Report No. 24286-PE Peru Restoring Fiscal Discipline for Poverty Reduction A Public Expenditure Review October 24, 2002 ~ The World Bank Peru Country Management Unit Poverty Reduction and Economic Management Department Latin America and Caribbean Region Inter-American Development Bank Country Division 5 (Peru) Regional Operations Department 3 Document of the World Bank and the Inter - American Development Bank CURRENCY EQUIVALENTS Exchange Rate Effective as of October 1, 2002 Currency Unit = Nuevo Sol US$1.00 = NS/.3.61 GOVERNMENT FISCAL YEAR January 1 - December 31 WEIGHTS AND MEASURES Metric System IBRD IDB Vice President David de Ferranti Manager, Regional Operations IIl Ciro de Falco Chief Economist Guillermo Perry Deputy Manager, Regional Operations III Maximo Jeria Country Director Isabel Guerrero Chief, Country Division 5 Ana Maria Rodriguez Lead Economist/Manager Sara Guerschanik Calvo Country Economist, Peru and Co-Task Manager Alberto Melo Task Manager Josd R. L6pez-C4lix Abbreviations and Acronyms ACE Community educational associations ISPs Institutos superior pedag6gicas ADEs Areas of educational developmet LAC Latin America and the Caribbean AEs Areas of execution LPTF Ley de Prudencia y Transparencia Fiscal BCRP Central Bank of Peru MAB Multi-annual budget CEC Consejos Escolares Consultivos MED Ministry of Education CETICOS Centers for Export, Transp ortation, MEF Ministry of Economy and Finance Industry, Commercialization and Services CG Central government MEM Ministry of Energy and Mines CLAS Local Communities of Health MINDEF Ministry of Defense Administration CONAM Consejo Nacional del Ambiente MINSA Ministry of Health (National Environmental Council) CONASEV Comisi6n Nacional Supervisora de MMM Marco Macroecon6mico Multianual (Multianual Empresas y Valores macroeconomic framework) CONITE National Commission for Foreign MTBF Medium-term budgeting framework Investment and Technobgy CONSUCODE Consejo Superior de Contrataciones y MTEF Medium-term expenditure framework Adquisiciones del Estado COOPOP National Office for Popular Cooperation NA Non-applicable COPRI Commission for Promotion of Pnvate. NFPS Non-financial public sector Investment CPS Consolidated public sector NGOs Non-governmental organizations CTAR Transitory Councils from the Regional NPS Non-personal services Administration DGAES Direcci6n General de Asuntos OECD Organization for Economic Cooperation and Econ6micos y Sociales Development DISAs Regional Directorates of Health ONP Oficina de Normalizaci6n Provisional (Pensions) DNPP Direcci6n Nacional del Presupuesto OSIPTEL Organismo Superior de Inversi6n Privada en Publico (General Directorate for the Telecomunicaciones Public Budget) DREs Regional Directorates of Education OSITRAN Organismo Superior de la Inversi6n en Infracstructura de Transporte de Uso Publico EBFs Extra-budgetary funds OUs Operational Units ENAHO Encuesta Nacional de Hogares (National PAAG Programa de Administraci6n de Acuerdos de Household Survey) Gesti6n ESSALUD Seguro Social de Salud (Health PAF Poverty Auction Fund Insurance) ETR Effective tax rate PARSALUD Programa de Apoyo a la Reforma del Sector Salud (Health Sector Reform Program) EUs Executing units PCM Presidencia del Consejo de Ministros (Premier's Office) FEF Fondo de Estabilizaci6n Fiscal PEAP Poverty Eradication Action Plan FIDE Intergovernmental Decentralization Fund PEM Public expenditure management FONAFE National Fund for Financing PESEM Plan Estratdgico Sectorial Multianual Government Enterprise Activity 111 Abbreviations and Acronyms FONAHPU National Savings Fund PETS Public Expenditure Tracking Survey FONAVI Fondo Nacional de Vivienda PRONAA Programa Nacional de Asistencia Alimentaria (National Food Assistance Program) FONCODES Fondo Nacional de Compensaci6n y PRONAMA Proyecto Nacional de Mantenimiento de Cuencas Desarrollo Social CHCS HidrogrAficas y Conservaci6n de Suelos (National Project for the Management of Water Sheds) FONCOMUN Fondo de Compensaci6n Municipal PSPs Protected social programs FSF Fiscal Stabilization Fund PSRL Programmatic social reform loan FONCOR Regional Compensation Fund PUP Planilla Jnica de Pagos (or Unified Payroll) FY Fiscal year SENSICO National Service for Construction Industry Training GDP Gross domestic product SIAFITO A simplified SIAF GFS Government Financial Statistics SIAF-SP Sistema Integrado de Administraci6n Financiera del Sector Publico (Integrated Financial Management System) GG General government SIS- SMI Integral and Mother and Child Maternal Insurances GOP Government of Peru SIMSIP Simulations for Social Indicators and Poverty HIPC Highly-indebted poor countries SMI Seguro Materno-Infantil (Mother and Child Health Insurance) IBRD International Bank for Reconstruction SPC Sector Publico Consolidado and Development IDB, IADB Inter-American Development Bank SUNAT Superintendencia Nacional de Administraci6n Tributaria (Superintendency for Tax Administration) IES Extraordinary Solidarity Tax (impuesto SUR Seemingly unrelated regression extraordinario de solidaridad) IGR Institutional and Governance Review UEs Unidades Ejecutoras IGV Value-added tax (in Spanish) UIT Unidades de Ingreso Tributario IMF International Monetary Fund USEs Unidades de Servicios Educativos (Educational Service Units) INA Iniciativa Nacional Anti-Corrupci6n USIS Unidades de Seguro Integral de Salud INADE Instituto Nacional de Desarrollo VDL Vaso de Leche (National Institute for Development) INEI National Institute for Statistics and WBI World Bank Institute Information IRR Internal rates of return ISC Selective tax on consumption or excise tax (in Spanish) Iv Contents Abbreviations and Acronyms ........................................................... iii List of Boxes, Tables and Figures in Text ......................................................... vi List of Background Papers...... ....................................................... ix Acknowledgments ................................................................... xi Chapter I. EXECUTIVE SUMMARY........................................................1 A. Rationale and Organization.........................................................1 B. M ain Findings and Recommendations................................................................. 2 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES...........................17 A. Long-Run Fiscal Trends................................................................17 B. The Current Fiscal Disequilibria 1999-200 1.............................................21 C. The Macroeconomic Consistent Expenditure Envelope 2002-03...................... .............23 D. The Fiscal Adjustment Effort....................................................................29 E. Revamping the Fiscal Rules........................................................ 33 Chapter . REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE...... .................... 37 A. Modernizing Financial Management.................................................. 38 B. Upgrading the Budget Management System............................................. 42 C. Building a Medium Term Expenditure Framework........................................47 D. Protecting the Budget as a Medium-Term Social Policy ............................................ 51 E. Budget Transparency ................................................................................................ 54 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE. ...................................57 A. Does Higher Government Expenditure Buy Better Results in Education and Health Care. ............57 B. Tracing Leakages of Public Funds in Peru-a Public Expenditure Tracking Survey....................... 60 C. Enhancing Targeting of Social Programs ............................................... 77 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION................. 83 A. The Decision to Decentralize: Promises and Risks ........... ................................... 83 B. The Deconcentration of Central-Government Spending and Transfers to the Municipalities .................... 86 C. The Emerging Legal Framework for Decentr lization in Peru Tracking...... ............................ 89 D. Entering into a Decentralization Path ................................................. 90 E. Decentralization in the Education Sector ............................................... 94 F. Decentralization in the Health Sector .................................................. 97 Chapter VI. UPGRADING THE CIVIL SERVICE Spending.and.Transfers............................................ 101 A. The Size and Employment Conditions of the Public Sector...................................101 B. Reforming the Legal and Institutional Framework....................... .................. 108 C. Additional Measures for Improving Human Resource Administration ............................112 D. The Education Sector ............................................................... 112 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION ................. ..... 117 A. The Costs of Weak Governance and Corruption ....... ................................... 121 B. Impact on Public Service Delivery .................................................... 122 C. A Policy Agenda for Improving Governancee A................ ................... ........................ 124 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES. ...................... 129 A. Background C....................I O............................................... 129 B. The Peruvian Taxation Regime ............................... ....................... 130 C. Comparison of Peru's Tax System to Minerals Tax Systems in Selected Countries......... ........... 130 D. Improving Mining Environment Management ................................. ........... 143 B IB L IO G R A P HY ........................................................................................................................................................ 145 ANNEX A. A Public Debt Sustainability Analysis for Peru............................................... ...................................... 151 ANNEX B. The Fiscal Effort Required for a Sustained Structural Deficit .................... .................... 167 ANNEX C. Questionnaire on Public Financial Management ........................... ........... 171 ANNEX D. A Public Expenditure Tracking Survey : Methodological Issues...........................179 ANNEX E. Public Enterprise Reform in Peru: Introducing Management Contracts at FONAFE... ......189 ANNEX F. A Methodology for Functional Reviews and for Analyzing Functional Prospects.........191 STATISTICAL APPENDIX ............................................................ 193 V Boxes, Tables and Figures in Text CHAPTER II Box 2.1 The Fiscal Prudence and Transparency Law ..................................................22 Box 2.2 Tax Regime and Proposed Changes ..............................................26 Box 2.3 Peru: Evaluation of the Transparency Law 2000-01 ................................33 Figure 2.1 NFPS Fiscal and Prim ary Balances ..................................................................................................... 17 Figure 2.2 Tax Revenue Composition, 1970-2001 ................. ...........................................18 Figure 2.3 Pro-cyclical Fiscal Stance....................................................................................... 20 Figure 2.4 Volatility in Fiscal Variables ............. .... .................................... ..................................20 Figure 2.5 Fiscal Trend Over Time ......................................................22 Figure 2.6 LAC: VAT Effectiveness .............. .......................................................................... 29 Figure 2.7 Central Government's Expenditure Composition.. .....................................30 Figure 2.8 LAC: Capital Expenditure (1998) ........................................................................................30 Figure 2.9 Structure of Central Government's Fixed Capital Formation.............................................................31 Figure 2.10 Social and Public Spending...........................................................................................32 Figure 2.11 Growth Rates in Social and Public Spending.........................................32 Figure 2.12 Actual and Potential Output....................................................35 Figure 2.13 Actual primary balance (% of GDP) and structural primary balance (% of potential GDP)........................................................35 Table 2.1 Key Macroeconomic Variables, 1969-1989.........................................18 Table 2.2 Elasticity to GDP, 1970-2000..................................................19 Table 2.3 Elasticities and Growth Volatilities for Peru (1990-2000)...............................20 Table 2.4 Key Econom ic Indicators..............................................................................................24 Table 2.5 Recommendations for Tax Reform............................................... 25 Table 2.6 Social Expenditures are Pro- or Counter-cyclical..................................... 32 CHAPTER 1H Box 3.1 Main IGR/PEM Issues........................................................37 Box 3.2 The Integrated Financial Management System (SIAF) of Peru........................................38 Box 3.3 Introducing the SIAF at the Ministries of Defense and National Security (Interior)..................... 41 Box 3.4 Peru: Principal Stages of the Budget Formation .....................................43 Box 3.5 Peru: Institutionalizing a Virtual Poverty Fund: A Look at Best-Practice Uganda...... ........53 Figure 3.1 Comparison of Central Government and NFPS Balance ................................47 Figure 3.2 Share of Social Expenditures to GDP Mngm t...........yste...............................................................48 Figure 3.3 Sectoral Budget Allocations of.Defense.and.National.Secu)...................................................48 Figure 3.4 Index of Budget Transparency ......... .....................................................................................55 Figure 3.5 Percent of Government Entities Publishing Information .at.Best-Practice.Ugad...............................56 Table 3.1 Peru: Status of Budget Management Benchmarks ................................................44 Table 3.2 Composition of CG Budget Sources ......................................... ................45 Table 3.3 Revised Projection of the 2002 Budget Allocations.....................48........ ............46 Table 3.4 Composition of Social Expenditure 199-02 Informatin................................ .........................................48 Table 3.5 Comparing Multi-Annual Budgets by Sector ........................................49 Table 3.6 Budget for Protected Social Program s ................................................................................................. 52 Table 3.7 Budget Transparency Ratings.................................. ...................................54 CHAPTER IV Box 4.1 MEXICO: Integrating Social Programs for Improved Efficiency - The Case of PROGRESA ............................................81 Figure 4.1 Lorenz Curves for Intergovernmental Transfers.................................................................................. 65 Figure 4.2 How the milk disappears in the Vaso de Leche Program .................... ............75 Figure 4.3 Lorenz Curves in Selected Social Programs .........................................78 Vi Boxes, Tables and Figures in Text Figure 4.4 Lorenz Curves for health expenditure.............................................79 Figure 4.5 Lorenz Curves for education expenditure..........................................79 Figure 4.6 Access to Social Programs, 2000................................................79 Table 4.1 a Regression Results for Education Indicators: Linear Regressions.........................58 Table 4.1b Regression Results for Health Indicators: Log-Log Regressions..........................59 Table 4.2 Peru: Efficiency of Public Expenditure in Reaching Social Outcomes 1990-1998 ..............60 Table 4.3 Per Capita Transfers to Municipalities............................................66 Table 4.4 Municipalities that do not know the arrival day of the transfer............................67 Table 4.5 Municipalities that are subject to CG supervision.....................................67 Table 4.6 Leakages in FONCOMUN and canon minero.......................................68 Table 4.7 Fraction of Transfers Used for Current Expenditure...................................68 Table 4.8 Beneficiary households that received training/information...............................70 Table 4.9 VdL Transfer Schedule.......................................................71 Table 4.10 Leakage #3: Municipality to Local Committees.....................................72 Table 4.11 Worst Offenders, Leakage #3...................................................73 Table 4.12 Leak 4: Vaso de Leche Program (at household level)..................................74 Table 4.13 Leak 5: Vaso de Leche Program (at beneficiary household level).........................75 Table 4.14 VdL Leakages.............................................................76 Table 4.15 Correlation Expenditure and Poverty by Department.....................................78 Table 4.16 Targeting by Food Assistance, Health and Education Programs, 2000.......................80 Table 4.17 Household Access to Social Programs by Poverty Level, 2000........................... 80 CHAPTER V Box 5.1 Lessons from Decentralization Experiences in Latin America .................... ........85 Box 5.2 The Decentralization Drive in Education in LAC: Some Cases...........................................................94 Box 5.3 Decentralization of the Health Sector in LAC Socia9....................... 8..................7 Figure 5.1 Municipal Revenues, 2001....................................................88 Figure 5.2 Central Government Transfers to Municipals, 2001 ..... ..............................88 Table 5.1 Budgeted Expenditure for the Public Sector for FY02 ..................................86 Table 5.2 Expenditure Coefficients by CTAR, 2001......................................... .....................87 Table 5.3 Expenditure by Department, 2000 ......... ...............................................................................87 CHA-PTER VT Box 6.1 The Sequencing of Civil Service Reform Measures ...................................116 Figure 6.1 International Government Employment Comparison ....................................................... 102 Figure 6.2 Corruption in Public Services...................................................103 Figure 6.3 Employment and Average Eaings of the Private and Public Sector ......................105 Figure 6.4 Permanent and Contractual Average Earnings............................................................... 108 Table 6.1 Number of Employees and the Average Monthly Wage Bill............................104 Table 6.2 Urban Public and Private Sector Employment and Average Monthly Earnings......................... 104 Table 6.3 Dispersion of Salaries Charged to the PUP by Occupational Group and Salary Grade ...........107 Table 6.4 Maximum, Average and Minimum Eaings by Occupational Group and Regimen ............107 Table 6.5 Hourly Earnings by Occupational Groups and Areas, Third Quarter, 1997 .............................. 114 CHAPTER VII Box 7.1 Regulation of Infrastructure in Peru bDpret............ ........................ . ............ 118 Box 7.2 A Multi-pronged Strategy for Addressing Corruption and Improving Governance .............125 Figure 7.1 Peru's Governance Indicators, 1998/2001 .........................................117 Figure 7.2 Obstacles to Doing Business.................................................. 119 Figure 7.3 Perceived Level of Corruption in Selected Agencies ....... .............................1.9 Vii Boxes, Tables and Figures in Text Figure 7.4 Bribes by Type of Activity........................ ...................................................... 120 Figure 7.5 Corruption in Public Service................................. ................................120 Figure 7.6 Average Percent of Income Paid in Bribes .....................................................121 Figure 7.7 Indices of Service Quality and Bribery................................................122 Figure 7.8a Quality of Public Services..............................................................................122 Figure 7.8b Quality of Health and Education Services ............................................ ................................123 Figure 7.9 Provision of Public Services - Local Governments vs. Other Government Agencies ..........................124 Figure 7.10 National and Municipal Agencies are ridden by different types of corruption.....................................124 Figure 7.11 Public Official's View on Reforms in Public Sector ..................................126 CHAPTER VIII Box 8.1 Creating Mining Stabilization Funds. The Ancash Model..................................142 Table 8.1 PERU: Description of the Existing M ining Fiscal System ...................................................................131 Table 8.2 Mineral Taxes in Peru and Other Countries ................. ........................................133 Table 8.3 Comparative Measures of Profitability and Effective Tax Rate for a Model Base Metal Mine in Selected Jurisdictions..................................134 Table 8.4 Comparative Economic Measures for a Model Gold Metal Mine in Selected Jurisdictions.....................................................134 Table 8.5 Copper Model: Tax System Sensitivity to Price and Cost Changes ........................134 Table 8.6 Availability of Tax Stability in Selected Jurisdictions................... ..............135 Table 8.7 Tax System Sensitivity to Income Tax Rate...................................... ..............136 Table 8.8 Income Tax Rates Applied to Mining Projects in Selected Jurisdictions ....................136 Table 8.9 Depreciation Applied to Typical Mining Equipment in Selected Jurisdictions. ................137 Table 8.10 Tax Sensitivity to Building Depreciation Rates........................................ ............137 Table 8.11 Loss Carry Forward/Back Policy in Selected Jurisdictions..............................138 Table 8.12 Tax Sensitivity to Loss Carry Forward Time Limit...................................................138 Table 8.13 Tax Sensitivity to Annual Allowed Closure Deduction Over the Mine Life.............138 Table 8.14 Tax Sensitivity to Reinvestment Allowance........................................139 Table 8.15 Presence of Mineral Royalty Tax Systems in Selected Jurisdictions .......................140 Table 8.16 Tax System Sensitivity to a Royalty Tax................................ .................140 Table 8.17 Tax System Sensitivity to Import Duty................................... ................142 Table 8.18 Tax Sensitivity to Eliminating Import Duty and Imposing a Royalty ....................................142 Viii Background Papers Background Papers - I Public Sector Employment in Peru, by Peter Gregory Prospects for the Decentralization of Health and Education in Per&, by Richard Moore Applying a Functional Analysis to Ministries' Decentralization, An Example, by Richard Moore Exoneraciones y Regimenes Tributarios Especiales en Per6, by Osvaldo Schenone Background Papers - H Voices of the Misgoverned and Misruled: An Empirical Diagnostic Study on Governance, Rule of Law and Corruption for Peru, by Francesca Recanatini La Deuda Corriente de los Municipios Peruanos, by Eduardo FernAndez LProtecci6n Presupuestaria? Los Programas Sociales Protegidos En El Per6, by Rossana Mostajo Background Papers - Il Peru - Propuesta de Reforma a las Reglas Fiscales, by Jos6 Valderrama Position of the Peruvian Taxation System as Compared to Mining Taxation Systems in Other Nations, by James Otto Peru: Fiscal Trends and the Fiscal Sustainability Issue, by Jos6 R. L6pez-Cdlix and Rashmi Shankar Background Papers - IV Public Expenditure Tracking Surveys (PETS) - Executive Summary, by Lorena Alcazar and Erik Wachtenheim A PETS in Central Government Transfers to Municipalities in Peru: A Detailed Look at the Vaso de Leche Program, by Lorena Alcazar and Erik Wachtenheim Background Papers - V Grado de Focalizaci6n de Algunos Programas Sociales, 1998-2000, by Juvenal Diaz Seguimiento del Desembolso de Recursos en el Sistema Educativo, by Apoyo Institute The Efficiency of Public Spending in Peru, by Quentin Wodon Background Papers - VI Public Expenditure Tracking Surveys (PETS) - The Education Sector in Peru, by Lorena Alcazar and Erik Wachtenheim ix  Acknowledgments This report is a joint effort of the World Bank and the InterAmerican Development Bank. Jos6 R. L6pez- Cdlix (WB-LCSPE) and Alberto Melo (IDB-RE3/OD5) are the task managers responsible for all the chapters and annexes. The report is an outgrowth of a collective effort including partial contributions and background papers by Osvaldo Schenone (consultant-section on taxation, Chapter II); Carlos Oliva (consultant-section on fiscal patterns, Chapter II); Nelson Shack and Oscar Pajuelo (General Budget Director and Accountant General of Peru-Annex C); Bruno Barletti and Lourdes Cueva (SIAF Coordinator and assistant-Section on transparency, Chapter III); Rossana Mostajo (consultant-section on protected programs, Chapter III); Lorena Alcazar, Erik Wachtenheim, Mitchel Seligson (consultants -section on public expenditure tracking survey, Chapter IV and Annex D); Gillette Hall (WB-Box on PROGRESA, Chapter IV), Juvenal Diaz (FONCODES -section on targeting, Chapter IV); Richard Moore (consultant- section on decentralization of social sectors, Chapter V and Annex F); Eduardo Fernandez (consultant- municipal debt, Chapter V); Peter Gregory (consultant-civil service, Chapter VI); Francesca Recanatini and Daniel Kauffman (WB-governance and corruption, Chapter VII); Jose Luis Gualsh (WB-box on regulatory agencies, Chapter VII); James Otto (consultant-mining taxation, Chapter VIII) and Felix Remy (WB- environment, Chapter VIII); Elaine A. Tinsley (WB-Statistical Appendix and thorough analytical and research assistance); Rashmi Shankar (consultant-research assistance on Annexes A and B); and Gabriel Ortiz de Zevallos (consultant-Annex E). Peer reviewers were Ritva Reinikka and William Dorotinsky (both WB), who provided very helpful and detailed inputs. The report also received numerous comments from Andrew Wolfe and Mauricio Villafuerte (IMF), Guillermo Perry, Mauricio Carrizosa, Sara Guerschanik Calvo, Yasuhiko Matsuda, Gillette Hall, Pierre Werbrouck, Quentin Wodon, Keta Ruiz, Marianne Fay, Harold Adelman, Norbert Schady, Evangeline Javier, Daniel Cotlear, Livia Benavides (all from the WB) and Gonzalo Afcha, Xavier Comas, Rafael de la Cruz, Gonzalo Deustua, Jaime Fernandez, Amanda Glassman, Susan Kolodin, Jorge Lamas, and Jaime Enrique Vargas (all from the IDB). It also greatly benefited from the field survey work of Instituto Apoyo in Peru, which did a magnificent job. Hazel Vargas provided the main administrative and production support, and careful editorial assistance. Silvia Marquina de Le6n, Rosalia Rushton, Marianella Rivadeneira and Patricia Chac6n Holt also provided production support. Ana Maria Arteaga and Ana Maria Angulo provided excellent support to field research in Lima. The report benefited from the active guidance and support of Isabel Guerrero (WB-Director), Ana-Maria Arriagada (WB-LCSHD Sector Director), Mauricio Carrizosa (WB-LCSPE Sector Manager), Ernesto May (WB-LCSPR Director), and Ana Maria Rodriguez (IDB, RE3/OD5 Chief). The team would like to express its sincere gratitude for the excellent cooperation provided by the Authorities throughout the process, especially during the team visits in Peru, and including the main mission of late November 2001. More in particular, the team would like to thank Mr. Roberto Dafiino, President of the Council of Ministers; Pedro Pablo Kuczynski, Minister; Kurt Burneo, Patricia Teuillet and Fernando Zavala, Viceministers of Economy and Finance; and Oscar Pajuelo, Accountant General, for their enthusiastic support. Many thanks also go to Veronica Zavala, Nelson Shack, Oscar Blanco, Fernando Lituma, Bruno Barletti, Carlos Ricse, Lourdes Cueva, Waldo Mendoza, Carmen Negron, Javier Abugattas, Lander Aleman, Jorge Estrella, Luis Paz, Pedro Francke, Beatriz Merino and Gustavo Perochena, Gilberto Moncada and Reynaldo Bringas who provided timely inputs and responded to numerous requests. The report has been subject to early and multiple consultation and dissemination efforts in Lima, Peru. In December 2001, at the closing of the main mission, draft preliminary conclusions were discussed with the authorities. In March 2002, a detailed proposal on the revamping of the Fiscal Prudence and Transparency Law was presented in a MEF's seminar on Fiscal Rules. A month later, a framework for decentralization was submitted in the FONCODES/Presidency joint seminar on the Decentralization of the Social Sectors. Finally, in August 2002, the full draft report was openly discussed with officials, think tanks, the media, donors, private sector, and civil society representatives in a seminar on Fiscal Discipline, Public Expenditure Efficiency, and Decentralization jo intly organized with MEF and IDB. Preliminary findings of this report and background papers have also been instrumental in the elaboration of components of the World Bank's Second Programmatic Social Reform Loan and the Country Assistance Strategy for Peru. X1  Chapter I EXECUTIVE SUMMARY A Comprehensive Reform Agenda of Public Expenditure Management For Peru "A successful economic policy needs two inputs: sound technical design and strong political support to bring it forward. A common mistake is to believe that both aspects are substitutes rather than complementary one to each other. In practice, this leads to wrongly believe that political power may offset poor technical design permanently. " Juan Carlos de Pablo "C6mo Fracasa Rotundamente un Ministro de Economia, 1988" A. RATIONALE AND ORGANIZATION 1.1 For the World Bank and the Inter-American Development Bank, it is a welcome privilege to provide the incoming Administration of a member country with joint and comprehensive diagnoses and policy recommendations in those areas that contribute most to enhance the client's development prospects.' One of these key areas is public expenditure management, which is the purpose of this report. As the Toledo Administration enters its second year in office, we hope that the content of this Public Expenditure Review (PER) will be useful for Peru to deal with some of the formidable challenges the country faces in its development agenda. 1.2 This chapter presents a synthesis of Peru's public expenditure reform agenda from the Banks' perspective. It is based on, and distills, the thematic chapters that make up this report. Chapters II-IV examine the core functions of public expenditure management: macro fiscal aggregates, resource allocations to strategic sectors, and microeconomic efficiency of public spending. Chapters V-VIII explore selected themes, such as the decentralization of public administration and the social-sectors; civil service reform; governance and corruption; and mining fiscal and environmental isues. These chapters are, in turn, supported by fifteen topic -specific background papers, including an innovative public expenditure tracking survey on municipal transfers, and numerous and valuable comments provided by authorities throughout disseminaton activities. 1.3 We do not attempt to provide full coverage of all areas affected by public expenditure. Expressing the strategic coincidence between the World Bank and the Inter-American Development Bank around the idea that poverty reduction is the paramount objective of their assistance efforts, the report deliberately focuses on social sectors spending and leaves aside infrastructure and other sectors, which have been the subject of more-specific sector studies in previous years. This report was finalized in June 2002, and discussed with Authorities in mid- August 2002. Policy developments that have taken place after that time are not reflected in it. In the last few years, several sector studies for Peru were published by both institutions in the areas of education, health, private sector participation and infrastructure, institutional governance, financial management, procurement and country strategy. However, the last WB Public Expenditure Review was produced in 1994. I PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION B. MAIN FINDINGS AND RECOMMENDATIONS 1.4 At the onset of the new century, Peru faces substantial development challenges. Significant achievements in terms of growth, poverty reduction and improvements in social indicators up to 1997 have been partially set back by ensuing persistent slowdowns in economic activity, fiscal indiscipline, political uncertainty, nstitutional weaknesses and poor governance. The momentum for reform, which used to be strong in the early nineties, ran out of steam in the late nineties, and genuine efforts by the new Authorities towards reinvigorating it are facing big social pressure for short-term, populist measures.2 Populism may perhaps bring palliatives, but is deleterious for the macroeconomic fundamentals of the country and, ultimately, for the standards of living. There is rather an urgent need to restore fiscal discipline and build momentum for often termed "second-generation" reforms, including those in public expenditure, which would provide longer-term sustainability to the strategy of poverty reduction. 1.5 The new Administration inherited a fiscal system under severe stress. Since 1999, Peru's fiscal stance remains fragile due to a combination of adverse external shocks, expansive fiscal policies, elections, and other unfavorable domestic developments featuring the transition to the post-Fujimori legacy. Led by a primary deficit that for the first time in more than a decade has become negative since 1999, the combined fiscal deficit significantly increased from 0.4 percent of GDP in 1997 to 3.2 percent of GDP in 2000, only to be slightly reduced to an estimated 2.5 percent of GDP in 2001. Low, volatile and procyclical tax revenue combined with a rigid public expenditure facing popular demands complicate efforts to finance an ambitious poverty reduction agenda, while achieving fiscal adjustment. Despite a multi-year pro-poor budgeting framework, social spending remains under-funded, diverted through multiple leakages, and mistargeted. Budget formulation authority's concentration in Lima coupled with discretional and decentralized budget execution through Transitory Councils of Regional Administration (CTARs), reduce the impact of social expenditure and the efficiency of service delivery. 