2002, the share of own revenue available to local government declined (Figure A.D.3). This causedan increasein vertical fiscal imbalances, and a higher share of subnational expenditures i s now financed by the center through transfers. Yet given the high 7. Eachyear roughly 150 of 220 municipalities are notified or alertedby the Comptroller's Office for inappropriate budget management practices. Most of them are able to justify budget management ex post, but a small number of municipalities are subject tojudicial processes. 8. However, monitoring of classifications for current and capital investment is always difficult to establish. In addition, the Ley Organica de Administracidn Financiera y Control (LOAFYC) states that municipalities can by request to the Ministry of Economy and Finance amend their budget classifications so that current spending is classified as investment expenditures, by, for instance, listing "recursoshumanospara inversion." 118 Volume I,Annex D proportion of transfers in local portfolios, mayors and prefects "do not feel the pain of the marginal tax dollar" they spend. This, in tum, does not create incentives for efficient, cost-effective, and fiscally sustainableservice delivery. Table A.D.2 Priorities of SubnationalInvestmentSpending, 2002 Service and Sector Amount of Investment inUS$ million InPercent of Total Investment UrbanPlanning and Cadastre 0.3 0.1 Busterminal 0.4 0.1 Slaughterhouses 0.5 0.1 Cemetery 0.8 0.2 Solid waste 1.7 0.3 Health 3.1 0.5 Markets 5.6 1.o Community Works 10.0 1.7 Recreation 16.6 2.9 Education 20.6 3.6 Water 29.8 5.2 Sewage 35.7 6.2 Streets androads 189.9 32.9 Other public works 261.9 45.4 577.0 100.0 Source: Author's calculationsbasedoninformationprovidedby MEF(2004). Figure A.D.3 Own Revenueand Expendituresper Levelof Government (inpercent) and Vertical FiscalImbalances, 1997 and 2002 100% 90% 80% 70% 0Municipalities 60% 50% Provincial Councils 40% 30% f3Central government 20% 10% 0% Own Expenditures Own Expenditures Revenue Revenue 1997 2002 Note: Incomplete reporting on subnational governments for 2002 (mostly on own revenue); in particular larger provincial councilshavenotreportedtheir ownrevenue, includingGuayasandManabi. Sources: Author's calculation,basedoninformationprovidedby MEF (2004) andCentralBank (2002). 119 VolumeI,Annex D D1.11 Third, the particular purposeof intergovernmentaltransfers remains vague. The criteria that are used for compensation do not clearly focus on either fiscal capacity or expenditure needs. By putting weight on Unsatisfied Basic Needs (NBI) andpopulation, transfers explicitly compensatefor neither low fiscal capacity nor expenditureneeds. The currenttransfer systemi s a hybridbetweenthese two options. This is not to say that transfers do not compensate for either low fiscal capacity or spending needs at all, butaclearerfocus couldestablishamorecoherentincentiveframework for subnationalgovernments. D1.12 Fourth,thetransfer systemrewardsthe inefficient distribution ofresponsibilities that existsacross levels of government. No clear delineation of responsibilities has emerged among municipalities, provincial councils, and the Central Government, and the large number of implementing agencies that exist. The.dual political structure on each subnational level, whereby elected and delegated authority coexist, is a major obstacleto accountability.' A series of autonomous agencies exist inparallel to sector ministries and subnational governments. Regional development authorities," with budgets that often exceed those of municipalities and provincial councils, execute many responsibilities that could be decentralized. This includes responsibility over ruralroads, secondaryroads, school buildings,hospitals, rural development programs, water and irrigation systems, and electrification programs, among others." Whereas municipalities are responsiblefor many urban services-streets, water and sewerage, electricity, markets-no clear delineation regarding provincial councils has emerged. This i s compounded by the fact that thejurisdiction of provincial councils is ambiguouswith regardto urbanandruralareas. Though it is constitutionally mandatedthat provincial councils are responsible for rural areas, they often provide services in urban areas. If involved in urban areas at all, the involvement i s based on informal arrangements and personal understandings with municipal authorities. The transfer system, however, rewards this inefficient institutional setup because it allows subnational government to pay for duplications andoverlapsinresponsibilities. D1.13 While the fiscal decentralization framework is questionable from many perspectives, ironically it may constitute a favorable framework for some services, particularly in the social sectors. Public services inhealth and educationtypically require a highlevel of transfers to compensate for positive externalities that are associated with service delivery, and that spill over into other local jurisdictions. Grants made available by the Central Government usually provide incentives for municipalities and provincial councils to spend in those sectors, even though the benefits-in terms of higher educational attainments (education) or disease control (public h e a l t h w o not accrue only to the population residinginthose areas. Yet the misfortune hereinis that the involvement of municipalities and provincial councils in these sectors is, as yet, minimal: both subnational governments spend only 0.5 percent of their investment budget on health, and 3.6 percent on education. In other words, less than 5 percent of subnational investment expenditures are destined to the social sectors (Table A.D.3). The challenge remains for local politicians to venture into these new, largely unexplored areas of service delivery. 9. Dual authoritiesexist on each subnational level: on the provincial level (appointed Governor and electedPrefect), on the cantonal level (appointed teniente politico and elected mayor), and on the parish level (appointed Jefe Politico and elected Presidente de la Junta Parroguial Rural). The Constitutional Reform of 1998 eliminates both the TenientePolitico andthe Jefe Politico, butthey still exist inpractice. 10. The more important regional development authorities include Centro de Reconversibn Econbmica del Azuay (CREA), Ceiitro de Desarrollo de la Cuenca del Guayas (CEDEGE), Centro de Reconversibn del Manabi (CRM), and Unidad de Desarrollo del Norte (UDENOR). 11. No data are available with regard to sectoral allocation of investments carried out by regional development authorities. 120 Volume I,Annex D Table A.D.3 Capital Spending of Municipalities and Provincial Councils inthe Health and Education Sectors, 2002 Health Education TotalInvestment Regions InUS$ million InPercent InUS$ million InPercent InUS$ million InPercent Costa 1.4 45.3 4.3 21.1 300.2 43.3 Oriente 0.6 19.6 5.8 28.4 87.9 12.7 Sierra 1.1 35.1 10.3 50.5 305.4 44.0 Total 3.1 100.0 20.4 100.0 693.5 100.0 Note: Costaincludesthe GalhpagosIslands. Source: Author's calculation,based on MEF (2004). D1.14 This picture is corroborated by taking a closer look at the education and health sectors separately. In the education sector, municipal involvement i s stronger in the pre-primary and primary levels compared to the secondary level. However, it remains very limited overall. In the education sector-at the pre-primary, primary, and secondary levels--only a minor share of teachers (0.4 percent of the total number of teachers), schools (0.4 percent of the total number of schools), and students (0.5 percent of the total number of students) are administered by municipalities (Table A.D.4). Both the Central Government and a large number of semi-autonomous or private entities deliver most of the country's educational services. To date, Ecuador's education system has either not taken advantage of municipalities, or has selected other service delivery models that emphasize deconcentration and school autonomy as a form of strengthening service providers directly.12 D1.15 This situation is similar in the health sector. Less than 1 percent of total public health spending i s executed by municipalities. Given their population size, the Quito and Guayaquil municipalities are outliers in terms the budget destined to the health sector. Their involvement far exceeds those of any other municipal government. Expenditures remain heavily centralized in the Ministry of Public Health, or are administered by a wide range of other autonomous or semi-autonomous institutions (Table A.D.5). As yet, however, and with few exceptions, no consistent involvement of subnational governments inthe health sector has developed. D1.16 Aligning increased transfers with expenditure responsibilities remains the most important challenge on the road to decentralized government, in particular in the social sectors. A future decentralization process in the social sectors would have a considerable impact on both national and subnational budgets. Municipalities and provincial councils and their associations have repeatedly refused to adopt new expenditure responsibilities commensurate with the new funds that are made available. Therefore, they mistakenly claim that any new responsibilities would need to be financed with additional transfers. This, under the present framework, would further deteriorate vertical fiscal imbalances and dependency rates of subnational levels of government. It would also most likely imply that the share of local current spending would increase, to come up with the needed wages for teachers, nurses, and doctors. In the education sector, in particular, improved service delivery may require emphasis on current as opposed to investment expenditures. Consequently, possibilities for fiscal adjustment on subnational levels may even further deteriorate as spending becomes more rigid. A deepening of decentralization-for example, inthe social sectors-within the current institutional setup i s therefore possible only ifsubnational governments first make optimal use of increased current transfers to finance new expenditure responsibilities. Only when they become supported by results-orientedand well- monitored mechanisms, and when fiscal space allows them to free additional resources, should 12. This includes models, such as Redes Amigas, supported by the international donor community (Inter-American Development Bank, among others; and previously, the World Bank). 121 VolumeI,Annex D Ecuadorian municipalities and provincial councils be exposed to an additional transfer push within sustainable levels. Table A.D.4 Administration of Schools, Teachers, and Students inthe Education Sector, School Year 2000-01 Education Administration Schools Teachers Students Level Number Percent Number Percent Number Percent Pre-primary Central Government 2,694 52.8 5,020 36.5 109,685 55.0 schooling Municipal Government 17 0.3 58 0.4 915 0.5 Other (private or semi- 2,388 46.8 8,677 63.2 88,988 44.6 autonomous schools) Total 5,099 100.0 13,755 100.0 199,588 100.0 Primary Central Government 14,004 77.7 55,472 65.5 1,436,124 73.5 schooling Municipal Govemment 74 0.4 455 0.5 10,063 0.5 Other (private or semi- 3,936 21.9 28,831 34.0 508,873 26.0 autonomous schools) Total 18,014 100.0 84,758 100.0 1,955,060 100.0 Secondary Central Government 1,938 55.8 51,988 64.1 662,666 69.2 schooling Municipal Government 7 0.2 178 0.2 3,134 0.3 Other (private or semi- 1,529 44.0 28,905 35.7 292,261 30.5 autonomous schools) Total 3,474 100.0 81,071 100.0 958,061 100.0 All education Central Government 18,636 70.1 112,480 62.6 2,208,475 71.0 levels Municipal Government 98 0.4 691 0.4 14,112 0.5 Other (private or semi- 7,853 29.5 66,413 37.0 890,122 28.6 autonomous schools) Total 26,587 100.0 179,584 100.0 3,112,709 100.0 Note: "Other" school types include escuelasfiscomisionales and clerical schools. Central Govemment schools includeescuelas fiscales. Secondary education equivalent to nivel medio (ciclo bcisico, post ciclo bdsico, bachillerato en ciencias, bachillerato te`cnico,postbachillerato). Source: Author's calculation based on information provided by the Sistema Nacional de Estadisticas Educativas del Ecuador (SINEC). Table A.D.5 Public Spending inthe HealthSector (inthousand USD),1997 TotalSpending Percentof TotalSpending Ministry of Health 176.3 47.3 Quito Municipality 0.6 0.2 Guayaquil Municipality 1.1 0.3 Other Municipalities 0.3 0.1 Other Institutions* 194.5 52.1 Total 372.8 100.0 Note: "Other institutions" include the Servicio de Erradicacidn de Malaria (SNEM), Instituto Nacional de Higiene * (INH), Centro Estatal de Medicamentos e Insumos MCdicos (CEMEIM), Patronato San JosC, universities, military, police, Seguro Social Campesino, Social Security Institute (IESS), and Military Social Security Institute (ISSFA). N o updated and more recent information exists. Source: Author's calculationsbasedon CuentasNacionales de Salud (1997:17). 122 Volume I,Annex E Alternatives for Cooking-Gas Subsidy Reform' Introduction El.1 The objective of this Annex is to promote the discussion of the cooking-gas subsidy reform in Ecuador. It i s divided in two parts. The first part presents a series of results on alternative reform scenarios, which stems from the World Bank's Ecuador Poverty Assessment. The second part evaluates the various mechanisms that could be used to implement reform. Diagnosis, Targeting Options and Impact E1.2 The gas subsidy, in its current format, i s universal (every household in the country has a right to this subsidy). Consequently, and given the fact that gas consumption is higher inricher households than in poorer ones, the subsidy follows a regressive distribution (richer households benefit relatively more from the subsidy than poor households (Table A.E.1). Table A.E.l Subsidy DistributionBy ConsumptionDeciles Consumption Percentageof Total Percentageof Total Deciles Consumption SpendinginSubsidy 1(poorer) 2 3 2 3 5 3 4 7 4 5 8 5 6 9 6 7 10 7 9 12 8 11 13 9 16 16 10(richer) 37 17 Source: Vos et alia (2003). E1.3 Indeed, most of the subsidy i s distributed to white and Mestizo households, while only a small portion reaches indigenous and Afro-Ecuadorian homes, since the latter are mostly poor and exhibit a low consumption of gas. This does not mean, however, that the subsidy i s not an important portion of these households' consumption, which explains the strong resistance to the reformor elimination of the subsidy within the indigenous movement, for example (Table A.E.2). E1.4 The subsidy i s not only targeted inadequately, but it i s also expensive. It cost the government of Ecuador approximately 0.9 percent of the GNP per year. All of the above points to a need to reform. Improving its targeting is an option, so as to provide more benefits for the poor, or considering its complete elimination i s another option. E1.5 A series of simulations were conducted with the purpose of evaluating the impact of various reform alternatives and/or eliminating the cooking gas subsidy. Following i s a description of the methodology used and a summary of findings. ~ 1. This Annex was prepared by Carolina SBnchez-PBramo. 123 Volume I,Annex E Table A.E.2 Distributionof the Cooking-GasSubsidy By EthnicGroup Ethnic Group Number ofpeople Percentageof Total Percentageof Total UsingGas Indigenous 824,189 6.8 42.9 Afro-Ecuadorian 267,196 2.2 89.5 Other 1,257,466 10.45 92.9 Source: Authors' calculations using the Population Census 2001. E1.6 Simulations are based on data obtained from the Encuesta de Condiciones de Vida 1999 (LSMS Survey 1999), which assumes a subsidy equal to US$3.4/liter (the difference between market price- US$S/liter-and the price consumers paid-US$l.6/liter-at the time these calculations were made). Although these amounts might have changed over the past months, rendering the estimated fiscal savings a mere approximation, the exercise continues to be informative with regard to the impact caused by the various policy options being considered. 0 Scenario 1: Current Situation 0 Scenario 2: Complete elimination of subsidy along with an equal drop in the price of total consumption inthe household 0 Scenario 3: Complete elimination of subsidy along with a distribution of generatedfiscal savings to poor households, inthe form of transfers 0 Scenario 4: Complete elimination of subsidy along with a distribution of generated fiscal savings to poor households in SelBen's quintiles 1and 2 0 Scenario 5: Retargeting subsidy to households in the System of Selection of Beneficiaries' (SelBen's) quintiles 1and 2 E1.7 The economic impact i s measured based on two variables: fiscal savings generated by the reform, and the effect of the latter on the monetary poverty rate and gap. E1.8 Under these assumptions, Scenarios 2 and 5 generate the largest fiscal savings (100 and 76.3 percent, respectively), and Scenario 3 reaches the lowest poverty rate, followed by the status quo (Table A.E.3). TableA.E.3 Effect of Various ReformScenariosfor the Cooking-GasSubsidy FiscalSavings Poverty Poverty (percentof subsidy) Rate Gap Scenario 1 0.0 49.7 19.3 Scenario 2 100.0 51.0 20.3 Scenario 3 0.0 47.0 17.4 Scenario 4 0.0 50.3 18.5 Scenario 5 76.3 51.0 19.7 Source: Authors' calculations using the LSMS Survey "Encuesta de Condiciones de Vida 1999" and the Population Census 2001. E1.9 While fiscal savings results are the obvious consequenceofthe described mechanisms, the impact of the different monetary poverty rate scenariosrequires a more elaborate explanation. 2. To carry on estimations, and with the purpose of simplifying some of the calculations needed for the simulations, it is assumed that: (i) are no changes in household consumption patterns in response to changes in (the there amounts of) subsidy (partial balance analysis); and (ii)the poverty line remains constant for all scenarios. The poverty line used is US$1.3 per person per day. 124 Volume I,Annex E E1.10 First, it is important to note that, although there is a correlation between monetary poverty and poverty measured according to the SelBen Index (quintiles 1 and 2), both rates are not the same, since they assess different dimensions of wellbeing. Monetary poverty reflects the lack in a household's income level and/or consumption, while SelBen's poverty measures structural scarcity (e.g., access to sanitation, electricity, and other services). El.ll Second, as a result, and given the fact that the analysis measures the impact of reforms on monetary poverty (and not on poverty measured by SelBen), the scenarios considering redistribution or retargeting mechanisms based on monetary poverty (Scenario 3) will have a greater effect on said poverty than those usingthe SelBen (Scenarios 4 and 5). E1.12 This does not mean that the SelBen is an inadequate tool for targeting. On the contrary, usingthis index provides two important advantages: e Transparency: The use of the SelBen index as a targeting instrumenthas contributed to its positioning as the main tool to select beneficiaries in government social programs-thus increasing transparency inthe distribution of resourcesallocatedto social assistance Cyclical Variation and Leakage: Given its nature, the SelBen Index i s less susceptible to cyclical changes in the economy than monetary poverty. This helps minimize any leakage created by cyclical changes in income and consumption. Inaddition, SelBen i s a less costly tool, as it does not require constant updates. E1.13 Given these arguments infavor of usingthe SelBen, it would be advisable to consider scenarios 4 and 5 as top choices. And, between those two, scenario 5 should have preference given its capacity to generate larger fiscal savings. Figure A.E.l shows the effect of applying scenario 5 to benefit distribution. It i s clear that this distribution would be significantly more progressive and pro-poor than the current one. Figure A.E.l Distribution of the Gas Subsidy Retargeted with SelBen I 09 0.8 0.7 B 0.6 ' 0.5 -Subsidy According to Selben E 45 degree 0.4 0.3 0 2 0.1 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Cumin Population Source: Author's calculation usingECV 1999. 125 Volume I,Annex E PossibleImplications of Offsetting the Retargetting of the Cooking-Gas with a BDHIncrease El.14 The Government of Ecuador has considered several alternatives for the cooking-gas subsidy reform, along with a possible increase in the Human Development Bonus (BDH: Bono de Desarrollo Humano) in an amount still to be determined. This section offers some comments about this option, comparing it to the scenarios considered above, and analyzing possible pros and cons with respect to its implementation. E1.15 Eliminatingthe subsidy with a simultaneous increase of the BDHcould be considered a variation of scenario 4 (eliminating subsidy along with a distribution of generatedfiscal savings to poor households inSelBen's quintiles 1and2), depending onthe amount offiscal savings devotedto the BDHincrease. E1.16 Using the BDHas a means for compensation provides some advantages. The main ones are the simplicity of logistics in this mechanism, which would enable applying (new) transfers almost immediately. The the fact that these are fixed transfers and not a single subsidy facilitates the task of estimating future costs for this measure. E1.17 However, it also presents some drawbacks. The first one has to do with coverage of the target population by the compensatory measure. According to the BDH's establishment decree, the program's target population are all households in SelBen's quintiles 1and 2-approximately 1,200,000 households. Inpractice, out of 1,050,000 beneficiary households currently inthe program, 850,000 belongto quintiles 1 and 2, and the rest are households in quintile 3-primarily seniors and/or physically challenged individuals. This suggests that: (i)not all households inquintiles 1and 2 are receiving the BDH, and (ii) not all BDH beneficiary households belong to quintiles 1 and 2. Thus, an increase in the BDH as a compensatory measure for the elimination of the gas subsidy will not benefit all households considered as poor by the SelBen. Only those which are currently part of the programand comply with the required co- responsibility conditions will benefit. E1.18 The second drawback to consider i s the loss of flexibility in identifying and determining the target population. The linkbetween subsidy reformand changes inthe BDHimply that the compensatory measure will only benefit quintiles 1and 2 of the SelBen. It is possible, however, for there to be a desire to extend coverage to other groups (Le., SelBen's quintile 3) in order to minimize a negative reaction from the population as a response to the elimination of the subsidy. This flexibility i s lost when implementing the change through the BDH, unless an increase incoverage of this program i s considered for the same group-a measure which, at present time, does not appear to be advisable, given its elevated fiscal cost, among others. E1.19 Lastly, it i s essential to note that there have been other attempts in the past to eliminate the gas subsidy through the creation of compensatory programs (in this case, the Solidarity Bonus). These attempts failed and gave place to a situation where the subsidy did not disappear and new liabilities emerged inthe form of transfers. The political economy to eliminate the subsidy i s complex and deserves additional actions. Even if the BDH amount i s increased, strong political opposition must not be discarded. Conclusion E1.20 There seem to be two viable action plans: the first one would consider eliminating the subsidy along with a simultaneous increase of the BDH, and the second one would consider retargeting the subsidy independently. Although implementation of both instances must be accompanied by a very good communications campaign, the steps to follow will depend on the action plan selected. E1.21 The first option will require: 126 Volume I,Annex E 0 A decision on the amount of increase in the BDHand cost of this measure, considering the number of programbeneficiaries-both current and potential; 0 A strong communications campaign; 0 Good planning and funding to implement a device to respond to an increase in BDH participation, which will be likely to occur once the measure i s announced. E1.22 The second option will require: 0 A decision about the beneficiary population(preferably usingthe SelBenIndex); 0 A strong communications campaign; 0 Development of a new targeted subsidy payment device. For such purpose, it will be necessary to: (i) provide recipients with an identification (ID, card or something similar), (ii)enable a payment method (direct deposit, ATM withdrawals with a card, reimbursable coupons). The political viability of the device should be determined on a smaller scale (pilot program) before implementing it at a national level 127 StatisticalAnnex Al. Non-FinancialPublic Sector (Nominal, GDP) A2. Non-FinancialPublic Sector (Real, Growth) A3. CentralGovemment(Nominal, GDP, Percentof NFPS) A4. Rest of GeneralGovernment(Nominal,GDP, Percentof NFPS) A5. Non-financial Public Enterprises (Nominal, GDP, Percentof NFPS) A6. Central Govemmentby SectoralClassification A7. Wages andSalaries of the Central Govemmentby Sectoral Classification A8. Goods andServicesof the Central Govemmentby SectoralClassification A9. Fixed Investmentof the CentralGovemmentby Sectoral Classification A10. Transfersof the CentralGovernment by SectoralClassification A11. EconomicClassification of Defense Expenditure(Nominal, GDP) A12. EconomicClassification of EducationExpenditure(Nominal, GDP) A13. EconomicClassification of HealthandSocialWelfare Expenditure(Nominal,GDP) A14. Numberof AscribedInstitutios to LineMinistries A15. Expendituresof Ascribed Institutionsto LineMinistries A16. Operations of Municipal& Provincial Councils A17. OrganizationReceivingEarmarkedRevenues A18. Institutions ReceivingEarmarkedRevenues A19. Social Expendituresas Percentof CentralGovernmentBudget A20. PersonnelExpendituresinEducation(in US$) A21. ExpendituresinEducationby Level (in MlnUS$) A22. BudgetIssuedFor Education(US $) A23. FinancingSourcesfor Education,2003 (inUS$) A24. Supply of public and privatehealthcenters, 1975-2000 A25. PreassignedSpendinginthe 2004 CentralGovernmentProforma(inThous US $) A26. Expendituresmanagedby the RegionalDevelopment Organizations,2003 A27. Military Expenditures A28. Salary Tables of the Non-FinancialPublic Sector, 2002-2004 A29. Direct Subsidiesto Water and SanitationSector A30. Illiteracy rate andyears of schoolingof adultpopulation (25 yrs and older), 1990-2001 A31. Ecuador: Population(25 years andolder) by completededucationlevel ( percentof total) A32. Ecuador: net enrolmentratesby educational level, 1990and 2001 A33. Ecuador:Years of schoolingby consumption quintiles and gender, 1995and 1999 A34. Ecuador: Years of schoolingby consumption quintiles and area, 1995 and 1999 A35. Ecuador: Net enrolmentratesby educationallevel and quintiles of p.c. consumption A36. Ecuador: Test scores for language andmathematics skills, by gender, regionandarea. A37. Ecuador: Intemalefficiency indicatorsfor primary education, 1995 and 2001 A38. Latin America: Intemalefficiency indicatorsfor primary education, 1990s A39. Ecuador: Differentialversus average unit costs inestimationthe public expenditure incidenceineducation, 1999 A40. Ecuador: "Marginal" expenditureincidence ineducation, 1995and 1999 A41. Ecuador: Cost-effectiveness modelestimates for net enrolment inprimary education A42. Determinants of years of schoolingand net enrolment in primary and secondaryeducation at the district (cantonal) level, 2001 A43. Determinantsof test scores for languageandmathematicsat primary (2nd and6thgrade) and secondary (10th grade) level by districts(cantons), 2001 A44. Retums to educationfor urbanworkers, 1990-2002(Specification 1) A45. Retums to educationfor urban workers, 1990-2002(Specification 2) A46 Profile of malnutrition, education, water andsanitationconditions, 2000 and2003 A47. Profile of knowledgeof family planningand sexually transmitteddiseases, 1999and 2000 A48. Profile of access to servicesof childrenage 0-5, 1999 129 A49. Profile of access to health services pregnant women, 2003 A50. Profile of vaccinationsof 2-5 year olds, 2003 A51. Profile of access to nutrition andhealthrelatedsocial assistance programs 2003 A52. Profile of coverage of healthinsurance, 2003 (% of population) A53. Profile of lifestylehabits, 1999 A54. Targetingefficiency of school mealsprogram(PAE), child andmatemalnutrition support (PA") and cashtransfer program(BDH), December 2003. A54a. Econometricresults of access to healthservicesmodel(1) A54b. Econometricresults of access to healthservicesmodel (2) AS5. Econometricresults of Infant Mortality (CPH) model A56. Unitcost estimates for public healthservices (in US$) A57. Ecuador:Budgetexecutionof institutional wages, central govemment A58. Activities andcomponentsof social programsby priority function A59. Numberof beneficiariesper priority social program, 2003 A60. Provisionof LMG servicesto infantsandpregnant women, 2003 A61. Budgetrevisionandexecutionof social institutions,2000-2002 A62. Priority social programsbudget and execution, 2003 A63. Track of Treasurytransfers to socialprograms, 2003 A64. Transfers andPayments of the "15 percent Law" to Municipalities, 2002 A65. Transfers andPayments of the "15 percent Law" to Municipalities, 2001 A66. Transfersof 15 percentLaw and Implications for Cash-Managementof the Central GovemmentBudget, 2002 A67. Transfers of 15 percent Law andImplications for Cash-Managementof the CentralGovemmentBudget, 2001 A68. RevenueBase for 15 percent-Lawwith andwithout PetroleumRevenue,2002 A69. RevenueBase for 15 percent-Lawwith andwithout PetroleumRevenue,2001 A70. Total Per Capita Investment of Municipalities andProvincial Counsils,2002 A71. Incidenceof Subnational GovemmentInvestmentSpendinginthe SocialSectors, 2002 A72. Incidenceof Subnational GovemmentInvestmentSpendinginthe Social Sectors, 2002 A73. Own Revenue andExpendituresper Level of Govemment, 1997 and2002 A74. Per CapitaInvestmentof Municipalities and ProvincialCouncils(in USD),2002 A75. IntergovemmentalTransfersto MunicipalitiesandProvincial Councils A76. InterestPayment SavingsfromAdditional Debt Repurchase 130 Al. Non-FinancialPublic Sector (Nomlnal, GDP) Average Average 1995 1996 1997 1998 1999 2000 2001 2002 2003 199S99 2WO-03 Mllllons of U S Total Revenues 4599 4656 4714 4027 3515 4126 4933 6271 6908 Petroleum 1329 1575 1270 913 1049 1460 1352 1390 1664 NonPetroleum 2717 2607 3134 3096 2334 2516 3475 4695 5090 dw VAT 621 633 779 831 591 893 1457 1667 1737 Operatingsurplus on the non-financial public enterprises 553 475 310 18 132 150 106 187 155 Total Expendltures 11 4604 5221 5220 5145 4165 3889 4653 6117 6585 CurrentExpenditures 3616 3773 3970 3990 3165 3095 3457 4536 5124 Interest 786 631 995 987 1183 1052 996 869 820 External 669 659 751 749 837 653 779 665 634 Domestic 117 173 243 238 346 199 217 205 186 Salaries 1426 1482 1542 1691 991 761 1163 1761 2288 Goodsand Services 305 533 600 569 397 410 579 901 948 Others lo99 927 833 744 594 871 719 1005 1069 Capital Expenddures 1188 1449 1250 1155 1000 795 1396 1582 1460 Fixedcapital formation 998 1265 1246 1143 982 782 1161 1417 1431 Central government 381 498 478 602 485 425 645 611 660 Non-financialpublic enterprises 389 544 486 199 245 83 143 236 254 Municipalgovernments 159 154 190 260 224 246 329 459 480 Others 69 69 91 82 27 29 44 111 38 Other caoitai 190 184 4 12 18 12 236 165 29 Balance -205 -566 -506 -1118 -650 237 80 154 323 Discrepancy2l 45 30 0 84 168 10 5 0 130 Overall Balance -250 -595 -506 -1202 -818 227 75 154 453 Prlmary Balance 536 236 489 -215 366 1280 1071 1023 1273 Inpercent of GDP Total Revenues 22.8 21.9 19.9 17.3 21.1 25.9 23.5 25.8 25.7 20.6 25.2 Petroleum 6.6 7.4 5.4 3.9 6.3 9.2 6.4 5.7 6.2 5.9 6.9 Non Petroleum 13.5 12.3 13.3 13.3 14.0 15.8 16.5 19.3 18.9 13.3 17.6 dw VAT 3.1 3.0 3.3 3.6 3.5 5.6 6.9 6.9 6.5 3.3 6.5 Operatingsurpluson the norrfinancial public enterprises 2.7 2.2 1.3 0.1 0.6 0.9 0.5 0.8 0.6 1.4 0.7 Total Expendltures 11 23.8 24.6 22.1 22.1 25.0 24.4 23.1 25.2 24.5 23.5 24.3 Current Expenditures 17.9 17.7 16.8 17.2 19.0 19.4 16.4 18.7 19.0 17.7 18.4 Interest 3.9 3.9 4.2 4.2 7.1 6.6 4.7 3.6 3.0 4.7 4.5 External 3.3 3.1 3.2 3.2 5.0 5.4 3.7 2.7 2.4 3.6 3.5 Domestic 0.6 0.8 1.o 1.o 2.1 1.2 1.o 0.8 0.7 1.1 1.o Salaries 7.1 7.0 6.5 7.3 5.9 4.8 5.5 7.2 8.5 6.8 6.5 Goods and Services 1.5 2.5 2.5 2.4 2.4 2.6 2.8 3.7 3.5 2.3 3.1 Others 5.4 4.4 3.5 3.2 3.6 5.5 3.4 4.1 4.0 4.0 4.2 Capital Expenditures 5.9 6.8 5.3 5.0 6.0 5.0 6.6 6.5 5.4 5.8 5.9 Fixedcapital formation 4.9 5.9 5.3 4.9 5.9 4.9 5.5 5.8 5.3 5.4 5.4 Centralgovernment 1.9 2.3 2.0 2.6 2.9 2.7 3.1 2.5 2.5 2.4 2.7 Non-financialpublicenterprises 1.9 2.6 2.1 0.9 1.5 0.5 0.7 1.o 0.9 1.8 0.8 Municipalgovernments 0.8 0.7 0.8 1.1 1.3 1.5 1.6 1.9 1.8 1.o 1.7 Others 0.3 0.3 0.4 0.4 0.2 0.2 0.2 0.5 0.1 0.3 0.2 Other capital 0.9 0.9 0.0 0.1 0.1 0.1 1.1 0.7 0.1 0.4 0.5 Balance -1.o -2.7 -2.1 -4.8 -3.9 1.5 0.4 0.6 1.2 -2.9 0.9 Discrepancy2/ 0.2 0.1 0.0 0.4 1.o 0.1 0.0 0.0 0.5 0.3 0.1 Overall Balance -1.2 -2.8 -2.1 -5.2 -4.9 1.4 0.4 0.6 1.7 -3.2 1.o Prlmary Balance 2.7 1.1 2.1 -0.9 2.2 8.0 5.1 4.2 4.7 1.4 5.5 Note: NominalGDP US$Mln 20196 21268 23636 23255 16674 15934 21024 24311 26913 Source: BCE, IMFfor 2003 11 Expenditureson aCCNed basis. 21 Includesreductionof personneland costs of strengtheningthe priiate financial system. 131 A2. Non-Financial Public Sector (Real, Growth) Average Average 1995 1996 1997 1998 1999 ZOO0 2001 2002 2003 1995-99 2000-03 Millions of 2000 US$ Total Revenues 3462 3408 3230 2864 3267 41265 3930 4465 4525 3246 4261 Petroleum lo00 1153 870 649 975 1460 1077 989 1090 929 1154 Non Petroleum 2045 1908 2148 2202 2170 2516 2768 3342 3334 2095 2990 olw VAT 467 463 534 591 549 893 1161 1187 1138 521 1095 Operatingsurpluson the non-financial public enterprises 416 347 212 13 123 150 84 133 101 222 117 Total Expenditures I / 3616 3822 3577 3660 3872 3889 3866 4355 4313 3709 4106 Current Expenditures 2722 2762 2721 2838 2942 3095 2754 3229 3357 2797 3108 Interest 591 609 682 702 1100 1052 794 619 537 737 750 External 504 482 515 533 778 853 621 473 415 562 591 Domestic 88 126 167 169 322 199 173 146 122 174 160 Salaries 1073 1085 1057 1202 921 761 926 1254 1499 1068 1110 Goods and Services 230 390 411 405 369 410 461 641 621 361 533 Others 828 678 571 529 552 871 573 715 700 632 715 Capital Expenditures 894 1060 857 821 929 795 1112 1126 957 912 997 Fixedcapitalformation 751 926 854 813 913 782 925 lo09 937 851 913 Centralgovernment 287 365 328 428 451 425 514 435 432 372 451 Non-financialpublic enterprises 293 398 333 142 228 83 114 168 166 279 133 Municipalgovernments 120 113 130 185 208 246 262 327 315 151 287 Others 52 50 63 58 25 29 35 79 25 50 42 Other capital 143 134 3 9 16 12 188 117 19 61 84 Balance -154 -414 -347 -795 -604 237 64 110 212 -463 156 Discrepancy2/ 34 22 0 60 156 10 4 0 85 54 25 Overall Balance -188 -436 -347 -855 -760 227 60 110 297 -517 173 Primary Balance 403 173 335 -153 340 1280 853 729 834 220 924 Percent change Total Revenues 6.2 -1.6 -5.2 -11.3 14.1 26.3 -4.8 13.6 1.3 0.4 9.1 Petroleum 4.1 15.2 -24.5 -25.4 50.1 49.8 -26.2 -8.1 10.2 3.9 6.4 Non Petroleum 9.2 -6.7 12.6 2.5 -1.5 16.0 10.0 20.7 -0.2 3.2 11.6 olw VAT 3.3 -0.8 15.3 10.7 -7.1 62.7 29.9 2.2 -4.1 4.3 22.7 Operatingsurplus on the non-financial publicenterprises -2.2 -16.5 -38.9 -93.9 850.7 21.6 -43.6 57.9 -24.0 139.8 3.0 Total Expenditures 1/ 13.7 5.7 -6.4 2.3 5.8 0.5 -0.6 12.7 -1.o 4.2 2.9 Current Expenditures 17.7 1.5 -1.5 4.3 3.7 5.2 -11.o 17.3 3.9 5.1 3.8 Interest 10.4 2.9 12.0 3.0 56.7 -4.3 -24.6 -22.0 -13.3 17.0 -16.0 External 6.2 -4.2 6.8 3.5 46.0 9.7 -27.3 -23.8 -12.2 11.7 -13.4 Domestic 42.5 43.7 32.0 1.3 90.3 -38.1 -13.1 -15.7 -16.7 42.0 -20.9 Salaries 8.6 1.1 -2.6 13.8 -23.4 -17.4 21.7 35.4 19.5 0 . 5 14.8 Goods and Services -22.9 69.9 5.4 -1.6 -8.8 11.1 12.5 39.0 -3.2 8.4 14.9 Others 68.4 -18.0 -15.8 -7.3 4.4 57.7 -34.3 24.9 -2.1 6.3 11.6 Capital Expenditures 3.2 18.6 -19.2 -4.1 13.1 -14.5 40.0 1.2 -15.0 2.3 2.9 Fixedcapitalformation -4.8 23.3 -7.8 -4.8 12.3 -14.3 18.2 9.1 -7.1 3.6 1.5 Centralgovernment -2.4 27.1 -10.1 30.6 5.4 -5.9 21.o -15.3 -0.6 10.1 -0.2 Non-financialpublic enterprises -17.4 36.0 -16.3 -57.5 61.O -63.7 37.4 47.7 -1.3 1.2 5.0 Municipalgovernments 21.6 -5.6 14.9 42.2 12.8 18.0 6.6 24.8 -3.7 17.2 11.4 Others 21.5 -2.6 24.4 -7.3 -56.3 15.1 21.1 122.8 -68.7 -4.1 22.6 Other capital 85.0 -6.1 -97.9 201.5 89.6 -25.2 1426.0 -37.5 -83.7 54.4 319.9 Balance -292.0 168.8 -16.2 129.2 -24.0 -139.3 -73.1 72.0 93.0 -6.8 -11.8 Overall Balance 587.5 131.7 -20.4 146.4 -11.1 -129.9 -73.7 83.5 170.7 166.8 12.7 Primary Balance -20.6 -57.1 93.6 -145.7 -322.3 276.4 -33.3 -14.6 14.4 -90.4 60.7 Note:GDP Deflator (2o00=1) 1.33 1.37 1.46 1.41 1.08 1.00 1 26 1.40 1.53 Source BCE I / Expenditureson accruedbasis. 21 Includesreductionof personneland costs of strengthening the privatefinancial system. 