1.6 Since his arrival on July 2001, President Toledo has proposed to take actions in the areas of macroeconomic stabilization; reopening of country's access to international financial markets; budget modernization and State decentralization; social policy; revamping of the armed forces, police and internal security services; and consolidation of Peru's democracy and social stability through consensus-building and transparent and participatory mechanisms. Despite its initial progress, Peru is not likely to meet these goals in the near term. Achieving success in the medium term will rather depend to a great extent of how quickly it restores fiscal discipline, shifts scarce spending to the Government's priorities, and ensures an efficient delivery of public services under a very complex and politically-driven decentralization process. We summarize the public expenditure management agenda in seven main messages: * Restoring fiscal discipline * Reorienting the budget toward pro-poor expenditure * Improving the efficiency of public expenditure * Addressing the promises and risks of decentralization * Upgrading the civil service * Improving governance and reducing corruption * Fine-tuning mining taxation and environment policies 2 In the context of this report, populism is defined as a conception and practice of economic policy that favors unsustainable expansionary fiscal and monetary polices as the main engine of GDP growth and employment creation. For a discussion of the concept, see Dornbusch and Edwards (1991). 2 Chapter I. EXECUTIVE SUMMARY 1.7 The storyline behind these seven themes is a bit long to tell, but simple. First, significant primary surpluses are the inescapable premise for sustainable fiscal and debt-management policies. They are also needed to lay the basis for viable counter-cyclical fiscal and social policies in the future. Second, shifting expenditure to pro-poor priorities is essential for achieving desired social outcomes in the medium-term. Budget reorientation requires not only to enhance transparency of budget cuts and reallocations during implementation, but to review inertial and in-year seasonal spending. Improving budget protection mechanisms of priority social programs is useful to consolidate an effective countercyclical social policy. Third, making most effective use of scarce resources is essential for a country in the midst of fiscal retrenchment. This implies reducing anti-poor disparities in regional budget allocations; eliminating duplications in social and basic infrastructure projects; minimizing leaks of public funds, particularly in food supplementary programs and at the bottom level of service delivery and, if feasible, considering their gradual downsizing or replacement by cash-transfer programs; and improving targeting of resources, specially outside Lima, for most social programs, especially those addressing the most vulnerable. Fourth, it is essential promptly to resolve fundamental questions of the decentralization process: in particular, what functions must be decentralized? And with what timing and outcomes in mind? An ill-designed decentralization process, especially if developed through brand-new regional governments, has considerable potential for provoking a severe fiscal disarray and macroeconomic destabilization. Thus there is a need to proceed very cautiously. Fifth, the Government needs to upgrade its human capital in charge of service delivery. The civil service regime is in total disarray and needs to be overhauled. A trustworthy public registry of civil servants must be established. The myriad of salary supplements and benefits in cash and in kind now prevailing must be consolidated and replaced by a more rational salary system. Sixth, improving governance and fighting corruption is essential not only to provide a better and less expensive public service delivery, but to fight income inequality and facilitate business activity and entrepreneurship in Peru. Local governments, in general, perform much worse than national agencies in service delivery quality, and extent of corruption and governance problems. Seventh, fine-tuning mining taxation, mining-canon transfers and mining-related environmental policies is needed not only to strengthen the global competitiveness of one of the main export sources of Peru through a stable tax regime, but to enhance transparency of transfers to local governments and minimize environmental damage. Conflicts between mining companies and the local populations will thereby be minimized, thus contributing to improve the quality of mining investment activities, the transparency of intergovernmental relations, and the relationship between foreign investors and local communities. 1.8 Around the seven organizing themes of this report, revolves a set of multiple and detailed diagnoses and policy recommendations, with some suggestions for their prioritization. The reader should, however, keep in mind a couple of limitations. First, there is no magic formula for achieving results. What is described here is essentially based on Banks' best-practice experiences elsewhere in Latin America and in the rest of the world, as well as on their informed views of Peru. Second, proposed sequencing of reforms is itself subject to political constraints. The World Bank and the IDB remain ready to assist the Government of Peru in implementing the PER's agenda. Finally, it is important to add that implementing these reforms can only be achieved through a new contract between government and citizens, which would elevate standards for government accountability, transparency, and civil society participation and oversight. Restoring Fiscal Discipline 1.9 Fiscal sustainability is perhaps the most important challenge of the current Administration. Peru needs to strengthen its structural fiscal position, not only because failure to do it jeopardizes the macro framework in the medium term, but also' because significant 3 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION contingent liabilities might arise from pensions, the financial sector, and sub-national government--registered cr non-registered-debts, and because sudden slowdowns or stops in external financing due to financial contagion from external (or regional) turbulence are not to exclude in the coming years. Fiscal deficits are explained by a combination of both structural and cyclical factors. On the structural side, they feature pro-cyclical and volatile patterns, mainly determined by the behavior of tax revenue and public expenditure. Over the last 30 years, Peru's non-financial public sector (NFPS) deficit experienced 5 cycles: 1970-79, 1980-85, 1986-91, 1992-96 and 1997-01 and averaged a high 5.8 percent of GDP. Such pattern resulted from long periods of fiscal expansion, followed by increasingly shorter periods of fiscal adjustment, particularly before the early nineties. The main component of the deficit was the Central Government (CG) imbalance, which represented about 3/4th of the average NFPS deficit. Most fiscal variables are pro-cyclical: they have significant positive correlation coefficients with the growth rate, and the cyclical components of the fiscal deficit and GDP growth are almost perfectly correlated. In addition to this, fiscal variables are extremely volatile, and this goes beyond their close correlation to GDP growth. Peru's growth rate volatility, measured by the standard deviation, is 33 percent higher than the average for Latin American countries (LAC), itself twice as high as the average in industrial economies. The volatility of fiscal variables, though, is much higher than the volatility of GDP growth: almost six times in the case of tax revenues; five times in the case of public salaries and transfers; more than seven times in the case of capital investments (the critical budget adjustment variable); and eleven times in the case of fiscal deficits. Tax revenue volatility is also explained by continuous changes in tax policy and administration; the hyperinflation of the late eighties and elections in the late nineties. Expenditure volatility is explained by erratic policies affecting the public payroll, transfers and capital investments. 1.10 Recent disequilibria are explained by expansive fiscal policies and correcting them is a priority task. Since 1997, trends in CG real revenue and real public expenditure have decreased and increased respectively. In 2001, the tax ratio fell below 12 percent of GDP, a dismal level by international standards and among the lowest in Latin America. As a result, primary balances have become negative since 1999, and fiscal deficits have deteriorated. Resulting public debt has increased as a percentage of GDP from 42.7 percent in 1998 to a high 46.5 percent in 2001, even though significant amounts of privatization proceeds have been used to finance expenditure and repay public debt. Although the public debt is mostly long-term and owed to external creditors, the domestic and short-term components of it have grown at a rapid pace in the last several years. In this connection, it must be borne in mind that domestic debt is exposed to currency risk and rapidly increasing debt service, which contributes to tighten scarce resources in a shrinking budget. Had Peru kept Government tax revenue and expenditure constant in real terms during the 1998-2001 period, it would have preserved a sound fiscal deficit below 1 percent of GDP. 1.11 What needs to be done to achieve fiscal discipline and, hence, macro sustainability? Six main policies are essential. First, and foremost, Peru needs a comprehensive tax reform. This reform should aim at improving the tax system's equity, neutrality, simplicity and revenue- yielding capacity. Reform should mainly be based on a combination of measures to broaden the tax base and enhance tax administration. Proposed measures to broaden the tax base suggest removing key tax exemptions (be it sector- or region-specific, or on interest and capital gains, or the exemption on import duties for CETICOS); introducing a presumptive income-tax method for the self-employed and a slightly wider three-level income tax rates structure for individuals; and unifying the consumption selective tax for vehicles. By complementing these measures with the introduction of a royalty tax on mining (offset by the elimination of the sector's import duties), and/or eventual modification of workers' participation, and the implementation of a major phased improvement in tax administration, the Government could raise additional resources approaching 4 Chapter I. EXECUTIVE SUMMARY 2 percent of GDP within. 2-3 years. The fact that the set of measures proposed does not modify the present tax structure and does not establish new taxes, except in the case of the royalty tax, makes it easier to gain political consensus for its enactment. The other pillar of tax reform is a marked improvement of tax administration by SUNAT, the collecting agency. In general, any improvement results from a combination of enhanced taxpayer services (with upgraded technology support), better auditing systems that focus on tax evaders, and better human-resource staffing. 1.12 The second area where a significant effort is required is on public expenditure. As capital investment continues to be the residual adjusting item in the budget, Peru now has one of the lowest capital expenditure ratios in LAC: less than 2.5 percent of GDP in 2001 and decreasing. Hence, expenditure reform should aim at reverting the present trend toward an increased current expenditure and a decreased capital investment. Excessive current spending could be cut in the area of defense and internal security in the short term (0.6 percent of GDP in 2002), and in wages and pensions in the medium term, supported by social security and civil service reform. If proposed increase in fiscal revenue materializes, then capital investment, especially in the social sectors, should be raised accordingly, thus contributing to fill the present gap that Peru has in social expenditure-- 1.2 percent of GDP in education and 0.2 percent of GDP in health--with respect to LAC averages. 1.13 A third area of policy reform is public debt sustainability. The public debt-to-GDP ratio of 46.5 percent in 2001 is too high for a country whose public debt is highly exposed to currency risk, and which faces potential contingent liabilities in dollars in the financial sector. Lowering such ratio to more sustainable levels is essential to face potential risks in the next decade. Assuming a conservative baseline scenario combining moderate growth and low inflation, a slight increase in external interest rates, decreasing, but still significant mixed concessionamultilateral financing, and a fiscal correction lowering the Non-Financial Public Sector (NFPS) deficit from 2.5 percent of GDP in 2001 to -1.4 percent of GDP in 2010, would bring the public debt-to-GDP ratio to a benchmark 36 percent of GDP at the end-of the decade. 1.14 A fourth area of policy reform relates to fiscal rules through the revamping of the Law of Fiscal Prudence and Transparency. As the Law is not a substitute, but a complementary tool to build credibility on a sound fiscal program, a prior condition for its official revamping requires, in the near term, thorough compliance with the benchmarks of the 2002-03 macroeconomic program, supported by an IMF Stand-By. Moreover, we do not expect a significant countercyclical policy as far as growth remains moderate and fiscal revenue is stagnant. So, revamping of the Law should proceed gradually, focusing on early credibility gains; prior institutional strengthening required by the Law's technical redesign and monitoring; consensus- building exercises led by Authorities; virtual monitoring indicators and reporting tools, including those for the Comptroller's and Congress oversight; and on introducing a set of sanctions for the case of non-compliance with numerical benchmarks and procedures. The design of a set of fiscal rules for incoming sub-national governments should impose hard budget constraints on them and justifies the upgrade of the incoming Law to one of Fiscal Responsibility. In addition to this, in the medium term, an effective countercyclical fiscal policy could be based on a structural primary balance benchmark. Such policy would require to make automatic fiscal savings in booms and dis-savings in the downturns, measured by the differential between estimated actual and structural balance. Under the baseline scenario described above, achieving a more sustainable public debt- to-GDP ratios by 2004-10 would entail a structural primary surplus ranging around 1 percent of GDP (roughly equivalent to around a 1 percent of GDP non structural deficit) during such years. Future savings woukl strengthen the role of the Fiscal Stabilization Fund. 5 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 1.15 A fifth area of policy reform regards contingent fiscal liabilities. Peru must start making transparent budgetary provisions to deal with their eventual materialization, which would make projected restoration of fiscal discipline a more difficult task. This report overviews three main sources of contingent risk: bank loans, pension payments and municipal debt. The Banks' broad estimates of the potential fiscal exposure to a systemic banking crisis is 4 percent of GDP, which is considered in the lower bound of comparable -mostly double -digit-world standards. Sound macroeconomic policies, close supervision of the financial sector and reduced exposure to external financial contagion can minimize this rsk. In addition, pension reform has generated two non-banking debts with the private sector which are actually paid through annual transfers. Official transfers to the social security represent 0.6 percent of GDP in 2002, and conservative projections indicate they will increase to about 1.0 percent of GDP per year in 2005-07. Pension reform should minimize this liability. Finally, registered municipal debt is not a threat for macroeconomic stability, but it is significant and explosive for some local,governments becoming non-solvent and illiquid. These local governments will have severe difficulties to assume additional responsibilities with decentralization due to their financial constraints. Nationwide total registered debt represents a third of the corresponding national debt (current national liabilities considered as short-term debt). In the short-term, identifying non-solvent municipalities and developing a plan for the gradual repayment of their registered debt, especially on pensions, is essential before these districts assume further responsibilities under the forthcoming decentralization process. In addition to this, the Fiscal Prudence and Transparency Law should put strict limits to new external and domestic debt, and prohibit sub national governments of acquiring contingent debts. 1.16 Finally, a sixth area regards public debt financing. Peru must complete its deficit- financing plan by meeting its privatization targets in the nearest term, sustaining its prudent external debt management strategy-including an enhanced access to international capital markets-and developing a market for long-term domestic public debt, issued in soles and inflation- indexed. Reorienting the budget toward pro -poor expenditure 1.17 The task of putting the economy on a sustainable, rapid and poverty reducing growth path has to be to be carried out with the help of a modern public expenditure management (PEM). Promising steps have been done by Peru over the past few years. This includes: the creation of an Integrated Financial Management System (SIAF, in its Spanish acronym) in 1998, the strengthening of accounting institutions and procedures, compatible with world standards, applied to the entire public sector since 1999; the development of a website Portal de Transparenci in 2000, which includes defense and national security spending for the first time from 2002; the passage of the Law of Fiscal Prudence and Transparency in 1999; the introduction of performance management contracts for State-Owned Enterprises under FONAFE in 1999; the adoption of multiple medium-term expenditure frameworks since 2000, based on 3-year Macro- Economic Frameworks and 5-year Multi-Annual Sector Budget (MAB) Plans, and carried out through a participatory methodology for multi-annual budgeting formulation since 2002; and the adoption of budgetary provisions for a set of Protected Social Programs (PSPS) as main counter- cyclical social policy since 2000. 1.18 The trend towards pro-poor spending and enhanced accountability and transparency seems irreversible, but consolidating these achievements in the medium-term means to solve five critical institutional weaknesses. These are: excessive budget rigidity, an incomplete SIAF, a disconnection between multi-annual and annual budgeting, a fragile institutional framework surrounding its counter-cyclical social policy, and an inadequate and user-unfriendly budget 6 Chapter 1. EXECUTIVE SUMMARY monitoring and reporting that affects public perceptions about the real degree of budget transparency. 1.19 Following a budget tracking test of the status cf PEM jointly developed by the World Bank and the IMF, Peru scores low in the area of budget formulation, average in the area of budget execution, and high in the area of budget reporting other than external. In the first area, the assessment recommends that budget classification be made fully consistent with the GFS definition of General Government and the quality of budget projections (especially in function and sector composition) be upgraded if social priorities are to be sustained in the medium term. The major constraint for budget formulation, though, is inertial expenditure. Peru ranks high among countries with the most inflexible budget not only in LAC, but worldwide. In addition to Brazil, India and a few HIPC countries, which have above 90 percent of their total expenditure fixed for particular functions, Peru has 91 percent and 86 percent of ordinary and total budget resources in 2002 assigned to inertial commitments. Budget rigidities originate from different sources including: wage and benefits. considered by the Constitution as inalienable rights; permanent contracts hidden as non personal services (no less than 40 percent of the total budget is allocated to such services, according to the Authorities' estimates); an inertial payroll under artificially prolonged investment projects; earmarked transfers to municipalities and the private sector, as the Vaso de Leche and Comedores Populares programs; and an earmarked use of the budget "contingency reserve" to fund the March school bonus, and July and December salary supplements (aguinaldos). Peru needs to reverse such a pervasive pattern. PER recommendations in this area include: an immediate general hiring freeze under the non-personal services category accompanied by strict supervision, on a case-by-case basis, of short-term renewal of existing temporary contracts, upon their prior revision; an urgent review of present policies regarding pension and wage benefits in a comprehensive civil service reform; streamline funds assigned to inefficient social programs-like Vaso de Leche or Comedores Populares; and make fully transparent the use of the contingency reserve, so as to let it play its true role, which is to fill un- predicted in-year budget needs, thus preventing it to fill recurrent needs. 1.20 SIAF core sub-systems are incomplete. The lack of a budget module, in particular, is a severe handicap. Contrary to similar systems in Latin America, Peru's SIAF is neither a budget formulation tool nor a budget assignment tool, but fundamentally a virtualpayment system. As a result, budget formulation and in-year modifications remain little known externally, centralized and decided at MEF's discretion, and only known ex-post, once they are monthly-registered in SIAF. Hence, a user-friendly budget module integrated to SIAF is an urgently needed policy tool to help assess consistency of budget proposals with macroeconomic conditions in early stages; facilitate monitoring of in-year budget modifications, consistent with inter- and intra- sector budget priorities in a participatory and transparent way; and allow Congress' legally mandatory oversight of the budget. SIAF also requires a multi-annual budget sub-module too; i.e one with capacity to track GOPs' compliance with multi-annual budgeting priorities and facilitate production of mandated bi-annual reports on compliance with MAB targets. Finally, other critical missing modules are the ones on human resources management, national assets, procurement, the national investment system, contingent liabilities risk, and regional and municipal financial management. The human resource module is essential for civil service reform. The one on national assets to have due registry of state assets. The module on procurement, starting point for &government, is critial for anti-corruption efforts, the most vulnerable area of governance. The module on public investment should help prioritize multi- annual investment plans. The module on contingent liabilities risk would help to register them and make proper budgetary provisions, while the one on municipal management would prepare SLAF support to an accelerated scheme of fiscal decentralization, as future regional and municipal governments require a tailor-made SIAF, whose development should take many years. 7 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 1.21 The budget tracking test also found that multi-year projections are disconnected from annual budget priorities and modifications, their quality requires significant improvement, and their monitoring an upgraded and regular reporting system. To such purpose, the creation of a budget module and refinement of the existing set of financial and physical indicators in SIAF should facilitate an user-friendly tracking of pro-poor expenditure and an effective results-based control system in budget execution. Complementary eternal reporting of budget execution, multi-annual budgeting and audits should be frequent, timely and comprehensive. Regarding reporting, the SIAF sub-system to track financial and physical performance indicators, now being developed, requires a thorough conceptual review before completion, because it is a critical tool for an outcome-oriented budget, and support from monitoring targets and performance management contracts . The present sub-system has too many useless and outdated indicators, and their use is poor. 1.22 The development of new SIAF modules requires detailed prioritization and permanent training. Authorities have good reasons to consider the creation of a university-degree in management of SIAF systems. Besides, associated systems could be developed outside SIAF, like a fiscal cadastre, or eventually be connected to a broader e-government strategy (like e- procurement or a geo-referenced expenditure map at district level, for targeting purposes). 1.23 These shortcomings have not prevented an important shift in the budget towards pro-poor expenditure in Peru. In spite of the recession and resulting fiscal adjustment, the share of social expenditure to GDP evolved from 3.9 percent of GDP in 1990 to 7.0 in 1999, and this share has remained steadfast up to 2002. The fact that the share of social expenditure to GDP has not declined is a positive outcome in a fiscally retrenched economy where total expenditure decreased from 19 percent of GDP to 17.8 percent of GDP in the period 1999-2002 (about one- seventh in dollar terms). Likewise, significant inter- and intra-sectoral pro-poor spending reallocation has begun to take place between 2000 and 2002. Budget for defense and national security has decreased from 2.9 to 2.1 percent of GDP; that for education, health.and social assistance has increased from 5.3 to 5.6 percent of GDP; and ESSALUD expenditure (non- targeted public health insurance administration) has decreased from 1.5 to 1.3 percent of GDP. In addition to this, the share of poverty reduction programs (consisting of both programs with universal-coverage spending and programs based on targeted spending) in social expenditure increased from 57 percent in 2000 to 60 percent in 2002. Poverty budgeting also includes an innovative effort to protect key social programs (PSPs) since 2000. This approach has consisted of assigning global budget allocations to non-salary current and capital spending grouped in 11 programs and distributed among eight social and basic infrastructure sectors. In 2001, budget resources equivalent to 12.7 percent of the ordinary budget and 1.9 percent of GDP were protected and executed. 1.24 In 2002, the Government decided to change its PSPs policy from merely budget protection when a downturn occurs, to sustained spending floors needed to reach Millennium development goals. This was accompanied by their improved focus, monitoring mechanisms and full transparency in their execution. The on-going restructuring of PSPs includes: expenditure allocation shifts in broad sector functions-education, health and sanitation, and social assistance; a reduced number (from eleven to six) of now core PSPs, grouped under three sectors: education (pre-primary, primary and secondary), health (collective and individual) and social assistance, a reduced budget as the amount of aggregate earmarked expenditure to this revised set of PSP represents 1.2 percent of GDP in 2002, down from the previous 1.9 percent of GDP in 2001 (this reduction of 0.7 percent of GDP makes room for increased budget flexbility); a set of monitoring indicators for specific projects grouped under the selected PSPs; and in a few projects, definition of some performance management contracts to improve their efficiency (e.g. the Integral and 8 Chapter I. EXECUTIVE SUMMARY Mother and Child Health Insurances-SIS-SMI). To introduce further flexibility, PSP budget also includes escape clauses to make provisions for budget under-execution or a major budgetary shortfall at the sector function level; and internal reallocations from under-performing to over- performing programs or projects, but maintaining global spending floors, at the program and project levels. To institutionalize and make its countercyclical social policy more transparent, Peru should enhance its de facto "virtual" protected budget policy following the example of Uganda. An enhanced mechanism would add two major features: the development of a permanent institutional framework in charge of PSPs policy, and the introduction of oversight capacity by sub national governments and civil society organizations, as well as of transparent monitoring and evaluation indicators, into PSPs implementation. Improving the efficiency of public service delivery 1.25 The efficiency of Peruvian public expenditure is below the Latin American average for health, but above average for education, and in relationship to world averages, similar results holds true. However, these results should be taken with caution. Indeed, Peru's level of efficiency in reaching health outcomes is below both the Latin American and the World averages. This is particularly true in malnutrition-height, under-five mortality and life expectancy. The range of values found suggests that Peru could and should do much better in health outcomes. And although outcome indicators for Peru's net primary enrollment is on par with other Latin American countries, whereas for net secondary enrollment is much above LAC average, its educational shortcomings are too severe in terms of the low quality of education services and poor teacher's qualifications, especially in rural areas. 1.26 Two major sources of inefficiencies are leakages of public resources and poor targeting of social programs, which ultimately contribute to poor governance and a corruption-prone management. 1.27 A highly decentralized budget execution such as Peru's raises the possibility of significant leakages in the flow of resources, thus reducing the extent to which GOP expenditure is able to produce better outcomes. By applying an innovative public expenditure tracking survey (PETS), it is possible to show that there are substantial leakages in the resources that municipalities receive under the Vaso de Leche program and then transfer to the ultimate beneficiaries, through several intermediaries. The actual amount received by final beneficiaries appears very much lower than the initial amount sent from the central government: on average 29 cents of each dollar initially transferred by the Central Government actually reach the intended beneficiaries! The survey surprisingly indicates that the leak i much higher in the bottom (VDL committees and households) levels rather than in the top (CG and municipalities-leaks) levels of the ladder, which not only demonstrates significant improvements in the official channeling of resources by the SIAF/MEF, but challenges the predominant view that local private organizations are more accountable in managing public resources than official ones. Leakages in municipality- level transfers are compounded by the generalized lack of audit controls, poor transparency and volatility. Finally, leaks affect more the poorest, urban and provincial municipalities, and their level appear similar among districts of different sizes and distances to the province. Authorities should pursue a twin strategy. In the short term, any effort leading to improve supervision, accountability and transparency of the flow .of resources should be pursued. This includes a regular and user-friendly publication of detailed transfers to municipalities by SIAF, capacity training at the local level in SIAF reporting, proper registration of beneficiaries, and strengthened budget and auditing procedures by the Comptroller's office. A mini-Stabilization Fund for canon minero transfers could contribute to minimize their annual volatility. In the medium-term, however, Authorities should consider replacing these programs by a voluntarily adopted cash- 9 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION transfer ("grants") program, following the successful model of the PROGRESA program in Mexico. Under such scheme, grants would be provided to mothers not only for food supplements, but also for basic education and preventive health, which, following a proper registration of beneficiaries, should also improve targeting of resources on beneficiaries at risk. 1.28 Poor targeting is another important shortcoming of the multiplicity of social programs in Peru. The general level of spending on targeted extreme poverty reduction programs increased from 1.1 percent to 1.7 percent of GDP between 1993 and 1998, but then remained constant up to 2002, which appears adequate due to the fiscal situation and program inefficiencies, and despite pressures to increase it . Poor targeting, however, deviates part of these outlays toward the non- poor. Targeting rates in social programs vary tremendously. Past findings indicate that with the exception of FONCODES and PRONAA, which have had about a 50 percent success rate, most targeted social programs have had rates reaching below 40 percent of the intended poorest. More recent findings on the food and nutrition programs confirm such picture. Targeting among the three main national programs-Desayunos Escolares, Vaso de Leche y Comedores Populares, also reflects prevailing weaknesses: in 1998-2000, Desayunos Escolares had the highest success rate of 70 percent, followed by Vaso de Leche and Comedores Populares, which had lower rates of 64 and 54 percent respectively. As a program is better targeted, the differential between its rural and urban success rates decreases: Desayunos escolares had a lower range (77 vis a vis 56 percent); than Vaso de Leche (77 vis a vis 50 percent) and Comedores Populares (72 vis a vis 37 percent). Finally, programs of temporary employment, like A Trabajar Urbano, have not started with the right foot, since it offers too high a wage rate to attract workers in the poorest deciles. Authorities should make intensive use of excellent targeting tools, especially the poverty map, already in their hands. The myriad of multiple social programs in Peru should be ranked and those with substantial mis-targeting be redesigned, merged or closed. Fiscal savings generated from such exercise should be used to increase coverage of more effective programs. Targeting criteria should be made more explicit and focused on populations extremely poor. In the case of A Trabajar Urbano, the immediate solution is to decrease the wage rate paid. A similar policy in A Trabajar Rural is not possible, since the individual income earned by a rural worker is often shared with other members of his/her community, and so the self-targeting mechanisms does not work in a similar way as it does with urban workers. In both cases, though, Authorities should be aware of the counter-cyclical and short-term nature of these employment generation programs: once the economy recovers its dynamism with high rates of growth, their budget resources should be phased out. Addressing the promises and risks of decentralization 1.29 In the past decade, Peru made irreversible progress towards redefining the role of the State but this agenda remains unfinihed. In the early-to mid-nineties, the size and scope of the public sector was reduced; the nature of government shifted from ownership and intervention to market regulation and provision of social services. However, in the mid-to-late nineties the State reform process lost steam and key tasks were not even initiated. Prominent among pending reforms set aside by the authoritarian leadership of the nineties was the decentralization of the State. 1.30 The current government has decided to reinvigorate the State modernization and decentralization agenda. This decision must be seen as part of a broader, deeply positive, trend toward democracy in the region. However, state decentralization is pregnant with both potentialities and risks. If appropriately carried out, it can, indeed, increase both the efficiency of government-service delivery and the responsiveness of government to citizens' needs. Nonetheless, decentralization also entails serious macroeconomic risks. Firstly, central 10 Chapter 1. EXECUTIVE SUMMARY government deficits can worsen if governments are unable to reduce expenditures or increase revenues to make room to accommodate the fiscal cost of additional functions and responsibilities transferred to sub-national governments. The Government may find itself decentralizing both revenues and expenditures, but being unable to reduce existing levels of central-government spending in those functions that have been decentralized. Secondly, attempts to eliminate rising deficits through higher taxes can fail or take longer than expected to materialize. Thirdly, decentralization can also hamper a government's ability to respond to economic shocks as it reduces central control over aggregate public -sector revenues and expenditures. Fourth, sub- national-government debt policies and indebtedness behavior can undermine overall fiscal discipline. Sub-national governments may end up borrowing excessively in the expectation that the central government will bail them out. In sum, decentralization has the potential to be a destabilizing factor for the already deteriorated Peruvian public finances and, ultimately, for the country's overall macroeconomic framework. The Latin American experience suggests, moreover, that decentralization should not be carried out without a coherent overall strategy and without a prior technical and institutional preparation at the sub-national levels of government. In the case of Peru, the fact that municipal governments are heavily dependent on central transfers justifies caution in the transfer of responsibilities to them. Taking into account that the degree of satisfaction in the provision of public services by local governments appears to be low, gradualness in the transfer of functions and a sizeable effort to strengthen the administrative capacity of sub national governments prior to the transfer of responsibilities seem to be necessary conditions of any reasonable decentralization strategy. 1.31 To enter the process of decentralization on a sound path is a critical public policy challenge for the Toledo administration. While Peru's initial efforts toward modernization and decentralization offer substantial promise and go, for the most part, in the right direction, the latent risk that fiscal discipline may turn out to be a casualty of the decentralization process will continue to be high until a clear framework for inter-governmental public finances is finally enacted and enforced. The guiding criterion for such a framework has to be the establishment of hard-budget constraints for the expenditure and-external and domestic-indebtedness decisions of sub-national governments. The immediate agenda includes issues of paramount importance such as the alternative options for intergovernmental fiscal relations; the budget and indebtedness constraints on sub-national governments; the possible functions and competencies to be decentralized; the characteristics and funding sources of the future Regional Compensation Fund and Intergovernmental Fund for Decentralization; and any possible changes to the transfers scheme for municipalities. It is also imperative that the hard-budget constraints for regional and local governments here advocated be an integral part of the future revamped Law of Fiscal Prudence and Transparency. Finally, a complete diagnosis of the financial status of municipalities, including its pension liabilities and other registered and non-registered debt, and a plan for both strengthening municipal tax collection and restructuring local government debt are essential before these are assigned new expenditure powers. Upgrading the Civil Service 1.32 Civil service reform is the missing link of the outcome- and poverty reducing- oriented modernization of the state in Peru and is in desperate need of a major overhaul. The past decade witnessed a severe deterioration of Peru's civil service. Hyperinflation of the late eighties and the beginning of the 1990's sharply reduced the real earnings of civil servants. A poorly conceived downsizing of the central government work force in the early nineties, which, in the end, amounted to a very modest reduction in the overall public sector, led, however, to the departure of the most capable civil servants. There is an inherited disregard for the provisions of the existing civil service law and its associated regulations. Civil servants are employed under a 11 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION variety of legal regimens with a wide dispersion in salaries for similar tasks. The existing formal salary structure has not been revised in over a decade, so base salaries are increased through resort to a multiplicity of supplements and benefits in cash and in kind. With the proclamation of a ban on hiring into permanent positions ten years ago, institutions have resorted to hiring on a contractual basis, a process that is not subject to any required procedures for recruitment and selection and has severe fiscal implications. All these factors have contributed to the existing legacy of a work force that is poorly prepared, poorly trained, poorly supervised, and hence of very low productivity. Consequently, there is an urgent need for a thorough reform to re-establish some semblance of control over the wage bill and to promote the creation of a more efficient government The current civil service Law has major weaknesses that argue for its replacement by a new framework designed to promote a higher level of efficiency in the delivery of government services 1.33 Any initiative designed to improve the performance of government must include a major overhaul in the institutional framework that governs public sector employment and the management of its human resources. Four actions are particularly relevant. First, the creation of a Civil Service Reform Commission should precede the establishment of an autonomous Directorate to administer the civil service. Second, developing a SIAF module on human resource management would greatly simplify registry of public servants and would be instrumental in the task of simplifying the remuneration package from the myriad of present benefits to a cash-equivalent amount. Third introducing pilot performance contracts in the health sector seems appropriate, given its recent successful experience with SMI, which makes it a natural candidate for later mainstreaming of such practice in the rest of the public sector. Finally, a new Law and Reglamenb should place emphasis on merit and specify determinants for compensation. 1.34 The education sector also faces serious problems with its civil service. There is general dissatisfaction with the quality of instruction and with the allocation of faculty resources among schools. The Ministry of Education is taking steps to improve the quality of instruction in the country's schools of pedagogy. However, the profession continues to attract individuals whose test scores rank at the bottom of the distribution of !udents entering institutions of higher learning. Improvement in the quality of instruction is likely to require a number of reforms. While the current level of salaries appears to be more than adequate to attract a surplus of candidates with minimal qualifications, the level and earnings trajectory over time are inadequate to attract more promising individuals. In addition to this, both the Ley Magisterial and the Ley General de Educaci6n are too rigid for most forms of decentralization and should be revised. 1.35 Reforms must focus on four aspects. On the institutional side, the current institutional morass in the education sector (financially depending on regional bodies, and normatively depending on sector ministries) should be unified under the decentralization design. On the recruiting policy, new entrants to the profession should be required to qualify by passing a rigorous examination. Successful candidates would be hired at a higher salary not much different than the current entry wage, but should be able to look forward to salary increases similar to those accruing to persons with similar skills and abilities in the private sector as long as performance is evaluated and meets defined standards. On the hiring levels, implementation of a new remuneration policy would be facilitated by moving to a system of contracting of teaching personnel in place of the currently generalized system of permanent tenure upon appointment. This would also have obvious fiscal savings. On the legal side, the new Ley del Profesorado and the Reglamento de la Ley del Profesorado should be based on teaching performance and effective attendance, rather than on a range of factors such as seniority, number of training courses 12 Chapter I. EXECUTIVE SUMMARY attended, papers presented at professional meetings, etc., none of which have any necessary impact on the quality of teaching. Improving governance and reducing corruption 1.36 Peru's picture in matters of governance is mixed. In the World Bank's scoring system for governance indicators, Peru's scores in Government effectiveness and regulatory quality have dropped significantly, even as the country moved from an authoritarian regime to one with a more vocal and active opposition. The score in corruption has remained approaching the LAC average. Peru's corruption appears particularly high in procurement contracts of public works. To offset such gloomy picture, Peru's scores in voice and accountability, political stability and rule of law-have improved, thus reflecting the political opening that is taking place. Significant challenges remain in these areas though. 1.37 Reducing corruption and improving governance requires a multi-pronged strategy that addresses the various forms of corruption - from state capture to administrative corruption - enhances political accountability and transparency, promotes a competitive private sector, strengthens institutional restraints, improves public sector management, and embraces civil society participation. It is more than just a matter of law enforcement. The evidence in this report also identified public officials' support for civil society oversight and merit-based civil service reform as anticorruption mechanisms. Public officials also favor strengthening public administration, in terms of simpler administrative procedures. They are also in favor of the introduction of a system of checks and balances and the oversight of public sector's activities by the civil society. 1.38 The Government first needs to build credibility on its intentions to fight corruption. In the survey, there is a significant discrepancy between users and public officials on their perception of the government's commitment to fight corruption. More than half of users, versus only 20 percent of public officials, reported that the government is not committed at all. This lack of trust is keenly felt among low-income users of public services. Conversely, the fact that a high proportion of public officials report preparedness in the public sector to fight corruption is an encouraging sign. The government needs to put the fight against corruption as a top priority. The church and the media should be potential contributors in the fight against corruption. 1.39 As a complement, the Government needs to continue enhancing transparency, especially through SIAF. Transparency is synonymous with openness. Introducing it requires that the Government provides the public-civil society organizations, the media, or anyone else who is interested-sufficient information about the budget and project activities, so as to serve as an effective check on abuses by government and public officials. Issuing a Freedom of Information Law that require governments to provide information to the public unless there is a valid reason (such as a threat to national security) for the information to remain secret would be welcome. 1.40 The Government needs to be accountable and reduce the potential gains for "captor" firms and politicians alike. Authorities should consider issuing an Ethic Code Law for public service with clear prohibitions of conflict of iterests, and an effective enforcement of both political and civil service positions. This is an essential corruption-deterrent tool. They should also demand transparency of the financing of political parties and campaigns making clear the links, both explicit and implicit, between politicians and the interests that support them. Mandated and publicized detailed reports of the finances of all political organizations, and identification of contributors and beneficiaries are tools to reduce the sources of state capture. Another possible measure would include banning the use of state resources for the incumbents' political 13 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION campaigns, limiting the amount that can be spent on political campaigns, providing public funding, and prohibiting certain types of entities from contributing to political campaigns. 1.41 The Government needs to promote a sound business environment by examining and simplifying the number of regulations and inspections. Clear rules defining when a regulation is appropriate, a sound analysis of regulation's impact on firms accompanied by a constant dialogue with the business community should be required to make regulatory reform sustainable. 1.42 Finally, a high-level steering committee, supported by a professional secretariat, can help to bring together representatives of governmental bodies, both central and local, to develop the specific action plans that would implement the governance strategy. By explicitly including representatives from outside government, such committees can further build credibility while mobilizing an important ally for reducing corruption. Fine-tuning mining taxation and environment policies 1.43 Given its size in the economy and social network, the mining sector is essential for public finance, structural reform and governance in Peru. The mining industry generates more than 45 percent of export earnings and contributes to about 5-7 percent of GDP. It employs over one- quarter million people and its tax exemptions account for about 0.4 percent of GDP, which is non negligible. Key structural and transparency issues have negatively affected the competitiveness and quality of insertion of mining with foreign investors and local society, particularly with the local communities since the late nineties. The mining reform brought about the effective opening of the sector to direct foreign investment and the shift of the role of the state from owner-operator to lessor-regulator. This was accomplished by reforming the legal and institutional framework, privatizing the mining state owned operations, and starting up an environmental management program. While the first three reforms, legal, institutional and privatization, have been widely considered successful, the sectoral approach of the environmental reform has proven to be controversial and is considered in need of adjustment. The perception of inadequate environmental control and of widespread conflict between mining companies and local populations is endangering the contribution and the stability of a key sector of Peru's economy. In addition to this, the highly centralized and non-transparent nature of transfers from the Central Government has created an issue of credibility with the local populations. Although the Mining Law establishes that 40 percent of the income tax generated by the mining operations-the canon minero-should go back to the municipalities (20 percent) and the regions (20 percent), and the report's findings show that its overall redistribution has de facto been progressive, there is a widespread perception that the local communities are not benefiting from it. Furthermore, the environmental information that the Ministry of Energy and Mines (MEM) has received from the mining companies has not been adequately released to the public, resulting in a feeling of mistrust against mining activities. Loss of competitiveness due to the taxation system is also an issue. To the extent that it affects rates of return on investment, the effective tax rate in a given country is a major factor influencing foreign investors' decision as to where to invest their capital. The report shows empirical evidence that supports the view that the current Peruvian mining tax system is among the more globally competitive systems worldwide and provides an appropriate balance between investors' needs and government needs. However, there is room for improvement. 1.44 Addressing major issues in taxation, environment regulations and the transparency of canon minero transfers is very important for preserving international competitiveness. Main tax policy changes proposed are to introduce a 3 percent royalty tax combined with the elimination of import duties and modifications to workers' participation; adjust by a premium raise of 5 percent all future stability agreements; extend the carry-loss forward time limit beyond 5 years; enhance 14 Chapter I. EXECUTIVE SUMMARY the system to register tax deduction of mine contributions to local communities and infrastructure; and publish canon minero monthly transfers to municipalities in SIAF, while providing training at the local level. Agreements on roles and responsibilities in the environment provisions are needed, accompanied by a new regulatory framework of environmental management, a new designated national authority with no conflicting interests and integrated with regional and local authorities; and with local participation-NGOs included-in local environmental monitoring and consultation mechanisms, so as to allow them to protect communities interests. Given the volatility of fiscal transfers, the 'creation of small mining stabilization funds, simulating the successful experience of the Ancash Model, is not to exclude. Assembling all the pieces of the agenda together 1.45 Peru's policymakers are well aware of the difficult fiscal situation, and the need for discipline. The agenda described above is large and requires prioritization and sequencing, possibly starting with tax reform and a modem and institutionalized pro-poor budgeting, followed by efficiency improvements in social programs and civil service reform-both critical inputs to consolidate expenditure management reform-and then completed by the decentralization, governance and anti-corruption programs, as well as by fine-tuned environment and taxation policies in the mining sector. Short-lived aggregate demand-boosts, like those advocated by populism will only delay needed reforms. Consistent choices have to be made and some efforts will likely have to be postponed. Most of the PER's agenda, however, is feasible provided that there is not only commitment of the Executive, but also support in the Legislature. Authorities are well aware of this. Last year budget presentation to Congress broadly covered each of three aspects of public expenditure performance. It singled out the central need to restore fiscal discipline for preserving price stability, readjust the pattern of expenditures toward priority social areas, and improve service delivery with efficiency gains. This report argues that for fiscal policy to become a powerful tool for Peru's development, designing an outcome-managed, poverty- oriented and consensus-built budgeting strategy is necessary. Both Banks are ready to continue assisting authorities in marshalling such efforts. 15  Chapter II PERU'S FISCAL CHALLENGES AND VULNERABILITIES 2.1 At the onset of the new century, Peru faces daunting fiscal challenges due to a combination of adverse external shocks, expansive fiscal policies, elections and other unfavorable domestic developments featuring the transition to the post-Fujimori legacy. During the early-to- mid nineties, volatile and pro-cyclical expenditure accompanying good times made the country's response to negative shocks more difficult to manage and vulnerable to a prolonged sbwdown, like the ones that have occurred since the end-of-the decade. Since 1998, the fiscal stance has deteriorated with successive negative primary deficits, in part due to a resilient fall in tax revenue, which reached a decade-low 11.9 percent of GDP in 2001. The ensuing combined fiscal deficit significantly increased from 0.4 percent of GDP in 1997 to 3.2 percent of GDP in 2000, only to be slightly reduced to an estimated 2.5 percent of GDP in 2001. Rigid and expansionary expenditure has also undermined fiscal stability. Ending privatization proceeds and high and dollarized public debt offers limited room for financing deficits in the years to come. The prolonged economic slowdown and political uncertainty resulting from external and domestic factors, including populism, also put additional stress on the fiscal balances. The fiscal rules enshrined in the Fiscal Prudence and Transparency Law of 1999 have not been fully.met. They are, though, no substitute for sound policy and require thorough revamping before they become a helping tool for consolidating credibility. This chapter analyzes Peru's main fiscal challenges and vulnerabilities for restoring fiscal discipline, including tax policy measures and new rules to be considered in a revamped Law. A. LONG-RUN FISCAL TRENDS' 2.2 Long-run trends in Peruvian fiscal policy can be broadly split in two contrasting periods: before and after the nineties. The average NFPS fiscal deficit was a staggering 5.4 percent of GDP in 1970-1990, and a much more moderate 2.4 percent of GDP in the nineties. Such outcomes resulted from long periods of fiscal expansion, followed Figure 2.1 NFPS Fiscal and Primary Balances by increasingly shorter periods of fiscal 8% (% of GDP) adjustment, particularly before the early nineties. The main component of such deficit was the Central 0 Government (CG) imbalance, which represented about 3/4th of the average -% NFPS deficit. The first period reflected -8% high deficits resulting from persistent macroeconomic disorders, negative -12% primary surpluses with either short- 1970 1975 1980 1985 1990 1995 2000 lived or half-hearted efforts at fiscal --. Fiscal balance - Primary balance adj.ustment. The lower deficits of the second period came from continuous primary surpluses that became the norm rather than the exception (Figure 2.1). This section mainly draws from Oliva (2002). PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 2.3 The impact of a severely deteriorated economy in the period 1970-1990 left a long lasting memory among Peruvians about the paramount need to preserve stabilization. As GDP growth deteriorated after 1975, the tax ratio deeply varied inside the 8-16 percent of GDP range, with the most striking fall taking place throughout Figure 2.2 Tax Revenue Composition, 1970-2001 the late 1980s, reflecting the 1%Percent of GOP hyperinflation crisis (Figure 2.2). As a 4to 6.0 percent of GDP and stayed high 18% 12% until the end of the decade. Such high 0% deficits were financed by resorting either % to domestic financing, leading to 6% ~ hyperinflation levels in the late eighties, o 60or to huge increases in external 2% indebtedness, with the ratio of public 0% . external debt to GDP increasing from 1970 1975 1980 1985 1990 1995 2000 about 14 percent in the mid- seventies to 13 Income El Imports 3ISC D IGV U Othher 56 percent in the late eighties (Table 2.1). Governments' heavy reliance on debt led to an explosive rising share of interest payments of total central government expenditure from 5 percent in 1970 to 40 percent in 1990. Only after 1990, the share of interest payments was dramatically reduced, following the restoration of normal relations between Peru and the international financial community, prompt payment of arrears to multilateral institutions, and debt restructuring agreements with Paris Club lenders, and other bilateral and private creditors2. By 2001, the share of interest payments had been reduced to 12 percent (Table A4). Table 2.1 Key Macroeconomic Variables 1969-1989 Period Fiscal Average inflation GDP Growth Public External Balance rate Debt/ GDP (%) 1969-1975 -3.9 10.6 5.5 14.1 1976-1979 -6.8 48.6 0.7 32.3 1980-1985 -6.0 94.2 0.5 41.5 1986-1989 -5.9 445.8 -1.5 56.2 2.4 Long-run changes in the tax revenue structure are relevant. The three most important tax categories' were the income tax, the general sales tax (IGV in its Spanish acronym), and the excise tax (or selective tax on consumption, ISC in its Spanish acronym) (Figure 2.2). Four sub-periods can be identified according to which of these three categories dominates. In the first period (1970-1975), the income tax matched the IGV; in the second period (1975-1985), it was the IGV that played the leading role; in the third period (1985-1993) the ISC prevailed; and in the fourth period (1994-2000), the IGV recovered its major significance. The declining trend 2 Relations between Peru and the international financial community deteriorated throughout the eighties. According to Larrain and Sachs (1991), "Peru first partially suspended its debt service in 1983, but the magnitude of the suspension was small. In 1984, renegotiations failed and the country started to accumulate more significant arrears. These increased dramatically after 1985, when president Alan Garcia announced the government's unilateral decision to limit the service of foreign debt to 10 percent of export revenues, a restriction that would apply in principle to all foreign creditors." Taxes on foreign trade transactions were excluded from the comparison. 18 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES of the income tax-which fell from a dominant share Table 2.2 Elasticity to GDP, of 35 percent in 1970 to 23 percent in 2000-and the 19702000 exceptional major role that the excise tax had between abetiet 1984-92, were the outcome of the excessive number of Total tax revenues 0.998 income tax and IGV deductions and exemptions, and Income tax 0.972 the serious deficiencies of tax administration. Import tax (tariffs) 0.972 Total IGV 1.007 2.5 Long-run trends in public expenditure are Internal IGV 0.999 determined by two structural factors, namely, the Imports IGV 1.024 role of the public sector in the economy, and output Total ISC 1.010 performance. Three sub-periods can be distinguished. Gasoline ISC 1.032 In the first one (1970-1978), total expenditure Other ISC 0.992 exhibited an increasing trend, rising from 15 percent to Total non financial expend. 0.962 about 20 percent of GDP -s explained by the steady Current 0.965 expansion of the role of the state in the economy Wage and salaries 0.926 during Velasco's Government and a strong economic Goods and services 0.995 4 expansion. The average ratio of public investment to Transfers 1.001 total investriient stood at a high 27.0 percente (Paredes Capital 0.944 and Sachs, 1991). The second sub-period (1979-1991) Investment 0.971 featured the opposite trend-i.e., central government expenditure declined as a share of GDP, which reflected the severe deterioration of tax collection and the economy. The third sub-period (1991- 2001) reverted to an accelerated expenditure expansion: from the deep trough of 14.5 percent of GDP in 1991, central government expenditure increased to an average of 17.6 percent of GDP in 1999-2001, close to the high levels reached during the era of strong state intervention, pushed by current spending rather than by capital investment (Table A5). 2.6 Fiscal variables were highly pro-cyclical before and during the nineties. Regression analysis over the entire period show that all tax and expenditure categories exhibit an almost unitary elasticity with respect to GDP (Table 2.2), i.e. were highly pro-cyclical.6 Most fiscal variables also have significant correlation coefficients (above 0.5) with the growth rate (Arias et al., 2000), and the cyclical components of the primary balance and GDP growth are almost perfectly correlated (Figure 2.3). 2.7 Fiscal variables were extremely volatile in both periods. The volatility of fiscal variables is associated to Peru's growth rate volatility (measured by the standard deviation). In the nineties, it was 33 percent higher than its average for Latin America (LAC), itself twice as high as in industrial economies (Perry et al. 2000). Unstable international export prices and terms of trade, financial shocks and limited openness in the financial system explain GDP growth volatility. The volatility of fiscal variables, however, appears much higher than the one for GDP growth in the 1970-1996 period: almost six times in the case of tax revenues, five times in the case of public salaries and transfers, more than seven times in the case of capital investments and eleven times in the case of fiscal deficits (Figure 2.4). In the nineties, tax revenue volatility is 4 Peru's economy grew an average annual rate of 5.5 percent between 1969 and 1975 (Table 2.1). 5 Compare with the value of the same indicator in a recent period (1999-2001) which was just 18 percent (Table A2). The difference was made, of course, by the privatization in the nineties of about 80 percent of the large number of state enterprises that the Velasco government bequeathed to its successors. 6 Regressions are estimated with a simple OLS model: nominal revenue and expenditure from each category was the dependent variable and nominal GDP was the explanatory variable. Variables for 1970-2000 were in logarithms and adjusted by a first-order autocorrelation coefficient to take into account serial autocorrelation. R-square exceeded 0.99 and the t-statistics indicated that the coefficients were different from zero at a significance level of 99 percent. 19 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION explained by continuous changes to tax policy and administration, the external environment,' the legacy of hyperinflation of the late eighties and elections in the late nineties. For its part, public expenditure volatility is explained by erratic policies affecting capital investment, debt payments, the public payroll and transfers: It reflects the continuous use of capital outlays as the typical pro- cyclical adjustment variable (elasticity of almost 2); the high variability of debt interest flows (elasticity of 7.1), and the swinging patterns of wage policy (it had an elasticity of -1.7) (Table 2.3). Figure 2.3 Pro-cyclical Fiscal Stance Figur 2.4 Volatility in Fiscal Variables (billions of 1994 soles) 13 NFPS Deficit 8, CG Defii 06 3~O VAT K4 INKT -2- IN W* # R 02 I 3 5 7 9 11 13 15 _12_ TR tax revenue VAT - value added Relative volatilty* 1970 1975 1980 1985 1990 1995 2000 YTincome tax W -wages expenditure ydiwel comoponent of prinmry balance MT- import tax IN - interest expenditure KX - capital expenditure cydical compont Of otput *Relative is ratio of standard deviations wrt GDP Table 2.3 Elasticities and Growth Volatilities for Peru (1990-2000)1 Absolute Relative to % of Total Elasticity t-ratio Volatility Growth Revenue / Volatilityb Expenditure GDP 5.0 1.0 100.0 Central Government Revenue 8.3 1.7 100.0 1.6 15.3 Expenditure 15.7 3.2 100.0 1.0 5.5 Current Revenue 7.9 1.6 98.8 1.5 14.5 Tax Revenue 8.0 1.6 87.8 1.4 11.6 Non-financial Expenditure 16.8 3.4 80.6 1.7 14.5 Current Expenditure 14.8 3.0 61.9 1.7 13.0 Capital Expenditure 32.0 6.4 18.7 1.9 3.6 Interest Expenditure 27.2 5.5 19.4 7.1 3.9 Wages 16.4 3.3 17.0 -1.7 -5.3 Subsidies and Current Transfers 10.2 2.1 31.8 1.2 11.7 Fiscal Balance 13.6 2.7 Primary Balance 20.3 4.1 a. Volatilities are computed as standard deviations over a 10-year period in annual percentage rates of growth. b. Volatility in the rate of growth of the fiscal variable divided by GDP growth volatility. 7 Tax revenue volatility originates in the external sector in a two-fold sense: while in the 1970-1990 period, the share of export taxes on total revenue was significant7, which made tax collection vulnerable to Peruvian commodity-export prices in the world markets; in the 1990s, the Peruvian economy became sensitive to capital inflows surging from the opening of the capital account. 20 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES 2.8 The biggest policy mistake (and lesson) of the early nineties is not to have taken the opportunity to save the transitory additional revenue generated with the economic boom. Since 1993, strong capital-inflows dominated Peru's economic performance and brought about a strong economic boom. Between 1993 and 1997, GDP grew at an average annual rate of 7.0 percent. Instead of developing a counter-cyclical fiscal policy, the expansive fiscal stance adopted magnified the boom and increased the risks associated with it. Hence, the overall fiscal deficit increased to 3.0 percent of GDP in 1995, prompting the current account deficit in the balance of payments to widen from 5.7 percent in 1994 to 7.7 percent of GDP in 1995, which made the country's external position more vulnerable. In 1996, the government had no choice, but to undertake a severe fiscal adjustment: the fiscal deficit was reduced to 1.1 percent of GDP in 1996 and to an average of 0.3 percent of GDP in 1997-98, a situation of virtual balance. B. THE CURRENT FISCAL D ISEQUILI1BRIA 1999-2001 2.9 In the late nineties, fiscal difficulties originated from adverse external developments, but were later accompanied by expansionary domestic policies. Financial contagion hit the Peruvian economy in mid-1998 as foreign banks drastically curtailed the availability of credit lines to the Peruvian banks in the wake of the Russian crisis of August 1998. The turmoil also affected the stock market resulting in portfolio outflows and a sharp drop in stock prices. The reduction of foreign credit lines to the Peruvian banks was especially deleterious, as the latter had relied heavily on them to finance the 1993-1997 credit boom. The ensuing credit crunch was a major blow for a real sector that was barely recovering from the effects of another shock: the El Nihfo phenomenon, which struck in the first half of the year. The capital outflows episode thus came to a close and the well-known vicious circle of credit reduction, fall in output and demand, lower sales, rising difficulties for the firms to serve their debts, deterioration of banks' loan portfolio, and further credit reduction set in. In 1998, the economy entered into recession with private investment significantly falling, both in real terms and as a percentage of GDP. Declining economic activity led to lower tax revenue and public expenditure, which also deepened the economic contraction. As a result, the overall fiscal deficit increased to an average of 2.9 percent of GDP in the 1999-2001 period. Three domestic reasons also contributed to this outcome. On the expenditure front, the political business cycle deteriorated the fiscal accounts, as in the run-up to the 2000 presidential elections, the central government expanded current non-interest expenditures from 11.8 percent of GDP in 1998 to 12.7 percent of GDP in 2000. On the revenue side, the strong decline in tax revenue from an average 13.8 percent of GDP in the 1995-1997 period to 12.3 percent in the period 1999-2001 was mainly associated not with the cycle, as the deviation of the cyclically-adjusted balance with respect to the observed primary balance was small, which would suggest that the effect of cyclical factors was small, but rather with a structural shift in tax revenue behavior caused by frequent changes to the tax regime. 