132 A3. Central Government (Nominal, % GDP, %NFPS) Average Average 1995 19% 1997 1998 1999 2000 2001 2002 2003 1995-99 2000-03 Mllllonr of U S Total Revenues 3129 3222 3448 3227 2668 3250 3839 4572 4771 Petroleum 1200 1480 1206 888 994 1397 1280 1363 1561 Non Petroleum 1929 1742 2242 2339 1694 1853 2559 3210 3210 Tax 1434 1407 1687 1998 1462 1623 2370 2746 2790 NonTax 340 212 230 178 138 237 139 340 375 Transfers 156 124 125 164 75 -7 50 122 45 Total Expenditures 3419 3726 3726 4186 3163 3230 3994 4757 5010 Current Expenditures 2696 2843 2947 3141 2427 2572 2555 3531 3693 Interest 622 772 939 941 1121 1009 938 823 627 External 505 604 703 708 781 812 728 621 614 Domestic 117 168 237 233 341 197 210 202 213 Salaries 1252 1317 1369 1510 898 707 lo88 1672 1864 Goods and Services 61 202 209 197 123 175 122 318 329 Others 527 276 9 65 206 542 300 351 257 Transfers 214 275 421 428 79 140 107 366 416 Capital Expenditures 723 883 779 1045 736 659 1439 1227 1317 Fixed capital formation 381 498 476 602 485 425 645 611 660 Others 233 -3 15 Transfers 173 232 301 443 251 234 562 618 642 CG Balance -269 -505 -278 -959 -476 19 -155 -185 -239 CG Primary Balance 332 266 661 -18 646 1029 782 638 588 Percent of GDP Total Revenues 15.5 15.1 14.6 13.9 16.1 20.4 18.3 18.8 17.7 15.0 18.8 Petroleum 5.9 7.0 5.1 3.6 6.0 6.8 6.1 5.6 5.8 5.6 6.6 Non Petroleum 9.6 6.2 9.5 10.1 10.2 11.6 12.2 13.2 11.9 9.5 12.2 Tax 7.1 6.6 8.0 8.6 8.9 10.2 11.3 11.3 10.4 7.8 10.8 NonTax 1.7 1.o 1.o 0.8 0.8 1.5 0.7 1.4 1.4 1.o 1.2 Transfers 0.8 0.6 0.5 0.7 0.4 0.0 0.2 0.5 0.2 0.6 0.2 Total Expenditures 16.9 17.5 15.8 18.0 19.0 20.3 19.0 19.6 18.6 17.4 19.4 Current Expenditures 13.3 13.4 12.5 13.5 14.6 16.1 12.2 14.5 13.7 13.4 14.1 Interest 3.1 3.8 4.0 4.0 6.7 6.3 4.5 3.4 3.1 4.3 4.3 External 2.5 2.6 3.0 3.0 4.7 5.1 3.5 2.6 2.3 3.2 3.3 Domestic 0.6 0.8 1.o 1.o 2.0 1.2 1.o 0.8 0.8 1.1 1.o Salaries 6.2 6.2 5.8 6.5 5.4 4.4 5.2 6.9 6.9 6.0 5.9 Goods and Services 0.4 1.o 0.9 0.8 0.7 1.1 0.6 1.3 1.2 0.8 1.1 Others 2.6 1.3 0.0 0.3 1.2 3.4 1.4 1.4 1.o 1.1 1.8 Transfers 1.1 1.3 1.8 1.8 0.5 0.9 0.5 1.5 1.5 1.3 1.1 Capital Expenditures 3.6 4.2 3.3 4.5 4.4 4.1 6.8 5.0 4.9 4.0 5.2 Fixedcapital fonation 1.9 2.3 2.0 2.6 2.9 2.7 3.1 2.5 2.5 2.4 2.7 Others 0.0 0.0 0.0 0.0 0.0 0.0 1.1 0.0 0.1 0.0 0.3 Transfers 0.9 1.1 1.3 1.9 1.5 1.5 2.7 2.5 2.4 1.3 2.3 CG Balance -1.4 -2.4 -1.2 -4.1 -2.9 0.1 -0.7 -0.8 -0.9 -2.4 -0.6 CG Primary Balance 1.6 1.3 2.8 -0.1 3.9 6.5 3.7 2.6 2.2 1.9 3.7 Percent of NFPS Total Revenues 68.0 69.2 73.2 80.1 76.5 76.8 77.6 72.9 69.1 73.4 74.6 Total Expenditures 71.2 71.4 71.4 61.4 76.0 63.1 82.3 77.8 76.1 74.2 79.8 Current Expenditures 74.6 75.4 74.2 78.7 76.7 83.1 73.9 77.8 72.1 75.9 76.7 Interest 79.1 92.9 94.4 95.4 94.8 95.9 94.1 94.6 100.9 91.3 96.4 Eaemai 75.5 91.7 93.5 94.5 93.2 95.2 93.4 93.5 96.9 69.7 94.7 Domestic 100.0 97.4 97.3 96.2 96.5 99.0 96.7 98.4 114.6 96.3 102.2 Salaries 87.8 68.9 68.8 89.3 90.6 92.9 93.6 95.0 81.5 89.1 90.7 Goods and Selvices 26.7 38.0 34.8 34.6 30.9 42.6 21.o 35.3 34.7 33.0 33.4 Others 47.9 29.6 1.1 8.7 34.6 62.2 41.8 35.0 24.1 24.4 40.7 Capnal Experdilures 72.4 69.8 62.5 91.5 75.0 84.2 124.0 86.6 92.0 74.3 96.7 CG Balance 115.9 84.8 54.9 79.8 58.2 8.5 .206.8 .120.1 100.0 78.7 -54.6 CG Primary Balance 62.0 113.4 135.4 8.3 176.6 80.4 73.0 62.3 5.8 99.1 55.4 Nontransfer expenditures/ NFPS expenditures 63.1 61.6 57.6 64.4 68.0 73.5 68.5 61.7 60.0 63.0 65.9 Source: BCE, MEF, and own calculations 133 A3b. Centnl Government(Real, X GDP, %NFPS) Avenge Avenge 1995 1996 1997 1998 1999 2ow 2001 2002 2003 1995.99 200003 Mllllonrof 2000 US$ Total Revenws 2356 2358 2363 2295 2496 3250 3058 3255 3125 2374 3172 Petroleum 903 1083 826 631 924 1397 1020 970 1023 874 1102 Non Petroleum 1452 1275 1537 1664 1575 1853 2039 2285 2102 1501 2070 Tax 1079 1030 1293 1421 1377 1623 1888 1956 1627 1240 1824 Non Tax 256 155 158 126 126 237 111 242 246 165 209 Transfers 117 90 86 117 69 -7 40 87 30 96 37 Total Expenditures 2573 2726 2554 2977 2940 3230 3182 3387 3282 2755 3270 CurrentExpenditures 2029 2081 2020 2234 2256 2572 2035 2514 2419 2124 2385 Interest 468 565 644 670 1M 2 1009 747 586 542 678 721 External 380 442 481 504 725 612 580 442 402 507 559 Domestic a3 123 162 le6 317 197 167 143 139 171 162 Salaries 942 964 938 1074 835 707 867 1191 1221 951 996 Goods and Services 61 146 143 140 114 175 97 226 216 121 178 Others 397 202 6 46 191 542 239 250 169 168 300 Transfers 161 202 288 304 74 140 e6 261 272 206 190 Capital Expenditures 544 646 534 744 684 659 1146 873 863 630 ea5 Fixedcapital formation 287 365 326 428 451 425 514 435 432 372 451 Others 0 0 0 0 0 0 185 -2 10 0 48 Transfers 130 170 206 315 233 234 447 440 421 211 3% CG Balance -216 -369 -191 -662 -442 19 *I24 -132 -156 -380 -98 CG primaryBalance 250 196 453 -13 Mx) 1029 823 454 385 297 623 FiscalLaw Indicators PrimaryExpenditures 2106 2162 1910 2306 1898 2221 2435 2801 2740 2077 2549 Percentchange Total Revenues 0.1 0.2 -2 9 8.9 30.I -5 9 6.4 -4.0 1.6 6.7 Petroleum 19.9 -23.7 -23 6 46.3 51.2 -27 0 -4.9 5.4 4.7 6.2 NonPetroleum -12.2 20.5 8 3 -5.4 17.7 100 12.1 -8.0 2.8 7.9 Tax 4.6 25.6 99 -3.1 17.8 163 3.6 -6.6 7.0 7.8 Non Tax -39.4 1.8 -20 0 1.2 65.1 -53 3 116.9 1.4 -14.1 38.0 Transfers -23.0 -5.0 35 7 40.4 -110.1 -6725 116.2 -65.9 -6.2 -183.1 Total Expendltures 6.0 -6.4 166 -1.2 9.9 -1 5 6.4 -3.1 3.7 2.9 CurrentExpenditures 2.6 -2.9 106 1 0 14.0 -209 23.5 -3.8 2.6 3.2 Interest 20.6 13.9 4 0 55.7 -3.2 -26 0 -21.6 -7.5 23.6 -14.6 External 16.4 6.6 4 6 44.0 12.0 -28 6 -23.7 -9.0 16.5 -12.3 Domestic 39.9 31.9 2 2 91.o -37.8 -15 1 -14.2 -3.0 41.3 -17.5 Salaries 2.3 -2.7 145 -22.3 -15.3 226 37.4 2.5 -2.0 11.8 Goods and Services 142.0 -3.4 -2 1 -16.5 52.9 -444 133.5 -4.6 29.5 34.3 Others -49.1 -96.9 631 0 314.6 183.3 -55 6 4.5 -32.6 199.9 24.9 Transfers 25.1 431 5 5 -75.7 69.0 -38 7 204.9 4.4 -0.5 64.9 Capital Expenditures 18.8 -17.4 39 3 -8.0 -3.8 74 1 -23.8 -1.2 6.2 11.3 Fixedcapitalformation 27.1 -10.1 306 5.4 -5.9 21 0 -15.3 -0.6 13.2 -0.2 Others -101.0 -644.7 -372.8 Transfers 30.6 21.1 53 1 -26.1 0.4 91 2 -1.6 4.4 19.7 21.4 CG Balance 69.5 -48.4 258 0 -35.2 -104.4 -7402 6.5 18.9 61.O -2046 CG Primary Balance -21.6 131.2 -1028 -4814.0 71.4 -39 4 -27.2 -15.2 .1201.8 -2.6 FlscalLaw Indicators PrimaryExpenditures 2.7 -11.7 20 8 -17.8 17.0 9 6 15.0 -2.2 -1.5 9.9 Source: BCE, MEF,and own calculations 134 A4. Rest of General Government (Nornlnal, GDP, Percent of NFPS) Average Average 1995 1996 1997 1998 1999 2Mx) 2001 2002 2003 1995-99 2000-03 Mllllonsof U S Total Revenues 1283 1375 1424 1503 978 989 1615 2422 2870 Oil-relatedrevenues 127 94 63 25 55 64 72 27 22 Non oil revenues 1155 1281 1381 1478 923 925 1545 2395 2E48 Contributionsto SocialSecurity 506 450 455 451 230 228 455 767 916 Interests and gains 85 149 102 92 123 62 91 158 209 Others 274 318 364 310 319 311 337 539 706 Transfers 291 363 439 625 251 324 663 931 1018 Total Expenditures 1186 12M 1391 1449 977 E14 1281 1992 2289 Current 937 1010 1106 1096 708 527 905 1255 1704 Wages 312 310 335 379 215 1EO 263 335 425 Goods andservices 224 331 391 372 274 235 457 583 E19 Interest 70 10 11 11 27 12 21 20 9 External 70 10 11 11 27 12 14 17 E IntemaI 0 0 0 0 0 0 6 3 3 Transfers 322 359 369 334 191 120 164 317 651 To publicsector 0 0 0 0 0 0 0 62 To private sector 322 359 369 334 191 120 164 317 589 Others 9 0 0 0 0 0 0 0 0 CapitalExpendtures 250 254 285 353 269 287 376 737 585 Fixed investments 228 223 281 341 251 275 373 570 561 Net lendingcmcessions -1 0 -8 1 11 -10 .1 159 11 Others 23 32 13 11 7 22 3 9 14 Balance 97 311 33 54 1 175 335 430 581 Percent of GDP Total Revenues 6.4 6.5 6.0 6.5 5.9 6.2 7.7 10.0 10.7 6.2 8.6 Oil-relatedrevenues 0.6 0.4 0.3 0.1 0.3 0.4 0.3 0.1 0.1 0.4 0.2 Non oil revenues 5.7 6.0 5.8 6.4 5.5 5.8 7.4 9.9 10.6 5.9 8.4 Contributionsto SocialSecurity 2.5 2.1 1.9 1.9 1.4 1.4 2.2 3.2 3.4 2.0 2.5 Interests andgains 0.4 0.7 0.4 0.4 0.7 0.4 0.4 0.6 0.8 0.5 0.6 Others 1.4 1.5 1.5 1.3 1.9 2.0 1.6 2.2 2.6 1.5 2.1 Transfers 1.4 1.7 1.9 2.7 1.5 2.0 3.2 3.8 3.8 1.8 3.2 Total Expenditures 5.9 5.9 5.9 6.2 5.9 5.1 6.1 8.2 8.5 6.0 7.0 Current 4.6 4.7 4.7 4.7 4.2 3.3 4.3 5.2 6.3 4.6 4.8 Wages 1.5 1.5 1.4 1.8 1.3 1.o 1.3 1.4 1.6 1.5 1.3 Goodsand services 1.1 1.8 1.7 1.6 1.6 1.5 2.2 2.4 2.3 1.5 2.1 Interest 0.3 0.0 0.0 0.0 0.2 0.1 0.1 0.1 0.0 0.1 0.1 External 0.3 0.0 0.0 0.0 0.2 0.1 0.1 0.1 0.0 0.1 0.1 Internal 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Transfers 1.E 1.7 1.E 1.4 1.1 0.8 0.8 1.3 2.4 1.5 1.3 To publicsector 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.0 0.1 To privatesector 1.6 1.7 1.8 1.4 1.1 0.8 0.8 1.3 2.2 1.5 1.3 Others 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 CapitalExpendaures 1.2 1.2 1.2 1.5 1.6 1.8 1.8 3.0 2.2 1.4 2.2 Fixedinvestments 1.1 1.o 1.2 1.5 1.5 1.7 1.8 2.3 2.1 1.3 2.0 Net lendingconcessions 0.0 0.0 0.0 0.0 0.1 -0.1 0.0 0.7 0.0 0.0 0.2 Others 0.1 0.1 0.1 0.0 0.0 0.1 0.0 0.0 0.1 0.1 0.1 Balance 0.5 0.5 0.1 0.2 0.0 1.1 1.6 1.8 2.2 0.3 1.7 Nontransferexpenditures' NFPS expenditures 18n 19.6 21.7 18.9 17.9 23.0 27.4 24.9 19.1 23.3 135 A5. Non-flnanclalPublic Enterprises(Nomlnal,GDP, Percentof NFPS) Average Average 1995 1996 1997 1998 1999 2000 2001 2002 2003 199599 2000-03 MlllloFsof US$ Operating Revenues 1645 1692 1828 974 618 639 746 937 1718 Operating Expenditures 1092 1218 1319 956 486 490 640 750 1627 Salaries 179 190 217 166 90 69 100 129 161 Contributionsto social security 24 23 24 17 10 9 9 14 16 Goods and Services 889 1005 1078 773 386 412 531 608 1451 Operating Balance 553 475 310 18 132 150 106 187 91 Non-operatingRevenues 145 133 239 170 46 57 47 77 176 Non-operatingExpenditures 321 235 323 202 io8 81 109 119 85 Interest 94 49 44 34 35 31 38 27 14 External 94 45 38 30 30 29 37 27 14 Domestic 5 7 4 5 2 1 1 0 Transfers to publicsector 131 93 39 40 38 9 5 2 2 Other 96 93 240 128 35 41 66 90 68 Capital Expenditures 389 544 486 199 245 83 143 236 210 Balance -12 -172 -261 ,213 -175 43 -100 -91 -28 Percentof GDP Operating Revenues 8.1 8.0 6.9 4.2 3.7 4.0 3.5 3.9 6.4 6.2 4.4 Operating Expenditures 5.4 5.7 5.6 4.1 2.9 3.1 3.0 3.1 6.0 4.7 3.8 Salaries 0.9 0.9 0.9 0.7 0.5 0.4 0.5 0.5 0.6 0.8 0.5 Contributionsto social security 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.1 0.1 0.1 0.1 Goodsand Services 4.4 4.7 4.6 3.3 2.3 2.6 2.5 2.5 5.4 3.9 3.2 Operating Balance 2.7 2.2 1.3 0.1 0.8 0.9 0.5 0.8 0.3 1.4 0.6 Non-operatingRevenues 0.7 0.6 1 .o 0.7 0.3 0.4 0.2 0.3 0.7 0.7 0.4 Non-operatingExpenditures 1.6 1.1 1.4 0.9 0.6 0.5 0.5 0.5 0.3 1.1 0.5 Interest 0.5 0.2 0.2 0.1 0.2 0.2 0.2 0.1 0.1 0.2 0.1 External 0.5 0.2 0.2 0.1 0.2 0.2 0.2 0.1 0.1 0.2 0.1 Domestic 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Transfers to public sector 0.6 0.4 0.2 0.2 0.2 0.1 0.0 0.0 0.0 0.3 0.0 Other 0.5 0.4 1.o 0.5 0.2 0.3 0.3 0.4 0.3 0.5 0.3 Capital Expenditures 1.9 2.6 2.1 0.9 1.5 0.5 0.7 1.o 0.8 1.8 0.7 BaIance -0.1 -0.8 -1.1 -0.9 -1.0 0.3 -0.5 -0.4 -0.1 -0.8 -0.2 Non-transfernon-operating expenditures/NFPSexpenditures 12.0 13.1 14.8 7.0 7.6 4 0 5.1 5.. 4.4 8 Source: BCEand own calculations 136 A8. Central Government by Sectoral Classification 1995 1998 1997 1998 1999 2000 2001 2002 (In millionsof US dollars) GeneralAdministration 101 117 50 40 14 34 24 47 Defense 499 483 588 622 311 340 443 669 PublicOrder 317 362 350 291 161 179 253 440 Education 586 676 605 707 391 357 568 920 Healthand Social Welfare 275 345 278 256 196 198 364 444 ProductiveActivities 169 235 213 109 227 56 304 210 infrastructure 272 311 354 435 242 163 408 402 Other EconomicServices 919 905 844 1048 1263 1509 1721 1195 Others 281 293 444 676 358 349 308 431 Total 3419 3726 3726 4186 3163 3185 4394 4757 (Inpercentof GDP) GeneralAdministration 0.5 0.6 0.2 0.2 0.1 0.2 0.1 0.2 Defense 2.5 2.3 2.5 2.7 1.9 2.1 2.1 2.8 PublicOrder 1.6 1.7 1.5 1.3 1.o 1.1 1.2 1.8 Education 2.9 3.2 2.6 3.0 2.3 2.2 2.7 3.8 Healthand SocialWelfare 1.4 1.6 1.2 1.1 1.2 1.2 1.7 1.8 ProductiveActivities 0.8 1.1 0.9 0.5 1.4 0.3 1.4 0.9 Infrastructure 1.3 1.5 1.5 1.9 1.5 1.o 1.9 1.7 Other EconomicServices 4.6 4.3 3.6 4.5 7.6 9.5 8.2 4.9 Others 1.4 1.4 1.9 2.9 2.1 2.2 1.5 1.8 , Total Central Government 18.9 17.5 15.8 18.0 19.0 20.0 20.9 19.6 (in percentof total) GeneralAdministration 2.9 3.2 1.3 1.o 0.5 1.1 0.5 1.o Defense 14.6 13.0 15.8 14.9 9.8 10.7 10.1 14.1 PublicOrder 9.3 9.7 9.4 7.0 5.1 5.6 5.8 9.3 Education 17.1 18.1 16.2 16.9 12.4 11.2 12.9 19.3 Healthand Social Welfare 8.0 9.2 7.4 6.1 6.2 6.2 8.3 9.3 ProductiveActivities 5.0 6.3 5.7 2.6 7.2 1.7 6.9 4.4 Infrastructure 8.0 8.3 9.5 10.4 7.7 5.1 9.3 8.4 Other EconomicServices 26.9 24.3 22.6 25.0 39.9 47.4 39.2 25.1 Others 8.2 7.9 11.9 16.1 11.3 10.9 7.0 9.1 Total Central Government 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (real percentchange) GeneralAdministration 13.4 -59.9 -16.6 -53.2 155.7 -44.2 75.0 Defense -6.0 14.0 9.8 -34.6 17.7 3.6 35.1 PublicOrder 11.1 -9.5 -13.6 -28.0 19.7 13.0 55.3 Education 12.3 -16.2 21.2 -27.7 -1.6 26.6 44.6 Healthand SocialWelfare 22.0 -24.6 -4.3 0.1 8.7 46.1 9.0 ProductiveActivities 34.8 -14.8 -46.8 170.7 -73.6 335.9 -38.4 Infrastructure 11.1 6.6 27.7 -27.3 -27.8 99.9 -12.0 Other EconomicServices -4.3 -12.7 28.9 57.4 28.5 -9.1 -38.0 Others 1.3 41.9 58.1 -30.7 4.7 -29.6 24.9 Total CentralGovernment 6.0 -6.4 16.6 -1.2 8.3 9.9 -3.2 Source: MEF 137 A7. Wages and Salaries of the Central Government by Sectoral Classification 1995 1996 1997 1998 1999 2000 2001 2002 (In millionsof US dollars) GeneralAdministration 14 10 11 17 2 1 2 3 Defense 373 379 437 491 270 200 266 356 Public Order 173 163 193 214 118 91 131 200 Education 335 456 450 499 308 297 458 722 Health and Social Welfare 214 167 163 137 83 70 97 190 ProductiveActivities 49 36 31 28 14 10 16 23 Infrastructure 54 42 34 35 20 17 21 58 Other Economic Services 39 24 25 66 6 4 6 13 Others 2 41 24 21 77 16 90 107 Total CG Wages and Salaries 1252 1317 1369 1510 a98 707 1088 1672 (In percent of GDP) GeneralAdministration 0.1 0.0 0.0 0.1 0.0 0.0 0.0 0.0 Defense 1.8 1.8 1.8 2.1 1.6 1.3 1.3 1.5 Public Order 0.9 0.8 0.8 0.9 0.7 0.6 0.6 0.8 Education 1.7 2.1 1.9 2.1 1.8 1.9 2.2 3.0 Health and Social Welfare 1.1 0.8 0.7 0.6 0.5 0.4 0.5 0.8 ProductiveActivities 0.2 0.2 0.1 0.1 0.1 0.1 0.1 0.1 Infrastructure 0.3 0.2 0.1 0.2 0.1 0.1 0.1 0.2 Other Economic Services 0.2 0.1 0.1 0.3 0.0 0.0 0.0 0.1 Others 0.0 0.2 0.1 0.1 0.5 0.1 0.4 0.4 Total CG Wages and Salaries 6.2 6.2 5.8 6.5 5.4 4.4 5.2 6.9 (in percent of total CG wages and salaries) General Administration 1.1 0.8 0.8 1.2 0.2 0.2 0.2 0.2 Defense 29.8 28.8 31.9 32.5 30.1 28.4 24.5 21.3 Public Order 13.8 12.4 14.1 14.2 13.1 12.9 12.0 12.0 Education 26.7 34.6 32.9 33.1 34.3 42.0 42.1 43.2 Healthand Social Welfare 17.1 12.7 11.9 9.1 9.2 9.9 8.9 11.4 ProductiveActivities 3.9 2.7 2.3 1.9 1.5 1.5 1.5 1.3 Infrastructure 4.3 3.2 2.5 2.3 2.3 2.4 1.9 3.5 Other Economic Services 3.1 1.9 1.9 4.4 0.6 0.6 0.6 0.8 Others 0.2 3.1 1.8 1.4 8.6 2.2 8.3 6.4 Total CG Wages and Salaries 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (real percent change) General Administration -31.4 -0.3 71.4 -85.2 -20.7 34.7 14.0 Defense -1.2 8.0 16.6 -28.2 -20.1 5.9 19.5 Public Order -8.3 11.0 15.2 -28.1 -16.9 14.6 36.6 Education 32.4 -7.6 15.2 -19.3 3.6 23.1 40.7 Healthand Social Welfare -24.0 -8.4 -12.7 -21.1 -9.1 9.8 75.9 ProductiveActivities -29.3 -19.2 -5.0 -36.0 -20.2 24.0 26.0 Infrastructure -24.9 -22.7 6.2 -24.8 -9.4 -2.4 150.4 Other Economic Services -38.8 -2.7 169.3 -88.6 -21.7 20.2 81.2 Others 1998.4 -44.5 -10.0 376.9 -78.2 358.0 6.1 Total CG Wages and Salaries 2.3 -2.7 14.5 -22.3 -15.3 22.6 37.4 Source: MEF 138 A8. Goods and Services of the Central Government bySectoral Classification 1995 1996 1997 1998 1999 2000 2001 2002 (In millionsof US dollars) General Administration 9.0 12.1 10.1 9.8 4.6 9.4 5.0 8.9 Defense 0.4 46.6 41.4 41.7 17.0 47.8 22.5 103.7 Public Order 11.8 41.1 44.3 36.3 21.o 20.2 17.9 43.2 Education 4.4 18.6 15.6 14.4 15.3 12.1 12.0 18.7 Healthand Social Welfare 26.4 35.6 42.4 51.7 24.7 48.8 23.0 54.4 ProductiveActivities 8.9 15.4 14.5 8.3 5.9 1.9 5.5 7.3 Infrastructure 6.7 11.6 10.9 6.3 4.5 1.5 4.2 5.5 Other Economic Services 4.8 6.1 8.4 8.5 15.0 18.5 7.5 15.2 Others 9.1 15.2 21.2 19.9 14.8 14.2 24.1 61.2 Total CG Goods and Services 81.3 202.3 208.8 197.0 122.8 174.5 121.7 318.0 (In percent of GDP) General Administration 0.0 0.1 0.0 0.0 0.0 0.1 0.0 0.0 Defense 0.0 0.2 0.2 0.2 0.1 0.3 0.1 0.4 Public Order 0.1 0.2 0.2 0.2 0.1 0.1 0.1 0.2 Education 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Healthand Social Welfare 0.1 0.2 0.2 0.2 0.1 0.3 0.1 0.2 ProductiveActivities 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 Infrastructure 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Other Economic Services 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.1 Others 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.3 Total CG Goodsand Services 0.4 1.o 0.9 0.8 0.7 1.1 0.6 1.3 (in percent of total CG goods and services) General Administration 11.1 6.0 4.8 5.0 3.7 5.4 4.1 2.8 Defense 0.5 23.1 19.8 21.1 13.8 27.4 18.5 32.6 Public Order 14.5 20.3 21.2 18.5 17.1 11.6 14.7 13.6 Education 5.4 9.2 7.5 7.3 12.5 7.0 9.9 5.9 Healthand Social Welfare 32.4 17.6 20.3 26.3 20.1 28.0 18.9 17.1 ProductiveActivities 10.9 7.6 6.9 4.2 4.8 1.1 4.6 2.3 Infrastructure 8.2 5.7 5.2 3.2 3.6 0.8 3.4 1.7 Other Economic Services 5.9 3.0 4.0 4.3 12.2 10.6 6.1 4.8 Others 11.2 7.5 10.2 10.1 12.1 8.1 19.8 19.2 Total CG Goods and Services 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (real percentchange) General Administration 30.8 -22.0 0.8 -39.1 122.2 -57.7 59.4 Defense 10616.3 -16.9 4.5 -46.6 202.3 -62.5 312.0 Public Order 239.9 0.9 -14.8 -24.4 3.5 -29.5 115.6 Education 310.3 -21.3 -4.1 38.5 -14.6 -21.1 38.8 Healthand Social Welfare 31.2 11.5 26.6 -37.6 112.6 -62.4 111.1 ProductiveActivities 69.3 -11.9 -40.3 -7.3 -64.9 129.2 17.0 Infrastructure 69.3 -11.9 -40.3 -7.3 -64.9 129.2 17.0 Other Economic Services 24.6 28.4 5.7 130.3 32.8 -67.9 81.8 Others 62.6 31.O -2.7 -2.7 3.1 34.9 127.2 Total CG Wages and Salaries 142.0 -3.4 -2.1 -18.5 52.9 -44.4 133.5 Source: MEF 139 A9. Fixed Investmentof the Central Governmentby Sectoral Classification 1995 1996 1997 1998 1999 2000 2001 2002 (In millions of US dollars) GeneralAdministration 0 0 2 2 8 3 10 10 Defense 0 12 36 3 0 1 2 43 Public Order 4 9 18 8 6 47 32 79 Education 18 15 15 19 8 11 9 30 Health and SocialWelfare 7 14 18 11 55 29 67 54 Productive Activities 102 71 106 66 195 33 246 146 Infrastructure 157 234 234 305 186 140 251 202 Other Economic Services 88 71 1 0 0 1 10 25 Others 5 72 48 187 2a 205 18 21 Total CG Fixed Investment 381 498 478 602 485 470 645 611 (In percent of GDP) General Administration 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Defense 0.0 0.1 0.2 0.0 0.0 0.0 0.0 0.2 Public Order 0.0 0.0 0.1 0.0 0.0 0.3 0.2 0.3 Education 0.1 0.1 0.1 0.1 0.0 0.1 0.0 0.1 Health and SocialWelfare 0.0 0.1 0.1 0.0 0.3 0.2 0.3 0.2 Productive Activities 0.5 0.3 0.4 0.3 1.2 0.2 1.2 0.6 Infrastructure 0.8 1.1 1.o 1.3 1.1 0.9 1.2 0.8 Other Economic Services 0.4 0.3 0.0 0.0 0.0 0.0 0.0 0.1 Others 0.0 0.3 0.2 0.8 0.2 1.3 0.1 0.1 Total CG Fixed Investment 1.9 2.3 2.0 2.6 2.9 2.9 3.1 2.5 (in percentof total CG fixed investments) General Administration 0.0 0.1 0.4 0.4 1.6 0.6 1.5 1.7 Defense 0.0 2.5 7.6 0.6 0.0 0.2 0.3 7.0 Public Order 0.9 1.7 3.7 1.3 1.3 10.0 5.0 12.9 Education 4.6 3.0 3.2 3.1 1.5 2.4 1.4 4.9 Health and SocialWelfare 1.9 2.8 3.8 1.8 11.3 6.2 10.4 8.8 Productive Activities 26.9 14.3 22.2 11.0 40.2 7.0 38.2 23.9 Infrastructure 41.1 46.9 48.8 50.7 38.3 29.8 38.9 33.1 Other Economic Services 23.1 14.3 0.2 0.1 0.0 0.1 1.6 4.2 Others 1.4 14.5 10.1 31.1 5.8 43.7 2.7 3.4 Total CG Fixed Investment 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (real percent change) GeneralAdministration 358.7 43.2 315.3 -61.O 165.3 -1.7 Defense 6938.4 172.4 -90.4 -100.0.. 