2.10 Real revenue dectreased and real expenditure increased Fiscal vulnerability becomes more apparent when tax revenue and CG expenditure are considered in real terms. In 1999, tax revenue fell to pre-1995 levels (below 12.5 percent), while expenditure rose considerably (about 19 percent of GDP). As a result, the CG (and combined general government) primary balance has become negative since 1999, and fiscal deficits have surged (Figure 2.5). An increased- mainly short and domestic-public debt (as a percentage of GDP from 42.7 percent in 1998 to 8 In a simple framework where tax revenues are a function of income, the cyclical effect can be seen as a movement along the revenue curve in response to a lower level of income. The deterioration of the tax system, by contrast, should be seen as a downward shift of the entire curve, entailing lower revenues at any level of income than in the situation prior to the deterioration. Our quantitative estimates indicate that the effect of the business cycle on the primary deficit in the period 1999-2001 was just worsening of 0.1 to 0.2 percent of GDP. 21 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION about 46.5 percent in 2001) and privatization receipts financed the deficit. Barely 8 percent of public Figure 2.5 Fiscal Trend Over Time debt is short-term and prospects for 20 improving public debt ratios in the -3 medium-term are favorable. 3 However, public debt is exposed to2.5 1 currency risk. Debt service, which 2.0 already increased from 3 percent of 1.5 GDP in 1998 to a projected 3.8 1.0 percent of GDP in 2002, will 0.5 8 contribute to tighten scarce resources 1991 1992 1993 1994 1995 199 1997 1998 1999 2900 2001 in a shrinking budget. Had Peru kept CG Deficit, LHS GG Deficit, UHS Government tax and expenditure - Real CG revenue Real CG expenditure constant in real terms during the 1998-2001 period, it would have preserved a fiscal deficit below Box 2.1 The Fiscal Prudence and Transparency Law 1 percent of GDP. The Peruvian Law was approved on December 1999 (Law No.27245) and its norms issued on April 2000 (Supreme Decree No. 039-2000- did nt fuly cmply EF). It was suspended in December 2001 (Law No. 27577). The 2.11 Peru di o ul opy Fiscal Stabilization Fund was slightly amended through the 2002 with its Fiscal Prudence and Budget Law (Law No.27573). The Law has 3 components:. Transparency Law of 1999. The Law. Numerical rules include (i) a zero fiscal deficit in general (over the established a set of procedural and cycle), but below 2 percent of GDP in 2000, 1.5 percent in 2001 and numerical fiscal rules (Box 2. 1). The I percent in 2002; (ii) an increase of the non -financial general government spending below 2 percent in real terms; (iii) electoral Law set five numerical targets for: (i) expenditure ceilings on the non-financial expenditure below 60 the overall deficit; (ii) the difference percent in the first 7 months and a public sector deficit below 50 between the inflation rate and the percent in the first semester. These rules have 2 exceptions allowing annual growth rate of consolidated for a fiscal deficit below 2 percent of GDP: one for emergency situations and another for recessions. general government's -non-financial Procedural rules require the introduction of transparency criteria expenditures; (iii) the total debt increase established in IMF Code of Fiscal transparency from 2002, including of the non-financial public sector; (iv) (i) a Transparency Decree of Acceso Ciudadano, detailing periodic the general government's non-financial information to be published in the Portal de Transparencia Econ6mica; (ii) a multi-annual budgeting framework; (ii) a mandated expenditures during the first seven pre-election fiscal report; (iii) an annual review of contingent fiscal months of an election year? as a risks and tax expenditure to be published in the annual budget; and percentage of budgeted non-financial (iv) a quarterly report on budget expansions and their implication expenditures of the general government over the macroeconomic framework. for the entire year; and (v) the overall The Fiscal Stabilization Fund was created by 75 percent of privatization proceeds, 50 percent of concessions, and excess,revenue deficit in the first half of each election (above 0.3 percent its average of past 3 years). Its resources should year as a percentage of the budgeted be mainly dedicated to poverty alleviation programs, and activated deficit for the year. Government's when current revenue is at least 0.3 percent of GDP below its average compliance with numerical targets was i h at3yas mixed, but the main target referring to the overall deficit was not met and the Law was suspended in December 2001. Although a case could be made that this was due to revenue-growth slowdown-related factors beyond the authorities' control, the fact is that Fujimori's late administration weakened the fiscal balances SDue to well-known political events in Peru, both 2000 and 2001 turned out to be Election years. 22 Chapter H. PERU'S FISCAL CHALLENGES AND VULNERABILITIES with new tax exemptions and electoral spending. Besides, the increase of non-financial expenditures in 2000 and 2001 built up upon their already high level reached in 199910 2.12 The present level of public debt needs to be reduced to more sustainable levels. Three years in a row of fiscal disequilibria have led to an increase in public indebtedness and a reduction of long-term public savings. The accumulated deficit for the period 1999-2001 amounted to 8.8 percentage points of GDP. Its financing came from domestic credit (4.6 percentage points), external credit (2.0 percentage points) and depletion of privatization proceeds (2.2 percentage points). The stock of total public debt increased from an already high 42.7 percent of GDP in 1998 to 45.8 percent of GDP at the end of 2001. While the net public external debt ratio increased only slightly1", the domestic public debt ratio rose from 6.3 percent of GDP to 9.8 percent of GDP. All domestic debt is in U.S. dollars. External debt represents around 70 percent of total public debt and has moderate risk, as virtually all of it is long-term. However, its service has rapidly increased from 3 to 4 percent of GDP between 1998 and 2002, amounting to 23 percent of total expenditure in 2002, which is a considerable burden on tight fiscal balances. Two additional vulnerabilities are the foreign-currency risk exposure of the domestic debt and the rollover-risk to changes in foreign investor confidence. Peru reopened its access to international markets with a US$500 million ten-year bond placement combined with a retirement of Brady Bonds amounting to US$930 million in 2002, which lowered its debt servicing costs, although at the expense of a shorter maturity. Given the low tax revenue situation and expenditure pressures, Peru requires not only fiscal adjustment, but also prudent debt management strategy. C. THE MACROECONOMIC CONSISTENT EXPENDITURE ENVELOPE 2002 -03 2.13 The medium-term outlook for the economy is favorable. As the focus of this chapter is on analyzing expenditure trends, the starting point for such analysis is the determination of an aggregate level of spending that is consistent with medium-term macroeconomic objectives (i.e. a sustainable fiscal deficit and public debt ratio) (Table 2.4). A base-case scenario assumes that real growth of about 4 percent a year and low inflation inside the 2-3 percent range a year would be achieved on the basis of the recovery of the mining sector, increased productivity, and increased private sector investment and capital inflows attracted by sound macroeconomic policies, political stability and the privatization program. It also assumes the combination of an increased export capacity, continuous implementation of structural reforms, a managed floating exchange rate regime and a successful fiscal adjustment, supported by the IMF program. Mining production should benefit from the first full-year operation of the Antamina project. Structural change should lead to complete the privatization program and improve regulations of private sector activities, develop public sector modernization and decentralization, strengthen financial supervision, and develop pension and social sector reform policies. In sum, under a low inflation and moderate growth scenario, present fiscal imbalances are expected to be gradually corrected provided the Government adopts an urgent tax reform, cuts unnecessary outlays, protects expenditure shifted toward social priority needs, minimizes the cost of potentially significant contingent liabilities--like those from pensions--and maintains a prudent debt management strategy. 2.14 This macroeconomic scenario is entirely consistent in the short term with the Stand- By arrangement agreed with the IMF. However, significant constraints remain, including limited policy credibility following a long period of political instability and a poor GOP track 10 In terms of GDP, combined general government non-financial expenditures in 1999 were the highest for the entire decade of the nineties (18.9 percent of GDP). ' From 34.7 percent of GDP in March of 1999 to 35.1 percent in December 2001. 23 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION record in complying with fiscal rules; policy uncertainty due to continuous changes in the tax regime and regulations that affect prospects for attracting foreign investment; fiscal adjustment delays due to Congress inherited populist-oriented and spending-prone fiscal decentralization measures; weak domestic consensus about further pension reform and the GOP's privatization plans; and a poor rule of law due to an inefficient judicial system and poorly managed local government institutions. Table 2.4 Key Economic Indicators 1996 1997 1998 1999 2000 2001p 2002e 2003e Real GDP Growth 2.5 6.7 -0.5 0.9 3.1 0.2 3.7 3.5-4.0 Inflation (% CPI average) 11.5 8.5 7.3 3.5 3.8 2.0 2.5 2-3 Key Macro Indicators (% of GDP) Gross Domestic Fixed Investment 22.5 23.8 23.5 21.7 20.1 18.3 18.8 19.2 Government investment 4.3 4.4 4.5 4.8 4.0 3.2 3.1 3.0 Private investment 18.2 19.4 19.0 16.8 16.2 15.1 15.7 16.4 Gross national savings 16.4 18.2 17.3 18.0 17.3 16.5 16.5 16.7 Current Account Balance -6.1 -5.2 -6.4 -3.7 -3.0 -2.1 -2.3 -2.5 Central Government (% of GDP) Current Revenue 15.7 15.6 15.7 14.5 14.7 14.1 13.9 14.1 Tax Revenue 13.9 14.1 13.8 12.5 12.0 11.9 11.9 12.4 Non-Tax Revenue 1.8 1.5 1.9 2.0 2.7 2.2 2.0 1.7 Current Expenditure 13.9 13.1 13.7 14.6 14.9 14.7 14.4 14.1 Interest 2.4 1.8 1.9 2.1 2.2 2.1 2.0 2.2 Interest on External Debt 2.3 1.7 1.8 2.0 1.9 1.9 1.7 1.9 Interest on Domestic Debt 0.1 0.1 0.1 0.1 0.3 0.2 0.3 0.3 Current Transfers 2.9 3.1 2.0 2.8 2.3 2.3 2.4 2.4 Consumption 8.6 8.3 9.8 9.7 10.4 10.3 10.0 9.5 Primary Balance 1.0 0.9 0.8 -1.0 -0.5 -0.6 -0.4 0.3 Overall Balance -1.4 -0.8 -1.1 -3.1 -2.7 -2.8 -2.3 -1.9 NFPS Balance (% of GDP) -1.0 -0.1 -0.8 -3.1 -3.2 -2.5 -2.3 -1.9 Privatization 3.9 0.9 0.5 0.8 0.8 0.6 0.6 0.5 NFPS Debt Public Sector Debt/ GDP 45.2 31.9 42.7 48.0 45.9 46.5 44.9 40.9 Public Sect. Debt Service/GDP 4.1 3.3 3.4 3.9 3.8 3.9 3.8 4.5 Memo: GDP Nominal (millions NS) 136,929 157,274 166,514 174,719 186,756 189,532 200,083 213,205 Source: WB staff estimates. Note: This Table has minor differences with Table A5, based on Central Bank data, and Tables 1-6 on Annex A, based on preliminary data. Tax revenue is net of tax on assets of public enterprises and of CG-IES payments. Recommendations. Authorities should > redouble efforts to build credibility before the Law is revamped; > offset several measures introduced by the transition government that had negative implications for the fiscal balances with revenue resulting from tax reform will; > set rules to prevent decentralization from increasing transfers to regional and local governments that are not offset by similar improvements in revenue collection; > stabilize the regulatory framework for private sector activities and strengthen the rule of law to diminish uncertainty that erodes the authority of rules and institutions that are essential to promote domestic and attract foreign investment. 24 Chapter H. PERU'S FISCAL CHALLENGES AND VULNERABILITIES Table 2.5 Recommendations for Tax Reform Revenue Recommended Measure impact (%of GDP) Income tax Elimination of exemptions on interest income and capital gains 0.27 Introduction of a three-level structure of rates for individuals 0.10 Elimination of sector- and region-specific exemptions 0.60 Value-Added Tax Elimination of IGV exemptions for Amazon region 0.25 Elimination of exemptions granted by Hydrocarbons Act 0.10 Elimination of the 50-UIT threshold for agriculture 0.30 Selective Tax on Consumption Elimination of ISC exemptions for the Amazon region' Elimination of exemption on fuels used by power industry2 Unification of ISC for vehicles at 25-percent rate 0.10 Import Tariffs Elimination of the CETICOS regime 0.13 Special Regime for Mining Introduce a system of mining royalties 0.10 Extraordinary Solidarity Tax (EST) Elimination of the EST -0.20 TOTAL IMPACT 1.65 The elimination of both the IGV and ISC exemptions for the Amazon region is estimated to have a total revenue impact of 0.10 percent of GDP. This is the reason why no revenue-impact figure is entered in this row of the table. 2The revenue impact of this measure as a percentage of GDP is a very small number. This is the reason no figure is entered in t his row. The revenue impact would depend on the characteristics of the royalties regime (see Chapter VIII). 2.15 Peru's most pressing fiscal challenge is to obtain primary surpluses, as negative primary fiscal balances are unsustainable. Recommendations: > In the short-term, under the base-case, the mid-term revised adjustment program aims a gradual reduction in the combined fiscal deficit to 2.3 percent in 2002 and 1.9 percent of GDP in 2003. Most of the adjustment would be achieved through cuts in public spending, while reorienting and protecting priority social programs, and tax reform, combining tax policy measures, changes to the exemptions regime and an improved tax administration (Box 2.2, Table 2.5). This scenario also assumes that the privatization of energy enterprises and the granting of a few concessions will materialize in 2002-03, averaging a minimum of 0.6 percent of GDP a year. Under these assumptions, a continuous implementation of a managed floating exchange rate and an open trade regime, the external current account deficit is projected to be about 2.5 in 2003 and the international reserve position to gradually decline to comfortable levels (7.5 months of imports by 2005). > In the medium-term, our estimates of the potential output and structural fiscal balance under the base-case scenario indicate that, in order to stabilize debt indicators, the present deficit in the non-financial public sector primary balance should rapidly shift to a surplus of 0.9 percent of GDP by 2004. The consolidation of this effort would require NFPS primary surpluses reaching levels between 1.0 and 1.3 percent of GDP during the second-half of the decade (Annex B). 2.16 Public dbt ratios should decrease to more sustainable levels in the medium term and debt currency risk should be reduced A public debt ratio of 46.5 percent of GDP is too high, especially for a country whose balance sheet contains a significant currency risk, which 25 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION makes it exposed to external shocks. Such Box 2.2 Tax Regime and Proposed Changes risk originates from the fact that most of Income tax on individuals. This tax is levied on four classes of income, namely: real estate, other capital assets; selfemployed labor; and wage labor. The Peru's public debt (external and domestic) is maximum marginal tax rate for net income greater than 54 Unidades de Ingreso in dollars, while the Government's earningsisbt 8)i 30 percentTe marginal rate for net income are mostly in soles (trade taxes are the only exempted. Dividends are not taxed. The interest income from savings with the tha i g national ftniancial system and capital gains from the sale of securtties and real estate revenue that is dollar-linked). Existing ae exempted. For the period 1997-2000, revenue from the income tax on dollar-denominatedS individuals was, on average, 1.0 percent of GDP(Kopits et al, 2000; and Schenone, dollr-de omin ted ebt lso lace 2001). Over 75percent of this revenue comes from wage-eamners' income and only authorities virtually against their own about 10 percent from the self-employed. This is tantamount to the existence of a high degree of horizontal inequity. The situation is worsened by underreporting of commitment to let freely float their own income from independent professionals' services and rents from real estate. The thuspresing them forcontnue considerable numberof exemptions also blunts the intended progressivity of this tax currency, category, which detracts of the system's claim to vertical equity. The exemption on intervening to avoid its depreciation. interest income from bank deposits is 0.1 percent of GDP (Kopits et al, 2000 and Schenone, 2001I). Recommendations: (i) Elimination of exemptions on all forms of capitol gains Recomme dationsincluding the sole of securities and real estate, on interest income andolther income Recom endatons:from financial assets; (ii) introduction ofo more gradual structure of marginalta > Under a baseline scenario, a debt r (e.g. wth three rates, 10, 20 and 30 percent) instead of the current two -rate structure; and (iii) introduction of presumptive-income methods for self-emploed sustainability analysis estimates the taxpayers. target public debt-to-GDP ratio to betax on corporate income. The taxable base for the corporate income tax is targ t p bli det-toGDP rato t be the worldwide income of enterprises legally regarded as domiciled in Peru. The reached in 2010 is 36.2 (Annex A, rate is 20 percent. Exemptions are numerous and incle: interest income from svnswith the national financial' system; interest income from registered Table 2). Such ratio would correspond instrments issued by corporations; capital gains from the sale of securities; to a fiscal effort of 1.6 percent of GDP" agriculture and/or transformation or processing of products described as to afiscl efortof .6 p rce tofGDP " native", and/or alternative crops in the Amazon region; enterprises operating in the in the structural primary surplus (1.9 Centers for Export, Transformation, Industty, Commercialization, and Services of G P i cycicaly ajused) (CETICOS in the Spanish acronym); and income from agricultural enterprises with percent sales under 50 UIT. Rate reductions and other tax benefits ae: Agriculture (a rate betw en 2 0 110. uch sti ate ake of 15 percent); income of enterprises engaged in agriculture, aquaculture, fisheries, betweentourism, and manufacturing activities associated with the processing, into account the debt-exchange of the transformation, and narketing of commodities from the aforementioned activities, the processing of forestry products and forest extraction (a rate of 5 pereent if the Peruvian Brady bonds with private business is located in the departments of Loreto and Madre de Dios, and in some districts of the Ucayali department, or 10 percent i f itsis located anywhere else in the creditors in 2002 and its slightly credtor in 002 and its lig tly Amazon region); accelerated depreciation is allowed in mining, hydrocarbons, shotend m turtie. I sutaied iscl tUrism, agriculture and agri-business; and tax-discount for re-invesunent in shortened maturities. If sustained fiscal t..s mining, education and manufacturing industry as well as for enterprises located in efforts are promptly developed, the the Amazon region. The effects of this type of tax benefits are well known: the neutrality of the tax system is impaired and the revenue-yielding capacity is public debt ratio to GDP is projected t pubic ebtrato o G P i prjecedto damaged. collection from the tax on corporate income fell firm 2.2 pereent of improveGDP in the first quarter of 1999 to 1.4 perent in the fourth quarter of 2001. The imprve ubsantally fro ab ut 6.5 fiscal cost of exemptions on deposits of corporations in the banking system is 0.2 percent in 2001 to 39.3 percent in 2005, percent of GDP. Moreover, in the case of foreign firms, a part of the benefits granted ends up as a transfer so the treasuries of foreign governments where the and then decrease until its benchmark firms in question have their headquartera. In many cases, this type of tax incentives ratioare no instruental in attaining the desired investment objectives and lack rati in 0 1. Th resltig pah to ard transparency as is is very difficult to quantify the benefits given so a particular sustainability in the public external debt sector or region of a country. Recommendations: (i) An across-the board elimination of sector- and region- service, however, is a longer one. As a specific exemptions is desirable; (ii) alternatively, if the authorities want to ta t GD det sevic sh uld subsidize spec ific productive sectors and/or regions of the country, explicit percentage totransfers should be included in the national budget; and () exemptions on all increase from about 3.8 to 6.0 percent forms ofcapitolgainsfromporolio investments must be eliminated. between 2001-05, and then decrease to I or t . Tariffs are levied on imports of goods on the basis of their CIF value. There are two basic tariff levels: a 12 percent general tariff (applicable to 4.6 percent of GDP in 2010. 5,780 customs items); and a special tariff of 20 percent (applicable so 752 items). In addition, there are two types of surcharges: one on agricultural products (surcharge rate of 5 percent) and one on meat products (surcharge rate of 10 ~' nde th exrem cae o a odiied percent). Only one exemption raises serious concern: the exemption from tariffs for >foreign goods introduced into the CETICOS that do not pay import duties and can baseline scenario with shocks derived be re-exported. This exemption impairs the horizontal equity of the tax system in that it provide for unequal treatment of enterprises that are otherwise equal Just from the impact of a sudden financing because they are located in diferent places. stop~~ ~~~~~~~~ (aranxhn ertedpeitoieomendation. The CETICOS regime must be abolished stop (a real exchange rate depreciationpage and the materialization of contingent liabilities contained in short-term debt), the target public debt-to- GDP ratio to be reached in 2010 would increase to 40.8 percent. Such ratio would require a fiscal correction in the structural primary surplus of 2.4 percent of GDP (cyclically adjusted) between 2001-10 (Annex A, Table 4). 2 On public domestic debt. Developing a market for domestic public debt issued in soles would contribute to reduce its currency risk and bring more credibility to the inflation 26 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES targeting policy. Despite being small,12 Box 2.2 Tax Regime and Proposed Changes (continuation) an inflation-indexed public debt in local The Extraordinary Solidarity Tax This is a tax on finns' payroll. The taxable currency, hopefully placed beyond base is the wage compensation to workers classified in tax-code categories 4 and 5, to excluding national-holidays and Christmas bonuses. It has long been recognized as short-term, would also contribute to aburden on enterprises' competitiveness and is being phased out. The current government reduced its rate from five percent to two percent. The government is reduce inflationary surprises. redu e iflatonay supries.seeking congressional approval for ita complete elimination. Recommendation. The government's decision to eliminate the Extraordinary > On public external debt. The initial Solidarity Tax is adequate. The General Sales Tax The General Sales Tax (IGV) has a aceof 18 percent and focus of the new Authorities has been is levied on the domestic sales ofmovable goods, the provision or use ofservices in on reducing rollover costs, widening Peru, construction contracts, the first sale of structures by their builders, and iprts of goods. The following goods are exempted: fish and senfood, cochinilla, maturities and refinancing debt with bullsemen, whole raw milk, potatoes and other root vegetables, dried vegetables, ot vegetables and fruits, cereal seeds, rice, raw coffee, tea, cocoa, unprocessed Paris Club creditors. Increased tobacco, wools and pelts, raw cotton, non-monetary gold, vehicles for diplomatic coordination between MEF and the use, books for educational institutions and the first sale of structures valued at less than 35 UITs. The following services are exempted: financial services, public BCRP, and strengthened management passenger transportation (except for sir transport), international cargo transport, live shows declared of cultural value by the National Cultural Institute, sale of beverage of the maturity and currency and food at snack bars and universities, construction and repair of foreign ships, composition of external debt would interes on transfersble securities issued by enterprises pursuant to the Securities Market Law, life insurance polices, postal services intended to complete such facilitate debt servicing. Approval of a services originating abroad, and exports of gocds and services. Apart from the aforementioned list of exemptions, there are a few special regimes for the IGV. General Debt Law in 2002 is expected First, the Hydrocarbons Act exempts the imports ofgoods and inputs required in the toexploation phase of oil contracts, and the sale of petrol, natural gas, and to dal wth ebt anag men isses. derivatives, for internal consumption, in the departments of Loreto, Ucayali, and Madre de Dios. The Amazon Investment Promotion Act of 1999 exempts from the IGV the sale of all goods and services for internal consumption as well as 2.17 Peru should start making construction contracts in the Amazon region. Furthermore, it allows a fiseal credit of 25 percent or 50 percent on the IGV for sales to the rest of the country. The provisions for contingent liabilities. Agriculture Act exempts from the IGV agricultural producem whose annual sales These could materialize in the near future, are below 50 tIT. The first sale of homes below US$35,000 is also exempted There is, finally, a special rate for rice (5 percent). In analyzing the fiscal impact of thus making more difficult projected the IGV exemptions we must distinguish between the general exemptions and the reco eryof udgt dscilin at special regimes. Defenders of the general exemptions contend that, by selecting recovery of budget discipline. There are a items ofpopular consumption, they introduces desirble element ofprogressivity in least three main sources of contingent risk this tax and a hence a positive re-distributive effect. It has been demonstrated, ,t: however, that, to the extent that they include items that are also consumed by the that could become a government liability: rich, exemptions to the value-dded tax are, at best powerless to generate positive non-performingre-distributive effects and, at worst, deepen the negative re-distributive effects that non-erfrmin bak lans, penion are an inherent feature of this tax. For their part, the special regimes damage the payments and municipal debt (not counting VAT's revenue-yielding capacity. In the case of the Amazon-region exemptions they not only deprive the national treasury from revenue in the Amazon region those that could originate from Banco itself, but open the door for VAT evasion.on the par of tax payer fr regions lon other than the Amazon region through the simple expedient of doing their shopping Agrario or public enterprises loa in the latter region. Finally, the exemption for agricltural produce whose annual guaranees).sales are below 50 UIT breaks its universality andneutrality. guarantees). Recommendations: (i) IG Vexemptionsfo r the Amazon region must be eliminated. (ii) IG aexemptions granted by the Hydrocarbons Law must be eliminated. and (iii) " The World Bank's broad estimates I tsei o-UlTthresoaldfr agricultural producers must be eliminated. of the potential fiscal exposure of a banking crisis is 4 percent of GDP, which is considered in the lower bound of comparabled-mostly double -digit-world standards. Moreover, official estimates put non-performing loans at about 10 percent of the Bank's portfolio in 2001, but such a ratio appears stabilizing in 2002. i The pension system is a potentially significant fiscal liability. Pension reform has generated two non-banking debts with the private sector which are actually paid through annual transfers: one is to cover the pension of public employees retiring under the public pension system and the other are bonos de reconocimiento of those who already migrated to the private pension system. Official transfers to the social security represent 0.6 percent of GDP in 2002, and conservative projections indicate they might increase to about 0.6-1.0 percent of GDP per year in 2005-07 (IN data and Arias et al. 2000). 12 Internal debt was very small during most of the nineties, as the GOP associated it to the decades of large deficits and inflation. Bonds were re-issued in 1999 to Support a bank merger, the purchase of the bad asset portfolio of a bank taken over by the government, and two bond-for-loan swap programs. In 2000 and 2001, new bond issues were made under the Financial System Consolidation Program. The BCRP has strict limits to purchase government debt. 27 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION u Registered municipal debt is not a threat for macroeconomic stability, but this is not an indication that the problem is not very serious at the local level. Indeed, it is significant and explosive for some localgovernments becoming non-solvent and illiquid like those from Arequipa and La Victoria (Fernandez 2002). These two local governments will have severe difficulties to assume additional responsibilities with decentralization due to their financial constraints. Nationwide total registered debt represents a third of the corresponding national debt (current national liabilities considered as short-term debt). Non-registered municipal debt adds up to such burden. Recommendations: > Peru should start making budgetary provisions for contingent liabilities, so a first step would be to make a comprehensive assessment of all potential sources and quantify them. > Financial contingent liabilities should be monitored through an adequate prudential supervision. Those eventually arising from a sudden stop of external financing due to contagion, can only be avoided with sound macroeconomic management (Annex A). > Reducing contingent liabilities in pensions will likely require more structural solutions. > Identifying non-solvent municipalities and their eventually non-registered short-term debt, as well as developing a plan for the gradual repayment of their registered debt are essential steps before districts assume further responsibilities under the forthcoming decentralization process. > The revamped Fiscal Prudence and Transparency Law should continue preventing Authorities to resort to the Central Bank for financing of its own expenditure, and add specific prohibition to incoming sub-national governments (or ceilings) of acquiring future debts (Chapter V). 2.18 There are also additional external and domestic risks to this scenario which would make Peru's growth potential uncertain, increase credit risk, reduce financing needed for a sustained social policy, or weaken the institutional le adership on social reform and fiscal decentralization. On the external front, a prolonged slowdown of the global economy, adverse regional developments or unexpected shocks in the international capital markets could lead to a significant fall in external demand for Peruvian exports, or in mineral prices and/or to private capital outflows. This could reduce exports dynamism and growth, destabilize the domestic currency, put strong pressure on both the public and the corporate sector which are heavily indebted in foreign currency, and impose additional stress on the banking sector. On the domestic front, the new Government and Congress could fail to agree in the implementation of the necessary policy measures needed for an effective fiscal adjustment and a financially sound state reform and decentralization process. Lack of consensus could also delay/revert the comprehensive legislative reform of the tax system to be introduced in 2002. Authorities should be supported in their efforts to: > build a contingent financing plan based on (i) the prompt materialization of external donor financing to the incoming Poverty Reduction Strategy; (ii) increased and enhanced access to international capital markets by further placement of sovereign bonds, and (iii) the completion of the privatization process. 28 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES > develop consensus-building mechanisms on fiscal issues, aimed at reaching a common-shared medium-term vision with the private sector and civil society. Such consensus should be adequately reflected in the revamping of the Law. D. THE FISCAL ADJUSTMENT EFFORT 2.19 Fiscal adjustment requires a certain combination of expenditure reduction-cum- revenue increase. In 2002, Peru's required revenue increase will not likely come out from a rapid recovery of the economy, as the cyclical component of the fiscal balance is very small (Figure 2.13). Besides, economic recovery has had limited impact on tax collection due to the prolonged debate that took place in Congress on the desagio tributario (a quasi-tax amnesty) and the granting of additional income tax exemptions. This points out to the need of a combination of expenditure cuts with structural remedies on the revenue side. A long-lasting fiscal adjustment requires measures to: (i) strengthen the tax regime and administration; and (ii) revert the trend toward increased current outlays, while curbing inefficient expenditure. Strengthening the Tax Regime 2.20 Continuous changes in the tax regime are weakening the successful tax reform of the nineties. In 1993, 80 Peru carried out a well-designed reform 70 a of its tax system: the reform aimed at 60 Chile simplifying the tax system, broadening the tax base, and increasing revenue yield (IMF 2000). The number of tax categories was reduced from almost * seventy to four basic categories, namely, 20 the general sales tax (a value-added tax); 10 5 1015 20 25 the income tax on individuals and R1le corporations; the selective tax' on consumption (an excise tax); and tariffs on imported goods. The income tax brackets were reduced from 5 to 3 and business income taxed at a flat rate of 30 percent. Dividend income was excluded from the tax base and net- wealth tax was eliminated. The minimum- income tax, based on gross assets, was maintained. The base for the value-added and excise taxes was broadened and most exemptions eliminated. Later, in 1994, the base for the income and value-added taxes were broadened further." However, since 1996 a long series of exemptions and other preferential regime measures were introduced, which severely eroded tax collection capacity. In spite of having a VAT collection effectiveness ratio that is average in LAC terms-about 50 percent (Figure 2.6)-the main tax breaks represent a minimum of 2 percent of GDP of direct tax losses (Table A21). Strengthening the tax regime requires a set of specific actions to improve the equity, neutrality, simplicity and revenue-yielding capacity (Box 2.2). A specific set of ensuing tax policy recommendations could generate an estimated potential revenue generating capacity of 1.65 percent of GDP (Table 2.5). 