90.9 1662.7 Public Order 129.0 93.7 -53.5 7.4 678.0 -45.4 119.8 Education -19.0 -2.7 27.2 -47.7 62.1 -35.8 192.8 Health and SocialWelfare 92.2 22.0 -39.9 578.3 -43.2 83.8 -28.2 Productive Activities -32.3 39.5 -35.4 285.3 -81.9 497.9 -46.9 Infrastructure 44.9 -6.4 35.6 -20.5 -18.9 42.8 -27.9 Other Economic Services -21.6 -98.5 -57.7 -83.3 955.6 1310.5 120.4 Others 1189.9 -37.2 301.9 -80.5 690.3 -93.1 4.5 Total CG FixedInvestment 27.1 -10.1 30.6 5.4 4.1 9.4 -15.3 Source: MEF 140 A10. Transfersof the Central Government by Sectoral Classification 1995 1996 1997 1998 1999 2000 2001 2002 (In millions of US dollars) General Administration 46 49 15 4 0 0 2 2 Defense 106 19 45 12 5 43 124 86 Public Order 100 76 58 4 5 20 57 101 Education 209 94 90 73 13 12 13 19 Health and Social Welfare 27 66 35 18 15 46 132 62 Productive Activities 9 73 58 2 1 9 9 20 Infrastructure 54 23 72 85 15 1 10 11 Other Economic Services 21 30 10 72 108 521 381 323 Others 171 122 226 268 125 29 80 93 Total CG Fixed Investment 741 551 609 538 285 681 808 718 (In percent of GDP) General Administration 0.2 0.2 0.1 0.0 0.0 0.0 0.0 0.0 Defense 0.5 0.1 0.2 0.1 0.0 0.3 0.6 0.4 Public Order 0.5 0.4 0.2 0.0 0.0 0.1 0.3 0.4 Education 1.o 0.4 0.4 0.3 0.1 0.1 0.1 0.1 Health and Social Welfare 0.1 0.3 0.1 0.1 0.1 0.3 0.6 0.3 Productive Activities 0.0 0.3 0.2 0.0 0.0 0.1 0.0 0.1 Infrastructure 0.3 0.1 0.3 0.4 0.1 0.0 0.0 0.0 Other Economic Services 0.1 0.1 0.0 0.3 0.6 3.3 1.8 1.3 Others 0.8 0.6 1.o 1.2 0.8 0.2 0.4 0.4 Total CG FixedInvestment 3.7 2.6 2.6 2.3 1.7 4.3 3.0 3.0 (in percent of total CG transfers) General Administration 6.1 8.9 2.5 0.7 0.0 0.0 0.2 0.3 Defense 14.3 3.4 7.4 2.2 1.6 6.3 15.4 12.0 Public Order 13.5 13.7 9.4 0.7 1.6 2.9 7.0 14.1 Education 28.1 17.1 14.8 13.5 4.5 1.7 1.6 2.6 Health and Social Welfare 3.7 12.0 5.8 3.3 5.1 6.8 16.3 8.7 Productive Activities 1.2 13.3 9.5 0.4 0.3 1.3 1.1 2.8 Infrastructure 7.3 4.1 11.8 15.9 5.1 0.2 1.2 1.6 Other Economic Services 2.8 5.4 1.7 13.4 37.8 76.5 47.1 45.0 Others 23.0 22.1 37.0 49.8 43.9 4.3 9.9 12.9 Total CG Fixed Investment 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 (real percent change) General Administration -71.1 -73.3 -97.5 186.9 658.5 -10.2 Defense -82.9 125.8 -72.2 -49.4 131.9 -38.4 Public Order -26.3 -28.9 -93.1 56.8 359.7 130.2 59.4 Education -56.1 -10.3 -16.3 -77.0 -1.9 -10.4 29.2 Health and Social Welfare 134.8 -49.8 -48.0 7.7 242.6 126.1 -57.7 Productive Activities 723.7 -25.8 -96.4 -42.6 1014.7 -20.6 97.8 Infrastructure -59.1 197.1 23.3 -77.8 -90.3 493.7 1.8 Other Economic Services 40.7 -68.2 641.8 95.4 420.5 -41.8 -24.2 Others -30.5 73.1 23.2 -38.9 -75.1 120.1 3.8 Total CG FixedInvestment -27.7 3.5 -8.4 -30.7 157.1 -5.5 -20.6 Source: MEF 141 A11. Economic Classificationof Defense Expenditure(Nominal, GDP) Prel. Average Average 1995 1996 1997 1998 1999 2000 2001 2002 1995-99 2000-02 (In millionsof USdollars) Total expenditures 499 403 500 622 311 340 443 669 Current 480 444 524 545 292 291 413 546 Wages and salaries 373 379 437 491 270 200 266 356 Goods and services 0 47 41 42 17 48 22 104 interest payments ... ... ... ... ... ... ... ... Transfers 106 19 45 12 5 43 124 06 Capital 20 38 64 77 20 49 29 123 Fixedcapital formation 0 12 36 3 0 1 2 43 Capital transfers 20 26 28 74 20 49 27 80 (In percent of GDP) Total expenditures 2.5 2.3 2.5 2.7 1.9 2.1 2.1 2.0 2.5 2.2 Current 2.4 2.1 2.2 2.3 1.7 1.0 2.0 2.2 2.3 1.9 Wages and salaries 1.0 1.8 1.8 2.1 1.6 1.3 1.3 1.5 1.9 1.4 Goods and services 0.0 0.2 0.2 0.2 0.1 0.3 0.1 0.4 0.1 0.2 Interestpayments ... ... ... ... ... ... ... ... Transfers 0.5 0.1 0.2 0.1 0.0 0.3 0.6 0.4 0.2 0.3 Capital 0.1 0.2 0.3 0.3 0.1 0.3 0.1 0.5 0.2 0.3 Fixed capitalformation 0.0 0.1 0.2 0.0 0.0 0.0 0.0 0.2 0.1 0.0 Capitaltransfers 0.1 0.1 0.1 0.3 0.1 0.3 0.1 0.3 0.2 0.2 Sources: MEF; and BCE 142 A12. Economic Classificationof Education Expenditure (Nominal, GDP) 1995 1996 1997 1998 1999 2000 2001 2002(p) Average Average 1995-99 200042 (In millionsof US dollars) Total expenditures 586 676 605 707 391 357 568 920 Current 548 568 556 586 336 321 484 759 Wages and salaries 335 456 450 499 308 297 458 722 Goods and services 4 19 16 14 15 12 12 19 Interestpayments ... ... ... ... ... ... ... ... Transfers 209 94 90 73 13 12 13 19 Capital 38 108 50 120 54 37 85 160 Fixedcapitalformation 18 15 15 19 8 11 9 30 Capitaltransfers 20 93 34 102 47 26 75 130 (In percentof GDP) Total expenditures 2.9 3.2 2.6 3.0 2.3 2.2 2.7 3.8 2.9 2.8 Current 2.7 2.7 2.4 2.5 2.0 2.0 2.3 3.1 2.6 2.4 Wages and salaries 1.7 2.1 1.9 2.1 1.8 1.9 2.2 3.0 2.0 2.2 Goods and services 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 Interestpayments ... ... ... ... ... ... ... ... ... ... Transfers 1.o 0.4 0.4 0.3 0.1 0.1 0.1 0.1 0.5 0.1 Capital 0.2 0.5 0.2 0.5 0.3 0.2 0.4 0.7 0.4 0.4 Fixed capitalformation 0.1 0.1 0.1 0.1 0.0 0.1 0.0 0.1 0.1 0.1 , Capitaltransfers 0.1 0.4 0.1 0.4 0.3 0.2 0.4 0.5 0.3 0.3 Sources: MEF; and BCE A13. Economic Classification of Health and Social Welfare Expenditure (Nominal, GDP) 1995 1996 1997 1998 1999 2000 2001 2002(p) Average Average 1995-99 2000-02 (In millions of US dollars) Total expenditures 274.8 344.6 277.6 256.0 196.2 198.3 363.7 443.7 Current 267.3 268.4 241.1 206.8 122.2 165.3 251.5 307.0 Wages and salaries 213.7 167.0 163.4 137.4 82.9 70.1 96.6 190.2 Goods and services 26.4 35.6 42.4 51.7 24.7 48.8 23.0 54.4 Interestpayments ... ... ... ... ... ... ... ... Transfers 27.3 65.9 35.3 17.7 14.6 46.5 131.8 62.4 Capital 7.5 76.2 36.4 49.3 74.0 33.0 112.3 136.7 Fixedcapitalformation 7.1 14.0 18.2 10.6 54.8 28.9 66.7 53.6 Capitaltransfers 0.4 62.2 18.2 38.7 19.2 4.0 45.5 83.1 (Inpercent of GDP) Total expenditures 1.4 1.6 1.2 1.1 1.2 1.2 1.? 1.8 1.3 1.5 Current 1.3 1.3 1.o 0.9 0.7 1.o 1.2 1.3 1.1 1.1 Wages and salaries 1.1 0.8 0.7 0.6 0.5 0.4 0.5 0.8 0.8 0.5 Goods and services 0.1 0.2 0.2 0.2 0.1 0.3 0.1 0.2 0.2 0.2 Interestpayments ... ... ... ... ... ... ... ... ... ... Transfers 0.1 0.3 0.1 0.1 0.1 0.3 0.6 0.3 0.2 0.3 Capital 0.0 0.4 0.2 0.2 0.4 0.2 0.5 0.6 0.2 0.4 Fixedcapitalformation 0.0 0.1 0.1 0.0 0.3 0.2 0.3 0.2 0.1 0.3 Capitaltransfers 0.0 0.3 0.1 0.2 0.1 0.0 0.2 0.3 0.1 0.2 Sources:MEF; and BCE 143 A14. Number of Ascribed lnstitutiosto Line Ministries 1995 1996 1997 1998 1999 2000 2001 2002 Number of institutions 65 54 51 51 46 42 41 46 Administrationand Planning 6 6 6 6 1 1 1 1 Environment 1 1 1 1 1 1 1 2 Justice 5 6 6 6 6 6 6 6 National Defense 1 1 1 1 1 1 1 1 ExternalRelations 0 0 0 0 0 0 0 1 Educationand Culture 10 10 10 10 10 8 8 8 Human Development 3 2 2 2 2 2 2 5 Work and Human Resources 3 2 2 2 2 1 1 1 Health 3 3 3 3 3 2 2 2 Agriculture 23 16 14 14 14 14 14 13 Energyand Mining 4 3 3 3 3 3 2 2 Industry 5 4 3 3 3 3 3 3 Informationand Tourism 1 0 0 0 0 0 0 0 Public Works and Communication 0 0 0 0 0 0 0 1 Housingand Urban Development 0 0 0 0 0 0 0 0 Source: MEF. A15. Expendituresof Ascribed Institutionsto LineMinistries (In percentof the sector's total expenditures) 1995 1996 1997 1998 1999 2000 2001 2002 Administrationand Planning 20.3 21.o 48.3 58.5 42.3 6.8 29.1 12.7 Environment 100.0 100.0 59.5 41.7 17.7 25.1 4.2 19.6 Justice 50.1 59.9 64.7 94.7 94.6 89.0 95.3 93.1 NationalDefense 0.6 1.o 6.0 0.6 0.5 0.4 0.6 0.8 ExternalRelations 0.0 0.0 0.0 0.0 0.0 0.0 0.0 11.7 Educationand Culture 3.5 2.1 2.7 2.1 2.1 1.2 0.9 0.5 HumanDevelopment 2.9 0.8 0.9 2.8 5.8 18.3 13.0 14.4 Work and Human Resources 54.9 63.0 65.7 63.2 64.6 7.6 58.9 51.3 Health 8.3 8.7 5.0 6.0 4.1 2.0 3.7 4.5 Agriculture 88.5 74.1 90.0 83.2 91.5 71.6 78.9 77.9 Energy and Mining 19.1 12.3 3.0 1.5 5.4 16.5 10.1 10.0 Industry 37.4 31.3 36.4 33.8 19.4 11.6 16.5 8.5 Informationand Tourism 50.3 0.0 0.0 0.0 0.0 0.o 0.0 0.0 PublicWorks and Communication 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.2 Total 16.1 17.2 20.1 15.5 21.o 11.2 19.4 19.5 Source: MEF. 144 A16. Operationsof Municipal & ProvincialCouncils 2001 2002 2003 2004 (In millions of US$) Revenue 814 1032 1079 1125 Current revenue 137 251 274 293 Interestand profits 1/ 54 40 60 63 Other 83 211 214 229 Transfers 678 782 805 832 . Capital transfersfrom budgetary operations 562 618 642 635 25% incometax sharing 30 38 54 Fondo SalvamientoCultural 24 24 27 30 From the rest of public sector 92 110 98 114 From FODESEC 39 73 80 90 From others 53 37 18 24 Expenditure 706 982 1102 1160 Current 376 495 621 644 Wages and salaries 108 137 169 171 Purchasesof goods and services 261 345 352 367 Interest payments 2 2 3 2 External 2 2 3 2 Current transfers 5 12 97 104 To IESS 4 7 10 11 To other public sector 0 3 27 29 To privatesector 0 2 60 64 Capital 329 487 480 516 Overall surplus or deficit (-) 109 50 -23 -35 Source: MEF; and IMFstaff estimates and projections 145 1 I --- t8 ;P 3 6 T X - A19. Social Expendituresas Percent of Central Government Budget Transfers to the Social Sector 1994 1995 1996 1997 1998 1999 2000 2001 Mlns US$ Education 464.5 562.1 624.2 563.3 638.0 368.2 310.7 492.8 Social Welfare 83.5 70.4 118.8 70.9 59.4 69.4 76.2 126.8 Work 4.6 7.0 10.3 9.8 9.0 5.4 6.3 5.6 Health 127.3 194.8 204.3 191.4 186.0 110.1 121.2 188.6 Urban Development 99.0 67.3 47.7 50.8 54.0 11.5 31.7 134.9 Total Social Budget 779.0 901.6 1005.4 886.3 946.5 564.5 546.2 948.7 Percentof CG Budget Education 19% 17% 18% 16% 19% 15% 11% 12% Social Welfare 3% 2% 3% 2% 2% 3% 3% 3% Work 0Yo Health 5yo 6% 6% 5% 5% 4% 4% 4% Urban Development 4% 2% 1Yo 1% 2% 0Yo 1Yo 3% Total Social Budget 31yo 27% 29% 25% 28% 22% 19% 23% Notes: Public Debt Interest I Social Budget 53% 90% 61Yo 89% 78% 120% 158% 193% Total Central Government Budget 2487.2 ,3311.3 3408.6 3595.3 3400.2 2532.9 2831.7 4196.5 Source: MEF. 148 A20. Personnel Expenditures In Education (In US$) 1995 199s 1997 1998 1999 2000 Ministry of Educationand Culture 204,160,433 422,769,422 350,793,654 454,234,386 308,753,551 245.079.025 National lnst of CulturalP a t n m y 584,828 426,704 382.182 399,948 240,455 206,772 Ecuad. MuseumNaturalC. 76.066 87,020 65,206 55,971 29,722 24,769 National Unit 01 Dance 54,931 68.205 58,310 50.872 27,912 25.928 National Systemof Archives and National Archive 53.197 60.663 54.356 56,433 38.230 22,579 National Systemof bbranes 6.434 11,419 10,185 10,255 4,763 2.216 NationalCouncilof Cutlure I 17,490 99,004 108,032 87,105 48.580 37,872 SENACM 39,679 32.364 256,860 10.847 National Council of Sports 46,189 73,219 83,876 73.858 51,654 38.900 OINACE 1,055,643 I.I60.315 1,040,912 1,081,383 320,680 549.625 DINADER 440.283 352,657 419.547 451,652 290,598 3,215 TOTAL 206,597,690 425,148,303 353,050,621 456,761,321 309,818,992 245,992,900 TOTAL EDUCATIONSECTOR 560,208,644 606,816,064 525,64037 623,959,767 370,794,090 283,509,503 New series total (UNICEF) 523,584,971 604,390,865 550,140,576 596,302,775 400,819,048 329,276,625 Newseries MINEF 562 624 563 638 368 311 %TOTAL EDUCATIONEXPENDITURE 36.9% 69.8% 67.2% 73.2% 83.6% 86.8% %of UNICEF total 39.5% 70.3% 642% 76.6% 77.3% 74.7% %of MINEFtotal 36.8% 68.1% 62.7% 71.6% 84.1% 79.296 Source: Ministenode Economiay Finanzas,1995-2ooO A21. Expenditures in Education by Level (In Mln US$) 1995 1996 1997 1998 1999 2000 2001 2002 Basic Hispanic Education 193.9 212.7 191.7 225.9 145.7 117.7 155.0 284.5 Bilingual Basic Education 7.9 9.8 9.5 11.8 7.7 6.8 9.3 16.0 Intermediate Education 194.0 249.6 203.9 227.7 154.1 118.6 159.3 304.1 Advanced Education 127.8 132.3 145.1 130.9 17.7 11.2 nd nd TOTAL 523.6 6044 550.1 596.3 325.1 254.3 323.6 W . 6 Share Basic Hispanic Education 37% 35% 35% 38% 45% 46% 48% 47% Bilingual Basic Education 2% 2% 2% 2% 2% 3% 3% 3% IntermediateEducation 37% 41% 37% 38% 47% 47% 49% 50% Advanced Education 24% 22% 26% 22% 5% 4% TOTAL 1w/. 1Wh 100% 100% 100% 1 W h 1W/. 1 W h Source: Ministeriode Economiay Finanzas,1395-2002 A22. Budget Issued For Education (US S) 1999 PersonelExoenses 309,816,993 Services 5,092,123 Subministries and Materials 7,978.724 Fumiture 247,366 Public Works 22,532,316 Transfers 25,125,959 TOTAL 370,794,088 Salaries (%) . 84% A23. Financing Sourceslor Education, 2003 (In US$) Financing Source I Distribution by Financing Source Own External Domlstlc Own External Domistlc Sector Fiscal Resources Credlt Credlt TOTAL Fiscal Resources Credlt Credlt Ministry of EducationandCutlure 760.2 23.7 15.3 31.3 830.5 91.5 2.9 1.8 3.8 National Inst. of Cutlural Patrimony 1.2 1.2 0.0 0.0 2 3 49.8 50.2 0.0 0.0 Ecuad. MuseumNaturalC. 0.2 0.0 0.0 0.0 0.2 93.8 6.3 0.0 0.0 National Unit of Dance 0.1 0.0 0.0 0.0 0.1 97.2 2.8 0.0 0.0 National Systemof Archives and National Archive 0.3 0.0 0.0 0 0 0 3 96.3 3.7 0.0 0.0 National Systemof Libraries 0.1 0.0 0.0 0.0 0.1 81.8 18.2 0.0 0.0 NationalCouncilof Culture 0.2 0.5 0.0 0.0 0.6 23.4 76.6 0.0 0.0 SENACYT 0.0 0.2 0.0 0.2 0.4 0.0 53.5 0.0 46.5 TOTAL 762.1 25.7 15.3 31.5 834.5 91.3 3.1 1.8 3.8 149 A24. Supply of publicand private healthcenters, 1975 2000 - Total In-patientcare Only out-patient Public Private Total Public Private Sub-total Public Private Sub-total Establishments per 100,000 inhabitants 1975 17.1 3.2 13.9 1980 19.5 2.7 22.2 1.8 1.6 3.4 17.7 1.1 18.8 1990 24.7 3.5 28.2 1.6 2.3 3.9 23 1.2 24.2 2000 24.6 4.5 29.1 1.4 3.3 4.7 23.2 1.2 24.4 Number of establishments 1975 1,182 224 958 1980 1,550 216 1,766 144 128 272 1,406 88 1,494 1990 2,533 359 2,892 169 235 404 2,364 124 2,488 2000 3,027 555 3,582 178 405 583 2,849 150 2,999 Source: INEC, Anuario de recursos y acfividadesde salud, various years. 150 A26. Expendltures managed bythe Regional Development Organlzatlons, 2003 Institutions Mlns US$ Central Government Entitles 171.29 Center for Economic Reconversion of Austro (CREA) 3.95 RehabilitationCenter of Manabi (CRM) 38.47 DevelopmentCommission Cuenca Rio Guayas (CEDEGE) 39.36 Subcom. Ecua.Des.Cuencas Puyango-Tumbez(PREDESUR) 20.56 RegionalCorporationof the Sierra Centro (CORSICEN) 3.81 Regional Corporation of the Sierra Norte (CORSINOR) 2.53 Regional Development Corporation of El Or0 (CODELORO) 7.09 Regional DevelopmentCorporation of Chimborazo(CODERECH) 1.30 Regional DevelopmentCorporation of Cotopaxi(CODERECO) 2.36 National Instituteof Galapagos (INGALA) 1.61 AfroecuadorianDevelopment Corporation 1.93 North Development Unit (UDENOR) 39.11 Ecuador Nationaland Town Development Council 8.71 Pueblo Montubio and Subtropical Zones Development Council (CODEPMOC) 0.50 Autonomous Entities 75.46 ExecutiveCorporationfor the Reconstruction of El NiAoAffected Zones 47.24 Eco-DevelopmentInstitute of the Amazonic Region (ECORAE) 6.09 Galapagos NationalPark 7.21 Group of Hydraulic Resources Jipijapa, Pajan Puerto Lopez 14.92 TOTAL 246.75 A27. Military Expenditures 1998 1999 2000 2001 2002 Percenf of Govemment Expenditures Ecuador 9.5 7.5 7.6 12.4 Bolivia 6.0 6.2 5.3 4.8 Chile 17.9 16.4 17.2 15.7 Colombia 9.0 9.0 8.5 11.5 Peru 8.8 8.4 8.0 8.3 LAC mean 6.0 5.8 5.4 5.7 Developedeconomies 6.1 6.0 6.1 6.3 Africa 14.0 14.2 13.4 10.4 Percent of GDP Ecuador 2.3 2.0 2.0 3.0 3.0 Bolivia 1.8 1.8 1.6 1.6 1.6 Chile 3.8 3.7 3.9 3.6 4.1 Colombia 2.5 2.7 2.6 3.7 3.7 Peru 1.7 1.8 1.7 1.7 1.6 LAC mean 1.5 1.6 1.5 1.6 na Developedeconomies 2.3 2.2 2.2 2.4 na Africa 3.5 3.5 3.5 2.8 na Source: International Institute for Strategic Studies (IISS)Yearbook 1997-2003 152 A29. Direct Subsidies to Water and SanitationSector 2001 2002 2003 Subsidies from Central Government A. MlDUVl 64,000 17,743 22,490 B. I.C.E. 50,000 46,000 45,000 Sub-totalsubsidiesfrom Central Govemment: 114,000 63,743 67,490 Subsidiesfrom Municipalities 32,000 N.A. N.A. Source: MlDUVl- SAPYSB, Estadode Ejecuccion presupuestaria,2002y 2003, and Ecuador InfrastructurePolicy TOTAL cl 146,000 63,743 67,490 Notes.World Bank2003 154 A30. Illiteracy rate and years of schooling of adult population (25 yrs and older), 1990-2001 ILLITERACY RATE YEARS OF SCHOOLING 1990 2001 1990 2001 BY PROVINCE Azuay 13.8 8.8 5.9 6.9 Bolivar 22.1 17.5 4.4 5.4 Caiiar 20.8 15.4 4.4 5.3 Carchi 10.1 7.2 5.4 6.3 Chimborazo 27.0 19.0 4.6 5.8 Cotopaxi 23.7 17.6 4.4 5.4 ElOro 5.9 5.5 7.1 7.5 Esmeraldas 14.5 11.6 5.7 6.2 Galapagos 2.9 2.7 9.2 9.5 Guayas 7.4 7.1 7.7 7.8 Imbabura 18.4 13.4 5.3 6.2 Loja 10.7 7.9 5.9 6.9 Los Rios 15.8 11.7 5.3 6.1 Manabi 15.5 12.5 5.3 6.1 Morona Santiago 12.3 10.0 5.3 5.8 Nap0 15.7 10.5 5.3 6.4 Orellana 13.5 9.2 4.9 5.9 Pastaza 14.3 10.1 6.1 7.1 Pichincha 7.3 5.5 8.5 8.9 Sucumbios 10.9 8.5 5.2 6.0 Tungurahua 14.0 10.0 5.9 6.8 Zamora Chinchipe 9.9 8.2 5.5 6.2 Non-delimited areas 15.8 12.4 4.5 5.1 BY GENDER Men 9.5 7.7 7.1 7.5 Women 13.8 10.3 6.3 7.1 BY AREA Rural 20.8 15.5 4.0 4.9 Urban 6.1 5.3 8.3 8.7 BY ETHNIC GROUP Indigenous ma. 28.2 n.a. 3.3 Blacks n.a. 11.6 n.a. 5.9 Other n.a. 7.4 n.a. 7.6 NATIONAL AVERAGE 11.7 9.0 6.7 7.3 Source: PopulationCensus, 1990 and2001. 155 A31. Ecuador: Population(25 years and older) by completed education level ( percent of total) COMPLETED COMPLETED PRIMARY SECUNDARY COMPLETED TERTIARY 1990 2001 1990 2001 1990 2001 BY PROVINCE Azuay 62.6 61.4 21.5 21.3 11.5 16.5 Bolivar 51.8 47.9 13.2 13.5 6.4 10.9 Caiiar 54.5 49.1 12.5 12.0 5.6 8.9 Carchi 60.4 57.8 15.4 13.9 7.3 11.0 Chimborazo 53.6 52.9 16.9 18.6 9.4 14.6 Cotopaxi 54.0 50.7 14.4 13.3 6.8 10.1 ElOro 74.3 71.2 23.1 20.6 12.3 16.2 Esmeraldas 53.6 56.1 17.3 18.3 11.1 14.7 Galapagos 86.4 85.6 35.4 34.0 20.9 29.5 Guayas 73.9 71.5 27.0 24.8 16.8 19.6 Imbabura 58.2 55.3 18.3 18.0 9.8 14.9 Loja 64.8 63.1 18.5 20.8 11.7 17.2 Los Rios 56.6 57.1 14.3 14.8 9.0 11.4 Manabi 54.1 54.0 16.6 17.1 10.0 13.4 Morona Santiago 63.1 54.6 14.1 12.1 5.7 10.4 Nap0 62.3 65.0 15.0 15.6 6.7 12.6 Orellana 58.1 58.9 10.8 10.7 4.3 8.7 Pastaza 67.2 65.7 23.2 19.1 10.1 16.4 Pichincha 78.7 77.9 36.2 32.6 22.6 26.9 Sucumbios 59.0 58.8 11.6 12.6 5.6 9.9 Tungurahua 66.0 61.6 19.8 19.1 11.3 15.7 Zamora Chinchipe 66.2 62.1 13.4 13.8 7.1 11.7 Non-delimitedareas 51.0 48.1 9.6 9.0 4.4 6.2 BY GENDER Men 69.0 66.6 24.4 22.7 15.9 18.7 Women 64.8 64.8 22.9 22.5 12.5 17.5 BY AREA Rural 48.8 45.4 8.3 8.7 4 6.3 Urban 78.3 77.7 33.0 30.5 20.4 24.8 BY ETHNIC GROUP Indigenous n.a. 31.5 n.a. 4.7 ma. 3.4 Blacks n.a. 55.6 n.a. 13.4 ma. 10.6 Other n.a. 69.0 n.a. 24.7 n.a. 20.2 NATIONAL AVERAGE 66.8 66.8 23.6 22.6 14.2 18.1 Source: PopulationCensus, 1990and 2001. 156 A32. Ecuador: net enrolment rates by educational level, 1990and 2001 NET ENROLMENT RATES PRIMARY SECONDARY TERTIARY 1990 2001 1990 2001 1990 2001 BY PROVINCE Pichincha 92.3 93.0 58.3 55.3 17.6 18.3 Azuay 90.9 92.7 37.3 43.3 11.8 16.5 Loja 89.7 92.4 36.3 42.3 11.8 15.2 Chimborazo 88.3 90.4 36.6 41.5 11.0 14.9 Tungurahua 92.6 93.6 43.4 45.2 11.0 13.4 Guayas 90.9 90.2 49.7 47.3 12.0 11.3 Imbabura 89.7 90.8 40.9 39.8 12.0 10.6 ElOro 91.6 92.9 47.8 50.8 8.5 9.3 Cotopaxi 88.7 88.9 34.1 36.5 7.4 9.0 Manabi 82.8 86.5 32.0 36.8 6.8 8.9 Caiiar 89.0 91.1 30.7 34.9 6.3 7.8 Bolivar 84.2 89.1 31.2 39.4 5.0 7.1 Pastaza 87.0 90.6 42.1 46.2 3.1 7.3 Carchi 90.9 91.6 38.2 44.9 6.4 6.4 Galapagos 94.4 95.0 48.9 61.1 2.5 5.6 Los "os 85.0 85.8 34.6 36.2 6.0 5.4 Zamora Chinchipe 86.9 90.7 26.3 39.1 2.9 5.3 Nap0 87.8 91.9 35.4 27.0 1.8 4.5 Esmeraldas 81.0 83.0 32.6 35.8 4.3 4.5 Morona Santiago 88.3 88.1 32.7 30.9 1.5 2.5 Sucumbios 84.5 86.6 21.8 35.3 0.5 2.4 Non-delimited areas 79.7 81.0 24.2 40.3 1.3 2.2 Orellana 85.6 87.3 17.9 31.5 0.4 1.5 BY GENDER Men 88.6 89.9 42.0 43.9 10.3 11.1 Women 89.2 90.4 44.1 45.4 11.3 12.6 BY AREA Rural 84.4 86.7 23.2 28.8 3.2 4.3 Urban 92.5 92.7 57.7 55.7 15.4 16.2 BY ETHNIC GROUP Indigenous n.a. 86.2 n.a. 22.6 n.a. 2.4 Blacks n.a. 84.4 n.a. 32.3 n.a. 4.5 Other n.a. 89.8 n.a. 43.2 n.a. 10.9 NATIONAL AVERAGE 88.9 90.1 43.1 44.6 10.9 11.9 Source: PopulationCensus, 1990and 2001. 157 A33. Ecuador:Years of schoolingby consumptionquintiles and gender, 1995 and 1999 Quintiles Sex 1995 1999 Poorest20 percent Females 3.5 3.2 Males 4.1 4.1 Total 3.8 3.7 SecondQuintile Females 4.7 5.0 Males 5.6 5.6 Total 5.I 5.3 Third Quintile Females 6.3 6.7 Males 6.5 7.0 Total 6.4 6.8 FourthQuintile Females 7.9 8.2 Males 8.0 8.4 Total 7.9 8.3 Richest20 percent Females 10.2 10.8 Males 10.7 11.6 Total 10.4 11.2 Total Females 7.0 7.3 Males 7.4 7.8 Total 7.2 7.5 Source: Calculations based on LSMS (ECV) surveys for 1995 and 1999. A34. Ecuador:Years of schoolingby consumptionquintilesandarea, 1995and 1999 Quintiles Area 1995 1999 Poorest20 percent Rural 3.2 3.1 Urban 5.2 5.3 Total 3.8 3.7 SecondQuintile Rural 4.3 4.2 Urban 6.2 6.5 Total 5.I 5.3 Third Quintile Rural 4.4 5.1 Urban 7.6 8.0 Total 6.4 6.8 FourthQuintile Rural 5.7 6.5 Urban 8.8 9.0 Total 7.9 8.3 Richest20 percent Rural 6.2 7.8 Urban 11.1 11.8 Total 10.4 11.2 Total Rural 4.4 4.8 Urban 8.8 9.2 Total 7.2 7.5 Source: Calculations based on LSMS (ECV) surveys for 1995 and 1999. 158 A35. Ecuador: Net enrolment rates by educational level and quintiles of p.c. consumption Primary Secondary University Quintiles Sex 1995 1999 1995 1999 1995 1999 Poorest20 percent Females 85.7 82.4 22.4 16.9 1.9 0.5 Males 81.7 83.4 15.1 21.0 1.7 2.8 Total 83.7 82.9 18.6 19.0 1.8 1.6 Second Quintile Females 92.6 92.3 44.7 43.6 5.5 5.2 Males 88.0 88.0 36.5 43.0 6.2 5.0 Total 90.I 90.0 40.5 43.3 5.8 5.I ThirdQuintile Females 89.4 92.5 , 56.5 57.4 8.2 9.2 Males 92.3 94.8 52.1 54.1 4.2 6.1 Total 90.9 93.7 54.3 55.8 6.3 7.6 Fourth Quintile Females 92.3 95.6 69.4 66.6 15.7 16.4 Males 91.5 93.4 64.9 73.0 8.6 18.9 Total 91.9 94.4 67.I 69.7 12.3 17.7 Richest 20 percent Females 91.9 92.4 74.2 77.3 20.3 36.3 Males 88.8 92.4 73.8 79.1 26.4 38.9 Total 90.4 92.4 74.0 78.I 22.9 37.6 Total Females 90.0 90.3 52.8 51.6 11.2 14.2 Males 88.1 90.2 46.4 51.4 9.5 15.4 Total 89.0 90.3 49.6 5z.5 10.4 14.8 Source: Calculations based on LSMS (ECV) surveys for 1995 and 1999. A36. Ecuador: Test scores for languageand mathematics skills, by gender, regionand area. BY GENDER 1996 Boys Girls Second grade Spanish language skills 10.26 10.64 Mathematics 9.43 9.14 Sixth grade Spanish language skills 10.74 11.59 Mathematics 7.35 6.98 Ninthgrade Spanish language skills 12.54 13.16 Mathematics 7.53 7.07 B Y REGION 1996 1997 2000 Costa Sierra Costa Sierra Costa Sierra Second grade Spanish language skills 10.00 10.74 7.57 9.25 8.97 10.15 Mathematics 8.74 9.75 6.59 8.14 8.03 9.15 Sixth grade Spanish language skills 10.65 11.51 8.38 10.48 9.58 10.09 Mathematics 6.60 7.57 4.11 5.80 5.74 6.45 Ninthgrade Spanish language skills 11.81 13.73 10.21 12.45 11.81 11.47 Mathematics 6.60 7.86 4.56 6.41 6.00 6.02 159 BY AREA and 1996 1997 Urban Urban Urban Urban PUBLICPRIVATE private Public Rural private Public Rural Second grade Spanish language skills 12.36 10.21 8.73 10.65 8.36 7.04 Mathematics 10.96 9.04 8.12 8.88 7.29 6.38 Sixth grade Spanish language skills 13.37 10.73 9.41 11.98 9.43 7.51 Mathematics 8.27 6.95 6.29 6.37 4.88 3.92 Ninth grade Spanish language skills 14.29 12.22 11.95 12.83 10.64 10.50 Mathematics 8.55 6.68 6.66 6.80 4.9 1 4.70 Notes: Test scores are on a scale of 20 with 13 considered as the pass grade. Ninthgrade refers to third year o f secondary school. Data have not been processed by the APRENDO program for all years at the given levels o f disaggregation. As indicated in the text, the national test score program was discontinued after 2000. Source: APRENDO. A37. Ecuador:Internal efficiencyindicatorsfor primary education, 1995 and 2001 Retentionrate (%) Years to complete Efficiency indicador(%) (5thgrade) school (5thgrade) 1995 2001 1995 2001 1995 2001 BY PROVINCE Azuay 92.68 96.27 6.46 6.56 92.83 91.49 Bolivar 82.04 75.14 7.14 7.10 83.98 84.47 Caiiar 86.66 83.34 7.21 7.13 83.24 84.18 Carchi 92.18 71.41 6.74 7.78 89.04 77.09 Cotopaxi 84.24 81.40 7.04 6.88 85.19 87.26 Chimborazo 80.70 82.96 7.28 6.97 82.44 86.08 ElOro 88.44 91.27 6.96 6.72 86.18 89.24 Esmeraldas 52.27 80.38 8.58 6.63 69.96 90.54 Guayas 79.81 74.57 6.69 6.93 89.70 86.53 Imbabura 93.32 86.60 6.82 6.84 88.02 87.66 Loja 87.35 55.55 6.70 9.25 89.58 64.89 Los Rfos 71.74 69.96 7.40 7.61 81.50 78.88 Manabi 87.89 71.42 6.38 7.22 94.02 83.11 Pichincha 90.46 94.77 6.61 6.40 90.71 93.69 Tungurahua 98.51 87.65 6.43 6.64 93.37 90.35 BY GENDER Men 84.13 81.50 6.75 6.84 88.90 87.74 Women 84.29 81.37 6.79 6.86 88.36 87.43 BY AREA Urban 87.69 84.58 6.56 6.68 91.41 89.78 Rural 76.34 74.16 7.32 7.30 81.95 82.14 Nationalaverage 84.22 81.44 6.77 6.85 88.63 87.59 Note: Efficiency indicator as defined by UNESCO stands for the ratio of the number of students of a particular cohort which completed a given level o f education with respect to the `ideal' number of students that would have reached that level with zero repetition and desertion. Source: SINEC data base. 160 A38. Latin America: Internal efficiency indicatorsfor primary education, 1990s Primary educationefficiency indicator ( percent) Andean countries Bolivia (1995-99) 81.5 84.5 83.0 Colombia (1991-95) 75.6 70.4 73.0 Ecuador(WB, 1995-99) 79.4 76.4 77.8 Ecuador(SINEC, 1997-2001) 87.4 87.7 87.6 Peru(1994-98) 87.6 88.2 87.8 Venezuela (1994-98) 94.3 87.6 90.8 Other Latin American countries Argentina (1995-99) 90.5 90.1 90.3 CostaRica (1995-99) 84.2 76.7 80.2 Dominican Republic (1994-98) 79.1 71.4 75.1 Mexico (1995-99) 89.5 87.5 88.5 Uruguay (1994-98) 88.4 93.2 90.8 Note: Efficiency indicator as defined by UNESCO stands for the ratio of the number of students of a particular cohort which completed a given level of education with respect to the `ideal' number of students that would have reached that level with zero repetition and desertion. Source: World Bank, World Development Indicators (2003). Second estimates for Ecuador are from SINEC data base. A39. Ecuador: Differential versus average unit costs in estimation the public expenditure incidence in education, 1999 PublicPrimary PublicSecondary PublicTertiary Private Tertiary unit cost estimation: differential average differential average differential average dgferential average Poorest20 percent 35% 30% 15% 10% 3% 2% 0% 0% Secondquintile 26% 26% 23% 22% 12% 8% 1% 1% Third Quintile 20% 21% 26% 26% 16% 13% 6% 5% Fourth Quintile 13% 16% 22% 27% 28% 35% 22% 14% Richest 20 percent 6% 7% 14% 15% 40% 42% 70% 79% Source: INEC, Encuestade Condicionesde Vida (ECV), 1999andeducationexpenditure data. A40. Ecuador: "Marginal" expenditure incidenceineducation, 1995 and 1999 Private Per capita Primary Secondary PublicTertiary University Total Education consumption Deciles 1995 1999 1995 1999 1995 1999 1995 1999 1995 1999 1995 1999 1 17.9% 15.8% 3.9% 3.3% 0.0% 0.0% 0.0% 0.0% 8.0% 6.1% 2.2% 1.9% 2 14.2% 14.3% 6.1% 6.3% 3.0% 2.4% 0.8% 0.0% 8.2% 7.2% 3.5% 3.1% 3 14.2% 12.3% 9.6% 10.7% 4.2% 3.3% 4.0% 1.2% 10.1% 8.4% 4.6% 4.0% 4 11.9% 13.7% 11.9% 11.1% 4.6% 5.0% 2.8% 0.3% 10.3% 9.3% 5.6% 4.9% 5 11.3% 11.1% 13.0% 12.7% 6.0% 8.0% 4.5% 0.6% 10.9% 9.7% 6.6% 5.9% 6 9.7% 10.1% 13.1% 13.6% 11.3% 4.8% 2.0% 4.5% 11.0% 9.5% 7.9% 7.1% 7 7.4% 8.3% 13.1% 13.2% 12.1% 15.7% 5.3% 3.4% 10.5% 11.0% 9.6% 8.8% 8 6.6% 7.2% 13.8% 14.3% 17.4% 19.3% 5.2% 11.0% 11.3% 12.8% 11.9% 11.2% 9 4.1% 4.5% 9.4% 9.4% 18.4% 20.4% 24.5% 29.1% 9.8% 12.6% 15.9% 15.7% 10 2.7% 2.6% 6.0% 5.5% 22.9% 21.2% 50.8% 50.0% 9.9% 13.4% 32.1% 37.3% Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% 100.0% Note: Datarefer to benefit incidenceineachyear usingaverageunit costs. The differencebetweenthe two years mightbeinterpretedas the "marginal" effect. Source: INEC, Encuesta de Condicionesde Vida(ECV), 1995 and 1999andeducationexpendituredata. 161 A41. Ecuador: Cost-effectivenessmodelestimates for netenrolment inprimary education (linear multinomial logit specification) Poor Non-poor Coefficient ey/ex Significance Coefficient ey/ex Significance Public primary schools Educationcosts -0.000027 -0.1917 0.0000 -0.000001 -2.5837 0.7130 Sex -0.5967 0.0114 0.2160 -0.3778 -0.0749 0.3980 Age 2.2607 -0.3932 0.1510 -0.7220 2.7793 0.4450 Age squared -0.1155 0.2491 0.2460 0.0569 -1.1970 0.2630 Educationlevelfather -0.0941 -0.1189 0.1270 -0.0618 -0.5482 0.4080 Educationlevel mother 0.3205 0.1282 0.0000 0.1544 -0.2880 0.2100 Location: Quito = 1 1.0787 0.0366 0.0320 1.1219 0.1464 0.1580 Region (Costa =1) -1.1744 -0.1365 0.0580 3.9791 -0.3372 0.0020 Students per classroom 0.1237 0.1130 0.0180 -0.0106 -0.7235 0.9350 Percentage of teachers with teaching qualification 6.0027 0.0806 0.0290 17.8841 1.6331 0.0000 Percentage of teachers with university degree -0.5159 0.0190 0.8410 -14.9457 -0.3393 0.0000 Percentage of teachers appointed by MoE -0.7781 -0.1414 0.7480 -13.6122 -0.3799 0.0000 Constant -15.1981 0.0330 0.0907 0.9880 Private primary schools Educationcosts 0.0000 1.3234 0.0010 0.0000 0.7400 0.0000 Sex -0.9755 -0.1890 0.0210 -0.2050 0.0207 0.6650 Age 3.3550 8.9130 0.0010 -1.1540 -0.8189 0.2700 Age squared -0.1804 -4.6386 0.0070 0.0785 0.3583 0.1700 Educationlevelfather 0.0442 0.8302 0.5750 0.0033 0.1547 0.9670 Educationlevelmother 0.2249 -0.4937 0.0250 0.1910 0.0884 0.0770 Location: Quito = 1 -0.3555 -0.2506 0.5340 0.2714 -0.0411 0.7480 Region (Costa =1) 0.6222 0.9598 0.4920 4.7639 0.1050 0.0000 Students per classroom 0.1235 0.1054 0.0370 0.0208 0.2062 0.8740 Percentage of teachers with teaching qualification 6.8636 0.8397 0.0110 15.5642 -0.4099 0.0010 Percentage of teachers with university degree -0.9919 -0.2450 0.7280 -14.2165 0.0665 0.0030 Percentage of teachers appointed by MoE 3.5209 1.1420 0.2750 -12.0046 0.0941 0.0010 Constant -25.9500 0.0000 -1.1238 0.8740 Source: Model estimations based on 1999 LSMS survey. See Appendix A.2 in Vos (2004a) for model specification. 162 A42. Determinants of years of schooling and net enrolment in primary and secondary education at the district (cantonal) level, 2001 Dependent variable: Years of schooling Model 1 Model 2 Model 3 Model4 Share of rural population -0.040962* 4.04001* -0.035997* Share of indigenous population -0.011259' Poverty incidence (UBN) 4.07831 * Regional dummy for Costa -0.184024 0.048407 -0.049981 0.74270* Regional dummy for Oriente 0.763725* 0.754012* 0.980182* 0.91284* Pupils per classroom(primary schools) -0.01281* -0.012131** -0.002955 Pupils per classroom(secondary schools) 0.001271 0.0000025 -0.000566 Share of one-teacher (unidocente) schools -0.001820 -0.004896 0.001853 Teacher quality' 0.077745*** 0.050195 constant 8.17978* 8.516025 7.971077 10.8162* R2 0.5579 0.5579 0.588 0.7716 Dependent variable: Net enrolment inprimary education Model 1 Model 2 Model 3 Model 4 Shareof ruralpopulation -0.085465* 4.071468* -0.043086* Share of indigenouspopulation -0.043853* -0.052403* -0.045286* Poverty incidence (UBN) -0.143479' Regional dummy for Costa -5.98924* -5.08484* -4.3 1605* -2.79523* Regional dummy for Oriente 0.035547 0.746331 1.33866** 0.702439 Pupils per classroom(primary schools) 4.053048* -0.058566** -0.043853* Share of one-teacher(unidocente) schools 4.051462* -0.054299* -0.030050 Teacher quality' 1.01943 * 0.876846* Constant 97.0141* 99.68792* 90.3573* 97.0328* R2 0.4047 0.4379 0.5130 0.5713 Dependent variable: Net enrolment insecondary education Model 1 Model 2 Model 3 Model 4 Share of ruralpopulation -0.313211* 4.313126* -0.26742* Share of indigenouspopulation -0.074687* -0.074705* -0.059888* Poverty incidence(UBN) -0.61184* Regional dummy for Costa -6.25430* -6.26186* 4.60912* 1.96075** Regional dummy for Oriente 3.07857*** 3.08184 3.92957* 4.30481* Pupils per classroom(secondary schools) 0.000542 -0.03023 -0.02551 Teacher quality' 1.86975* 1.5099* Constant 57.7269* 57.7091* 41.2119* 69.8136* R2 0.4392 0.4884 0.7493 Notes : All estimates have been corrected for heteroscedacity. level and *** significant *0.4392 Significant at 99 percent level; ** significant at 95 percent at 90 percent level. 1. Teacher quality proxied by average salary levelof teachers indistrict. This variable correlateshighly with alternative specifications such as the share of teachers with secondary or university level training and shares o f teachers with 15-30years of experience. Source: INEC, Population census 2001;SINEC database. 163 A43. Determinants of test scores for language and mathematics at primary (2nd and 6th grade) and secondary (10th grade) level by districts (cantons), 2001 Spanish Mathematics Spanish Mathematics Test scores: Primary Zndgrade Zndgrade 6th grade 6th grade Shareof ruralpopulation -0.013134 -0,011256 -0,011256 -0.002711 Share of indigenouspopulation 0.00291 0.012759 0.012759 -0.01360 Regionaldummy for Costa -1.04343 -0.899487 -0.899487 -1.01049 Regionaldummy for Oriente -0.94261 -0.929692 -0.929692 1.52098 Pupilsper classroom(primary schools) -0.02082 0.008524 0.008524 0.01744 Share of one-teacher (unidocente) schools -0.00384 -0.003536 -0.003536 -0.03992 Teacher quality' -0.28453 -0.373924 -0.37392 -0,12737 Average level of schooling 0.38626 0.572619** 0.572619** 0.167951 Constant 11.0784* 8.32515* 8.32515* 7.10771* R2 0.3183 0.3103 0.3619 0.4172 Spanish Mathematics Test scores: Secondary 10thgrade 10thgrade Shareof ruralpopulation -0.024371 -0.009358 Share of indigenouspopulation -0.024581 0.005553 Regionaldummy for Costa -1.466913 -0.713785 Regionaldummy for Oriente 0.215342 -0.640394 Pupilsperclassroom(secondary schools) -0.005824 0.006055 Teacher quality' -0.531906 0.015544 Average level of schooling 0.493665** 0.282452 constant 14.55702* 4.160791*** R2 0.3869 0.2577 Note : All estimates have beencorrectedfor heteroscedacity. at 95 percent level and *** significant * Significantat 99 percent level; ** significant at 90 percent level. 1. Teacher quality proxiedby averagesalary level of teachers indistrict. This variable correlateshighly with alternative specifications such as the share of teachers with secondary or university level training and shares of teachers with 15-30years of experience. Source: APRENDO and SINEC database. 164 A44. Returns to education for urban workers, 1990-2002(Specification 1) Dependent variable: 1990 1995 1999 2002 Log hourly labour income Age 0.04663* 0.037823* 0.035451* 0.031806* Age squared -0.00044* -0.00037* -0.000304* -0.000272* Sex (female = 1) -0.24029* -0.219613' -0.167533* -0.191946* Years of schooling 0.060783* 0.063502* 0.075719* 0.072984* DummyEmployers 0.159022** 0.484643* 0.477405* 0.206899" DummySelf-employed -0.080048 0.022223 0.061917 -0.13599** Dummy Public sector workers -0.022409 -0.089235 -0.009430 -0.02894 Dummy Private sector workers -0.125576*** -0.151246 -0.124045 -0.28861 * Dummy Modem sector workers 0.154993* 0.275220* 0.295771* 0.284010* Dummy Informalsector workers -0.048390 -0.139169' -0.195299* -0.041285 Dummy Quito 0.122607* 0.259545* 0.25 13267* 0.307813* Dummy Guayaquil 0.05215* 0.089547* 0.112876* 0.066625* Dummy Cuenca 0.042701*** 0.199734* 0.099603* 0.201051* Constant 4.38338* 6.25446* 7.123794* -1.36919* Heckmanselection equation Age 0.05313* 0.062282* 0.077416* 0.115522' Age squared -0.00067* -0.000775* -0.000956* -0.001381* Sex (female = 1) 0.237919* 0.249121* 0.104719* -0.458433* Years of schooling -0.02426* -0.019044* -0.010509* -0.03 1387* Dummy Employers 3.381763* 3.164547* 3.004328* 1.927083* Dummy Self-employed 3.349655* 3.146664* 2.878786* 1.821808* DummyPublic sector workers 3.549583* 3.588317* 3.58531* 2.261496* DummyPrivate sector workers 3.285143* 3.479024* 3.143168* 2.002983* Dummy Modemsector workers -0.54169* -0.865282* -1.172938* -0.093226 D u m y Informalsectorworkers -0.17838* -0.196854* -0.168075* 0.694701* Dummy Quito -0.13566* 0.1811485* 0.1330385* -0.07867** Dummy Guayaquil 0.13424* 0.141310* 0.0541961 -0.018018 Dummy Cuenca 0.13178* -0.279711* 0.0454622 -0.066061 Number of children0-5 years 0.05309* 0.037097** 0.0203716 -0.00404*** Number of children 6-10 years -0.04046** -0.048430** -0.056745* -0.00148 Dummy head of household 0.375907* 0.420933* 0.311681* 0.619840* Constant -2.647381* -2.78977* -2.8 1467* -2.49129* /athrho -0.47782* -0.447623* -0.428941* -0.320253 * /Insigma -0.317001* -0.332464* -0.239392* -0.237471* Waldtest of indep.eqns (prob chi 2) 0.0000 0.0ooo 0.0000 ..... 0.0000 * significantat 99 percent,**significant at 95 percent, *** significant at 90 percent *Estimates havebeencorrectedfor heteroscedasticityandfor Heckmanselectionbias. Source: Vos (2004a). 165 A45. Returns to educationfor urbanworkers, 1990-2002(Specification2) Dependentvariable: 1990 1995 1999 2002 Log hourly labour income Age 0.048161* 0.039633* 0.037307* 0.036810* Age squared -0.000460* -0.000394* -0.00033* -0.000340* Sex (female = 1) -0.236599* -0.216669* -0.171522* -0.200075* DummyPrimary 0.286526* 0.273503* 0.263963* 0.187704* Dummy Secondary 0.577270* 0.580138* 0.563814* 0.484728* Dummy Tertiary 0.944181* 0.975027* 1.148927* 0.995283* DummyEmployers 0.193955** 0.494273* 0.499188* 0.258925* Dummy Self-employed -0.070897 0.008055 0.065918 -0.111301 Dummy Public sectorworkers -0.00368 -0.089786 0.013894 0.014428 Dummy Private sector workers -0.122936*** 0.167746*** -0.115617 -0.253482* Dummy Modem sector workers 0.194408* 0.31137* 0.328558* 0.318270* Dummy Informal sector workers -0.037836 -0.127038* -0.160156* -0.012676 Dummy Quito 0.125470% 0.262345* 0.257789* 0.320304* Dummy Guayaquil 0.047675* 0.087757* 0.120849* 0.070033* Dummy Cuenca 0.050948** 0.205858* 0.101773* 0.207237* constant 4.360915* 6.285155* 7.22305* -1.29527* Heckmanselection equation Age 0.059702* 0.068339* 0.083218* 0.119148* Age squared -0.000746* -0.000846* -0.001028* -0.001432* Sex (female = 1) 0.227881* 0.240854* 0.097601* -0.463672* Dummy Primary 0.109432*** 0.095117 0.153950** 0.0620591 Dummy Secondary -0.103152 -0.154748** -0.094921 -0.338876* Dummy Tertiary -0.13227*** -0.13398*** -0.000510 -0.360222* Dummy Employers 3.360116* 3.154139* 3.002508* 1.918642* Dummy Self-employed 3.349083' 3.146859* 2.884234* 1.831032* Dummy Public sector workers 3.530839* 3.576834* 3.587225* 2.265916* Dummy Private sector workers 3.298993* 3.493261* 3.164747* 2.029496* Dummy Modem sector workers -0.568170* -0.870465* -1.17843* -0.105790*** Dummy Informal sector workers -0.181692* -0.184898* -0.15662* 0.708992* Dummy Quito -0.144294* 0.178434* 0.13487* -0.07571*** Dummy Guayaquil 0.1295324* 0.143907* 0.05934*** -0.008542 Dummy Cuenca 0.1236273* -0.290042' 0.04078 -0.066344 Number of children0-5 years 0.039160** 0.0232076 0.00871 -0.00394*** Number of children6-10 years -0.032906*** -0.047958** -0.05848* -0.00133 Dummy head of household 0.352657* 0.396010* 0.290610* 0.60817* Constant -2.93900* -3.00897* -3.02183* -2.64863* /ahrho -0.473026* -0.456165* -0.419226* -0.29488* /Insigma -0.311069* -0.326662* -0.237280* -0.230680* Waldtest of ind. Eqs.Rob chi2 0.0000 0.0000 0.0000 0.0000 * significant at 99 percent,** significant 95 percent, *** significant 90 percent at at *Estimates havebeencorrectedfor heteroscedasticityandfor Heckmanselectionbias Source: LNEC, UrbanHousehold(LabourForce) Surveys, 1990-2002. 166 A46. Profile of malnutrition, education, water and sanitation conditions, 2000 and 2003 Access to Housing Popula- Stunting Under- Education safe Access to overcrow- Popula- Popula- tion (low height weight (low women 15- water sewerage dedness tion tion shares for age) (% weightfor 49 yrs (% of (% of (% of shares shares (women of children height)(% (years of popula- popula- popula- (house- (0-5 year 15-49 age 0-5) children0-5) schooling) tion) tion) tion) holds) olds) years) Ethnicity by languageofhouseholdhead Non-indigenous 23.7% 11.5% 9.4 83.2% 83.9% 34.7% 90.0% 81.9% 91.8% Indigenous 35.1% 11.9% 6.5 77.1% 59.6% 48.9% 10.0% 18.1% 8.2% Total 25.8% 11.6% 9.2 82.6% 81.4% 36.1% 100.0% 100.0% 100.0% Ethnicity by self- definitionof householdhead Indigenous 49.9% 15.5% 5.2 75.5% 52.1% 53.0% 8.5% 8.2% 7.1% Black (Afro) 15.8% 8.2% 8.5 76.9% 78.2% 50.0% 4.4% 9.8% 4.0% Mestizo 25.0% 12.0% 9.4 83.2% 84.1% 34.2% 78.6% 69.0% 80.5% Caucasian 22.5% 9.9% 10.3 87.5% 87.8% 29.6% 8.2% 11.9% 8.2% Other 21.8% 1.8% 9.9 92.0% 91.5% 37.0% 0.3% 1.1% 0.2% Total 25.8% 11.6% 9.2 82.6% 81.4% 36.1% 100.0% 100.0% 100.0% Sex of householdhead Male 27.2% 11.1% 81.5% 81.3% 36.8% 81.8% 52.2% Female 24.2% 12.1% 87.6% 82.1% 33.2% 18.2% 47.8% Total 25.8% 11.6% 82.6% 81.4% 36.1% 100.0% 100.0% Area of residence Urban 18.5% 8.9% 10.3 95.1% 93.7% 31.0% 66.0% 58.7% 70.1% Rural 36.0% 15.3% 6.4 58.4% 57.7% 46.2% 34.0% 41.3% 29.9% Total 25.8% 11.6% 9.2 82.6% 81.4% 36.1% 100.0% 100.0% 100.0% Incomedeciles 10%poorest 33.4% 16.1% 6.2 65.9% 54.8% 62.8% 9.7% 15.9% 8.0% 2 35.2% 15.3% 6.6 69.8% 65.1% 57.9% 10.1% 13.6% 9.1% 3 34.6% 18.6% 7.3 73.2% 72.0% 53.1% 10.2% 12.0% 9.2% 4 23.6% 10.3% 7.7 78.2% 77.0% 47.8% 10.0% 11.7% 9.5% 5 24.8% 10.3% 8.3 80.8% 82.7% 40.5% 10.1% 8.7% 10.3% 6 22.6% 9.2% 8.6 84.9% 84.6% 36.2% 9.9% 10.4% 10.0% 7 18.0% 8.1% 9.6 89.1% 90.4% 29.7% 9.8% 7.2% 10.3% 8 19.7% 6.5% 10.2 91.6% 92.5% 17.8% 10.3% 8.0% 11.1% 9 13.2% 4.0% 11.6 95.0% 96.8% 12.3% 9.9% 7.9% 10.7% 10%richest 8.9% 6.6% 13.3 97.1% 98.7% 4.4% 10.0% 4.7% 11.9% Total 25.8% 11.7% 9.2 82.6% 81.5% 36.2% 100.0% 100.0% 100.0% Region Costa 20.2% 10.4% 53.7% Sierra 32.4% 13.3% 38.8% Amazonla 31.2% 10.7% 7.6% Total 25.8% 11.6% 100.0% Source: INEC, EMEDINHO (special module of household survey), 2000 for malnutrition indicators; INEC, Sistema Integrado de Encuestas de Hogares (SIEH), special module on access to social services, December 2003, for other indicators. 