14 ' In the institutional domain, the Superintendency for Tax Administration (SUNAT) was created with effective administrative and budgetary autonomy. Personnel policies were reformed so as to get highly qualified staff and endow them with up-to-date computational equipment. The tax code was modified to enhance SUNAT's ability to audit taxpayers, increase fines and penalties, simplify procedures, and strengthen tax courts. 14 Both proposals, the tax regime changes and the set of tax policy recommendations are based on Schenone (2002). 29 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Reverting Expenditure Figure 2.7 Central Government's Expenditure Composition Trends 2.21 Current expenditure 70% is increasing explosively. 6 /* Since the mid-nineties, Peru's current expenditure has been increasing at an accelerated pace. It rose from about 50 percent of total Central Government (CG) expenditure in 1990 to almost 73 percent in the early 2000s (85 percent for the General Government (GG) __ teq_-_CpiW_=_da,_C t in 2001). These very high levels for Peruvian standards were characteristic of strong state intervention in the seventies (Figure 2.7). One could argue that the levels of current and capital expenditure are broadly in line with LAC regional averages (Table A29); but what is worrisome are its increasing trend and present composition. In 2001, as a share of total expenditure, Peru's 39 percent wage bill was second highest in LAC, and as a share of GDP, Peru's 6.6 percent figure was higher than the 6.3 percent LAC average. These ratios suggest an increasing sense of overstaffmg and do not include additional "services" hired as temporary contracts under goods and service categories (Chapter VI). It also suggests a high level of budget rigidity (Chapter III). Such a rising trend would not have been possible unless capital expenditure and interest payments had not decreased their share below 15 percent. Contrary to the early nineties, interest payments have been below the region's average. For its part, Peru's capital expenditure (3 percent of GDP) is among the lowest in LAC (Figure 2.8) and keeps decreasing (see Table A5). Indeed the budget priority in the nineties was not to increase the capital stock. Recommendation: Figure 2.8. LAC: Capital Expenditure (1998) > The proposed comprehensive (%ofGDP) reform of the state should examine proposals to combine a Dom. Republi gradual reduction of current expenditure, especially wages and hoer salaries, while developing a more qualified and efficient labor force under a civil service reform par"M (Chapter V). laracWer Mexi4or 2.22 Public investment remains ay low and only recently has given Trn&Tob priority to the social sectors. In the Ha nineties, capital expenditure was more 1 3 6 9 12 15 18 volatile and pro-cyclical than all the rest of expenditures categories (Table 2.3). In addition to this, the average level of central government capital expenditures as percentage of GDP was below 2.8 percent in recession years and 4.0 percent in booming years. 30 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES 2.23 These outcomes reflect Figure 2.9 Structure of Central Government's Fixed Capital Formation that public investment continues 1991iM-lti to be, to a major extent, the residual item in the budget, i.e. the one that bears the brunt of iff. fiscal adjustment efforts. The composition of fixed capital formation has suffered important changes. Whereas investment in 71 economic infrastructure played a dominant, but declining share as a percentage of total capital investment, falling from 72 percent in 1990 to 54 percent in 2000; the share of investment in social infrastructure increased from 12 percent to 19 percent of total expenditure (Figure 2.9). The disaggregated analysis of the internal decomposition of investment in economic infrastructure shows that, by far, the biggest fall was in the share of public investment devoted to agriculture, which more than halved from 46 percent to 19 percent of total capital formation; whereas the energy and mining sector increased its share from almost nil in 1990 to 10Q percent in 2000. For its part, the transport and communication sector essentially maintained its share as roughly a quarter of total capital formation (Table A7b). The doubting share of investment in social infrastructure is a very positive trend inherited from the nineties. This effort is also reflected by the pattern of total social expenditure, which seemingly rose from 3.9 percent of GDP in 1990 to 6.9 percent in 2000 and remain steady until 2002 despite the economic slowdown.G5 Recommendation: > Peru has the capacity to gradually fill its financial gap in the social sectors by raising its historically under-funded education and health expenditure ratios of 2.9 and 3 percent of GDP in 2000, to at least the average LAC ratios of 4.1 and 3.2 percent of GDP in the next 5 years. 2.24 The observed increase in fixed capital formation of the "general sectors"-mainly including general public services like military expenditures-is a matter of concern. As a percentage of total capital investment, such share almost tripled from 5.6 percent to 14.8 percent, reflecting, to a considerable extent, the leading role assigned to military expenditures within total public expenditures. In a similar vein, in 1998-2000, defense and national security expenditure averaged 2.9 percent of GDP. Such a relatively high level of military expenditure does not seem to be justified anymore by the needs of a country in the aftermath of the successful peace process with Ecuador and far away from the civil conflict that afflicted it in the mid-nineties. Moreover, such expenditure turned out to be corruption-prone, featuring lack of any control. Recommendations: Authorities should redress this situation by > reducing resources devoted to military outlays to the budgeted 2.1 percent of GDP in 2002, and shifting part of those savings to social expenditure; and i completing the application of Sik subsystems to military expenditure, to make it more transparent and accountable. r Authorities indicated that such increase could partly be due to a reclassification of line items grouped under social expenditure. 31 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Figure 2.10 Social and Public Spending Figure 2.11 Growth Rates In Social and Public Spending 20000- (1994 mln NS) 40 2000 30 15000- 0 Q. 10000. 2 10 5000 0 .9 -10 a) 91 92 93 94 95 96 97 98 99 00 01 02 0 92 93 94 95 96 97 98 99 00 01 02 ---Social Spending - Public Spending -- Social Spending -Public Spending 2.25 Developing effective counter- Table 2.6 Social Expenditures are Pro-or Counter-cyclical? cyclical fiscal and social policies Point a pioity epecaly i godElasticity (1991 - % change in elasticity Point elasticity remains a priority, especially in good(1998) (2001) times. Both policies are, in general, Total Expenditures ... -0.28 complementary. The most significant Social increases in total (and social) spending Expenditures 1.19 -1.50 -8.32 -0.17 took place in the early-to-mid nineties, when growth was high. Then, in the Health 1.15 -5.58 -1.87 -5.83 late nineties, real social spending Nutito 2.35 0.81 13 -. remained fairly constant even during Social Security 1.47 -0.75 -17.01 -3.34 the economic downturn of 1998 and, Other Social later, in 2001 (Figures 2.10-2.11). How Investment 1.66 -2.75 -61.87 2.32 expasive wer totf an soI a. Elasticities are with respect to total real expenditure of the consolidated public expansive were total' and socia sector expenditure despite fiscal contractions b. Significant at 1 is an empirical question. The percent changes and point elasticities of total and social expenditure were negative in 1998 (Table 2.6). For the entire period of 1991-2002, the level of social expenditures was not protected vis--vis fiscal adjustments. The elasticity of all categories of social expenditures was positive and equal to or higher than unity-which reveals their pro-cyclical behavior (Table 2.6). These findings are all supportive of pro-cyclical policies. N However, on the one land, Authorities have developed the basic institutional framework required for a countercyclical fiscal policy with the creation of the Fiscal Stabilization Fund as part of the Law of Fiscal Prudence and Transparency. Short of resources, however, the Fund has rather played a marginal countercyclical role (para 2.29) Li On the other hand, since 2001, and in the midst of another economic downturn, Peru's government decided for a countercyclical social policy, as can be inferred from the non- negative growth of real social spending, wlen compared to the actual decrease in real total expenditure (Figure 2.11). In a similar pattern, point elasticities for social, health, nutrition, justice and social security expenditure in 2001 were also negative, indicating that in a year of fiscal contraction, expenditure in these sectors actually increased.16 These results for fiscal 16 Elasticities were calculated such that =XiX, where q is the elasticity of expenditure in the i sector, N is the percentage change in real expenditure in the ith sector and X equals the percentage change in total real expenditure. The higher the magnitude of the elasticity, the greater the responsiveness of that category of expenditure to variations in total expenditure. Sectors that are favored in periods of fiscal contraction of total government expenditure have elasticities of less than unity, that is, they would be expected to experience reductions of a proportionately smaller magnitude than the total expenditure cutback, or even real increases if the elasticity is found negative. An elasticity of 32 Chapter II. PERU'S FISCAL CHALLENGES AND VULNERABILITIES contraction appear robust to period changes used to estimates, but also appear asymmetric as, with the exception of pensions, social assistance and primary education, they have values below unity in the 1998-2000 period of budget increases, reflecting a less than proportionate response to an increase in total expenditure (Table A24). E. REVAMPING THE FISCAL RULES 2.26 Peru needs a fiscal rule to Box 2.3 Peru: Evaluation of the Transparency Law 2000-01 help stabilize its fiscal balances in 2000 2001 Law Acua Law Oserve the medium-term, reduce its volatile ii as /. o. and pro-cyclical fiscal deficit-bias, CD . 1 5/ . 2.3% and consolidate credibility. As Difference between annual mentioned above, Peru is under growth rate of considerable fiscal stress with only a gene mild counter-cyclical social policy in financial expenditures and the early 2000s. Pro-cyclical and inflation less than deficit-biased fiscal policies can be 2% 2.Oh 0.6/ 2.0/ -5.0% explained by the combination of faulty ;ii4ebt inrease espeiall ~~Elecion(adjusted) no measures, especially during Election reater than NFPS years, fiscal conservatism and weak F On previous requirements. Issuing a new law should be preceded by a building process of (i) institutional strengthening regarding the human and technical inputs required by the design and implementation of the Law; (ii) consensus-building participatory mechanisms led by Authorities, with a draft-design of the Law being published in advance to receive comments through the web; (iii) enhanced SIAF-supported and user-friendly budget transparency (Chapter III), so as to allow open access by Congress and civil society to monitor implementation of the Law, once it is approved; and (iv) a comprehensive evaluation of contingent liabilities, so adequate and precautionary budget provisions can be made under the new rule. > On procedural norms. From the recent experience, it is clear that (i) budget reporting needs-human resources and technical tools-have to be significantly strengthened; (ii) the multi-year budgeting framework also requires improvement (Chapter III); (iii) introduction of a set of sanctions for non-compliance is unavoidable; and (iv) upgraded oversight capacity to evaluate and enforce sanctions by the Comptroller's office is needed. > On numerical targets. Two main measures aiming to simplify the framework for quantitative benchmarking are proposed. The first measure would consist of targeting an estimated structural NFPS primary surplus of 1 percent of GDP (Annex B) by 2008-10, which would be roughly equivalent to targeting a structuralNFPS deficit close to 1 percent of GDP. These targets are set in structural terms and do not take into account the cyclical component of the deficit. Progress toward meeting such target should be done through incremental steps. Following Banks' detailed estimates of the potential output and the structural deficit (Figures 2. 12-2.13) and assuming a prompt recovery under a baseline Figure 2.12 Actual and Potential Output Figure 2.13 Actual primary balance (% of GDP) and structural o (billions of 194 Primary balance (% of potential GDP) l3a 1 011 11 11 121. 90. 700. 90 91 92 9 9 95 9 97 9 99 00 01 02 90 91 92 93 94 95 96 97 98 99 00 01 -- AzUW Output - ...PotentialOutput .. alA b am .... .PotendialbaLan scenario since 2002, adjusting for the cyclical component would require a primary surplus of about 0.9-1.3 percent of GDP as early as 2004 and until end-period (Annex A)." Although broad convergence prevails between the Banks' and the Authorities' estimates, an eventual official benchmark would need to be thoroughly reviewed and published by the Authorities. ' The estimated cyclical component of the fiscal deficit is small and varies between 0.1-03 percent of GDP between 2000-02, but might increase a bit, specially if recovery takes place (Annex B). 1 Notice that Banks' estimates are particularly close to ones made by Deutsche Bank (2001) of 0.8 percent of GDP, the Central Bank of Peru (2002) of 1.3 percent of GDP, and Jim6nez/MEF (2002) of 1.1 percent of GDP. All of them do not include, though, any provision for the materialization of contingent liabilities. 35 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION The second measure would propose to eliminate some of the less-important numerical rules. Not only a few of them overlap with the main rule, but are ineffective, as it happened to those applied during Elections years: they did not take into account strong end-year public spending seasonality. > On fiscal rules for sub-national governments. Introducing fiscal rules at the sub-national level is a critical priority for the revamped Law, given the pace of the decentralization process under the present Administration, and justify its upgrade to a Responsibility Law. Although there is no consensus on pre-setting limits or ceilings for specific fiscal indicators at the sub-national level, the vulnerability of the overall fiscal situation requires a hard budget constraint, in terms of a modified golden rule-to include capital expenditure--or zero operational deficit. Rules should also be more restrictive during the last year of the administration to avoid electoral cycles, and be more flexible in the presence of exogenous and unexpected shocks. Tight non-contingent municipal debt restrictions are needed. Finally, as there is little knowledge, but partial evidence (Chapter V) about the existence of a significant contingent municipal debt, debt restructuring deals (supported by its previous comprehensive evaluation) should precede the adoption of any local government rule. > On the Fiscal Stabilization Fund. There is no sense to expect a significant countercyclical policy as far as growth remains low and fiscal revenue is stagnant. Additional features of the to-be-revamped Law include: (i) redesigning an effective special clause for Election years, as the three past Election episodes provoked a clear fiscal deterioration; (ii) making room for changes that will affect the structural balance over time, including those affecting the productive capacity of the economy, external prices of main mining exports, and the taxation structure; and (iii) strengthening existing inflation targeting mechanisms in consistency with a revamped fiscal rule, and supported by enhanced coordination mechanisms between the Central Bank and the MEF. 36 Chapter III REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE 3.1 Reorienting the budget toward effective poverty reduction requires as a pre-condition an overall reform effort of public expenditure management (PEM). Such reform includes a modern financial nanagement system; a pro-poor and performing budgeting system; a clear definition of medium-term social objectives; an effective tracking system of the ,shift of budget allocations toward these priorities; and transparent information access to an outcome-oriented budget. Since the late-nineties, Peru has made promising inroads in all these areas and, although remaining challenges are still significant, its implementation record provides a solid foundation for further developments. This chapter assesses progress in pro-poor budgeting management, whereas next chapter deals with factors limiting efficiency in public sector delivery. The final chapters examine prospects for second-generation institutional reforms in decentralization, civil service, governance, and mining sector policies. 3.2 Tangible progress in public expenditure management Box 3.1 Main IGR/PEM Issues (PEM) in Peru over the past few years is impressive. I The IGR identified several factors includes: affecting financial accountability in PEM (W13, 2001 b): * creation of an Integrated Financial Management System Q Excess dominance by the (SIAF, in Spanish) in 1998; Executive through the use of (SIAF,Presidential Decrees: between 1994 * development of an upgraded budget management system and 2001, Congress passed 1,152 since 1999, with improved transparency of budget laws or legislative resolutions, while execution supported by the website Portal de the President issued 870 (86 percent Transparencia in 2000, which includes for the first time Urgency) Decrees; defense and national security spending and a participatory authonaentral m in method for budget formulation scheduled from 2002; Economy and Finance, and * passage of the Law of Fiscal Prudence and Transparency Presidency-to maximize and creation of a Fiscal Stabilization Fund (FSF) in 1999; discretionary control of budgetary * strengthening of accounting institutions and procedures, cosoliedbudget0(excludingedet compatible with world standards, applied to the entire service) assigned to these two public sector since 1999; institutions represented 63 percent; * introduction of performance management contracts of 0 Insufficient Congress oversight State-Owned Enterprises under FONAFE in 1999; of budget execution; O3 Local governments still do not * adoption of a medium-term expenditure framework in have the necessary checks 2000, based on a Multi-year Macro-Economic Framework and balances to implement for 2001-03 and a Multi-Annual Sector Budget (MAB) decentralization properly; and Plan for 2001-05; and L Severe institutional disorder in a state in which highly modem * adoption of budgetary provisions for a set of Protected institutions coexist with antiquated Social Programs (PSPs) in 2000. entities, and multiple social agencies overlap their service delivery efforts. 3.3 In spite of ongoing reforms persisting weaknesses remain. Some of them were identified in the Peru Institutional and Governance Review (IGR) (Box 3.1) and undermine budgetary transparency andaaccountability. These include: 37 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION o an incomplete SIAF; L excessive concentration of budget authority in a few ministries and budget rigidity reflected in inertial spending; L a disconnect between multi and annual budgeting, the latter being amended by continuous Presidential decrees; D a fragile and uneven institutional framework for developing an effective medium-term social policy and for decentralization; and o an inadequate and user-unfriendly external budget monitoring and reporting, which does not facilitate its virtual oversight by authorities themselves, Congress, and civil society. A. MODERNIZING FINANCIAL MANAGEMENT 3.4 Peru has developed a well-performing integrated financial management system It aims to increase the control, efficiency, accountability, and transparency of public spending, and support the future decentralization to line ministries and regional and local governments of the primary responsibility for managing resources and improving service delivery in Peru (Box 3.2).1 Box 3.2: The Integrated Financial Management System (SIAF) of Peru SIAF was created by Vice-ministerial Resolution No.029-98-EF/II on October 22, 1998 as the mandatory internet- based system for the recording of all data related to revenue and expenditure execution that are carried out by the Executing Units (UEs) of the public sector budget, at the national level, and according to the procedures and norms approved by MEF: * It covers all Central Government and most decentralized entities: a total of 580 EUs (including Defense and National Security since 2002), all public enterprises, Supervisory bodies (SUNAT, Customs, Regulators, municipal enterprises), and 23 municipalities (June 2002). * It contains four modules: Treasury-which receives budget ceilings prepared by the commitment calendar of the Budget Office and pre-approved by the Cash Oversight Committee-Accounting, Financial Statement Reporting, and Public Debt, the latter managed as an EU. Public Enterprises only use the financial accounting module, * It has very strict budget execution control mechanisms. Once payments have been previously authorized by Treasury through the system and are daily registered in SIAF, then Banco de la Nacion proceeds to alert and transfer funds to the EUs' bank accounts or issue checks. * The expansion to include financial accounting since 2000 has provided reliable and timely information to the Accountant General who is now responsible for preparing financial statements. Recording of transactions at the EUs is organized in two parts: administrative (commitment, accrual, payment and disbursement), and accounting (under standard categories). To record a commitment, it is necessary to use a specific expenditure classification, a financing source, and an objective corresponding to the expenditure. The system then verifies that the operation is in accordance with budget limit criteria and commitment calendar. An accrual is done only when a commitment exists and the EU receives the verification that the goods or services have been provided according to the "best date" (prompt payment-no later than one-month) concept. After full payment, the registry is updated with the correct disbursement information provided by the Banco de la Nacion. * All expenditures, revenues, and complementary operations use a SIAF Matrix Operations Table that integrates budget classifications with the Government's chart of accounts. At present, it combines efforts in the areas of Treasury, accounting, and debt management. Its role begins once the annual budget is approved and monthly ceilings are authorized by the Budget Office and Cash Oversight committee. SIAF is tied up with Treasury to register all operations (revenue, expenditure) approved for the budget execution cycle through 580 Executing Units (EUs) dependent on the budgetary envelopes (pliegos) of various ministries, regional Directorates (CTARs) and public institutions. SIAF contains 3 operational elements: (i) single registry of Box 3.2 is based on WB/IADB (2001). 2 The Peruvian public sector is divided into 5 categories of entities: the Central Government (CG) with 580 EUs, the National Fund for Financing Government Enterprise Activity (FONAFE) with 42 entities, the General Directorate for 38 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE each EU transaction, and virtual transmission of this information to the Budget Office, the Treasury Office and the Accountant General; (ii) cash payments management, but only for those financed by Treasury ordinary resources; and (iii) financial reporting. In three years, SIAF has consolidated itself as the instrumental financial management tool of the public sector. 3.5 Overall Efficiency Gains. After three years of reform efforts, quantifiable savings have emerged from several outcomes: (i) duplication of financial transactions is eliminated as payments are entered once and instantly in a reliable database system; (ii) poorly controlled quasi-extra-budgetary funds are significantly reduced (para. 3.12); (iii) electronic transfers are replacing the use of checks, thus reducing payment lags, transaction costs and risks of fraud or lost/stolen checks; and (iv) unused funds are virtually eliminated, as once a payment is authorized by the Cash Oversight Committee, the banking transfer is executed within 72 hours and remains available only for one-month. 3.6 Overall Effectiveness Gains in Budget Management. Outputs arising from a better mix of public resources consist of (i) the elimination of arrears on floating debt to suppliers, a major budgetary distortion in the past; (ii) an enhanced cash management that prevents shortages of counterpart funds to external projects; (iii) an improved level of budget execution by the Central Government; (iv) control for the first time of the previously sensitive spending "black hole" ministries, Defense and National Security, now fully adopting SIAF system and procedures; (v) enhanced expenditure control and transparency due to a unique database which offers virtual information through the web; (vi) the simplification of the functions of accounting and auditing, which reduces the scope of fraud and corruption; and (vii) an enhanced public debt management, as the system also prevents external arrears. 3.7 Despite its impressive progress, Peru SIAF still faces significant challenges: . SIAF core sub-systems are incomplete. The lack of a budget module, in particular, is a severe handicap for improving budget efficiency, participation, and transparency. Contrary to similar systems in Latin America, the Peruvian SIAF does not have a budget module, but fundamentally works as a virtual payment system. As a result, Peru's budget formulation and in- year global budget modifications remain little known externally, centralized and decided at MEF's discretion, and only known ex-post, when they are registered monthly in SIAF. Despite individual ministries providing inputs to overall budget formulation and same in-year modifications, they do it on a bilateral basis with MEF and have little control over ex-post decisions regarding charges to their global allocations. Recommendations: > A user-friendly budget module is urgently needed to allow its virtual access with no restrictions. It would help assess consistency of budget proposals with macroeconomic conditions in early stages; build a stronger link between budget formulation and multi-annual strategic priorities; facilitate monitoring of in-year budget modifications consistency with inter- and intra-sector budget priorities in a participatory and transparent way; and allow Congress' legal mandatory oversight of the budget execution. > SIAF's budget module requires a multi-annual sub-module too, i.e. one with capacity to track GOPs' compliance with multi-annual budgeting benchmarks. On the one hand, SIAF only the Public Budget (DNPP) with 99 entities, 1826 municipalities and 101 public beneficiaries. The CG has 26 sectors, divided into 141 envelopes (pliegos) comprising 580 EUs, which accounts for about 60-65 percent of the public budget. 39 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION registers current-year transactions. On the other hand, the Law of Fiscal Prudence and Transparency requires bi-annual reports on compliance with Multi-Annual benchmarks, but such commitment has not been fulfilled. The separate sub-module should allow a historic database, virtual monitoring of compliance with annual executed expenditure and help produce mandated biannual reports on compliance with MAB targets. > Among other possible SIAF modules, the one on human resources management is critical. Others that could gradually be developed are national assets, procurement, national investment system, contingent liabilities risk, and regional and municipal financial management. The human resource module is essential for civil service reform. The one on national assets to have due registry of state assets. The module on procurement, starting point for e-government, is critical for anti-corruption efforts, the most vulnerable area of governance. The module on public investment should help prioritize multi-annual investment plans. The module on contingent liabilities risk would help to register them and make proper budgetary provisions, while the one on sub-national management would prepare SIAF support to an accelerated scheme of fiscal decentralization, as future regional and municipal governments require a twin-decentralized SIAF and training, whose development should take place in many years. > The existing SIAF sub-system to track financial and physical performance indicators, a critical tool for an outcome-oriented budget, supported by benchmarks - final and intermediate -- and performance management contracts, requires a thorough conceptual review before completion. The pilot sub-system has too many useless indicators and their inputting and tracking is poor. > Due to their complexity and the already over-committed limited technical and human resource capacity of the SIAF team at MEF, all development of new SIAF modules requires detailed prioritization and permanent training Authorities should support and outsource the creation of a superior university-degree in management of SIAF systems. > Finally, international experience indicates that two important modules, like a fiscal cadastre (essential for a predial municipal tax), or an e-government website (including e-procurement or e-geo-referenced expenditure map at district level, for targeting purposes) should be developed outside, but with link facilities to SIAF. Control of decentralized budget execution by SIAF is partly diluted beyond EUs through encargos.3 Grosso modo, in 2001, a little less of thirty five percent of the budget was not devoted to salaries, financial obligations, or pensions. At least 12 percent (NS. 1.4 Billions) of such share of the budget was executed through encargos. The encargo is an ad hoc mechanism. It transfers funds from the EUs to a myriad of at least 50 thousand Operational Units (Unidades Operativas-- OUs) dispersed nationwide (including schools and health posts) and, most of them, outside SIAF control. Operationally, it works as follows: Following reception of funds, if the EU decides to commit an encargo to an OU, it registers it as a transfer in SIAF. Such transfer is endorsed in a bilateral institutional agreement that has no legal implications. Ensuing budget execution control by SIAF then faces multiple deficiencies: (i) late registration, for only by end-year the UEs could receive a general report of the use of those funds; (ii) no direct data entry, as most OUs do not use SIAF software; (iii) poor auditing, as not all EUs have adequate administrative capacity to Among other control failures from SIAF, only one is significant: internal encargos, also known as anticipos, whose 2001 amount (NS/.1.3 Billion) was very close to the one of encargos, and are mainly used for travel advances. Other leaks like petit cash used as revolving cash fund are less significant and are later duly registered. 40 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE internally audit their OUs; and (iv) misallocation of resources, as there are no expenditure tracking mechanisms set to verify that resources have effectively been assigned to the original purpose of the transfer. The use of encargos is generalized in the health, transport, agriculture and defense sectors, which use them to speed-up execution of their investment program. The abuse of encargos has obstructed and delayed the proper application of the SIAF Treasury module at the ministry of Defense. However, since 2002, it has become the pilot ministry for urgently reengineering encargos registry by a SIAF2 special module (Box 3.3). Lessons from this pilot ministry should be replicated. Box 3.3 Introducing the SIAF at the Ministries of Defense Recommendations: and National Security (interior) Until December 2000, the 5 EUs of the Ministry of Defense > The SIAF2 special module should be (MINDEF) did not register their transactions in SIAF. An expandedadministrative mechanisms, known as libramiento, allowed expaded hrouh al acive Us; MIN'DEF's UEs to transfer their budget resources to about 870 > an administrative reorganization of opened banking accounts which worked as authentic "black the EUs should be implemented to boxes": Resources in these accounts did not belong to support it (see below); and Treasury, were neither monitored by the Cash Oversight > quarterly deadlines should be set for Committee nor audited, and end-year non-spent funds were t never returned to Treasury. In January 2001, hDwever, all all OUs to submit full and transparent MINDEF's EUs were suddenly incorporated into SIAF and financial statement reports. libramientos were eliminated. As such decision was political, and no previous technical preparation was done, all EUs were SIAF has reproduced the State's broad allowed to operate their new SIAF-registered banking accounts under a transitory mechanism known as encargo to OUs. But, institutional disorder into its own a few control issues appeared: internal structure as no homogeneous 9 About half of MINDEF's EUs operations (Army, Navy criteria for defining an EU prevail. and Air Force) were executed through encargos to OUs; Below budget envelopes (pliegos), there 9 OUs did not properly register their direct purchase of are 580 EUs nationwide. In theory, an EU goods and services, did not use the SIAF or, in the case of is the lower level defined by the MEF as some located outside Lima, did register them, but manually. is te loer lvelOn January 2002, to face these problems, SIAF moved to: the one with full administrative capacity Maintain a limited number of EUs; to register the use of funds into the 0 Eliminate OUs operating in the same office of the EU; system.4 In practice, budget size, * Design and install a tailored-made pilot SIAF2 special geographic location and political module in 100 OUs executing encargos; affiliation appear as guiding principles to Develop an accelerated training program in SIAF2; detemin it cratio. Bcomng n E Reduce the number of OUs located outside Lima and determine its creation. Becoming an EUthem; is equivalent to have direct control of the 0 Replicate such exercise at the Ministry of Interior. checking account, which makes it a well- As a result: (i) 100 percent of defense and national security desired milestone to achieve by OUs, expenditure will be managed through SIAF by end-2002; (ii) regional Directorates or ministerial units the SIAF2 module will be expanded through all active OUs; Uniddes e Sevicos Eucatvos (iii) an administrative reorganization of the Army will be like the Unidadesimplemented to support it; and, for the first time, (iv) also by (USEs). All Regional Administration end-2002, all five EUs of MINDEF will submit full and Transitory Councils (CTARs) existing transparent financial statement reports. nationwide are considered as budget envelopes (pliegos), and each budget envelope (pliego) has the capacity to propose the creation of an EU and assign its budget. However, no standard minimum requirements for the creation of an EU exists, which contributes to their lack of homogeneity, wide dispersion of the budget amounts-an average of between NS/14.1M in Uyacali to NS/122M in Lima-per unit (Table A45), and perhaps unnecessary bureaucracy. For instance, the Ministry of Education has 72 EUs, b The authority of the EU is significant: Whereas the Pliegos are in charge of budget formulation and sector policy norms, the EU has the capacity to make budget commitments, order payments and execute spending. Whereas budget changes between EUs require a resolution from the head of the Pliego, authority to approve changes within an EU (e.g. from one program to another or within programs) may be delegated from the head of the EU to the OU. The EU is also free to reassign resources among objects of expenditure items, except when required for compensation purposes. 41 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION and this includes up to 8 different management models in a menu of widely heterogeneous institutions with different hierarchical responsibilities, like Regional Directions, Sub-Regional Directions, and USEs. Recommendations: > Taking advantage of the incoming decentralization process, it would be advisable to have a gradual reordering of EUs, both at the regional and district level. In fact, Supreme Decree No.001-2002-PCM opens the door to their restructuring. This could sound very technical, but it is very important as it shapes incentives faced by those delivering goods and servic es, and the oversight/accountability arrangements (checkbooks) of those units below EU status. This is a sensitive matter, though, as new regions could assume that EUs reordering is taking away their actual budget resources. Hence, this process could be bmand-driven by the to- be-decentralized social sector ministries at the regional level; and/or supply-driven by the Executive. In both cases, it would be suitable to examine and enforce a set of standard minimum norms before a unit reaches (or is confirmed) as having a EU status. > Complementary to the above, multiple types of EUs (and SIAF software) could be considered, depending on their budget size (usually correlated with population), location and functions. SIAF requirements could be stricter as their budget gets larger, as provincial districts are involved, or as more complex functions are developed. A simplified SIAF- software (it could be called SIAFITO) could be considered in support of small budget or distant EUs. SIAFITOS could be easier to handle by regional and local governments than the standard SIAF. B. UPGRADING THE BUDGET MANAGEMENT SYSTEM 3.8 Peru is upgrading its budget management system at an accelerated pace. Recent innovations in budget practices includes lengthening the time horizon of budgeting from a single - year to the medium-term, setting explicit fiscal targets that limit aggregate expenditure and developing baseline projections for estimating the impact of proposed policy changes on future budgets. The budget process has 4 stages: formulation, approval, execution and evaluation (Box 3.4). 3.9 Following a standard assessment of the Peruvian PEM system, in consultation with country authorities, results indicate that it requires some upgrading (Table 3.1 and Annex C). The assessment covers 15 indicators covering critical elements of PEM systems deemed necessary for tracking expenditure. Seven of the indicators relate to budget preparation and four each to execution and reporting. The total number of satisfactory benchmarks met, 8 in the case of Peru, can be viewed as an indicator of the quality of a country's PEM system. According to IMF/WB standards, a country is deemed to require "little or no upgrading" if at least 11 of the benchmarks were met; "some upgrading" is required if between 8 to 10 benchmarks, and "substantial upgrading" is required if less than half (7 or fewer) of the benchmarks were met. 3.10 Peru scores unsatisfactory in some areas of budget formulation and execution, and satisfactory in most areas of budget reporting, other than external. 42 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE Recommendations: > Budget classification should be made fully consistent with the GFS definition of General Government; > The quality of budget projections (especially in function and sector composition) needs to be upgraded if social priorities are to be sustained in the medium term; > Multi-year projections also require significant improvement and adequate integration with the annual budget process. > The creation of a budget module and refinement of the existing set of financial and physical indicators in SIAF should facilitate a user-friendly tracking of pro-poor expenditure and an effective results-based control system in budget execution. > Sector tracking surveys should be developed randomly on a regular basis; > External reporting of budget execution, multi-annual budgeting and audits should be frequent, timely and comprehensive. > Full adoption of the IMF Code of Fiscal Transparency requires a thoroughly defined and supported technical assistance program. Box 3.4 Peru: Principal Stages of the Budget Formation Ministry of Economy Pliego Others Level of and Finance Dsawegation Convey initial ceiling Current and capita expenditures The head establishes the priorities The ceiling is distributed By target and among the pliegos specified Congress approves the e v law (the executive A all evels of submits the projects disaggregation previously approved. by the Counsel of Ministers) Current and capital Convey allocations expenditures Trimester of expenses (day 19 of the month before By target and the end of the Plan trimester expenses specified 5 trimester (day 26 of the month before the end of the trimester)' By pliego and executing unit By sector and Approve calendar of program commitments (up to By generic expense the 5 of the month cateories being approved) Assign calendar of commitments among suboroirams' Create Report on Iea Physical Progress > and Financing of Goals 1. Discretion of the Pliego's Budgetaryunit to decide how much to assign to each executing unit. 2. Additional discretion when there is no executing unit. Source: Normas Presupuestales 43 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table 3.1 Peru: Status of Budget Management Benchmarks Budget Management Benchmark Status of Compliance: S=Satisfactory; P=Partial; U=Unsatlsfactory Formulation-Comprehensiveness 1. Composition of the budget Meets IMF-Govemment Finance (P) The annual budget covers Central entity Statistics definition of general Government expenditure. Coverage ofthe government consolidated general government budget has minor discrepancies and does not include regulatory agencies, local governments, and social funds. 2. Limitations to use off-budget Extra or off-budget expenditure is not (S) Quasi-etra-budget expenditure is about transactions substantial 18 percent in 2002, mostly linked to external credit, and rapidly declining. 3. Reliability of budget as guide Level and composition of outturn is (P) The level is close, below 10 percent to outturn "quite-close" to budget deviation in 2001; but its composition still shows significant differences. 4. Data on external financing Both capital and current donor-funded (S) External credit is duly registered, except spending included for a few donor off-budget accounts. Form ulation-Classification 5. Classification of budget Functional/program information (S) The budget contains proper functional transactions provided and program classification. 6. Identification ofipoverty- Identified through use of a classification (P) Broad and detailed (beyond the usual reducing expenditure system (e.g., a virtual poverty fund) social sectors) mapping of budget for poverty-reducing expenditure is possible, but it is neither virtual, nor user-friendly; and requires continuous consistency checks between SlAF and the Budget Directorate. Formulation-Projection 7. Quality ofmultiyear Projections are integrated into budget (U) Multi-year projections are not projections formulation and close-fit to actual integrated into the budget, are subject to numbers significant modifications each year, and their execution has a poor fit to actual numbers. Execution-Internal Control 8. Level ofpayment arrears Low level ofarrears accumulated (5) SIAF requires payment arrears to be cleared within one month. 9. Quality of internal/exteral Effective interal/exteral audit (P) Good quality of internal audit, but audit infrequent and poor quality (e.g. no program costs are reported) of external audits (WB/mDB 2001). 10. Use oftracking' surveys Surveys used on a regular basis (U) Non-existent until recently2 Execution-Reconciliation 11. Quality of fiscal/banking data Reconciliation of fiscal and monetary (S) Satisfied permanently reconciliation data carried out on routine basis Reporting 12. Timeliness of budget reports Monthly expenditure reports provided (P) Regular internal reports are provided within 4 weeks of end-of month within 4 weeks of end-of the month; but external reports are annual, 6 months after year-end. MEF has announced quarterly budget reports in the web since mid-2002. 13. Classification used for budget Detailed functional reporting derived (S) Existing functional reporting derives tracking from classification system from the classification system. Its program and project decomposition could be improved. Reporting-Final Audited Accounts 14. Timeliness of accounts closure Accounts closed 2 -months before year- () Accounts close one-month after year- end end 15. Timeliness of final audited Audited accounts presented to (S) Audited accounts submitted on June accounts Legislature within one year 30th 1. Tracking is defined here as the mere identification and reporting ofbudgeted and executed outlays. Tracking surveys are usually needed, when the financial management system does not have the capacity to regularly ideRetgl, monitor and report expenditure at the lowest levels of the Government, e.g. districts (IMF/WB, 2002). 2. Public Expenditure Tracking Surveys (PETS) are rare in LAC: Only Honduras (for service quality) and now Peru (for quantitative leaks and service quality) have developed them (see Chapter y). 44 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE 3.11 This assessment should however be used with caution, since not all indicators are of equal importance. Its bias in favor of process measures (e.g. audit reports) in detriment of impact outcomes (e.g. objected audit reports) is usually not optimal in a country context and may miss critical issues of a PEM, like the ones outlined below. 3.12 Although the share of quasi- Table 3.2 Composition of CG Budget Sources EBFs in Peru is rapidly declining, its (Percent) importance is significant for a single 2001l 2001P 2002b category: external credit (Table 3.2).' Ordinary 80.7 70.7 75.1 Following the elimination of FONAVI, Canon/Sobrecanon* 0.5 0.6 0.6 the Housing EBF which accounted for Customs 0.4 0.4 0.0 about 6 percent of the budget in 1999, the Self-Raised Funds 6.8 6.5 6.9 Central Government budget coverage of External Credit* 7.7 16.9 13.2 all resources has increased from 65 Grants 0.4 0.7 0.2 percent in 1999 to 82 percent in 2002. Contrib. to Funds 3.5 4.2 3.9 Self-raised revenue by local governments, Toure MEF. Customs proceedings, and grant funds are now controlled through the single *quasi.EBFs; b budget; p preliminary Treasury account managed by the Banco de la Naci6n. However, three main sources of quasi-EBFs remain: (i) external credit accounts (in local currency), which manage salary payments to top-ranked public servants serving external projects and whose eventual late registration might affect reporting on executed expenditure; (ii) contribution to funds (especially ESSALUD); and (iii) the canonsobrecanon revenues from mining, oil, gas and other natural resource activities. Recommendations: S The total reduction in the number of Extra-Budgetary Funds (EBFs) would not only improve control over the level and quality of public expenditure, but also reduce the cost of "sobregiro" fees that Treasury pays when seasonal shortages of tax revenue collection by SUNAT lead to insufficient cash availability to cover all public payments. SOther less visible EBFs, like tax expenditures, should continue to be registered from 2003 onwards. >' Accounting of external credit should be fully made in SIAF and have timely and transparent reporting. 3.13 The budget paradox: budget formulation remains centralized in MEF, while budget execution * decentralized While budget formulation is centralized and a high proportion of budget expenditure remains under the authority of government agencies located in Lima, roughly fifty-six percent of CG non-previsional, non-financial budget is executed through the regions in a decentralized fashion (Table A39). Hence, although larger and decentralized entities appear to be tightly controlled politically and may operate as means for distributing political spoils (World Bank, 2001 a), they preserve enough spending authority to modify the budget at their discretion. In OECD terms, EBFs comprise all government transactions that operate outside the normal budgetary process. Peruvian "quasi" EhFs are considered as such since they are monthly capped and reported in SIAF as part of the overall budget envelope, which is unusual for such type of resources, but conform to other EBFs features: their effective allocation takes place only if they become available; funds are managed through commercial bank accounts separate from the central government budget account; and, despite standardization efforts, still might have separate accounting systems and classification. 45 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 3.14 Inertial expenditure is a Table 3.3 Revised Projection of the 2002 Budget major constraint for budget (Millions of NS and % of Total Expenditure) formulation. Peru is ranked among Structure of Expenditures All Sources Ordinary Resources countries with the most inflexible Expenditures by Category 21,790 59.9% 17,290 63.3% budget not only in LAC, but also Fixed Component 16,820 46.2% 14,844 54.3% worldwide. In addition to Brazil, Goods and Services 3,200 8.8% 2,170 7.9% India and a few HIPC countries, Investment 1,330 3.7% 868 3.2% which have above 90 percent of their Financial Investments 171 0.5% 139 0.5% total revenue fixed, Peru has 91 Other Current Expenditures 1,967 5.4% 1,770 6.5% percent and .86 percent of ordinary Other Capital Expenditures 25 0.1% 25 0.1% and total budget resources in 2002 Personal and Social assigned to inertial commitments Obligations 9,272 25.5% 9,016 33.0% (Table 3.3). Budget rigidities Contingency Reserves 856 2.4% 856 3.1% originate from different sources Flexible Component 4,970 13.7% 2,446 8.9% including: (i) wage and benefits Goods and Services 1,714 4.7% 1,124 4.1% considered as derechos adquiridos Investment 2,933 8.1% 1,258 4.6% by the Constitution; (ii) permanent Other Current Expenditures 36 0.1% 31 0.1% contracts hidden as non personal Other Capital Expenditures 288 0.8% 33 0.1% services (NPS)-no less than 40 Structural Obligations 13,982 40.1% 9,438 36.7% percent of the total budget allocated Financial Expenditure 7,787 23.1% 4,495 18.7% to NPS, according to Authorities' Previsional Expenditure 6,195 17.0% 4,943 18.1% estimates); (iii) inertial payroll under TOTAL 35,772 100.0% 26,728 100.0% prolonged investment projects; (iv) Source: MEF. earmarked transfers to municipalities and the private sector, as the Vaso de Leche and Comedores Populares; and (v) misuse of the budget contingency reserve, which finances March school bonus, and July and December salary supplements (aguinaldos). The high share of inertial current expenditure explains not only its rapid expansion, but also the complexity that attempting to reverse such pattern will require. Recommendations: > An immediate general hiring freeze under the NPS category is critical, accompanied by a case-by-case short-term renewal of existing temporary contracts, upon their prior revision; G authorities had no option, but to review their present policies regarding pension and wage benefits in a comprehensive civil service reform (Chapter VI); > funds assigned to Vaso de Leche and Comedores Populares need to be streamlined (Chapter IV; and >' the use of the contingency reserve requires review, so as to fulfill its true role: filling unpredicted in-year budget needs.I 3.15 End-year excessive seasonal spending is another critical shortcoming of budget execution. Notwithstanding the purpose of the creation of SIAF-which is to enhance and improve budget control--the persistent pattern of excessive deficits in the last quarter of the last six years, coupled with small budget surpluses in the first quarter, remains a constant and fiscally damaging seasonal concern-whether they occur in an election year or not. (Figure 3.1). Such a budget behavior is not Peru-specific only. Most public budgets in LAG feature such fiscal impulses to cover not only for end-year cumulative and additional payments, but for building fiscal provisions for the slowly-opening expenses of the next year's first-quarter. A similar pattern is also typical of the LAG political business cycle, and Peru is no exception (L6pez-Clix, Authorities reported that such review was achieved in the formulation of the 2003 budget. 46 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE 2001). However, deficits appear Figur 3.1 Comparison of Central Govt and NFPS Balance for both, Central Government (CG) and Non-Financial Public Sector (NFPS) and following SIAF 2% creation, their size is hard to justify for other reasons than fiscal 0 indiscipline. In year 2000, the quarterly deficit developed since i t_ the third quarter, lasting until end- -6% year; while in 2001, it grew with -8% respect to 2000 (in particular the one of the Central Government). a CG Balanm m NFPS Balanm Recommendations: > Given its persistence throughout so many years, the first step for the GOP (and its Cash Management Committee) is to assess the deep implications that this seasonal episode is having on fiscal imbalances and monetary policy (possibly affecting short-term interest rates). > A thorough review of cash management and budget planning for the closing fiscal year is needed to examine structural-revenue and expenditure-factors that might explain end-year cumulative deficits and determine effective measures aimed to smooth them. > Under the present situation, setting and meeting realistic revenue projections and tight quarterly expenditure ceilings looks critical to make room for end-year impulses. > Until this seasonal pattern is significantly corrected, meeting mid-year tighter expenditure ceilings, especially in an Elections year, is particularly meaningful for fiscal discipline. 3.16 As PERU PEM modernizes, it will need to adopt further contemporary innovations in budget practices, both in terms of rules and procedures that will be required eventually under the new MF Code of Budget Transparency. Recommendation: > Defining procedures for identifying the government's exposure to contingent liabilities, particularly those associated to the financial system, pensions and public enterprises' guaranteed debt, as well as other fiscal risks; and > Accrual accounting for certain commitments could cautiously be attempted at least on the expenditure side.' C. BUILDING A MEDIUM TERM EXPENDITURE FRAMEWORK 3.17 In the past two years, there has been an important shift in the budget towards pro- poor expenditure. In spite of the recession and resulting fiscal retrenchment, the share of social expenditure to GDP has remained constant between 1999 and 2002 (Figure 3.2). Likewise, 7 The IMF Code of Fiscal Transparency does not require accrual accounting. Countries are allowed to go at their own speed in its adoption. Its implementation continues to have mixed results with Australia experiencing particular difficulties and New Zealand and Brazil claiming to have had a relatively smooth conversion passage (Woods, 2002). Among opposing "big bang" and "incremental" approaches, the latter looks closer to the Peruvian way to PEM reform, as the country has, de facto, modified accrual, since it already captures some commitments upstream in the expenditure chain. Accrual accounting may confuse the definition of fiscal deficit among civil society and other stakeholders and must be developed with caution. 47 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION significant inter- and intra-sectoral pro-poor spending reallocation has begun to take place between 2000 and 2002. Budget for defense and national security decreased from 2.9 to 2.1 percent of GDP; that for education, health and social assistance increased from 5.3 to 5.6 percent of GDP; and ESSALUD expenditure (non-targeted public health insurance administration) decreased from 1.5 to 1.3 percent of GDP (Figure 3.3). On the other hand, the share of poverty reduction programs (consisting of programs with universal coverage--education and health- spending and targeted spending) to social expenditure increased from 57 percent in 2000 to 60 percent in 2002 (Table 3.4). Figure 3.2 Share of Social Expenditures to GDP Figure 3.3 Sectoral Budget Allocations 3.5 20 100% 13% 11% 3.0 280%% 2.5 15 24 OAdministrative 2.0 60%: - -L_ 1.5 1Defense and 10 public security 0.5 , q , 506% - Social 0.0 .= 5 20% a Economi 1997 1998 1999 2000 2001 2002 m ii Total Expenditure, RHS ma Social Expenditure, RHS 13% - - & - Education, LHS --- Health and Water, LHS 0% - Social Assistance, LHS ---Defense/ Security, LHS 1997-2001 2002-2006 Table 3.4 Composition of Social Expenditure 1999-02* (Millions Nuevos Soles) Categories 1999 2000 2001P 2002e Universal Coverage-Education & Health- 8,378 8,804 9,113 10,170 Programs Targeted Programs (Extreme Poverty) 3,257 3,043 2,979 3,531 Non-Targeted Social Programs 7,909 8,928 8,960 9,175 ESSALUD 2,489 2,801 2,666 2,695 Pensions 5,330 6,127 6,294 6,480 Fonavi 90 1 0 0 0 Total Social Expenditure and Pensions 19,544 20,775 21,051 22,877 Memo: Poverty Reduction Programs (1+11)/Total (%) 59.5 57.0 57.4 59.9 Extreme Poverty Reduction Programs (11)/Total (%) 16.7 14.6 14.1 15.4 Non-Targeted Programs/Total (%) 40.5 43.0 42.6 40.1 Source: MEF. * Includes ESSALUD, FONAVI and pensions. Ppreliminary. e estimated. 3.18 Peru has made promising steps towards a medium-term budgeting framework (MTBF). As part of the Law of Fiscal Prudence and Transparency, since 2000, the GOP has published Multi-Annual Budgets (MABs), themselves based on Sector Strategic Plans (known as PESEMs). MABs have introduced the initial framework for a shared strategic vision in the on- going overhaul of the budgetary process, emphasizing a poverty focus, output driven program budgeting, transparency in the use of budget information and, less but not least, political prioritization in the medium-term. MABs provide the bridge between a sound technical preparation of annual budgets and political priorities of multi-annual budgets. Rolling five-year 48 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE sector budgets are published by Executive Decrees and made consistent with also rolling three- year monetary programs. These allocations provide an indicative analysis of ex-ante expenditure ceilings reflecting inter- and intra-sector policy priorities at the formulation stage; and set referential medium-term broad performance indicators based on a few monitoring benchmarks. Revised and approved before the annual budget proposal to Congress is actually sent by the Executive, MABs are intended to be living processes, adjusted in consistency with macro aggregates certified by the Central Bank, and supported by inter-ministerial teams, under the leadership of the Ministry of Economy and Finance (MEF) and the Central Bank. 3.19 Revisions by the new Authorities of the 2002-06 budget also reflect a shift toward pro-poor and decentralized expenditure, and a search for a more efficient safety net system The first 2001-05 MAB was modified to accommodate the difficult transition period during the electoral cycle and reflect the priorities of the Toledo Administration. Major changes in multi- annual budget allocations 2001-05 and 2002-06 were: (i) a reduction of about 10 percent in the overall expenditure amount; (ii) a significant reduction to defense and security outlays; (iii) a slight increase to health expenditure; (iv) a significant increase to the safety net programs grouped under the Presidency, supporting the new countercyclical Program "A TrabajaP' and the decentralization process; and (v) an electoral campaign-promised salary bonus of 50 soles to public servants, and a conversion of temporary to permanent teachers' hiring status. In sum, when compared to 1997-2001, the 2002-06 MAB projects a higher share allocated to social spending (adding up for about 56 percent of the budget), offset by lower shares allocated to defense and security (falling to about one fifth of the total budget) and to public administration (falling to 11 percent) (Figure 3.3). Such priorities are being revised under the national dialogue, but the intended shift toward pro-poor and decentralized expenditure looks irreversible. 3.20 Significant inter- and intra-sectoral pro-poor spending reallocation has also taken place between 2000 and 2002. The budget for defense and national security decreased from 2.9 to 2.1 percent of GDP. The budget for education, health and social assistance increased from 5.3 to 5.6 percent of GDP. Complementary to these major shifts, ESSALUD expenditure (non- targeted public health insurance administration) decreased from 1.5 to 1.3 percent of GDP. In a similar vein, the share of poverty reduction programs (consisting of programs with universal coverage--education and health-spending and targeted spending) to social expenditure has increased from 57 percent in 2000 to 60 percent in 2002 (Table 3.5). Table 3.5 Comparing Multi-Annual Budgets by Sector (Billion NS and %) Sector 2001-05 2002-06 Absolute Variation Agriculture 1,085 3.23 1,010 3.37 -74 Transport, Communication & Housing 2,154 6.41 2,062 6.88 -92 Mining and Energy 506 1.50 435 1.45 -71 Education 8,906* 26.49 4,662 15.56 -4,244 * Health 2,863 8.51 3,024 10.09 161 Presidency (Social Assistance) 4,701 13.98 6,745 22.51 1,994** Justice 586 1.70 553 1.84 -33 Judicial Court 804 2.40 711 2.37 -92 Defense 4,476 13.31 3,210 10.71 -1,266 Security 4,375 13.01 2,812 9.38 -1,564 Others 3,169 9.42 4,744 15.80 1,624 TOTAL 33,625 100.00 29,968 100.00 -3,657 * Reflects an overestimation of the education payroll originally budgeted in the MAB, but not approved, to hire teachers outside Lima by the Ministry of Education. **Includes projected support to " A Trabajar" and to regional and local governments. 49 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 3.21 MABs' framework, however, has four weaknesses when compared to international standards: Li Low technical content. The quality of MABs' projections have two important limitations: (i) they barely project budget sector institutional allocations, but excludes more familiar ones like economic or functional classifications; and (ii) their forecasting horizon-5 years-is too long to be credible; and its consistency with monetary aggregates (and ceilings) published by the Central bank is too broad. u Weak link to the annual budget MABs are produced 3 to 4 months before the draft-new fiscal-year budget appears, which should allow time for fine-tuning, but yet they are not the main basis for formulating or determining actual modifications to annual budget allocations. Moreover, broad institutional allocations assigned ex-ante to strategic priorities per sector are modified ex-post throughout the year with little consideration to the original priorities; their link with the planning function of the Investment Office is non-existent; and, as a matter of fact, their actual benchmarks are so numerous and inadequate that compliance by public institutions is not regularly evaluated. o Fragile institutionalization. MAB's institutional framework, which was born as part of an unfulfilled Fiscal Law, is weak and should be strengthened. Because it is an internal MEF exercise, it bears fundamental review of such questions as: Why are MABs produced? Who are the users of this information? How is this information relevant to policy design and implementation? o Public's misperception about MAB's transparency. As a MEF-led exercise and despite the Portal's existence, MAB's transparency is also 'constrained by public misperceptions about the budget. The fact that a module on multi-annual budgeting is not yet part of the SIAF, and that mandated biannual evaluations of the MABs are neither regularly prepared nor published, contribute to such misperceptions. 3.22 In the medium-term, MABs should consolidate their role in determining the strategic priorities of annual budget allocations. To achieve this, authorities expect to carry out certain improvements to the MABs' framework from 2002 onwards. Recommendations: > The annual budget investment formulation is following an innovative participatory process and, from 2003 onwards, MABs will broadened to take into consideration regional priorities; > MAB's 5-year horizon should be reduced to 3 in order to make it consistent with the 3-year horizon macroeconomic program; > Revenue estimates should be opened to external and hopefully independent review to gain credibility; > MABs' link to annual budget priorities should be strengthened and mandated reports prepared twice a year to correct for deviations; > SIAF should design and implement MAB's performance monitoring indicators, and the Ministries' budget execution should regularly be monitored against their benchmarks; > MABs's reporting should be regularly published on the Web, to enable Congress and civil society to monitor and oversee expenditures-both programmed and executed-thereby ensuring both accountability and compliance with the annual budget. This is especially critical in light of national dialogue; 50 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE > As SIAF improves its monitoring indicator system, a very basic costing per unit arrangement system could be developed for selected poverty benchmarks; > Finally, projected performance-based contracts in the public sector by end-2002-starting with pilot social ministries, programs, or projects-will require SIAF-monitored benchmarks in accordance with agreements under the MABs. D. PROTECTING THE BUDGET AS A MEDIUM-TERM SOCIAL POLICY 3.23 Poverty reduction efforts in the nineties also included an innovative, but hardly known emphasis on protecting priority social expenditure. In the nineties, social expenditure in Peru was mostly pro-cyclical, such that their spending rose during periods of economic growth, but fell during economic downturns when such interventions were most needed. In this context, the fact that the last two GOPs have actively protected social spending in a fiscally retrenched economy, with total expenditure decreasing from 19 percent of GDP to 17.8 percent of GDP in the period 1999-2002 (about one-seventh in dollar terms), is exceptional. Indeed, the share of social expenditure to GDP, which evolved from 3.9 percent of GDP in 1990 to 7.0 in 1999 and then has remained steadfast up to 2002, is a positive outcome. 3.24 Social programs were highly pro-cyclical until 2000, but anti-cyclical afterwards, when as part of actions leading to an effective social protection strategy, authorities guaranteed floor budget financing for a series of Protected Social Programs (PSPs), initially conceived during periods of recession.8 This approach consisted of assigning a global budget allocation to non-salary current and capital spending for broadly selected 11 programs, distributed among eight sectors (functions): education (pre-primary, primary and secondary); health (collective and individual); social and community assistance; transport; justice and peace; water and sanitation; agriculture; and energy. In 2001, the minimum aggregate budget allocation was about US$1.0 billion, equivalent to 12.7 percent of the ordinary budget and 1.9 percent of GDP. Preliminary projections for end-2001 budget execution show that the GOP complied with such commitment by transferring an aggregate amount of US$1.1 billion, equivalent to 13.1 percent of ordinary budget or 1.9 percent of GDP, and slightly above its target despite recessionary trends in 2001. Such anti-cyclical behavior in social expenditures was atypical in Peru, particularly in light of the significant economic slowdown all year due to unfavorable external and internal events. Moreover, in 2002, the GOP proposed an aggregate allocation of 2.0 percent of GDP to the PSPs (Table 3.6). 3.25 Meeting PSPs' expenditure floors was met in 2001, but not without difficulties. Protected expenditure was maintained despite a fiscal retrenchment of about 0.8 percent of GDP. However, a few individual programs exhibited significant variations between target and actual expenditures, e.g., under-execution in education and over-execution in health, social and community assistance, and agriculture. Over-execution was mainly due to a moderate expansion of rural health programs and school breakfast programs, as well as unexpected expenses associated with natural disasters like the Arequipa earthquake, or prevention work required to deal with potentially adverse effects of El Nifio on agriculture production. Under-execution in education programs was needed to make room for converting teachers from contratados to regularizados by the Paniagua Transition Government. As the definition of the set of individually protected programs was too broad, this prevented to assess the efficiency of individual programs in accomplishing their ultimate outcomes. Other actions include: introduction of a countercyclical workfare program; improved targeting and consolidation of food/nutrition programs; and reduction of the public pension system deficit, while exploring options for expanding social insurance options for the poor who are currently employed in the informal sector. 