167 A47. Profile of knowledge of family planningand sexually transmitteddiseases, 1999 and 2000 Population that knows Knowledgeof Women that Women HIV/AIDS how to know about that do not canbe protect Population Population family usefamily transmitted against shares shares(15 planning planning to childat HIVIAIDS (Women15- years and methods methods birth duringsex 45 years) older) Ethnicityby languageofhouseholdhead Non-indigenous 85.8% 41.1% 74.0% 59.1% 94.9% 81.9% Indigenous 33.8% 82.1% 54.0% 44.6% 5.1% 18.1% Total 83.2% 43.4% 71.4% 57.2% 100.0% 100.0% Ethnicityby self-definitionofhouseholdbead Indigenous 31.5% 24.0% 8.2% Black (Afro) 72.2% 52.8% 9.8% Mestizo 73.9% 59.1% 69.0% Caucasian 75.6% 62.1% 11.9% Other 63.1% 64.0% 1.1% Total 71.5% 57.3% 100.0% Sex of householdhead Male 72.4% 58.3% 0.0% 52.2% Female 83.2% 43.4% 70.5% 56.1% 99.9% 47.8% Total 83.2% 43.4% 71.4% 57.2% 100.0% 100.0% Area of residence Urban 91.8% 36.5% 79.4% 62.7% 64.9% 58.7% Rural 67.1% 55.7% 55.7% 46.3% 35.1% 41.3% Total 83.2% 43.4% 71.4% 57.2% 100.0% 100.0% Incomedeciles 10%poorest 48.0% 71.2% 58.1% 46.1% 7.8% 15.9% 2 67.9% 61.4% 59.6% 45.7% 8.8% 13.6% 3 79.3% 49.1% 61.1% 53.1% 9.4% 12.0% 4 80.5% 38.4% 68.9% 56.2% 8.7% 11.7% 5 85.9% 40.6% 69.7% 54.4% 10.0% 8.7% 6 85.8% 41.5% 72.9% 60.8% 10.4% 10.4% 7 90.4% 40.4% 74.3% 58.3% 10.9% 7.2% 8 92.8% 34.7% 71.2% 59.9% 11.4% 8.0% 9 92.8% 37.0% 80.2% 66.0% 10.9% 7.9% 10%richest 93.0% 32.1% 79.3% 65.4% 11.7% 4.7% Total 83.1% 43.4% 71.6% 57.4% 100.0% 100.0% Source: INEC, ECV (LSMS) 1999 for family planning indicators and INEC, EMEDINHO (special module of household survey), 2000 for HIV/SIDA indicators. 168 A48. Profile of access to services of childrenage 0-5, 1999 Medical Medical assistance Population shares Prevalence of assistancefor Repiratory for respiratory (children 0-5 diarrhoea diarrhoea diseases diseases years) Ethnicity by language of householdhead Non-indigenous 26.2% 32.4% 58.7% 29.7% 90.6% Indigenous 25.4% 23.2% 47.2% 10.8% 9.4% Total 26.1% 31.6% 57.6% 28.3% 100.0% Ethnicity by self-definition of householdhead Indigenous Black (Afro) Mestizo Caucasian Other Total Sex of household head Male 26.2% 31.8% 57.3% 30.8% 51.3% Female 26.0% 31.4% 58.0% 25.7% 48.7% Total 26.1% 31.6% 57.6% 28.3% 100.0% Area of residence Urban 26.2% 33.7% 59.5% 34.8% 53.4% Rural 26.0% 29.2% 55.6% 20.3% 46.6% Total 26.1% 31.6% 57.6% 28.3% 100.0% Income deciles 10%poorest 25.5% 15.7% 50.0% 12.3% 14.1% 2 31.6% 28.3% 55.8% 13.1% 13.2% 3 29.8% 34.3% 62.6% 23.7% 12.3% 4 26.8% 32.3% 53.7% 36.6% 11.1% 5 26.1% 40.0% 63.9% 27.1% 11.1% 6 26.1% 29.9% 54.9% 31.7% 10.0% 7 25.3% 35.7% 65.1% 25.8% 8.7% 8 21.6% 41.3% 59.8% 51.9% 7.4% 9 24.7% 40.7% 60.1% 37.7% 6.1% 10% richest 14.2% 32.7% 53.3% 51.9% 5.9% Total 26.1% 31.6% 57.6% 28.3% 100.0% Source: INEC, ECV (LSMS), 1999. 169 A49. Profile of access to healthservices pregnant women, 2003 Population Birth Birth shares of Population delivery Birth delivery womenof shares of Birth in delivery without 12years women of delivery public in profess- Coverage Fiveor andolder 12years in hospital private ional vaccina- more who are and older hospital or hospital medical tion prenatal pregnantor whoare orhealth health orhealth assis- against medical breast breast centre centre centre tance tetanus controls feeding feeding Ethnicity by languageof householdhead Non-indigenous 86.6% 66.3% 20.4% 13.4% 71.8% 58.1% 89.2% 88.3% Indigenous 64.8% 54.4% 10.5% 35.2% 63.7% 25.6% 10.8% 11.7% Total 84.3% 65.0% 19.3% 15.7% 71.0% 54.3% 100.0% 100.0% Ethnicity by self-defidtionof household head Indigenous 59.3% 51.2% 8.1% 40.7% 60.6% 23.5% 10.6% 11.8% Black (Afro) 82.2% 76.2% 6.0% 17.8% 74.2% 44.3% 4.7% 4.4% Mestizo 87.7% 66.6% 21.1% 12.3% 72.5% 59.6% 76.9% 76.9% Caucasian 85.6% 61.2% 24.4% 14.4% 68.5% 51.5% 7.6% 6.7% Other 81.0% 81.0% 0.0% 19.0% 81.0% 86.4% 0.2% 0.3% Total 84.2% 65.0% 19.2% 15.8% 71.1% 54.2% 100.0% 100.0% Sex of householdhead Male 0.0% 0.0% Female 84.3% 65.0% 19.3% 15.7% 71.O% 54.3% 100.0% 100.0% Total 84.3% 65.0% 19.3% 15.7% 71.0% 54.3% 100.0% 100.0% Area of residence Urban 90.0% 66.6% 23.4% 10.0% 70.9% 66.5% 59.5% 58.5% Rural 75.9% 62.7% 13.2% 24.1% 71.1% 37.1% 40.5% 41.5% Total 84.3% 65.0% 19.3% 15.7% 71.0% 54.3% 100.0% 100.0% Incomedeciles 10%poorest 77.2% 69.3% 7.9% 22.8% 76.8% 34.6% 13.8% 15.9% 2 80.5% 71.4% 9.1% 19.5% 72.7% 49.1% 12.6% 14.1% 3 80.9% 65.1% 15.1% 19.1% 65.9% 48.5% 13.1% 12.6% 4 87.3% 72.0% 15.3% 12.7% 72.2% 51.1% 10.2% 10.9% 5 85.6% 66.3% 19.3% 14.4% 65.7% 55.9% 9.9% 9.6% 6 86.3% 65.9% 20.4% 13.7% 73.9% 55.9% 10.6% 10.2% 7 82.5% 62.6% 19.9% 17.5% 71.6% 62.3% 8.3% 7.8% 8 86.6% 61.4% 25.2% 13.4% 71.4% 71.5% 8.4% 7.1% 9 89.8% 49.6% 40.2% 10.2% 61.9% 78.4% 7.2% 6.2% 10%richest 96.7% 46.9% 49.8% 3.3% 66.4% 79.9% 5.8% 5.5% Total 84.2% 64.8% 19.4% 15.8% 70.8% 54.4% 100.0% 100.0% Source: INEC, Sistema Integrado de Encuestas de Hogares (SIEH), special module on access to social services, December 2003. 170 A50. Profileof vaccinations of 2-5 year olds, 2003 Has vaccination Hascompletedoses Populationshares card of all vaccinations BCG DPT (children2-5 years) Ethnicityby languageofhouseholdhead Non-indigenous 87.2% 43.6% 98.4% 68.6% 89.3% Indigenous 87.1% 43.7% 96.9% 70.9% 10.7% Total 87.2% 43.7% 98.2% 68.9% 100.0% Ethnicityby self-definitionofhouseholdhead Indigenous 86.8% 45.4% 97.7% 73.5% 9.6% Black (Afro) 89.3% 45.8% 98.7% 66.9% 5.7% Mestizo 87.4% 43.5% 98.2% 68.3% 76.7% Caucasian 84.1% 40.4% 99.1% 69.9% 7.7% Other 90.1% 71.3% 100.0% 86.4% 0.3% Total 87.2% 43.7% 98.2% 68.9% 100.0% Sex of householdhead Male 87.1% 42.7% 98.0% 67.9% 50.7% Female 87.3% 44.6% 98.5% 69.9% 49.3% Total 87.2% 43.7% 98.2% 68.9% 100.0% Area of residence Urban 87.2% 42.1% 98.9% 66.7% 63.1% Rural 87.3% 46.3% 97.1% 72.7% 36.9% Total 87.2% 43.7% 98.2% 68.9% 100.0% Income deciles 10%poorest 84.6% 44.0% 97.7% 71.3% 13.4% 2 87.2% 43.2% 97.1% 69.4% 13.1% 3 85.0% 38.1% 98.4% 65.2% 12.6% 4 86.8% 41.8% 97.9% 69.5% 11.1% 5 88.9% 44.2% 97.6% 73.6% 11.0% 6 88.6% 46.3% 99.1% 67.2% 9.7% 7 88.7% 44.1% 98.2% 68.5% 8.8% 8 86.6% 44.8% 99.4% 71.4% 7.6% 9 85.9% 51.1% 99.6% 68.8% 7.6% 10%richest 94.5% 40.5% 99.2% 57.6% 5.2% Total 87.2% 43.6% 98.2% 68.8% 100.0% Source: INEC, Sistema Integrado de Encuestas de Hogares (SXEH), special module on access to social services, December 2003. 171 I 3 c A52. Profile of coverage of health insurance, 2003 (% of population) Farmers Health Publicsocial Insurance health (Seguro Any formof Privatehealth insurance Social Other health Population healthinsurance insurance (IESS) Campesino) insurance shares (total) Ethnicityby languageofhouseholdhead Non-indigenous 18.8% 2.0% 11.3% 4.1% 1.4% 91.5% Indigenous 15.6% 1.2% 5.9% 7.6% 0.9% 8.5% Total 18.5% 1.9% 10.9% 4.4% 1.3% 100.0% Ethnicity by self-definitionofhouseholdhead Indigenous 14.4% 0.8% 5.7% 7.3% 0.6% 7.5% Black (Afro) 18.6% 1.4% 9.1% 6.9% 1.2% 3.9% Mestizo 22.5% 1.9% 15.2% 4.0% 1.4% 80.2% Caucasian 27.4% 4.0% 18.6% 3.3% 1.5% 8.1% Other 16.4% 1.O% 8.2% 7.2% 0.0% 0.3% Total 22.1% 2.0% 14.5% 4.3% 1.4% 100.0% Sex of householdhead Male 19.3% 2.0% 11.3% 4.4% 1.6% 50.0% Female 15.2% 1.8% 8.2% 4.1% 1.0% 50.0% Total 17.2% 1.9% 9.7% 4.3% 1.3% 100.0% Area of residence Urban 18.5% 2.6% 13.3% 0.8% 1.8% 66.0% Rural 14.8% 0.4% 2.8% 11.1% 0.4% 34.0% Total 17.2% 1.9% 9.7% 4.3% 1.3% 100.0% Incomedeck 10%poorest 9.7% 0.1% 1.1% 8.4% 0.1% 9.7% 2 8.8% 0.4% 1.5% 6.8% 0.1% 10.1% 3 9.4% 0.2% 1.9% 6.9% 0.3% 10.2% 4 9.4% 0.4% 3.8% 4.8% 0.3% 10.0% 5 9.8% 0.5% 4.3% 4.0% 1.O% 10.1% 6 12.7% 1.1% 6.9% 3.8% 0.9% 9.9% I 15.9% 1.2% 9.8% 3.6% 1.3% 9.8% 8 20.0% 1.2% 14.3% 2.1% 2.4% 10.3% 9 29.6% 2.6% 23.2% 1.3% 2.6% 9.9% 10%richest 47.4% 10.9% 30.9% 1.1% 4.5% 10.0% Total 17.3% 1.9% 9.8% 4.3% 1.3% 100.0% Source: EWC, Sistema Integrado de Encuestas de Hogares (SIEH), special module on access to social services, December 2003. 173 A53. Profile of lifestyle habits, 1999 Shareof Shareof population Shareof Populationshares populationthat consumingalcohol population (15 years and smokesdaily regularly practicingsports older) Ethnicity by languageof householdhead Non-indigenous 8.6% 24.6% 29.4% 94.3% Indigenous 1.0% 20.8% 20.4% 5.7% Total 8.2% 24.3% 28.9% 100.0% Ethnicity by self-definitionof householdhead Indigenous Black (Afro) Mestizo Caucasian Other Total Sex of householdhead Male 13.8% 42.5% 46.0% 48.9% Female 2.7% 7.0% 12.5% 51.1% Total 8.2% 24.3% 28.9% 100.0% Area of residence Urban 9.1% 25.1% 32.3% 61.6% Rural 6.7% 23.1% 23.4% 38.4% Total 8.2% 24.3% 28.9% 100.0% Incomedeciles 10%poorest 3.4% 19.8% 19.5% 7.9% 2 5.4% 19.3% 22.1% 8.6% 3 7.7% 21.4% 25.5% 9.1% 4 6.4% 22.4% 26.8% 9.4% 5 6.4% 27.0% 27.0% 9.9% 6 7.7% 21.7% 25.4% 10.0% 7 9.0% 24.8% 29.1% 10.7% 8 8.2% 25.8% 32.5% 10.7% 9 10.3% 28.4% 35.5% 11.8% 10%richest 13.4% 28.0% 38.6% 11.9% Total 8.1% 24.2% 28.9% 100.0% Source: INEC, ECV (LSMS) 1999. 174 A54. Targeting efficiency of school meals program (PAE), child and maternal nutrition support (PANN) and cashtransfer program (BDH), December 2003. Targeting errors if eligiblepopulation is incomepoor Papilla- Bebida- Indicator PAE BDH PA" PA" Efficiency 65.4% 59.6% 61.2% 62.3% Leakageeffect 34.6% 40.4% 38.8% 37.7% Undercoverage 30.9% 37.9% 38.9% 32.5% Exclusion error 32.7% 23.4% 31.2% 36.0% Inclusion error 7.7% 9.7% 12.6% 10.5% Targeting errors if eligible population is poor according to SELBENindex Papilla- Bebida- Indicator PAE BDH PANN PA" Efficiency 74.4% 68.8% 77.5% 78.0% Leakageeffect 25.6% 31.2% 22.5% 22.0% Undercoverage 32.7% 42.1% 38.4% 31.7% Exclusion error 34.2% 22.2% 40.5% 47.1% Inclusion error 5.7% 7.3% 7.3% 6.2% Source: INEC, SIEH survey, December2003. 175 A54a. Econometric results of access to healthservices model (1) Multinomial Logistic Regression Number o f obs = 8734 Wald chi2(28) = 2111.08 Prob > chi2 0.0000 Log pseudo-likelihood = -7436.9461 Pseudo R2 --- 0.2122 .............................................................................. c h i l d 1 Robust delivery 1 Coef. Std. E r r . P>lZ/ [95% Conf. I n t e r v a l ] -------------+---------------------------------------------------------------- 2 Public medical assistance during child delivery motherage -.0497466 .0335403 -1.48 0.138 -.1154844 .0159913 motheragesq .0009839 .0006055 1.63 0 . 1 0 4 -.0002028 .0021706 ethnic -1.114562 .1541768 -7.23 0.000 -1.416743 -.8123808 motherschool . l o 9 1 5 9 3 .0123994 8 . 8 0 0.000 ,084857 .1334616 segiess ,4543054 .1762758 2.58 0.010 . l o 8 8 1 1 2 .7997997 segother 1I1111 -.4090823 . l o 3 9 0 8 5 -3.94 0.000 -.6127393 -.2054253 premature .4569984 .169017 2.70 0.007 ,1257312 .7882656 prenatal .9960061 .0755828 1 3 . 1 8 0.000 .8478666 1 . 1 4 4 1 4 6 tasin99 3.447864 ,6209034 5.55 0.000 2.230916 4.664812 tasper99 .0193296 .0075508 2 . 5 6 0 . 0 1 0 ,0045303 .034129 urban .480766 .0746713 6 . 4 4 0.000 .3344129 .6271191 coast .5514504 .1165911 4 . 7 3 0.000 . 3 2 2 9 3 6 1 .7799647 s i e r r a .7181279 . l o 4 8 1 5 4 6.85 0.000 .5126934 .9235624 logcons -cons I 11III1II .8418111 .069812 12.06 0.000 .704982 .9786402 -11.85295 .9525744 -12.44 0.000 -13.71996 -9.985937 -------------+---------------------------------------------------------------- Private medical assistance during c h i l d delivery motherage -.0796691 .0433046 - 1 . 8 4 0 . 0 6 6 -.1645447 .0052064 motheragesq II .0016382 . 0 0 0 7 8 5 1 2 . 0 9 0.037 .0000994 .003177 ethnic -1.560406 .325238 - 4 . 8 0 0.000 - 2 . 1 9 7 8 6 1 -.922951 motherschool . 1 4 0 5 7 1 .0143748 9.78 0 . 0 0 0 .1123968 . 1 6 8 7 4 5 1 segiess -.4449881 .2005398 -2.22 0.026 -.8380388 -.0519374 segother -.1710756 .137003 -1.25 0.212 -.4395965 .0974453 premature .5367914 .193296 2.78 0 . 0 0 5 .1579382 .9156445 prenatal .8463122 . l o 5 8 5 9 6 7 . 9 9 0.000 . 6 3 8 8 3 1 1 1 . 0 5 3 7 9 3 tasin99 -1.439944 .8779082 -1.64 0.101 -3.160613 .2807242 tasper99 .0038656 .0086115 0.45 0 . 6 5 4 -.0130126 .0207439 urban .9210909 . 0 9 1 4 1 3 1 10.08 0.000 .7419245 1.100257 coast 2.168203 . 1 6 2 6 9 6 1 13.33 0.000 1 . 8 4 9 3 2 4 2 . 4 8 7 0 8 1 s i e r r a .955125 .155626 6.14 0.000 .6501036 1 . 2 6 0 1 4 6 logcons .............................................................................. -cons 1 111I1111111I 1 . 6 0 5 7 2 3 . 0 8 9 1 1 1 1 18.02 0.000 1 . 4 3 1 0 6 9 1 . 7 8 0 3 7 8 -22.71929 1.219182 -18.63 0.000 -25.10885 -20.32974 (Outcome delivery== No professional a t t e n t i o n during d e l i v e r y i s the comparison group) Marginal effects on the probability of public medical assistance during child delivery Predicted p r o b a b i l i t y = .5163273 ................................. v a r i a b l e dy/dx X motherage -. 0022378 2 5 . 5 9 2 1 motheragesq .0000363 696.126 ethnic* -.141728 .036849 escolam .009289 8 . 1 7 5 2 9 segiess* .081466 segother* -..1593452 0808517 .093355 premature* ,0375999 .061837 prenatal* ,1527258 .818675 tasin99 1 . 0 4 5 1 4 1 . 0 4 5 5 1 1 tasper99 .004333 .915242 urban* ,0049252 .549393 coast* - ,1513633 . 4 4 8 0 9 1 s i e r r a * ,0563149 .483088 logcons .0049399 1 2 . 4 6 9 8 176 Marginal effects on the probability of private medical assistance during c h i l d delivery Predicted p r o b a b i l i t y = ............................................................................... ,24761028 variable 1 dy/dx X ---------------------------------+--------------------------------------------- motherage -.0084823 2 5 . 5 9 2 1 motheragesq 1I .0001794 696.126 ethnic* ] - ,1352754 .036849 motherschool .0122325 8.17529 segiess* II -.1203158 .OB1466 segother*l .0179069 .093355 premature*j .0397861 .061837 prenatal*/ .818675 tasin99 -.,04201641 0 9 0 6 3 . 0 4 5 5 1 1 tasper99 I1 -.70017511 .915242 urban* 1 . l o 8 5 0 9 7 .549393 coast* 1 .3389076 , 4 4 8 0 9 1 sierra* ] .OB5662 .483088 logcons 1 .1915215 1 2 . 4 6 9 8 ( * ) dy/dx i s f o r discrete change of d m y variable from 0 t o 1 13arginal effects on the probability of no professional medical assistance during child delivery Predicted p r o b a b i l i t y = .23606242 variable 1 dy/dx X ---------------------------------+--------------------------------------------- motherage IIII I ,0107202 2 5 . 5 9 2 1 motheragesq - .0002157 696.126 ethnic* .036849 motherschool -..2770034 8.17529 segiess* -.0215215 0390294 ,081466 segother*l .0629448 .093355 premature* 1 .077386 .061837 prenatal*l ,818675 tasin99 , 0 4 5 5 1 1 tasper99 .915242 urban* .549393 coast* 1I11 -.-1947422 -- .1134349 -.-.3360777 002582 . 4 4 8 0 9 1 sierra* j -..1875442 -.1419769 .483088 ............................................................................... logcons 1 1964615 1 2 . 4 6 9 8 ( * ) dy/dx i s f o r discrete change o f durruny variable from 0 t o 1 Source: Vos et al. (2004b) 177 A54b. Econometricresults of accessto health services model (2) (differentiating for poverty status) Multinomial Logistic Regression Number o f o b s = 8789 W a l d c h i 2 ( 3 0 ) = 2149.62 P r o b z c h i 2 Log pseudo-likelihood = -7552.6731 __--________-----__----------------------------------------------------------- Pseudo R2 ---- 0.0000 0 . 2 0 5 1 child I R o b u s t delivery 1 C o e f . S t d . E r r . 2 P>lZ1 [95% C o n f . Interval] --------------+---------------------------------------------------------------- Private medical assistance during child delivery m o t h e r a g e 1 -.027721 . 0 3 7 8 8 1 1 -0.73 0.464 - . l o 1 9 6 6 6 .0465247 m o t h e r a g e s q .0006387 .0006903 0.93 0.355 -.0007142 .0019916 ethnic -.5256076 .3373288 -1.56 0 . 1 1 9 -1.18676 .1355447 m o t h e r s c h o o l .0531941 .0140692 3 . 7 8 0 . 0 0 0 .025619 . 0 8 0 7 6 9 1 s e g i e s s -.8448353 .1197283 - 7 . 0 6 0 . 0 0 0 -1.079498 -.6101723 s e g o t h e r 11I1 1II11 ,3265989 ,127186 2 . 5 7 0 . 0 1 0 ,0773189 .5758788 p r e m a t u r e . l o 1 5 2 3 1 .1219678 0.83 0 . 4 0 5 -.1375294 .3405755 prenatal -.1265036 . l o 8 5 8 6 4 - 1 . 1 7 0 . 2 4 4 - . 3 3 9 3 2 9 1 .086322 t a s i n 9 9 -4.569829 .7523565 - 6 . 0 7 0.000 - 6 . 0 4 4 4 2 1 -3.095237 t a s p e r 9 9 -.0168079 .0053851 -3.12 0.002 -.0273625 -.0062533 urban .4594418 .0807476 5.69 0 . 0 0 0 .3011794 , 6 1 7 7 0 4 1 coast 1.543108 .1513095 1 0 . 2 0 0 . 0 0 0 1 . 2 4 6 5 4 6 1.839669 s i e r r a .228362 .1448796 1 . 5 8 0 . 1 1 5 -.0555967 .5123207 povertycons .3165306 .2297867 1 . 3 8 0.168 -.1338431 .7669042 n o p o v e r t y c o n s -cons I 11III .3338368 .2164033 1 . 5 4 0.123 -.0903059 .7579795 -5.597298 2.734849 -2.05 0 . 0 4 1 -10.9575 -.2370924 -------------+---------------------------------------------------------------- No professional medical assistance during child delivery m o t h e r a g e I .0492707 ,033322 1 . 4 8 0 . 1 3 9 -.0160393 ,1145807 m o t h e r a g e s q -.0009394 .0006006 -1.56 0 . 1 1 8 -.0021166 .0002378 e t h n i c 1.124449 .1522932 7.38 0 . 0 0 0 .8259601 1 . 4 2 2 9 3 9 m o t h e r s c h o o l - . 1 2 1 4 1 1 .0120753 -10.05 0 . 0 0 0 -.145078 -.0977439 segiess -.5581175 .1777254 -3.14 0 . 0 0 2 -.go6453 - . 2 0 9 7 8 2 1 segother I1111 .4649545 . l o 2 2 9 4 1 4.55 0.000 .2644617 .6654473 p r e m a t u r e -.4936718 .1652666 -2.99 0.003 -.8175885 -.1697552 prenatal -1.02866 .0749116 -13.73 0 . 0 0 0 -1.175484 -.8818358 t a s i n 9 9 -3.655743 .6276646 -5.82 0 . 0 0 0 -4.885943 -2.425543 t a s p e r 9 9 - . 0 1 8 6 1 .0075802 -2.46 0 . 0 1 4 -.0334669 - . 0 0 3 7 5 3 1 urban -.5872361 .0733392 - 8 . 0 1 0 . 0 0 0 -.7309782 -.4434939 coast -.4207808 .1177154 -3.57 0.000 -.6514988 -.1900628 s i e r r a -.6049926 . l o 6 3 1 8 3 - 5 . 6 9 0.000 -.8133728 - . 3 9 6 6 1 2 5 povertycons -.084268 .027974 - 3 . 0 1 0.003 -.1390961 -.0294399 n o p o v e r t y c o n s -cons 1 1111III11 -.1419281 .0261029 - 5 . 4 4 0.000 -.1930887 -.0907674 2.874505 .5638197 5 . 1 0 0.000 1.769438 3 . 9 7 9 5 7 1 Outcome d e l i v e r y = P u b l i c m e d i c a l assistance during child delivery i s the c o m p a r a t o r group Marginal effects on the probability o f public medical assistance during c h i l d delivery P r e d i c t e d probability = ............................................................................... .51150358 variable I dy/dx X m o t h e r a g e -. 0024683 2 5 . 6 0 3 6 m o t h e r a g e s q .000033 696.765 ethnic* -.1396402 .037478 m o t h e r s c h o o l .0080147 8.17574 segiess* 11II1 .081904 segother*l -..1676768 0983826 .094395 p r e m a t u r e t i ,0376237 .061826 prenatal*l ,1552804 .818107 t a s i n 9 9 1 . 0 3 0 3 1 1 .045518 t a s p e r 9 9 ,0044197 .914973 urban* . 5 4 8 2 1 coast* 5 5 9 9 7 1 . 4 4 5 7 0 1 sierra* 1III1 -..10 1 6 2 2 1 1 .485221 povertycons -..0440194 0301833 5.64377 nopovertycons 11 -.0253598 6 . 7 4 8 1 ( * ) dy/dx i s f o r discrete change o f d m y variable f r o m 0 t o 1 178 motherage - .0081345 25.6036 motheragesq ,0001758 696.765 e t h n i c * -.1458584 .031478 m o t h e r s c h o o l .0172113 8.17574 s e g i e s s * 1111I -.1107326 .081904 s e g o t h e r 2 * / .0306878 ,094395 premature* I .0466071 ,061826 p r e n a t a l * / .818107 t a s i n 9 9 1 -..0496072 6387214 .045518 t a s p e r 9 9 - .0020413 .914973 urban* .1192334 .54821 c o a s t * .3204807 ,445701 s i e r r a * .0785841 .485221 p o v e r t y c o n s .0643637 5.64377 nopovertycons I1I111 .071046 6.7481 ( * ) dy/dx i s f o r d i s c r e t e change of d m y v a r i a b l e f r o m 0 t o 1 m r g i n a l effects on the probability of no professional medical assistance during child delivery P r e d i c t e d p r o b a b i l i t y = ............................................................................... .23856136 v a r i a b l e I dy/dx X ---------------------------------+--------------------------------------------- motherage .0106029 25.6036 motheragesq - .0002087 696.765 e t h n i c * .2854987 ,037478 m o t h e r s c h o o l 8.17574 s e g i e s s * 111II -.