51 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table 3.6. Budget for Protected Social Programs Non-Salary Current and Capital Budget (Million Soles) PSRL I PSRIL II Program Sector Target* Executed* Target** Approved* Code 2001 2001P 2002' 2002 026 Pre-Primary Education 132 66 230 160 027 Primary Education 467 152 229 255 028 Secondary Education 467 107 212 175 063 Collective Health 196 420 432 479 064 Individual Health 471 660 835 847 014 Social & Community 315 414 383 609 Assistance*** 052 Transport 686 503 NA 385 002/022 Justice & Peace 559 583 NA 655 052 Water and Sanitation 89 67 NA 79 009 Agricultura 85 606 NA 310 035 Energy 74 73 NA 54 TOTAL 3,541 3,650 2,321 4,008 % of Ordinary Budget 12.7 13.1 8.7 15.0 % of National Budget 9.9 10.2 6.5 11.2 % of GDP 1.9 1.9 1.2 2.0 Memo: GDP 189,800 189,800 201,002 201,002 Ordinary Budget 27,812 27,812 26,728 26,728 National Budget 35,712 35,712 35,772 35,772 * Includes ordinary resources; **Includes ordinary and external resources under restructured PSPs; *** Vaso de Leche resources are excluded and, under restructured PSPs in 2002, those corresponding to Comedores Populares andA Trabajar. Source: Bank Staff estimates supported by MEF data. NA: Non Applicable. ' forecast. "preliminary. 3.26 The Government has agreed to go beyond the initial countercyclical role of the PSPs-- i.e. providing budget protection when a downturn occurs-to sustain their spending floors in the medium-term while improving their focus, efficiency, monitoring mechanisms, and transparency. The ongoing restructuring of PSPs will be accompanied by expenditure shifts in broad sector functions, as both will have a central role in a poverty reduction strategy. Recommendations: In order to institutionalize this key social policy, the GOP has announced several intentions: > At the sector function level, to assign a broad aggregate amount of earmarked resources to education, health and sanitation, and social assistance. This amount is considered not only as a budget-protection tool when a downturn occurs, but as the sustained floor required to make progress towards Peru's medium-term development goals.9 The aggregate 2002 budget allocation for the three priority social-sector functions--education, health and sanitation, and social assistance-is at least similar to its 2001 level in real terms and shows an increase from 5.3 percent of GDP in 2001 to 5.6 percent of GDP for these sectors. In addition to this, in 2003, at least a similar aggregate expenditure floor should be agreed, in real terms, in the approved Budget Law. 9 Peru has defined a set of about 20 medium-term development goals. 52 Chapter III. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE > At the program level, reduce the number of virtually selected PSPs to six (Table 3.6), and group under three sectors: education (pre-primary, primary and secondary), health (collective and individual) and social assistance. The amount of aggregate earmarked expenditure to this revised set of PSP is set at 1.2 percent of GDP in 2002, seemingly down from the previous 1.9 percent of GDP in 2001.1o This reduction of 0.7 percent of GDP makes room for increased budget flexibility. In addition to this, in 2003, a similar aggregate expenditure floor should be agreed, in real terms, in the approved Budget Law. > At the project level, set monitoring Box 3.5 Peru: institutionalizing a Virtual Poverty indicators to specific projects grouped Fund: A Look at Best-Practice Uganda under the selected PSPs and, in a few The virtual Poverty Action Fund (PAF) is focused on implementing cases, add performance management the Government of Uganda's highest priorities within the Poverty contracts to improve their efficiency Eradication Action Plan (PEAP). The PAF is funded by a cobination of HIPC debt relief, donors--general or sector- (e.g., the Integral and Mother and Child earmarked-support, and Government's own resources. The Health Insurances-SIS-SMI). 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 separate Fund, but a subset of the overall > To introduce budget flexibility on budget; (iii) the Government commits not to cut its funds; (iv) all executed PSP expenditure, by including expenditures fall under full Congress and auditor general oversight; escae causs a thesecor uncion (v) it is managed by sector and ministerial level Working Groups; escapeand (vi) PAF programs have a structured and participatory level to make provisions for budget institutional framework that ensures that they are properly planned, under-execution or a major budgetary budgeted and implemented. This includes: Eligible Criteria: A program qualifies if it meets 4 criteria: it is part shortfall; and at the program and project of the PEAP; it directly reduces poverty (with involvement by the levels, by allowing internal reallocations poor); it delivers a service to the poorest 20" of the population; and from under-performing to over- has a well-developed implementation plan (with costs, outcome, and output targets clearly identified). Programs are reviewed once a performing programs or projects, but year.. maintaining global spending floors. Administration: 5 percent of PAF resources are set aside for improving program effectiveness and transparency; Size: at least a constant proportion of the original budget; > To design and develop a tailor-made Reporting: An overall quarterly report by central or local and user-friendly SIAF-supported governments should be produced by the Ministry of Fin ance and distributed countrywide. Biannual sector performance evaluations monitoring and evaluation tool. are also required. Audits should cover all Central Government-led and at least 60 percent of district-led programs. Sector expenditure > To consider the possibility of converting racking studies are required when there is inadequate audit information. Civil society independently monitors reviews. the de facto "virtually" protected budget Safeguards: (i) in case of budgetary shortfalls, under-allocation to into an Enhanced Virtual Poverty Fund. PAF should be lower than cuts in non -PAF programs; (ii) in case o uder-performance, funds can be reallocated to other PEAP This enhanced mechanism would programs or to PAF programs in the following fiscal year. consist of tagging certain budget lines Pvr that meet or approximate the definition of poverty-reducing spending (the so-called "virtual" fund) and draw principally from the experience of Uganda's Poverty Action Fund (Box 3.5). It is rapidly expanding through Highly Indebted Poor Countries (HIPC). 'I 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 1O The revised set of programs, however, increases its global allocation as a percentage of GDP from 1.0 in 2001 to 1.2 in 2002. The reduced level of protection in the revised set is the direct result of the narrow number of PSPs considered. 1 This ceilings for budget reassignments are required, on the one hand, to maximize the effective use of protected funds that could be left unused by under-performing projects in priority programs; and, on the other hand, to guarantee a minimum level of budget execution in priority programs. 12 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 is committed to channel cash-flow savings on interest payments resulting from the reduction of the stock of debt under HIPC toward PAF programs. 53 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION (Mauritania, Mozambique and Tanzania); or producing additional Table 3.7 Budget Transparency Ratings information on all poverty-reducing (November 2001) LAC spending (Honduras) (IMF, 2002). Peru Mean Often, virtual funds are believed not ArealScore A B A-B GnrlKnowledge (Overall Score) 3.4 3.6 -0.2 to be needed in countries that Formulation already have a system of program Legislative powers 3.8 3.4 0.4 Civil society participation 1.6 1.8 -0.2 classification, or are well on the way Budget is inertial 2.1 2.5 -0.4 to setting up such systems. Budget follows M-T priorities 1.6 2.4 -0.8 However, in the case of Peru, notModification H wvrintecsofPr,nt Legislative participation 3.0 2.9 0.1 only does such virtual Fund exists, Awareness of the civil society 1.7 2.2 -0.5 de facto, with the support of the Execution End-year impact evaluation 1.5 2 -0.5 SIAF and the decision of the GOP's Monitoring on physical targets 1.8 2.2 -0.4 authorities to tag protected Budget Transparency Indexes (General) 2.7 3.0 -0.3 Formulation 2.5 2.7 -0.2 spending, but also the Fund's recent Approval 2.4 2.6 -0.2 successful experience proves that Execution 2.4 2.6 -0.2 Fiscalization 1.9 2.3 -0.4 virtual tagging is not the only Data Quality sufficient condition for sustaining Trustworthiness of NE data 2.4 3.2 -0.8 pr-orepedtr n h eim Independence from executive 2.1 2.7 -0.6 pro-poor expenditure in the medium Supervision term. An Enhanced Virtual Poverty Openness to bidding prices 2.4 2.8 -0.4 Fund, tailor-made to Peruvian Knowledge of public salaries 2.4 2.6 -0.2 Trustworthiness of internal audits 1.7 2.6 -0.9 needs, would essentially require Comptrollers office fights corruption 1.9 2.6 -0.7 adding two new features: (i) the Sanctions for misuse applied 2.7 3.1 -0.4 Reporting Timeliness development of its institutional Formulation 1.9 2.0 -0.1 framework providing open access Approval 2.2 2.8 -0.6 Execution 1.8 2.3 -0.5 and reporting; as well as (ii) the Fiscalization 1.5 2.0 -0.5 definition of monitoring indicators Also includes Argentina, Peru, Mexico, and Chile. and evaluation mechanisms in a Source( Moron (2001) participatory way. Such a proposal would definitely strengthen a medium-term social policy with one of the finest tools for a solid poverty reduction strategy, supported by SLAFY's full development and expansion throughout the public sector. E. BUDGET TRANSPARENCY 3.27 There is no question that Peru's Authorities have made a huge amount of progress toward making the budget process more transparent and accountable to the civil society (Table 3.7). In 2000, a few social programs were first in posting on the Web their district-by- district expenditures. Then, two popular infrastructure programs--Caminos Rurales and Rural Electrification-also published on the Web the areas covered by their work. In February, MEF launched the Economic Transparency website (Portal de Transparencia Economica), including detailed information on budget execution. This provided citizens and civil society organizations with the necessary tools to begin exercising control over social and decentralized programs-to verify that expenditure outlays are actually spent as indicated; whether targeting is appropriate; and whether there appears any politically motivated wrongdoing in the use of public resources. Finally, the rapid expansion of SIAF throughout all Government will allow, in an unprecedented move for LAG countries, full disclosure of the budget of the ministries of Defense and National Security by end-2002. SCAF is also getting ready to support decentralization to subnational governments. 54 Chapter Ill. REORIENTING THE BUDGET TOWARD PRO-POOR EXPENDITURE 3.28 In spite of such efforts, Peru 3.28 In piteof uch ffots, eruFigure 3.4 Index of Budget Transparency still ranks lowest in LAC in terms of Average classification on a scale of ito 10 overall budget transparency. Comprehensive national surveys 8 developed in 5 LAC countries found that Peru ranks in fifth place below Chile, Argentina, Brazil and Mexico 4 (Figure 3.4.). The survey was developed by the Universidad del 0 Pacifico (Moron, 2001). It approached Chil I Argentina Mexico Peru Congressmen from all political parties, Source: Mor6n (2001) journalists, think-tanks, and NGO members. Why, in spite of multiple recent efforts, did Peru obtain such a low ranking? A first look at the multinational survey shows detailed features about Peru's legal framework and, more important, prevailing perceptions on budget transparency existing in the civil society. 3.29 Peru ranks favorable in terms of the legal framework surrounding budget transparency. The survey underscores existing norms that are in favor of the budget's public knowledge and mechanisms like the Portal, which promotes a transparent management of public expenditure, or SIAF, which improves budget execution. Budget formulation, though, faces one obstacle: inertial spending, which prevents an adequate fit between budget allocations and priorities. For its part, budget auditing is affected by the diminished role of the Comptroller's Office in controlling public expenditure. Furthermore, such Office has no legal authority to issue sanctions against public servants found guilty of wrongdoing. Finally, the survey also finds that no minimum reporting format requirements exist in SIAF, thus preventing a user-friendly reading of the budget database. 3.30 In spite of a positive legal framework, the major obstacle to budget transparency is that the public's perceptions remain skeptical in Peru. The budget transparency ratings allows a detailed reading of the issues found by the survey (Table 3.7), classified per stage: formulation, in-year modifications, and execution; or per area: overall transparency indexes, data quality, supervision and reporting timeliness. Our analysis of selected perceptions is associated either to a higher differential (gap) between Peru and the rest of LAC, or to a low score obtained in the survey:13 L Budget formulation. Perceptions suggest that the Peruvian Congress has above-average leverage to be able to influence budget formulation; and others affirm that the budget dbes not follow medium-term priorities. The first finding is surprising since about 69 percent of surveyed responses also consider that Congressmen do not have the capacity to assess the budget. " In-year modifications. The public's perception is that the Peruvian Congress has a higher participation in budget in-year modifications than it is assumed in the rest of LAC; but that. civil society's awareness of such adjustments is significantly below other countries. o Budget execution. Monitoring and evaluation of public expenditure in Peru appears as having a significant lag when compared to LAC regional terms. o Overall transparency indexes. The public's perception is quite homogeneous in the gap found associated to the different stages of the budget process, with the lowest score assigned to budget fiscalization. 1 Scores are the average total response in a 1-5 rating. 55 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION o Data quality. The public clearly mistrusts INEI data, as well as its independence from the Executive. INEI ranks low in regional terms. o Supervision. The public's perception assigns the lowest supervision ratings to internal audits and the Comptroller's Office. The degree of openness to bidding processes and sanctions for misuse of public funds are perceived as requiring improvements. o Reporting timeliness. Budget formulation shows a rating that is close to LAC. The public's perception points to significant gaps to fill in fiscalization, approval, and execution. 3.31 In the interim, the GOP is 3.1 I teitei,teffOst Figure 3.5 Percent of Goverment Entities making extraordinary effortsPublishing Information improve transparency. An internal survey developed by the MEF on reporting, and the degree of compliance with directives referring to budget transparency in all public sector entities, shows impressive progress in compliance (Table A25 Info on Investment and Figure 35). In 2002, the only remaining reports to be published on Info on Contracting the Web are: (i) MEF-the and Acquisitions consolidated balance of the public sector, the balance of the Fiscal Strategic Plans Stabilization Fund, and the evaluation of the financial and physical indicators; (ii) FONAFE-audits of IMMinistries 03CTARS 03superintendencies @Public Enterprises financial statements and results of the Source: MEF. evaluation of management indicators; (iii) ONP (Pensions)-consolidated financial statements of both the provisional reserve and the National Savings Fund (FONAHPU). Regarding the degree of compliance with directives, the only common area where generalized weaknesses are found is in the hiring of public employees and procurement contracts. For their part, CTARs and most public entities do not reveal their strategic plans; whereas public enterprises still do not reveal economic information related to their investment projects. 3.32 Recommendations. A set of additional actions is proposed to enhance transparency. E Strengthen training and advice with improved learning tools for budget oversight by Peruvian Congress and civil society. The recent creation of a joint MEF-Congress Commission of Economy is a positive step in the right direction. > The budget for internal auditing and for the Comptroller's office should be increased. Lack of due fiscalization and corruption-prone practices are recurrent themes in this report. > SLAF reports should be made user-friendly, 14 and nationwide capacity building for its use should be considered a permanent priority. > A special effort should be developed in outputs reaching populations outside Lima, especially in regard to transfers to subnational governments, as they will be critical in supporting the decentralization process. > Accomplish the ongoing official plans to make public sector publications fully available. 14 SIAF officials reported to have completed a Portal Amigable (special module) last August. 56 Chapter IV IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE 4.1 Does higher public expenditure buy better results in social outcomes? In Peru, there have been little empirical efforts to evaluate the evidence supporting such a common belief and assess the efficiency in the use of increased resources. However, it is commonly recognized that reducing unproductive outlays, increasing social expenditure, and improving its efficiency and quality are critical elements of countries that have achieved macroeconomic stability and have decided to pursue poverty reduction actively. Leakages and poor targeting are two critical shortcomings that affect the efficiency of pro-poor expenditure. On the one hand, a highly decentralized budget execution brings the possibility of significant leakages in the flow of resources toward the final beneficiaries of social program. Leaks may undermine social policy by their ability to prevent a critical input-public expenditure-to produce desired social outcomes. The introduction of an innovative tool-a Public Expenditure Tracking Survey (PETS)-reveals major leakages in revenue transfer mechanisms toward local governments and downwards, in.particular for the Vaso de Leche program, the most important food supplementary program in Peru. The same national survey indicates that the most significant leaks are not at the origins of the chain, between the Central and local governments, which is a very positive achievement for the financial management system, but at the mid- to lowest levels ofthe chain, as resources get closer to beneficiaries. Transfers to local governments are also compounded by insufficient audit control and poor transparency. On the other hand, poor targeting deviates pro- poor outlays toward the non-poor. Targeting rates in social programs vary tremendously in Peru and their degree of regressivity/progressivity provides a mixed picture. Household survey-based findings show that targeting is adequate for Desayunos Escolares, but inadequate for others like ESSALUD and Comedores Populares. This chapter first addresses the empirical evidence supporting the rationale for increasing expenditure in the education and health sectors briefly; then goes on length to fully describe the main findings of Peru's PETS; and finally assesses the quality of targeting in pro-poor expenditure in Peru. A. DOES HIGHER GOVERNMENT EXPENDITURE BUY B ETTER RESULTS IN EDUCATION AND HEALTH CARE? 1 4.2 Increased public expenditure in health and education improves both access to and attainment in schools, and reduce mortality rates for infants and children There is increased recognition that expenditure allocations in favor of education and health can boost economic growth, while promoting equity and reducing poverty. The rationale for higher public spending in education is often based on its impact on individuals' lifetime incomes (i.e., the social rate of return, highest in primary education, then in secondary). Similarly, the rationale for increased spending in health is justified on the basis that it reduces the impact of diseases on the productive life years of the population. Empirical evidence, however, has not been conclusive, especially if public resources are used inefficiently and inequitably, and public expenditure crowds out private spending on the social sectors. Using a model and a cross-sample database developed by Gupta, Verhoeven and Tiongson (1999), data from Peru are added to a sample of 'This section draws on Gupta, Verhoeven, and Tiongson (1999), who kindly shared the IMF database and allowed us to include Peru in their consistent dataset in order to facilitate the estimates leading to the findings hereby submitted. 57 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION around 50 countries, and ensuing estimation followed an OLS and two-stage least squares (2SLS) linear regression technique (Tables 4.la-b).2 Highly robust results for education and health indicators show that: c Total education spending has mixed coefficients: a statistical significant one as a determinant of enrollment rates in secondary education, but insignificant ones as a determinant of enrollment rates in combined gross primary and secondary education. In the 2SLS gross secondary education regression, the coefficient of combined primary and secondary education spending is also statistically significant.' Overall, the Fstatistics is significant at the 1 percent level. Other variables whose coefficients also appear as statistically significant are population, child mortality rate, income per capita, and urbanization. u Total health spending has a statistically significant effect on both outcomes, infant and child mortality. Its coefficient appears weakly significant in the 2SLS infant and child mortality regressions. Other variables whose coefficient appears statistically significant are: adult illiteracy rate and income per capita. Table 4.1a: Regression Results for Education Indicators: Linear Regressions' ENROLLMENT RATES Gross Primary and Secondary Gross Secondary OLS 2SLS OIS 2SLS (weighted (weighted)' (weightedj (weighted)' Constant 53.81 36.28 23.04 ***31.10 (42.65) (27.38) (18.92) (11.25) Primary and secondary education -.19 -.20 .12 *.22 spending (% of total educ. spending) (.30) (.31) (.11) (.11) Education spending 1.18 .80 ***2.43 *1.86 (percent of GDP) (1.47) (1.52) (.95) (1.08) Population aged 0-14 .17 .85 **-.76 ***-.77 (percent of population) (.98) (.77) (.37) (.26) Child mortality rate **-.12 ***-.21 ***-.01 *-.07 (per thousand of children 0-5 years) (-.06) (.07) (.03) (.03) Income per capita in PPP termi 1.82 -.086 ***3.07 **1.65 (2.20) (1.32) (1.16) (.81) Urbanization (percent of population) .53 **.71 ***.45 ***.45 (.37) (.32) (.15) (.13) Adjusted R-squared 53.78% 48.46% 78.15% 81.05% Number of observations 43 42 44 43 F-statistic ***17.45 ***12.29 ***60.40 ***81.64 P-Value 0.00 0.00 0.00 0.00 Source: World Bank estimates. a. Robust standard errors are in parenthesis: *** indicates significance at the 1 percent level, ** significance at the 5 percent level, and * significance at the 10 percent level. b. By adult illiteracy. c. Instruments used: aid in percent of government expenditures, military spending in percent of government expenditures, share of unallocated education spending and total government spending. d. Multiplied by 1000. 2 OLS are corrected for heteroskedasticity and 2SLS is used to address the problem of reverse causality (i.e higher spending on primary education may have a positive effect on enrollment, but a higher demand for primary education, reflected in higher enrollment rates, may also provide a push for higher spending). We also applied a third method of estimation, the seemingly unrelated regression (SUR) procedure, and our findings remained robust to changes in specification, instruments, and weights defined. In evaluating the regression results, it should be borne in mind that some degree of multicollinearity among variables affects the standard errors of coefficients. Fortunately, the variables for overall sectoral spending are generally not correlated with other independent variables, except health spending with adult illiteracy. White's technique is used to correct for heteroskedasticity and Sargan's test is used to assess 2SLSL specification. 3 Regressions do not permit to draw up conclusions about changes in the level of spending on primary and secondary education as opposed to the share of such spending in total expenditure. 58 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE Table 4.1b: Regression Results for Health Indicators: Log-Log Regressions' Infant Mortality Child Mortality OLS 2SLS OLS 2SLS (weighted)b (weighted)b Constant ***5.64 ***5.49 ***6.44 ***6.31 (.89) (.87) (1.13) (1.15) Health spending -.11 *-..14 .11 *-..14 (percent of GDP) (.09) (.08) (.076) (.07) Adult illiteracy rate ***.35 ***.39 ***.36 ***.39 (percent of population 15 or older) (.07) (.07) (.08) (.08) Income per capita in PPP terms **-.30 **-.34 **-.37 **-.39 (.12) (.13) (.15) (.15) Urbanization (percent of -.19 -.12 -.24 -.19 population) (.18) (.22) (.19) (.20) Access to sanitation .06 .08 .10 .11 (percent of population) (.10) (.11) (.12) (.13) Adjusted R-squared 75.75% 79.89% 77.97% 78.85% Number of observations 31 29 31 29 F-statistic 27.49 50.99 46.10 65.00 P-Value 0.00 0.00 0.00 0.00 Source: World Bank estimates. a. See footnote I in Table 4.1 b. Instruments used: aid in percent of government expenditures, military spending in percent of government expenditures, and total government spending. 4.3 The policy implications of such findings are obvious. Recommendations: > The GOP has made the right decision in deciding to increase the share of public resources allocated to education and health. However, caution is required in using these figures to estimate budgetary resources needed for achieving specific targets in social areas, as the verified lack of significance for some critical coefficients (e.g., for combined primary and secondary education spending on gross primary and secondary enrollment rates) confirms that there are also other important determinants to take into account. > There are mutually reinforced positive implications between social outcomes: e.g., a higher illiteracy rate has a positive impact on health outcomes; ditto for a lower mortality rate on higher education enrollment. > As the study reported on below shows, increasing resources is not enough. Improving efficiency through the reduction of leaks and improvements in targeting in order to assure funds actually reach their intended beneficiaries is critical for achieving desired outcomes. 4.4 Further analysis shows that the efficiency of Peru's public expenditure is below Latin America's average for health, but above average for education, and in relationship to world averages, similar results holds true, but these indexes should be taken with a grain of salt. 13 Peru's level of efficiency in reaching health outcomes is below both Latin America's and the World's averages. Two indexes-technical and relative-measure the efficiency of public expenditure on health among Latin American countries during 1990-98 (Table 4.2). These indexes combine five outcomes: life expectancy, infant mortality, under-five mortality, height-for-age malnutrition, and weight-for-age malnutrition. The technical efficiency index varies from 76 to 94 percent, well below LAC and world averages in most health indicators, particularly in malnutrition-height, under-five mortality, and life 59 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION expectancy. The relative efficiency index is below 100 in most cases (except malnutrition- weight), which reflects that compared to LAC benchmarks and given its inputs-health spending and a time-trend (a proxy for technological progress)-Peru could do better in improving its health outcomes (Jayasuriya and Wodon, 2001).4 E Outcome indicators for Peru's educational inputs look acceptable in comparison to LAC and world averages, but severe quality and equity shortcomings remain. Net enrollment rates in primary and secondary education are the outcome indicators for the education sector; and education spending, adult literacy (which is not an outcome, but an input since we are looking at performance in primary and secondary education), and a time- trend are inputs. Peru's technical efficiency index for net primary enrollment is 77 percent, on par with other Latin American countries; whereas for net secondary enrollment is 72 percent, much above LAC average of 51 percent. In spite of this progress, severe educational shortcomings are still relevant in terms of the very low quality of education services and teacher's qualifications, especially in the rural areas; and in terms of the high inequality of resources devoted to private and public education (Apoyo Institute 2002b, Wodon 2002). Table 4.2 Peru: Efficiency of Public Expenditure in Reaching Social Outcomes 1990-1998 Relative to Relative to Relative to Technical Technical Technical Efficiency LAC World Efficiency LAC World Efficiency LAC World Life Expectancy Infant Mortality Under Five Mor(ality Peru 81.6 97.7 100.9 94.5 98.6 98.8 89.9 98.1 98.3 Latin America 83.4 103.2 95.9 100.3 91.6 100.2 World 80.9 96.9 95.6 99.7 91.4 99.8 Malnutrition -Height Malnutrition - Weight CombinedHeaith Peru 76.2 89.2 91.5 94.3 101.3 107.7 87.3 97.1 99.5 Latin America 85.4 102.5 93.1 106.3 89.9 102.4 World 83,3 97.6 87.6 94.1 87.8 97.6 EDUCATION Net Primary Enrollment Net Secondary Enrollment Combined Education Peru 77.1 100.9 104.1 71.8 141.0 132.8 74.4 116.9 116.2 Latin America 76.4 103.1 50.9 94.2 63.7 99.4 World 74.1 97.0 54.0 106.2 64.1 100.6 Source: Jayasuriya and Wodon, 2001. The value 100 reflects the average for the LAC region. B. TRACING LEAKAGES OF PUBLIC FUNDS IN PERU-A PUBLIC EXPENDITURE TRACKING SURVEY5 4.5 This section focuses on the leakages of public funds through municipalities (districts) in Peru. "Leakage" is defined as the portion of public funds that do not reach their ultimate targeted beneficiary, but instead is diverted for other purposes, including private gain or 4 The Technical Efficiency index values depict country level efficiency of spending, with a value of 100 indicating that a country has reached the maximum possible outcome given its inputs. The Relative Efficiency index measures each country's spending efficiency relative to a Latin American regional benchmark efficiency, with the benchmark being the straight average for the countries in the sample. The inputs taken into account for the analysis are the level of health spending, adult education levels (as measured by the share of the adult population which is literate), and time. Sensitivity tests have been performed with additional input variables, such as GDP per capita, but the relative ranking of various countries does not change much. 5 This section draws on the findings of a national survey jointly developed with Apoyo Institute (2002a). 60 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE other potentially legitimate, but clearly unintended purposes. This particular study of leakages is different from the study of corruption per se. Studies of corruption examine the bald diversion of public funds and the taking of bribes by public officials that are both clearly illegal and fraudulent in intent (Rose-Ackerman 1999; Seligson 2002; Treisman 2000). Research on leakages, instead, begins by asking the question: Why do public expenditures often not produce concomitant increases in social outcome indicators?' While there are many factors that go into the answer to that question, only recently has it been appreciated that part of the explanation lies in the fact that institutional factors, as well as local organization constraints, or private gain prevent that some public funds never in fact reach their intended targets. This "leaking away" of public funds in Peru is the subject of the present investigation. 4.6 Work on leakages of public expenditure is in the pioneer stage worldwide. It builds on the seminal work developed by the World Bank in Africa, more particularly in Uganda. Reinikka and Svensson (2002) found that only 13 percent of the non-wage expenditures made by the central government were received by the local schools in Uganda. The study of Peru deepens the approach followed in the Uganda study, mainly because it is able to trace leakages at each level in the chain from the first emission of public funds at the central level, down to the consumer at the level of the household. The Uganda study looked only at the national/individual leakage, and thus was unable to attribute leakages to each stage in the chain. As a result of this more comprehensive and disaggregated focus, surprising findings presented here emerge, especially because it is possible to identify the specific steps where main leakages occur and quantify them. Moreover, the Peru PETS, because it pinpoints the locus and key causal factors responsible for the leakages, gives policy makers clear direction for dealing with the problem.' In doing so, it looks at how resources are procured and distributed, and examines both the provider and household behavior, which allows to identify how much the government spend on the wrong goods or wrong people, and infer about those instances of the chain where a reasonable presumption of corruption could be pinpointed as "worst offenders." Finally, the Peru PETS is preceded by a thorough discussion of the system of Central Government transfers to the municipalities, and its shortcomings. 4.7 The approach taken to measure leakages is to employ survey instruments at each level in the process of transference of government funds from the central authority down to the household. The study employed data on 120 municipalities of the 1828 municipalities in Peru. Data were obtained from the Central Government on four transfers to municipalities managed by the GOP: FONCOMUN, Canon Minero, Canon/Sobrecanon Petrolero, and the Vaso de Leche Program. Since all but the last of these programs do not extend below the level of the municipality, the deepening of the research on leakages is on the last of the four, the so-called "Glass of Milk" program. It is in tracing the flow of funds in this program that the research attempts to make its most innovative, but not unique, contribution. Using survey data at the level of the municipality, at the level of the local milk distribution committees, and, finally, at the level of the beneficiary household, it is possible to trace the flow and leakage of central funds from the top of the chain to the last link at the bottom. The methodology is very complex, not only because it involves multi- level comparisons, but because the input itself is transformed from cash to commodities as the funds move from the top to the bottom, and as "the commodity itself' 6 A recent parametric model (SIMSIP) developed by the World Bank (Wodon et al., 2001) to estimate the projected impact of fiscal inputs on selected social outcomes, especially the Millennium Goals, does not take into account country differences in leakages of public spending in its estimates and assumes them constant in its projections. If developed regularly and in several countries, PETS could not only modify SIMSIP results, but set baselines for countries with similar levels of leakages. 