-.025226 0569442 .081904 s e g o t h e r * / .0676949 .094395 p r e m a t u r e * / -.0842307 .061826 p r e n a t a l * l -.2048876 .818107 t a s i n 9 9 1 -.3915898 ,045518 t a s p e r 9 9 -.0023783 .914973 u r b a n * .54821 c o a s t * ,1644836 .445701 s i e r r a * --- ,1354545 .485221 p o v e r t y c o n s -.,1226036 5.64377 nopovertycons 1111II -.0341803 0456862 6.7481 ( * ) dy/dx i s f o r d i s c r e t e change o f d m y v a r i a b l e f r o m 0 t o 1 Source: Vos et al. (2004b) 179 ASS. Econometricresults of Infant Mortality (CPH) model COX Proportional Hazard Regression-Breslm method for t i e s F a i l u r e c o n d i t i o n i s d e a t h b e f o r e 12 months and a n a l y s i s t i m e i s age i n months o f t h e c h i l d . No. o f s u b j e c t s 8333.41 Number o f obs = 8654 No. o f f a i l u r e s -- 229.71 Time a t r i s k = 267638.22 Wald c h i 2 ( 2 6 ) = 1302.98 Log p s e u d o - l i k e l i h o o d = -1828.0596 P r o b > c h i 2 --- =------_--- 0.0000 _-----_--- I n t e r v a l ] s e x h .3300083 .1345469 2.45 0.014 .5937155 e t n i a .0797408 .2 957927 0.27 0.787 -..0663012 5000022 .6594838 m u l t i .4165362 .279883 6 1.49 0.137 -. 1320256 .9650979 prema tu .1628445 7.37 0 . 0 0 0 .881238 1.519576 p r i m o -.1.200407 5664898 .1736109 -3.26 0.001 -.9067609 - ,2262187 p r i v a t e - .7020839 .226774 -3.10 0.002 -1.146553 -.2576149 p u b l i c -.2783172 .1586849 -1.75 0.079 -.5893338 .0326994 p r e n a t a l -.4767503 ,1670435 -2.85 0.004 -.8041495 -.149351 b r e a s t -3.133959 .145596 -21.53 0.000 -3.419322 -2.848597 a n t i c o n -.0029464 .1610362 -0.02 0.985 -.3185715 .3126787 motherage - .0279377 ,071583 -0.39 0.696 -.1682378 .1123625 motheragesq .0005065 .0012899 0.39 0.695 -.0020217 motherschool - .0561199 .0226313 -2.48 0.013 - .lo04765 -..0030346 0117633 h h s i z e -.1832374 .0449701 -4.07 0 . 0 0 0 - .2713771 -.0950977 p o v e r t y .271631 ,1768258 1.54 0.125 - .0749411 .6182031 w a t e r .1937337 .1763515 1 . 1 0 0.272 -.1519088 .5393762 s a n i t a t i o n ,28795 .2100322 1.37 0.170 ,6996055 u r b a n .1791172 0.43 0.667 -.-.1237055 273909 .4282176 c o a s t -..0771543 .2365184 -1.33 0.183 - ,7784489 .1486861 s i e r r a -.3148814 3915519 .2206265 -1.77 0.076 - .8239719 .0408681 immune -2.032153 .6001818 -3.39 0.001 -3.208488 - ,8558186 b1994 -.2984598 .3192781 -0.93 0.350 - .9242334 .3273137 b1995 -.3734531 .3146699 -1.19 0.235 -. 9901948 .2432886 b1996 .0829271 .282448 0.29 0.769 - .4706609 .6365151 b1997 -.2489747 .2956106 -0.84 0.400 -.8283609 .3304114 b1998 .0116956 .2905883 0.04 0.968 -. 557847 .5812382 Marginal effects on the probability of infant mortality a f t e r COX P r e d i c t e d p r o b a b i l i t y = .00154343 sexh* .0005103 .507978 b r e a s t * - ,0297797 .958641 h h s i z e - .0002828 5.41863 e t n i a * .035386 p r i m o * -..0001278 ,29172 mot h e r a g e -. 0007875 0000431 25.5714 motheragesq 7.82e-07 I 695.004 m u l t i ' .0007935 .012168 p r e n a t a l* - .0008658 .821355 w a t e r * .000287 ,718845 u r b a n * -..0001186 ,552542 0004805 .450579 s i e r r a * c o a s t * - .0006038 .481283 s a n i t a t i o n * .862462 p r i v a t e* -..0004018 .284827 public* -.0009509 0004221 .425276 prematu* .061897 immune -..0033264 0031365 .542454 a n t i c o n * -4.55e-06 .810469 p o v e r t y * ,472818 m o t h e r s c h o o l -..0004236 8.20421 b1994" -.0000866 .184845 b1995' -.0004209 0005145 .180456 b1996* .188627 b1997* -..0001314 0003566 .18946 b1998* .0000181 .183673 180 A56. Unit cost estimates for public healthservices (inUS$) Estimatedunit costs Beneficiaries inUSdollars unit Comment in2003 2002 2003 General medical services -Personnel cost IESS Nurses per workerhonth 414.93 665.25 Auxiliary personnel per workerhonth 338.26 542.33 Medical doctors (4 hour shift) per workerhonth 451.29 733.17 Medical doctors (6 hour shift) per workerhonth 582.36 933.69 Dentists per workerhnonth 393.47 630.84 - Personnelcost MoH Residentdoctor assistant per workerhonth 324.00 519.46 Doctor(tratante 15) (8 hr) per workerhonth 702.00 1,125.51 Doctor (tratante 15) (6 hr) per workerhonth 527.00 844.93 Doctor (tratante 1) (8hr) per workerhonth 379.00 607.64 Nurse (2) (6 hr) per workerhnonth 25 1.OO 402.42 Mean auxiliary healthworker (8 hr) per workerhonth 204.00 327.07 Constructioncost of health centres Municipalhospital unit 250 m2 construction, 325,000 equipment and furniture Health centre unit 150m2 construction, 169,500 equipment and furniture Clinic unit 200 m2 construction, 236,000 equipment andfurniture Maintenance and operation cost health centres Hospitals andclinics per healthunit Own estimate 135,079 Health centres per healthunit Own estimate 20,404 Special programs - BonoDesarrollo Humano (BonoSolidario) person/year Includes administrative 186.92 overhead andtargeting 1,273,346 costs - Matemidadgratuita beneficiary1year Includes administrative 14.94 overhead and cost 1,999,867 - Immunizationprogram medicalpersonnel (PAI) beneficiarylyear Budget dividedby 10.97 vaccinatedchildren 948,049 - Mobile health units per health Operating cost 29,300.00 unitlyear per beneficiary/ Per medical consult 5.09 consult 182.562 - Subsidies onhealthinsurancepremium Per Assumptionby authors 1.oo targetedat poor beneficiaryhnonth - Subsidies onhealthinsurancepremiumfor Per Assumption by authors 1.oo non-poor beneficiaryhonth - Subsidies onhealthinsurancepremiumfor uer Assumption by authors 1.oo indigenouspopulation beneficiarylmonth Source: Vos et al. (2004b) and SIISE, based on budget data. 181 A57. Ecuador: Budget Execution of Institutional Wages, Central Government 1/ 1995- 1995- 2000- Sector 1995 1996 1997 1998 1999 2000 2001 2002 2002 1999 2002 Legislative ... ... ... ... 1.42 0.94 1.15 1.14 1.16 1.42 1.08 Judiciary 1.22 1.37 1.30 1.00 1.02 0.82 1.24 0.90 1.11 1.18 0.98 Administration 1.35 1.42 1.36 1.91 1.17 0.93 1.61 1.23 1.37 1.44 1.26 olw Office of the President 1.27 1.70 1.82 1.32 1.04 0.87 1.61 1.21 1.35 1.43 1.23 NationalSecurity Council (Cosena) 1.53 1.31 1.35 1.03 1.16 0.87 1.45 1.17 1.23 1.28 1.16 Staff Development Service(Senda) 1.35 1.37 1.25 3.27 ._. ... ... ... 1.81 1.81 ... CONAM 5.00 5.19 10.28 6.00 6.84 4.37 6.16 4.64 6.06 6.66 5.06 Planning 1.30 1.27 1.15 2.04 ... ... ... ... 1.44 1.44 ... Environment ... ... ... ... 0.99 0.27 1.08 0.84 0.80 0.99 0.73 Interior 1.24 1.24 1.24 1.20 1.07 0.91 1.11 1.07 1.13 1.20 1.03 olw Police 1.24 1.22 1.23 1.22 1.11 0.97 1.07 0.98 1.13 1.20 1.01 Defense 0.98 1.14 1.22 1.25 1.02 0.98 1.10 1.02 1.09 1.12 1.03 ExtemalAffairs 1.06 1.07 1.00 1.14 1.86 0.98 0.91 1.04 1.13 1.23 0.97 Finance andeconomic services 1.38 1.45 1.30 2.97 0.16 0.05 1.18 1.48 1.25 1.45 0.91 Social Sectors 1.36 1.98 1.11 1.03 1.27 0.88 1.25 1.26 1.27 1.35 1.13 olw Education 1.20 2.52 1.04 1.06 1.32 0.88 1.22 1.27 1.31 1.43 1.13 Social Welfare 2.08 1.28 1.22 1.06 1.18 1.10 1.13 1.42 1.31 1.37 1.22 Health 1.64 1.24 1.36 0.93 1.10 0.87 1.36 1.21 1.21 1.25 1.15 Labor 1.21 1.37 1.23 1.02 1.15 0.75 1.22 2.92 1.36 1.20 1.63 olw ProfessionalTraining Service ... ... ... ... ... ... ... 2.03 2.03 ... 2.03 Agriculture, Fishing andLivestock 1.33 1.46 1.27 1.36 1.17 0.86 1.34 0.95 1.22 1.32 1.05 Energy 1.56 1.20 1.16 1.24 1.12 0.99 1.16 1.08 1.19 1.26 1.08 Industry 1.06 1.24 1.30 1.07 1.37 0.86 1.60 1.03 1.19 1.21 1.16 Tourism 0.73 0.20 1.02 1.01 0.11 0.04 2.38 1.26 0.84 0.61 1.23 Transport andCommunications 1.27 1.20 1.30 1.07 1.08 0.88 1.27 1.08 1.14 1.18 1.08 olw Directorate of Civil Aviation ... ... ... ... ... ... ... 1.11 ... ... 1.11 HousingandUrbanDevelopment 1.81 1.97 1.28 1.04 1.01 1.16 0.83 0.84 1.24 1.42 0.94 OthersGeneralServices 2/ 1.88 0.99 4.58 0.18 1.34 1.63 1.62 1.14 1.67 1.79 1.46 TOTAL 1.20 1.47 1.18 1.15 1.11 0.83 1.17 1.14 1.16 1.22 1.05 Notes: l/Refers to the ratio of devengado to the initial planned budget. Shaded areas indicate overspending/underspendingabove 15%. 2/ Includethe Electoral and Constitutional Courts. Source: Ministry of Finance. 182 A58. Activities and components of social programs by priority function Sector EDU DIS CIS HOU DPC O W Education 1 Mejoramiento de laeducacionintercultural bilingue X 2 Desarrollo e implementacion de la educacion basica X 3 Desarrollo e implementacion del bachillerato X 4 Descentralizacion y desconcentracion del MEC X 5 Ecuador Educa X 6 Programa nacionalde educacion pre-escolar (PRONEPE) X 7 Programa de Ailmentacion Escolar (PAE) X 8 Redes amigas X Social Welfare 9 Bono de Desarrollo Humano X 10 Credit0 Productivo Solidario X 11Direccionde proteccion de menores X 12 Direccion Nacional de Discapacitados 13 Fondo de desarrollo de lajuventud X 14Fondo de Inversion Social de Emergencia (FISE) X 15 Comision Nacional de Discapacitados 16Programa de Alimentacion y Desarrollo Comunitario (PRADEC) X 17 Gerontologia 18 Nuestros Niiios X 19 Operacion de rescateinfantil (ON) X 20 PROLOCAL X 21 PRODEPINE X 22 SELBEN X Health 23 Control de Malaria X 24 Dengue X 25 Tuberculosis X 26 Cuidado Matemo Infantil X X 27 Programa de Alimentacion y Nutricion Nacional (PA") X X 28 Plan Ampliado de Inmunizaciones (PAI) X 29 Programa de micronutrientes X 30 Medicamentos Genericos X 31Aseguramiento social basico X 32 Unidades moviles X Housing and urban development 33 Agua potable y saneamiento ambiental (PRAGUAS) X 34 Mejoramiento de barrios X 35 Programa de agua potable y saneamiento ambiental X 36 Programa de construccion y rehabilitacion de agua y saneamiento X 37 Vivienda Bono Solidario X 38 Vivienda Campesina X 39 Vivienda Urbana S N X 40 Vivienda urbano marginal X Labor 41 Capacitacion laboral X 42 Compras estatales X 43 Mediacion laboral X 44 Erradicacion del trabajo infantil X 45 Perfiles profesionales X Notes: EDU=Education reform; DIS=Welfare aid to disabled and aging people; CIS=Child and infant support (food and development); HOU=Low-cost housing; DPC=Disease prevention and control; OTH=Aid to rural and indigenous households, and other general purpose aid. Source: Secretaria Tecnica del Frente Social (STFS). 183 A59. Number of beneficiariesper priority social program, 2003 ProgradGod and targeted beneficiaries Outcome Bono de Desarrollo Humano (BDH) To transfer acash stipendto 1,280,000 poor people Cash stipends were provided to 1,043,826 mothers, as recordedby SELBEN (1,025,882 mothers, 226,848 elderly people, and 8,840 disabledpeople 244,234 elderly people, and 9,884 disabledpeople) Programa de Alimentaci6nEscolar (PAE) To provide free breakfastand lunchto 1,450,000 Intheregion Sierra, lunchwas providedto 520,000 childreninschool age (between5 and 14years old), children for 120days whereas inthe Coast, lunch for 160days was provided to 751,710 children andbreakfastto 371,821 children, for 40 days. Programade Alimentaci6n y Desarrollo Comunitario (PRADEC) To feed 118,221children ages between2 and 6 Feedingwas provided to 160,125 children in years old and 87,963 middle and high schoolers primary school age and 87,963 middle andhigh ages between7 and 14years old, to feed 207,536 schoolers, andto 149,135 elderly peopleand28,729 elderly peopleand 30,000 disabledpeople, and train disabledpeople.A total of 43,087 small 60,000 peopleinthe developmentof community entrepreneurs were trained incommunity projects development projects. Programade MatemidadGratuita y Atencionala Infancia (LMG) To provide matemal care anddelivery services to Services were provided to 4,315,215 pregnant and 4,613,515 pregnantand fertile women andprimary fertile women health care servicesto infants ages below 5 years old PlanAmpliado de Inmunizaciones(PAI) To apply 5,264,99 1vaccines to vulnerable children, 5.1 million vaccines were applied to the targeted infants, and pregnant women. Vaccines to prevent population groups. By type of vaccines, services polio, diphtheria, measles, tetanus, and yellow were provided as follows: 355,000 doses of BCG, fever, among others. 522,349 DTP, 657,110 pentavelante, 1'382,750 OPV, 90,558 SRP, 518,332 DTchildren, and 1'085,912 DT women, and 157,882 doses for Source: STFS. 184 A60. Provision of LMGservicesto infantsand pregnant women, 2003 Projected Numberof Population beneficiaries Coverage Pre-birthcontrols Firstpre-birth control 371379 318471 86% Subsequent controls 568210 529029 93% Dental care to pregnant women 371379 209528 56% V M tests inpregnant women 371379 18198 5% Normal labor 371379 102201 28% Cesareas 371379 30376 8% Obstetric emergencies Toxemia 371379 4888 1% Hemorrage of first half of pregnancy 371379 23135 6% Hemorrage o f second half of pregnancy 371379 2024 1% Hemorrage o f labor and puerperio 371379 2764 1% Sepsis 371379 660 0% Post DmUm Postbirthcontrol 371379 112731 30% Familv counseling Counseling and prevention 3990840 660154 17% Sterilization 2770464 15694 1% Vasectomy 500 147 29% Earlv detection of cervical uterino cancer 1757110 156344 9% Sexual transmission diseases Siphilis 4249006 2219 0% Gonorrea 4249006 2360 0% Genitalherpes 4249006 807 0% HPV 4249006 1090 0% Newbornbabies care Newborn babies care, normal 378807 117007 31% Newborn babies care. Dathological Intermediate care 259427 11479 4% Intensive care 259427 2749 1% Care to infants ages < 5 vears old Infants < 1year old n/a 1097323 n/a Infants between 1and 5 years n/a 1451253 n/a Hospital complications n/a 30324 n/a Blood and hemoderivatives Blood components 371379 9412 3% Total blood 371379 8699 2% 16% Source:Programa de Matemidad Gratuita y Atencion a la I n f a n c i M H . 185 A61. Budget revision and execution of socialinstitutions, 2000-2002 (Inmillions of US dollars) InitialBudget Revised budget Devengado 2000 Education 356.4 307.5 283.7 Social Welfare 248.9 87.7 54.2 Health 125.8 115.5 103.2 Total 731.1 510.7 441.1 2001 Education 456.2 519.5 492.8 Social Welfare 187.7 151.7 126.8 Health 151.8 191.6 188.6 Total 795.7 862.8 808.2 2002 Education 601.8 733.5 694.3 Social Welfare 217.6 115.2 76.0 Health 311.1 314.6 259.0 Total 1,130.5 1,163.3 1,029.3 Source: Ministry of Finance. A62. Priority socialprograms budget and execution, 2003 (Inmillions of US dollars, except otherwise noted) Programmed Executed Approved Revised Sector/Program budget budget Change % Devengado Transferred Ministry of Social Welfare Bono de DesarrolloHumano 1/ 203.1 164.6 -19% 161.8 161.8 PAE 30.7 17.0 -45% 14.2 15.2 PRADEC 12.2 10.0 -18% 7.0 7.0 Ministry of Health PA" 2000 5.7 5.7 -1% 5.7 5.7 PA1 10.0 9.0 -10% 9.1 9.1 Ley de Matemidad Gratuita 19.9 25.5 28% 23.9 22.1 Ministry of Education Investment expenditure 2/ 90.8 98.4 8% 26.0 26.1 1. TOTAL 372.4 330.2 -11% 247.7 247.0 Notes: 1/IncludesBecaEscolar. 2/ Includes Secretaria Nacionalde Deportes. Source: Ministry of Finance 186 A63. Track of Treasury transfers to socialprograms, 2003 (In million of US$) Treasury transfer to Ministry transfer to Ministry owes to Program Ministry Program Program Bono de Desarrollo Humano 1/ 162.2 161.8 0.4 PAE 17.2 14.2 3.0 PRADEC 8.6 7.0 1.6 P A " 2000 1.4 1.4 0.0 PAI 11.1 9.1 2.0 Ley de Maternidad Gratuita 2/ 7.1 7.1 0.0 207.6 200.6 7.0 Sources: MEFand STFS. 187 Data Annex A64. Transfers and Payments of the "15 percent Law" to Municipalities, 2002 Payments Source: Based on informationprovidedby MEF(2004). A65. Transfers and Payments of the "15 percent Law" to Municipalities, 2001 30,000 T Q 25,000 S p! L 2 20,000 v = S g 10,000 0 3 s I- 5,000 Jan Jan Feb Feb Mar Mar Mar Mar Mar Jun Jun Jul I Jul Aug Aug Sep Sep Cct Cct Oct Nov Nov Dec Dec Dec I /I I II I II 111 IV v I II II I II I II I II 111 I II I II 111 Payments Note: No payments were executedin months of April andMay. Source: Basedon informationprovidedby MEF(2004). 188 A66. Transfers of 15 percent Law and Implications for Cash-Management of the Central Government Budget, 2002 70 c 1.1 Planned transfer according to 15%-rule -1.1 Executed transfer Monthly Payments Source: Based on informationprovidedby MEF(2004). A67. Transfers of 15 percent Law and Implications for Cash-Management of the CentralGovernmentBudget, 2001 120,I 1 E2 100 A Budaet I 1-1 Planned L 3 80 transfer according to Ef v 0 60 15%-rule 3 Executed -.-.-a c 40 -1- 0 Transfer e 2o 0 ) , , , I J MonthlyTransfers Source: Based on informationprovidedby MEF (2004). 189 A68. RevenueBasefor 15 percent-Lawwith andwithout PetroleumRevenue,2002 450 I --- 400 Revenue Base 350 300 250 200 --- Revenue 150 Base without petroleum Months Note: Dotted line represents trend-line (moving average). Source: Basedon information provided by MEF (2004). A69. REVENUEBASEFOR 15 PERCENT-LAWWITH AND WITHOUT PETROLEUM REVENUE,2001 450 -.- Revenue h E 400 Base 350 3 300 v '3 250 200 -1- Revenue 150 Base without petroleum Months Note: Dotted linerepresentstrend-line (moving average). Source: Basedon information provided by MEF(2004). 190 ,70. Total Per Capita Investment of Municipalities and ProvincialCouncils (inISD), 2002 300 7P 250 I f2 200 5 L 150 n v) 3 100 50 0 Provinces I Source: Basedon informationprovidedby MEF(2004) and INEC (2001populationdata). A71. Incidence of Subnational Government Investment Spendinginthe Social Sectors (based on NBIIndex), 2002 c 100 C 90 .w u) 80 8 70 *..- = a > Bs c 60 -.-m $ 5 50 40 0 0 30 v) 20 .c 0 10 8 0 0 20 40 60 80 100 Population: from poor to rich Note: Socialsector investment spendingincludes:healthandeducation expenditures. Source: Basedon expendituredata providedby MEF(2004); the UnsatisfiedNeedsIndicator (Necesidades Bdsicas Znsatisfechas, NBI) basedon SIISE (2001); populationdata from INEC, based on the 2002 Census. 191 A72. Incidence of Subnational Government Investment Spending inthe Social Sectors(following 0 20 40 60 80 100 Population: from poor to rich Note: Social sector investment spendingincludes: healthand educationexpenditures.Subnationalgovernment includes bothmunicipalities andprovincialcouncils. No Consumption-basedPoverty Index is available for the Morona Santiago, Napo, Pastaza,ZamoraChinchipe, Sucumbios, and Orellana provinces and hence they are not includedinthis analysis. Source: Basedon expendituredata provided by MEF (2004); the consumption-basedPoverty Index developed by Carolina Sinchez-PAramo(World Bank, EcuadorPoverty Assessment 2004); andpopulation data from INEC, basedon the 2002 Census. A73. Own Revenue and Expenditures per Level of Government (in percent), 1997 and 2002 1997 2002 Expenditures Own Revenue Expenditures Own Revenue Central government 95.03 86.01 99.92 88.08 Provincial Councils 0.32 2.87 0.00 4.13 Municipalities 4.65 11.12 0.08 7.79 Total 100 100 100 100 Source: Basedon information provided by Ministry of EconomyandFinance (2004) and Central Bank (2002). 192 A74. Per CapitaInvestmentof MunicipalitiesandProvincialCouncils(inUSD),2002 Region Province Per capita Per capita Total per capita investment in investment in investment health education Costa 0.04 0.11 19.80 ~~ ElOro 0.18 0.03 8.92 Esmeraldas 0.00 0.03 3.53 Guayas 0.00 0.07 25.74 Los Rios 0.01 0.13 12.21 Manabi 0.02 0.06 17.77 Galapagos 0.00 0.00 42.24 Oriente 0.03 0.18 37.34 Francisco de Orellana 0.00 0.00 7.55 Morona Santiago 0.12 0.33 39.09 Nap0 0.00 0.13 54.12 Pastaza 0.00 0.13 35.60 Sucumbios 0.05 0.20 25.98 Zamora Chinchipe 0.00 0.28 71.56 Sierra 0.09 0.10 8.81 Azuay 0.03 0.06 23.33 Bolivar 0.04 0.06 11.15 Canar 0.25 0.25 6.77 Carchi 0.00 0.21 8.69 Chimborazo 0.00 0.02 4.93 Cotopaxi 0.33 0.08 3.46 Imbabura 0.22 0.10 13.49 Loja 0.07 0.08 19.60 Pichincha 0.00 0.02 3.38 Tungurahua 0.02 0.06 12.76 Nationalaverage 0.05 0.13 15.62 Source: Basedon informationprovided by MEF(2004) and INEC (2001population data). 193 A75. Intergovernment, TI - s to M inicipalitiesand Provinci Councils Fecha Montoflo Fuente Criterios us0 Fondo de Salvamentodel Patrimonio 13ldicl87 0.03 Localidadesespecticulos Restauracidn, Cultural pcblicos :onservacidn y proteccidnde bienes histdricos, artisticos y cultwales " _ 0 1 t I Presupuestodel Fondo lnversiones , de Emergencias Nacionales(FONEN) Fondo de Desarrollo Provincialde 12lsep189 7 5% I 1% de transaccionesde 20% Consejo Provmcial Saneamiento Bolivar operacionesde credit0 16%Guaranda ambiental I en monedanacional 16%Chillanes Alcantarillado 16% Chimbo Desarrollo urban0 16% San Miguel 16%Ecbeandia Decreto legislativo de asignacionesa 12/sep189 5% I Facturasde energia de 50% municipios de 100%para favor de las provincias de Aznay, INECEL aempresas kzuay, Caiiar y Morona infraestructura Caiiar y Morona Santiago eltctricas por generacidn Santlago de centralesde Pisayambo, Pantey 10% CREA 80% ejecucidnde &wh obras 20% forestacidn ~ Fondo de SaneamientoAmbiental, 18/dic/89 10% Ingresos Autondad 20% Consejo Provmcial Riego, drenaje y Vialidad y Riego de la Provincia de El Portuana Puerto Bolivar vialidad - _ _ _I Oro (FONDORO) ~- I 1% de operacionesen ~~ I -II_ I 5% 60% municipios Saneamientoy monedanacional 20% municipios en vialidad Funcidnde poblacidn Fondo de Desarrollo Provincial 14lmarl9 2% comentes 47 5% aporte a1capital Respaldarcr6ditos (FONDEPRO) 1 1 Ingresosdel totales presupuesto del BdE del BdE 47 5% Consejos Obras de desarrollo I provinciales e INGALA I 25% por poblacidn y II 75% equitativo 0.5% CONCOPE CONCOPE Fondo de Desarrollo Seccional 2lImayB 2% Ingresoscomentesnetos (FODESEC) 1 distnbuido del presupuesto 2% Mun~cipioscapitales de provincia 25% Quito, 25% Guayaquil, 50% otros equitatlvamente " .- . ._., . . . . 98% 20% COn%JOS 70% gasto comente/ provmciales (60% inversidn poblacidn, 20% NBI, 30% inversidn zona 20% Eficiencia rural administrativa y eficacia fiscal) ~~-- . .. .. - _-_ I ^^ 75% municipios: 60% 40% BdE Fondode municioios (60% mversidn poblackn, 30% NBI, 10 eficiencia administrativa y eficacia fiscal: 40% BdE .. . ~ . II" 5% emergencias Organismos seccionales 3,000 Previo a distnbucidn de millones mgresos petrolerosque recibe el Estado Programade Vialidad Rural de la I1/jun/90 10,000 PresupuestoGeneraldel Contrapartecdditos Fideicomiso cdditos Provincia de Manabi millones Estado BdE BdE hasta 2002 Otras asignaciones presupuesta- nas Fondode Riego de la Provinciade 16lagol90 1,000 PresupuestoGeneraldel Planprovincial de Cotopaxi millones Estado riego desde 1991 Fondopara el Sector Agropecuariode 18lenel91 12 5% 1% operacionesde 30% Conseio Camlnos vecmales, la Provincia de Cbimborazo cddito en moneda 1I Provincial ' mgacidn, forestacidn nacional 1 20% municioio de j Mercadozonal y Riobamba ' saneamiento .". 1.- I unbiental ~ 194 .. . . . 50% equitativo entre Centros de acopio y -- restode cantones saneamiento ambiental Fondosde Desarrollo de las Provincial 122 676 3lmayi91 2.5% Facturaci6nDOT servicios 50% Consejo Obrasde de la Regi6n Amaz6nica Provincial infraesuucturaurbana 20% capital de y rural de Sucumbios, . . . , petrolerosaPetrcecuador de las empresas , nacionaies . . .. ." . -.~ .... provincia Napo, Pastaza, Morona Sanuago, ~ . " 1.5% I Facturacibnpor servicios 30% equitativo entre petrolerosaPetroecnador consejos provinciales ZamoraChmchipey de empresasextranjeras restantes. Orellana . . 40 S. 71ago189 US$5 ctvs POI bamlde petrbleo EquitativoparaNapo, 50% obras de 248 transportadoporel Esmeraldasy mfraestructurade 10s oleoducto Sucumbios mmcipios 50% obras de infraestructurade 10s consejosprovmciales 10 30 21lsep192 US$ 10 Por baml de petr6leo BdE reparte 80% proyectosde :tvs ' prcducido en la regibn 30% consejos vialidad y amaz6mca provinciales saneamiento 60% municipios (55% ambiental equitahvo, 45% 20%gastocomente poblacibn) -- 10%Fondoregional Fondo de Desarrollo de la Provinciadi 145 899 23lmar19 15% I 1% operacionesde 25% Consejo Estudios, Pichincha 2 I cddito en moneda Provincial consuuccibn, nacional 25% Quito mejoramientode 50% equitativo entre caminos vecinales y el restode cantones obras de infraesuucturaen -- parroqniasurbanasy rurales Fondode Desarrollo de la Provinciadi 146 899 231mad9 15% Diferencial cambiano de 20% Consejo Obrasde vialidad e carchi L transaccionessemanales Provincial infraesmctura urbana del BCE 27.5% T d c h Y mal 15% 1% operacionesde 16.5% MonrJfar cddito en moneda 11%Espejo nacional 9% Bolivar 9% Mira -- 7% Dacha Distribuci6ndel 15%del Presupuesto sJn 21 20lmar19 15% Ingresosdel presupuesto 70% municipios (50% Planesde desarrollo del Gobiemo Central I del gobiemocentral con NBI, 40% poblaci6n, econ6mic0, social y excepcibnde 10s 10% equitativo) cultural provenientesde cdditos 30% consejos intemosy extemos provinciales (50% -- NBI, 40% poblaci6n, 10% snpeficie) Sustitutiva de la ley que crea el Fondo 92 335 91jud98 100% ' Impuestodel 1% a 70% municipios Obrasde vialidad de Vialidad de la Provincia de Loja compra-ventade 30% consejo Equip0 caminero (FONDVIAL) -- 1vehiculos usados orovincial Creacibnde CORPECUADOR 10 S. 7laeoi98 100% Peajevias rehabilitadas Reconsuucci6nde Renta liquida anualdel zonas afectadas por el Fondode solidandad. 1 Fenbmenode ElNSo Donacionesy Inversi6n subvenciones proporcional al d ~ o Pmcipacibn del Estado 11 por el fendmeno de El en mcremento de NSo exportaciones petroleras t I Exportacionesde banano 1 Prestamosanombre del $ j Estado Prbstamosanombre de Corpecuador Asignaciones - - - ji presupuestanas Source: CONAM (2000): Nuevo /lodelo de Gesti6npara el Ecuador. 195 Table A76 Interest Payment SavingsfromAdditional Debt Repurchase 2003 2004 2005 2006 In US$ millions Additional debt repurchase 77 266 165 199 Interest payments 815 811 848 844 Scheduled interest payments 815 843 900 920 Reduced interest payments 32 52 76 Inpercent of GDP Interest payments 3.0% 2.8% 2.8% 2.6% Scheduled interest payments 3.0% 2.9% 2.9% 2.8% Reduced interest payments 0.1% 0.2% 0.2% 4.7% 5.0% 5.0% 4.8% Note: NFPS Primary balance 196 BIBLIOGRAPHY Antinolfi, Gaetano and Todd Keister. 2001. "Dollarization as a Monetary Arrangement for Emerging Market Economies." Review. Federal Reserve Bank of St. Louis, November/December, Volume 83, NO.6:29-39. Artana Daniel, Ricardo L6pez Murphy, and Fernando Navajas. 2003. "A"Fisca1 Policy Agenda" in Pedro Pablo Kuckzynski and John Williamson, Editors. After the Washington Consensus.Restarting Growth and Reform in LatinAmerica. Institute for InternationalEconomics, Washington D.C. Banco Central del Ecuador 2003. Quarterly Report, Quito. Beckerman, Paul. 2002. "Longer-Term Origins of Ecuador `Predollarization' Crisis." in Paul Beckerman and AndrCs Solimano eds. ,Crisis and Dollarization in Ecuador. Stability, Growth and Social Equity. The World Bank, Washington, D.C. Berg, Andrew and Eduardo Borensztein. 2000. "The Pros and Cons of FullDollarization." IMFWorking Paper WP/00/50. ResearchDepartment, InternationalMonetary Fund, March, Washington, D.C. Braconier, Henrik and Tomas Forsfalt. 2004. "A New Method for Constructing a Cyclically Adjusted Budget Balance: The Case of Sweden." The National Institute of Economic Research, Working Paper No. 90, Stockholm. Calvo, Guillermo. 1999. "On Dollarization," University of Maryland, April 20, College Park, MD. Chang, Roberto (2000). "Dollarization: A Scorecard." Economic Review. Federal Reserve Bank of Atlanta, Third Quarter, Vol. 92, No. 1, pp. 1-11. Collier, Paul and Anke Hoeffler. 2002. "Military Expenditure: Threats, Aid, and Arms Races." World Bank Policy ResearchWorking Paper No. 2927. Washington, D.C. Conaghan, Catherine M. 1994. "Loose Parties, `Floating' Politicians, and Institutional Stress: PresidentialisminEcuador, 1979-1988," inJuan J. Linz and Arturo Valenzuela, Editors, The Failure of Presidential Democracy. Baltimore, Maryland: The Johns Hopkins University Press. Conaghan, Catherine M. 1995. "Politicians Against Parties: Discord and Disconnection in Ecuador's Party System." in Scott Mainwaring and Timothy R. Scully, Editors, Building Democratic Institution: Party Systems in Latin America. Stanford, California: StanfordUniversity Press. Coppedge, Michael. 1997. "A Classification of Latin American Political Parties," Working Paper No. 244, South Bend, Indiana: Kellogg Institute for International Studies, University of Notre Dame. Cuentas Nacionales de Salud. 1997. De la Torre, Augusto, Roberto Garcia-Saltos, and Yira Mascar6. 2001. "Banking, Currency, and Debt Meltdown: Ecuador Crisis inthe late 1990s." October mimeo. Donders, Jan and Caroline Kollau. 2002. "The Cyclically Adjusted Budget Balance: The Brussels' Methodology." CPB Report, 4. Eifert, Benn, Alan Gelb, and Nils Borje Tallroth. 2002. "The Political Economy of Fiscal Policy and Economic Management in Oil Exporting Countries." World Bank Policy Research Working Paper No. 2899. World Bank, Washington, D.C. Fierro-Renoy, Virginia and Mariana Naranjo (2003). Ecuador: sostenibilidadfiscal y desarrollo humano, 1970-2001. Quito: UNDPAJNICEF. Frank, Jonas. 2004. "Revenue Trends in Ecuador," Contribution to Ecuador Public Expenditure Review. The World Bank. Washington D.C. 197 Fretes-Cibils, Vicente, Marcel M. Giugale, and JosC Roberto L6pez-Cfilix, Editors. 2003. Ecuador-An Economic and Social Agenda in the New Millennium. The World Bank, Washington, D.C. Gelb, Alan and Jorge Marshall-Silva 1988. Ecuador: Windfalls of a New Exporter," in Alan Gelb and " Associates. Oil Windfalls: Blessing or Curse?New York: OxfordUniversityPress. Gbmez, CarlosR. and Eloy Vidal. 2004. "Telecom." BackgroundPaper for Ecuador-Creating Fiscal Spacefor Poverty Reduction-A Fiscal Management and Public Expenditure Review (Volume 11) Hagemann, Robert. 1999. "The Structural Budget Balance: The IMF's Methodology," IMF Working Paper No. 99/95. Washington, D.C. Hodrick, Robert, and EdwardPrescott. 1997. "Postwar U.S. Business Cycles.'' Joumal of Money, Credit and Banking, 29(1): 1-16. InternationalMonetaryFund2000. Ecuador. SelectedIssues and Statistical Annex, Western Hemisphere Department,August 18, Washington, D.C. International Monetary Fund 2003a. Ecuador: 2003 Article IV Consultation, Request for a Stand-By Arrangement and Approval of an Exchange Restriction. IMF Country Report No. 03/90. Western HemisphereDepartment and Policy Developmentand Review Department, March 14, Washington, D.C. International Monetary Fund 2003b. Ecuador. First Review Under the Stand-By Arrangement and Requests for Modifications and Waiver of Non-Observance and Applicability of Performance Criteria. IMF Country Report No. 03/248. Western HemisphereDepartment, July 23, Washington, D.C. InternationalMonetary Fund, 2003. "Informes sobre la Observancia de Cbdigos y Normas. Mbdulo de Transparencia Fiscal." Report elaboratedby Julio Viiiuela, Jan Singh and Esteban Vesperoni. IMF. WashingtonD.C. International Monetary Fund, 2003b. Country Report No. 03/233. Trinidad and Tobago: Selected Issues and Statistical Appendix. WashingtonD.C., July. Izquierdo, Alejandro 2002. "Sudden Stops in Capital Flows: Ecuador's Performance and Options," ResearchDepartment, Inter-AmericanDevelopmentBank, October 11,Washington, D.C. Jficome, Luis I.2004. "The Late 1990s Financial Crisis in Ecuador: Institutional Weaknesses, Fiscal Rigidities, and Financial Dollarization at Work". IMF Working Paper WP/04/12, Monetary and Financial SystemsDepartment, January, Washington,D.C. Jones, Mark P. 1995. Electoral Laws and the Survival of Presidential Democracies. University of Notre Dame Press, NotreDame. Judge, Lindsay andJeni Klugman, 2003. "The Budgetprocess inBolivia," Contributionto Bolivia Public ExpenditureReview, Mimeo, The World Bank, WashingtonD.C. Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobath. 2002. "Governance Matters 11: Updated Governance Indicatorsfor 2000/01." Policy ResearchWorking Paper Series No. 2772, World Bank, Washington, D.C. Kopits, George, E.Haind1, E. Ley and J. Toro, 1999. Ecuador, Modemization of the Tax System. InternationalMonetaryFund,Dept. of Public Finances, November,WashingtonD.C. Latin American Index of Budgetary Transparency, 2003. "Annual Report," Coordinated by Poder Ciudadano, Quito. Lecaros, Fernando. 2004. "Electricity." BackgroundPaper for Ecuador-Creating Fiscal Spacefor Poverty Reduction-A Fiscal Management and Public ExpenditureReview (Volume 11) 198 Lbpez-Chlix, JosC R. 2003. "Maintaining Stability with FiscalDiscipline and Competitiveness," inFretes- Cibils, Vicente, Marcelo M. Giugale, and Jos6 R. L6pez-Chlix. Ecuador: An Economic and Social Agenda in the New Millennium. The World Bank, Washington, D.C. LSMS. 1994. Ecuador Living Standard Measurement Survey, The World Bank, Washington D.C. Mainwaring, Scott and Matthew Soberg Shugart, eds., 1997. Presidentialism and Democracy in Latin America. New York: Cambridge University Press. Marcel, M.,M. Tokman, R. ValdCs, and P. Benavides. 2001. "Structural budget balance: methodology and estimation for the Chilean Central Government 1987-2001 ." Cepal. Serie Seminarios y Conferencias. Santiago. Mejia-Acosta, AndrCs 1999. "Explaining `Camisetazos': The Logic of Party-Switching in the Ecuadorian Congress 1979-1996." Paper prepared for the 57" Annual Meeting of the Midwest Political Science Association, Chicago, Illinois, April 15-17. Melo, Albert0 2003. La Competitividad de Ecuador en la Era de la Dolarizacidn: Diagndstico y Propuestas. Working Paper No. C-101, Competitiveness Studies Series, Research Department and Regional Operations Department 3, July, Washington, D.C. Pachano, Simbn, Maria Caridad Araujo, AndrCs Mejia Acosta, Anibal PBrez-Linhn y Sebastihn Saiegh 2004. Political Institutions, Policymaking processes, and Policy Outcomes in Ecuador. FLACSO. Latin American ResearchNetwork, Inter-AmericanDevelopment Bank, f r s t draft, January. PAHO. 1994. Annual Report, Geneva. PanAmericanHealth Organization PAHO. 2002. Health Report, Geneva. PanAmericanHealthOrganization. Powell, Andrew (2002). "Ensuring a Sound and Financial Systemina Dollarized Economy. Assessment and Prescriptions for Ecuador." Paper submitted to IDB seminar on "Dollarization inEcuador. Policies to Ensure Success." Mimeo. Washington, DC. Reis. 2003. "Ecuador Post-Dolarizacion, Recomendaciones de Politica Fiscal.'' Report to proyecto SALTOAJSAID, Mimeo, Quito. Rofman, Rafael P. 2004. "Pensions." Background Paper for Ecuador-Creating Fiscal Spacefor Poverty Reduction-A Fiscal Management and Public Expenditure Review (Volume 11) Rogers, Halsey, JosC L6pez-Chlix, Nancy Cbrdoba, Nazmul Chaudhury, Jeffrey Hammer, Michael Kremer and Karthik Muralidhharan, 2004. "Teachers Absence, Incentives, and Service Delivery in Ecuadorian Primary Education: results from a National Survey," World Bank, CEDATOS and HarvardUniversity, Mimeo. Washington D.C. Salazar, 2004. "Informe del Gasto Ptiblico en Ecuador." Report deliveredto the PER misibn. Quito. SALTO/AID, 2003. "Medicibn del Gasto Tributario en Ecuador. Paper by Jerbnimo Roca and Hugo Vallarino, Mimeo, Quito. Schenone, Osvaldo. 2003. Tax Policy and Administration inFretes-Cibils, Vicente, Marcelo M.Giugale, and JosC R. L6pez-C6lix, editors. Ecuador: An Economic and Social Agenda in the New Millennium. The World Bank, Washington, D.C. Schick, Allen, 1998. "A Contemporary Approach to Public Expenditure Management," The World Bank Institute, Washington. D.C. Shepherd, Jorge A. 2004a. "Ecuador: Public Expenditure Structure and Trends, 1995-2003." Background paper for Ecuador's Public Expenditure Review, March 9, Washington, D.C. 199 Shepherd, Jorge. 2004b. "Social Sectors Spending in Ecuador," Report to Ecuador Public Expenditure Review, Mimeo, Washington D.C. Sistema de Informacih Municipal, 2001. Cuantas de Gobiernos Seccionales. Mimeo. Quito Sotomayor, Maria Angelica. 2004. "Water and Sanitation." BackgroundPaper for Ecuador-Creating Fiscal Spacefor Poverty Reduction-A Fiscal Management and Public Expenditure Review (Volume 11) STFS-Secretaria Tecnica del Frente Social. Several documents. Traa, Bob 2003. "The Balance Sheet as an Indicator of Sustainable Fiscal Policies," in International Monetary Fund. Ecuador: Selected hues and Statistical Appendix. WIF Country Report No. 03/91, Western Hemisphere Department, April, Washington, D.C. United Nations Devopment ProgrammeAJNICE, 2003. Ecuador: Sostenibilidad Fiscal y Desarrollo Humano, 1970-2010. Quito Vos, Rob, Juan Ponce, Mauricio Lebn, Jose Cuestay Wladymir Brborich, 2003. "Quien se Beneficia del Gasto Social en Ecuador?' Estudios e Informes del SIISE-STFS No.4, Secretaria T6cnica del Frente Social, Quito. Vos, Rob and Juan Ponce. 2004. "Education." Background Paper for Ecuador-Creating Fiscal Space for Poverty Reduction-A Fiscal Management and Public Expenditure Review (Volume 11) Vos, Rob, Jose Cuesta, Lucio L e h , and MauricioRosero. 2004. "Health." BackgroundPaper for Ecuador-Creating Fiscal Spacefor Poverty Reduction-A Fiscal Management and Public Expenditure Review (Volume 11) World Bank 1984. Ecuador: An Agenda for Recovery and Sustained Growth. Country Study, PUB-5094, Washington, D.C. World Bank 1991. Ecuador: Public Sector Reforms for Growth in the Era of Declining Oil Output. Country Study 9596, Washington, D.C. World Bank 1993. Ecuador. Public Expenditure Review: Changing the Role of the State. Report No. 10541-EC, August 13, Latin America and the Caribbean Regional Office, Washington, D.C. World Bank, 1996. Ecuador Poverty Report. The World Bank. Washington D.C. World Bank. 1998. Public Expenditure Management Handbook, The World Bank, Washington D.C. World Bank. 2001. "Honduras, Improving Access, Efficiency and Quality of Care inthe Health Sector." The World Bank, Washington D.C. World Bank 2003a. Implementation Completion Report on a Loan inthe Amount of U S $151.52 Million to the Republic of Ecuador for a Structural Adjustment Program. Report No. 27471, Latin America and the Caribbean Region, December 29, Washington, D.C. World Bank. 2003b. "Ecuador, Fiscal Consolidation and Competitive Growth," Loan Document. The World Bank, Washington D.C. World Bank, 2003c. "Country Assessment and Action Plan for HIPC," Mimeo, Washington D.C. World Bank 2004a. Ecuador. Development Policy Review. Sustainable Growth, Social lnclusion and Governance: TheRoad Ahead. Report No. 27443, The Latin American and Caribbean Region, March 5, Washington, D.C. World Bank. 2004b. Ecuador Poverty Assessment. Green Cover Mimeo. The World Bank. Washington D.C. 200 World Bank. 2004c. Making Services Work for Poor People. The World Bank. Washington.D.C. World Bank, 2004d. WorldDevelopmentIndicators. 2003. The World Bank, WashingtonD.C. World Bank and Inter-AmericanDevelopmentBank 1997, Honduras, Public Expenditure Management for Poverty Reduction and Fiscal Sustainability. Report No.22070. The World Bank. Washington D.C. World Bank and Inter-AmericanDevelopmentBank 2002. Peru. Restoring Fiscal Disciplinefor Poverty Reduction. A Public ExpenditureReview. ReportNo. 24286-PE, October 24, Washington, D.C. World Bank and Inter-AmericanDevelopment Bank 2003, Peru, RestoringFiscalDiscipline for Poverty Reduction:A Public ExpenditureReview. The World Bank.WashingtonD.C. World Bankand Inter-AmericanDevelopmentBank 2004. "CFAA Report on Ecuador," Paper Submitted to ROSC, CFA andPERmeeting, Quito, May 20-21. World Bank and Inter-AmericanDevelopment Bank, 2004. Ecuador Country Financial Arrangements Assesment. The World Bank. Washington. D.C. World Bank and Inter-American DevelopmentBank 2004. Repliblica del Ecuador. Prdcticas y Sistemas de Administracidn Financiera Gubemamental. Technical Background Document for Ecuador's Country FinancialAccountability Assessment, April, Washington,D.C. World Economic Forum. 2001. The Global Competitiveness Report 2001-02. New York: Oxford University Press. WHO, 2003. WorldHealthReport. World Health OrganizationGeneve. Yepes, Gomez, and Maria Carvajal, 2002. "Plan Nacional de Desarrollo del Sector Agua Potable y Saneamiento Bisico." Quito 201