7 The study gets as detailed as to identify the entities presumed as "worst offenders" in producing leakages (and their estimated amount)! 61 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION actually becomes commodities, since the program is not limited to milk or milk products alone, despite its name. The product is then transformed at the household level, as the food products are mixed with other foods before being served. Yet, despite this complexity, it has been possible to determine the relative magnitude of leakages at each level. 4.8 The survey findings do send up an important warning signal: leakages in Peru are significant and far more pervasive and extensive at the bottom of the chain than at the top. From the entire amount of public funds intended for the Vaso de Leche program, barely 29 percent get to their intended beneficiaries. This does not mean that 71 cents from each dollar are fully lost in corruption costs. Rather, the diverted resources get leaked away through a combination of off-budget administrative costs; expenditure on non-eligible products; h-kind deliveries to non-beneficiaries; fees for overpriced items; and, last but not least, sheer corruption. Results also challenge the predominant view of the last decade that organizations that are closer to the people perform necessarily better in service delivery. This is not necessarily the case if the program does not have a proper design or if the local organizations are not transparent and do not practice accountability, which seems to be the case of the Vaso de Leche Committees in Peru. These are hybrid organizations where both the government and the beneficiaries are represented. They are made up of three government representatives and three representatives of the beneficiaries of the Program, democratically elected by the rank and file. The government representatives are the Mayor of the municipality; another municipal official; and a representative of the Ministry of Health. Additionally, a representative of the Association of Agricultural Producers of the region is entitled to participate in the Committee. The relationship between these committees and the direct beneficiaries is characterized by at least two features. First, there is asymmetric information: final beneficiaries have limited access to the information about the decisions made by the committees. The beneficiaries also have limited information about how much resource they are entitled to receive from the Program and which procedures should they employ to secure them. Second, there is lack of transparency and accountability of the committees vis-a-vis both the beneficiaries and the upper echelons of government. In addition to these two features there likely is, in many committees, a somewhat low level of management capacity. The fact of the matter, anyway, is that committees so dominate the running of the Program at the local level that they may divert resources from their original purpose, without being held accountable or sanctioned for doing so, since both the higher official authorities (say, the MEF) and the intended beneficiaries do not know about it. The committees then dispose of the resources at their own discretion and sometimes end up vitiating, even unwittingly, the Program's expected effects. This arrangement is, of course, a far cry from the desirable participatory setting where the citizens of a community could directly observe, talk to, and even argue with, those providing them key services and where they would be able to hold those individuals and institutions accountable for their actions. On the contrary, in the Vaso de Leche Program we have the case of hybrid committees (made up of both government and elected representatives) placed in direct control of a development program, which, due to lack of accountability and transparency, can distort its goals and/or become rent-seekers benefiting not the collectivity but their own narrow interests. This, of course points to the fact the Program has severe design problems. These committees are beyond the common citizens' reach and are frequently dominated by self-serving, rent-seeking, self-styled "representatives" of the Program beneficiaries. The evidence amassed in this study enables us not only to directly compare diversions (referred to here as "leakages") of public resources for private gain or for a distorted purpose at each level of the public assistance "food chain," but also to conclude that, in this case, the lower we go "down the chain," the greater the diversion. Thus the conventional belief that every local body is necessarily more accountable than the national and public authorities is turned on its head. 62 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE 4.9 Anticipated main findings above presented deserve some caveats. The survey was conducted in a country that in the past three years has undergone a restructuring of the way public expenditures are ninaged and controlled by the integrated financial management system (SIAF) (Chapter III). For that reason, it may be the case that our central finding is one that is not easily replicated elsewhere. SIAF allows that nearly all central government expenditures get to the municipalities for which they were intended, and our findings confirm that most arrive without extensive delays, although their volatility and supervision remain somewhat of a problem in some instances. Where the leakages are extensive, however, is below the level of the municipality. Thus, in Peru, the leakages increase exponentially as central funds move away from the capital and move down to the committees and families. Future studies in other countries in which central funds are less well controlled may well find that leakages remain extensive at the local level, but perhaps would be far greater at the upper levels than they are in Peru. In both cases, leakages are extensive and have serious negative implications for development. In addition to this, a tracking survey is needed because a municipal SIAF is not installed in all municipalities yet. So, even though transfers from the Central Government to municipalities are well recorded, there is no further virtual control of what happens inside the municipality once it receives its transfers and moves it downwards. PETS methodology does not allow measuring other types of leaks such as inflating the prices of milk products, but indirectly through comparison of price variation for the same milk or milk-derived product, including overpricing with respect to supermarket retail prices. B.1 The Significance of Intergovernmental Transfers 4.10 . The Government of Peru has committed itself to improving the efficiency of its social spending and the quality of the provision of social services at the local level, including its nutritional programs. It has recognized that an increased social spending needs to be decentralized and has realized that this implies delegating more budgetary responsibilities t> Regional Units of Ministries (particularly Education and Health), and efficient mechanisms to transfer resources to local governments. While these improvements are important, they may not necessarily translate into actual increases in public funds reaching their intended destinations. That is because the GOP lacks a solid baseline to evaluate the quality, efficiency, and efficacy of public expenditure below the national level. In fact, very little is known about how resources are channeled (particularly outside of Lima), and even less is known about how much of these resources initially allocated actually get spent in their original purpose, what percentage really reaches their intended beneficiaries, or what are the magnitudes of transfer delays. Supervision is also very poor. For instance, we found that in 78 percent of the municipalities visited the central government had not carried out any supervision regarding the use of resources in the Vaso de Leche program. Moreover, in rural areas, only 14 percent of the municipalities had any supervision. 4.11 This study applies Public Expenditure Tracking Surveys (PETS) (i) to detect, analyze, and quantify the leakages and delays in the transfer of public expenditure as well as (ii) to assess their effects on municipal service delivery deficiencies. In this section, we deal with the latter topic. Then, in next section, a quantitative PETS is developed collecting information currently non-existent and otherwise very difficult, if not impossible, to systematically obtain. The information is surveyed at the different levels involved: central government, decentralized government unit (if any), final service units as well as any intermediate units. 4.12 In Peru, public resources are distributed by two mechanisms: 1hose that are centrally allocated and administered through branch offices of the central government, and those that are transferred to local governments (municipalities). The education budget is an example of the 63 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION former, while the Vaso de Leche program is an example of the latter. In both cases, there is a considerably long chain of intermediaries between the original central government budgeting office and the intended recipient. Findings presented here focus on the latter mechanism, in contrast to the work in Uganda. 4.13 It is difficult not to overstate the importance of transfers to municipal governments. For the districts outside of Lima, transfers-on average-represent 72 percent of their total income and, among the districts of the poorest stratum, they can represent in excess of 90 percent of total income. Central government's main four transfers include FONCOMUN and Vaso de Leche (VdL) for all municipalities, and Canon Minero and Canon/Sobrecan6n Petrolero for provinces and districts in regions where mining and petroleum products are extracted or the mining and petroleum company headquarters are located. 4.14 In 2001, these four major central government transfers totaled 1.9 billion Nuevos Soles (roughly $560 million).! Total transfers to municipalities in 2001 can be broken down in three major ways: (i) by separating Lima from the rest of the country; (ii) by dividing the sample into urban vs. rural and population size; and (iii) by level of poverty (Annex D). o The largest of the four transfers is the Fondo de Compensacidn Municipal (FONCOMUN), which accounts for 1.4 ($413 million) of the 1.9 billion Nuevos Soles in 2001 (in some districts it represented above 90 percent of total income). o The second largest central government transfer is the Vaso de Leche transfer, which totaled 330 million Nuevos Soles ($97 million) in 2001. By law, approximately 7 percent of public social spending in Peru is dedicated to nutrition programs. Much of this effort involves the Vaso de Leche program. This transfer, unlike the others, is earmarked specifically for use in the purchase of VDL products. This program is very important: excluding Lima, the municipalities in our survey reported a total of 645,346 direct beneficiaries; or, expanding this to the national population, equals 3,693,406 (2,207,209 being children between the ages 0 to 6) which would suggest a rough coverage of 92 percent for children between the ages of 0 to 6 pregnant, and nursing mothers. o The third largest of the four major transfers is the Canon/Sobrecanon Petrolero that totaled 128 million Soles (roughly $37 million) in 2001. The importance of the total figure is misleading, however, at the local level. For municipalities that are eligible to receive this transfer, it can represent as much and, in some cases, more resources than the FONCOMUN. L Of the four transfers, only the canon minero is not variable month-to-month. It is the fourth largest transfer with about US$24 million distributed in 2001. 4.15 How meaningful are these transfers to the individual Peruvian? On a per capita basis, FONCOMUN transfers-the largest of the programs-average $8.57 in Lima per year and $18.61 per year in the rest of the country. In a country in which the GNP per capita (Atlas method) is in the neighborhood of $2,080, the largest of the transfers (FONCOMUN) amounts to no more than nine tenths of a percentage of GNP per capita. A similar comment applies to the canons. Yet, these calculations are somewhat misleading since the funds are designated for the poor-not the entire population; and, since the poor receive the highest portion of the transfers, transfers are higher on a poor per-capita basis. In addition, such comparisons are misleading in the case of Vaso de Leche. The cash value of those funds is not the only factor to consider as the I A comprehensive and detailed description of the transfer mechanisms is in Annex D. 64 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE transfer provides, in theory at least, key nutritional supplements for children, whose nutritional status during childhood could impact their future health and productivity. 4.16 Expenditure on intergovernmental transfers shows a significant degree of progressivity. Using Lorenz curves, the highest degree of progressivity happens with the distribution of the canon minero, followed by Figure 4.1 Lorenz Curves for Intergovernmental Transfers FONCOMUN and Vaso de Leche, which exhibit almost a similar - distribution as a social program with universal 2 coverage (Figure 4.1). These results are consistent with the laws that govern them, as well as with the findings of our survey. According to the 0% 20% 4 60% 80% 100% legislation, all but canon Cumulative population (poverty 20-tiles) petrolero transfers are to be CVasoLeche - - Canon Foncom - 45 distributed according to per capita population, adjusted for poverty levels.'o This is especially so for the FONCOMUN allocation formula, which counts each rural resident (who are usually the poorest in the country) twice as much as each urban resident. This should mean that the transfers would be higher in the rest of Peru than in Lima, and also would be higher in the more impoverished areas than in the less poor areas. In practice, the FONCOMUN per-capita contributions are clearly far higher in the rest of the country than in Lima. The same pro-poor poverty bias appears in the canon minero and, to a much lesser extent, in the Vaso de Leche program (Table 4.3). Recommendation: > The formula (and criteria) used for determining the canon minero and FONCOMUN transfers do not need any change and should be kept. > As transparency about the amount and timeliness of transfers is the main issue, it is to the interest of the Authorities to have them fully available to the public and the direct beneficiaries through SIAF and the Portal de Transparencia. 4.17 Volatility of most transfers is significant. One of the most serious long-term problems faced by local governments in Latin America is the consistent lack of reliability of central government transfers. In many countries, such transfers often result in arrears. Volatility is calculated as the standard deviation of the annual percentage changes in the transfer amounts. While volatility is not directly a leakage issue, it does make planning difficult and does cause suffering when milk and other foodstuffs are not &blivered on time. In Peru, using the new financial management system that is now in place, volatility has been reduced, but not eliminated, as the results of our survey show. In the worst case, volatility for the Vaso de Leche transfer, outside of Lima, often exceeds 10 percent with the poorest districts averaging over 15 percent (Annex D). 9 Since these are progressive distributions, curves are above the 45 degrees line. Otherwise, standard regressive Lorenz curves are depicted below it. 10 The canon/sobrecanon petrolero is distributed by other criteria, but by introducing the urban/rural factor, indirectly takes into account poverty as p art of the criteria. 65 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Recommendation: > SlAF expansion among municipalities should continue, but alternatives to diminish the remaining volatility of transfers should be explored. Structural volatility in the canon minero transfers (related to international mineral prices) may require a Stabilization Fund (Chapter VHI). Table 4.3 Per Capita Transfers to Municipalities in 2001 (in U.S. dollars) FONCOMUN Canon Minero Canon/Sobrecanon Vaso de Leche Petrolero PERU 15.35 1.20 12.51 3.73 Lima 8.57 0.09 NA 3.99 Urban 8.33 0.09 NA 4.00 Rural 25.24 0.19 NA 3.33 No. ofobservations 177 171 NA 177 Rest ofPeru 18.61 1.89 12.51 3.60 Less poor 14.38 1.55 10.97 2.96 Poor 18.94 2.07 11.16 3.54 More poor 22.54 1.99 19.47 4.35 Urban 15.46 1.54 10.22 3.14 Rural 22.73 2.25 17.37 4.21 Small 31.97 1.84 48.15 4.37 Medium 20.05 1.77 19.40 4.13 Large 16.28 1.92 10.62 3.39 More accessible 17.33 1.81 9.81 3.39 Less accessible 23.72 2.15 20.90 4.47 Non-provincial capital 16.98 1.48 11.95 3.73 Provincial capital 21.60 2.69 13.09 3.41 No. ofobservations 1641 1296 142 1641 4.18 Understanding of the amount of the transfers at the local level is insufficient and knowledge about the arrival day of the transfers is poor. Overall, most of the officials interviewed in the municipalities surveyed claim to have a reasonable understanding of the various transfer programs. However, the same is not true at the neighborhood level. For example, the survey found that 90 percent of the municipalities in the Lima area and 79 percent in the rest of the country claimed to know the allocation criteria used for the FONCOMUN program. Yet, the survey also found that when questioned, only 11 percent of the municipal cfficials in Lima-who earlier claimed to have knowledge of the criteria-actually did. In the rest of Peru, surprisingly, the knowledge was higher among those who claimed to know, as 67 percent actually did. As for the amount of transfers expected from FONCOMUN, the knowledge base is more reasonable as only 5 percent in Lima and 15 percent in the rest of Peru claimed not to know. In poor and rural areas outside Lima, however, this percentage of uncertainty increased to nearly one-third. In the case of Canon/Sobrecanon Petrolero transfers, there is considerable uncertainty as to the expected amounts, with the majority in the rest of Peru districts not knowing. Knowledge of the date of arrival of the transfers was far weaker in Lima with 40 percent of the municipalities not knowing. In the rest of Peru, 33 percent did not know. Similar percentages are found for FONCOMUN and canon minero (Table 4.4). Recommendation: > The GOP's decision to include monthly transfers to each municipality in a user-friendly format in SLAF from 2002, also including the methodology used for their estimates, should be 66 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE complemented with adequate permanent Table 4.4 Municipalities that do not know the arrival day of training and information to major and the ter local authorities (e.g., Vaso de Leche Canon Committees). Canon Sobrecanon E ..............................F o nc o n .M inero Petrolero ina 40 4.19 Supervision of transfers to Urban 42 NA NA municipalities is extremely poor. On Rural 38 NA NA average, only six out of ten of the No. Observations 20 NA. NA municipalities are audited with regards to FONCOMUN and canon minero, and barely est of Peru 33 35 40 Not Poor 27 29 2 about a third of those eligible for the Poor 49 63 94 canon/sobrecanon petrolero are supervised by Extreme Poor 38 30 40 some central government entity. Furthermore, Urban 61 87 35 the central government audits are reaching Rural 28 28 47 only the more accessible districts while leaving the poorer, mid-size, urban and more Medi 39 36 31 remote districts wholly unsupervised! (Table Large 45 62 44 4.5) CG supervision is not only rare, but also More accessible 18 21 15 irregular, as the bulk of audits in about 80 Less Accessible 66 66 64 percent of FONCOMUN and canon minero Non-Provincial capital 34 33 41 cases (43 percent for the canon petrolero) are Provincial capital 28 50 29 done on a yearly basis (Apoyo Institute 2002). Recommendation: Source: Survey among municipalities > Auditing procedures to districts should be overhauled, including training, management reform, and more periodic internal and external audits by the Comptroller's Office. This reform is particularly urgent in the case of the canon petrolero. 4.20 Leakages in the transfer Of Table 4.5 Municipalities that are subject to C supervision FONCOMUN and the canon minero (in percent) appear very small. This leak is defined as Canon / Canon Sobrecanon the percentage of transfer reported by Ministry Foncomun Minero Petrolero of the Economy and Finance that is Rest of Peru 61 61 34 unaccounted for by the municipality. Not Poor 75 75 .34 Leakages amount to 1.5 percent in Limaand Poor 27 29 16 Extreme Poor 52 42 58 0.5 percent in the rest of Peru for 3532 31 21 FONCOMvUN, which is the largest program, -Rural 67 65:, 55 but rise to 7.1 percent in the Canon Minero Small 67 program in Lima (essentially driven by to two Medium 30 28 31 tLarge 58 50 51 More accessible 178 71 23 Peru for this program. (Table 4.6) These so- Less accessible 26 19 46 called small leakages are tolerable and can be Non-Provincial safely assumed to be mainly due to reporting capital .62 61 30 errors (round-off) or simply. bad recollection Provincial capital 58 56 65 due to poor or non-existent records at the No. Observations 100 73 33 municipality. SCanon/sobrecanon petrolero figures provided by the MEF were not trustworthy and complete, so were not included, but point out to an area of priority reform. 67 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Recommendations: Table 4.6 Leakages in FONCOMUN and Canon Minero > SIAF proposed expansion at the municipal (in percent) level should be able to trace these leaks on FONCOMUN Canon Minero a regular and virtual basis. LrbA 0.5 5.88 Rural 3.66 8.90 > Registry of the canon/sobrecanon Petrolero No. ofObservations 18.0 17.0 transfers must be overhauled. REST OF PERU 0.45 0.70 4.21 Important shortcomings surround Poor 0.11 0.03 the lack of evidence to determine how much More Poor 1.02 2.47 of the transfers of Foncomun, Canon Urban 0.00 0.00 Minero, and Canon Petrolewo actually go Rural 0.89 1.13 towards current expenditures. Last year Small 1.56 2.76 FONCOMUN law required a ceiling of 30 Medium 0.35 0.06 percent devoted to current expenditure (100 Large 0.00 0.00 percent devoted to capital expenditure and More Accessible 0.39 0.40 milk products in the cases of canon minero and Less Accessible 0.65 1.38 Vaso de Leche transfers). In 2002, Non-Provincial capital 0.83 1.10 FONCOMUN restriction was eliminated, and Provincial capital 0.00 0.00 the GOP left open to each municipality to No. ofObservations 96 64 decide on how best combine its resources. Survey findings show that the percentage assigned to current expenditure from FONCOpenN resources varies between 27- 41 percent, but such evidence Table 4.7 Fraction of Transfers Used for Current Expenditure is non-conclusive as a significant number of Unrestrictive Restrictive Definiti on b respondents (more than half De5.8ition8 in the case of the canons) did No. of No. of not know how the transfers _______ Percent Observations Percent Observations were used; other respondents FONCOMUN 41 61 27 61 accounted for more than 100 Canon/Sobre 35 16 29 16 percent of the transfer, and Canon PetroleroII yet about another one-third of Canon minero 8 45 0 45 respLesst Poord 0.000.0 respndens culd nly a Employee Payrolls (white and blue collar), Pensions, Road Maintenance, account for less than 70 Sanitation, Other Current Expenditures. percent of canon b. Employee Payrolls (white and blue collar), Pensions, Other Current minero/petrolero -funds (in Expenditures implicit violation of the law) (Table 4.7). However, these finidings show an important shortcoming in transfer mechanisms- that is, the readiness of municipalities for having any idea about the exact use of funds and complying with required current expenditure/capital ratios. Recommendations: >- If current expenditure/capital ratios are to be respected, present procedures for their auditing and accountability should be upgraded; otherwise, due to the lack of GOP's enforcement capacity, the present ratios should be eliminated and left open to municipalities to decide on how best to allocate their own resources. 68 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE B.2 Leakages in the Vaso de Leche Program 4.22 The Vaso de Leche program is the only one of the four main transfers to municipalities that can be traced from the top of the chain to the bottom. It targets as direct beneficiaries children six years old or younger, as well as pregnant and nursing mothers. However, the Law under which the program operates also allows for leftover resources to be used for children between the ages of seven and thirteen, the elderly, and those suffering from tuberculosis. The transfer criteria from the central government to the municipality are based on per-capita poverty formulas. At the municipal level, the local government is required, via special committees set up for the purpose, to use 100 percent of the funds into milk products, which must be overwhelmingly produced nationally. These committees are nearly ubiquitous, with 98 percent of the urban municipalities and 95 percent of the rural municipalities having them, according to the survey. The products should be purchased via competitive bidding, which is supposed to help insure employment of the lowest price criterion. However, the study found that while bidding was predominant, 19 percent of the products purchased were done through other mechanisms, and some excessively high prices were also found. 4.23 Despite its name, the program called Vaso de Leche appears in fact including milk, milk products, or milk substitutes, and other products such as oatmeal, quinua, and other grains. This flexibility in the program produces the unfortunate effect of reducing both the protein and calcium intake of the beneficiaries since milk and milk products contain the highest levels of these nutrients in comparison to grains. The fieldwork determined that only 15 percent of all municipalities distribute milk alone, with the vast majority "diluting" the milk with the distribution of cereal, a combination of milk and cereal, or distributing cereal only. Once these products are purchased, they are transferred to the next level down in the chain: to the local committees or clubs comprised of mothers, which are neighborhood or village-based volunteer groups. These local groups then distribute the "milk" on some sort of regular cycle (daily, weekly, monthly, bimonthly) depending on local circumstances, presumably based on the legal criteria mentioned above, as well as locally determined criteria for need. Within the recipient household, presumably the "milk" is then fed to the children and mothers for whom it was designated. As we shall see, much of the above is more theoretical than real, as the fieldwork for this research determined. 4.24 Leakages in the Vaso de Leche Program occur at many levels, but measurement of these leakages is an extremely complex task. Perhaps the major complexity emerges from the law itself. According to the law, the foods can be distributed to beneficiaries in prepared form. This could mean, for example, mixing of powered milk into a cereal or other cooked product. It would be virtually impossible for any study to then measure with exactitude how a given amount of milk input arrives in the stomach of the beneficiary. But, more importantly, from a practical point of view, distribution committees often cannot reasonably prepare the food since the beneficiaries are pre-school children whose parents cannot transport them on a daily basis to a central distribution point. Consider the mother who is nursing two pre-schoolers, and whose partner works outside the home. She cannot reasonably be expected to visit a central kitchen each day to feed her children. In addition to this, and more important according to our findings, the overhead costs of preparing the food, including distance, time, materials and spoilage for unconsumed food, deter many committees from attempting to follow the law. As a result, 60 percent of the committees in the sample -do not prepare the food and distribute it unprepared. For the purposes of the study, this is a plus, since it allows us to more precisely measure the distribution, since we can more easily count cans of milk, pounds of cereal, etc. However, it brings an additional challenge in that many of these products are marketed in units that are not 69 PERU: RESTORING FISCAL DISCIPLINE FORPOVERTY REDUCTION easily divisible. For example, if a household is Table 4.8. Beneficiary households that received entitled to 1.5 cans of milk based on the number of training/information children, the committee could not reasonably be (in esNo expected to open a can and divide it and pour the Urban 34 66 remainder into a glass for another beneficiary Rural 20 80 family. The result is that individual families will Stratum receive more or less than their exact ration of Milk Least poor 36 64 and other products, a factor which makes Poor2 calculation of leakages at the household level even Mo ccessible 8 more complex. Less accessible 15 85 More accessible 32 68 4.25 The problem of food distribution and Source: Survey among households in the rest ofthe preparation is exacerbated by the widespread countoyf February2002. absence of knowledge of the municipalities about the program, (as well as committees) and no effective training of the mothers. The survey found, for example, that only 20 percent of the municipalities in the rest of the country (and 43 percent in Lima) were familiar with the central government criteria for allocation of the "milk." At the level of the committee, barely two percent of them in the rest of the country outside of Lima have knowledge of the allocation criteria used by municipalities (and 5 percent in Lima). And, at the level of the mothers, only 27 percent of them reported having received training in the preparation of the "miWk and 26 percent reported receiving training on its proper allocation within the household. Most disturbing was that the level of training declined as poverty levels increased (Table 4.8), so that training was lowest where it was needed the most. "Milk" Leakage Stage : Central Government to Municipality 4.26 A first, and very smatl, leakage occurs during the transfer of the Central Government to the municipalities. u It represents on average 0.06 percent in Lima and 0.02 in the rest of Peru, which could be determined by rounding and recording errors. Thus, at the top level, where one often assumes the greatest level of corruption (and therefore the greatest leakage) the leakage is virtually non-existent. This is a major accomplishment for the SIAF system! However, considerable volatility remains in Vaso de Leche transfers outside Lima. Volatility in 2001, calculated as the standard deviation of monthly percent. changes of Vaso de Leche transfers, was virtually zero in Lima, but in the rest of Peru it averaged 11.6 percent, and a high of 15.4 percent in the poorest areas (para 4.19). Hence, the less accessible the area, the more volatility there were at this level. Despite this volatility, none of the municipalities in Lima and only 1.7 percent in the rest of the country were unsure of the amount of Vaso de Leche funds that they would be receiving. However, in the municipalities of Lima, 40 percent have no knowledge of the date in which they will receive their VdL transfer, while 31 percent of the districts outside of Lima claim the same problem, a surprising result given the relative simplicity of the transfer mechanism at the central government. In Lima, 21 percent of the municipalities suffered delays of seven or more days, while in the rest of Peru this level reached 25 percent. Furthermore, given that these delays imply that children and other beneficiaries are kept expecting food, a major basic need, the large percentage of municipalities whose arrival time variation is more than 7 days is quite serious (Table 4.9). 12 Exact formulas applied to estimate each leakage are in Annex D. 70 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE "Milk" Leakage Stage 2: Unaccounted for Conversion of Transfer to Products 4.26 Once the transfer reaches the municipality, a second leakage occurs when the funds are converted to products to be given to the local committees. From the municipal level onwards, the transfer of resources for the VdL program becomes in-kind transfers such that no subsequent stages of execution receive money but rather receive the transfer in-kind. Our field work team was instructed to get prices and quantities of VdL product purchases made by the municipality in December 2001 and to verify this information via signed contracts, purchase orders, or receipts. The quantities were in most cases obtained from the municipality's distribution roster (padr6n municipal), which includes the amounts allocated and distributed to each mothers committee within the municipality's jurisdiction. This leak is defined as the percentage of the amount transferred to the municipality from the Central Government for the month of December 2001 that is unaccounted for by the total expenses of the municipality for that month (in terms of products purchased for the VdL program). 4.27 Leakages found at this stage Table 4.9: VdL Transfer Schedule were also quite small. In Lima, they Municipalities Arrival time variations appear to have amounted to 3.03 percent wno o of the totals transfer, whereas in the rest entwrrivl da of Peru they amounted to 0.63 percent. (%) 1-2 Days 2-7 Days 7+ Days We say "appear" because of the larger LIMA 40 26 53 21 Urban 42 27 55 18 urban districts surveyed in the province Rural 37 25 50 25 of Lima-which all have populations No. Observations 20 20 20 20 exceeding 200,000-most refused to RST OF PERU 31 43 32 25 provide our team with any price NotPoor 25 0 29 71 information or price-related Poor 45 49 51 0 Extreme Poor 35 45 23 33 documentation. This refusal supports the Urban 57 0 43 57 qualitative information collected by our Rural 26 47 31 21 Small 29 58 0 42 team at later stages of the execution path, Medi 30 47 43 10 that suggest there is considerable misuse Large 42 17 64 18 of funds at the municipal level within More accessible 67 45 33 22 Less Accessible 13 41 32 27 these districts. We were, however, able Non-Provincial to document a number of worst-case capital 31 42 32 26 offenders. We found one municipality in Provincial capital 32 48 33 19 Lima in which this leak was 18 percent of Soervaong municialitie Sourc: Suvey mon unicipalities the transfers and another where it was 15 percent, again, keeping in mind that most larger municipalities refused to cooperate with us on obtaining this data. In the rest of Peru, we found 4 municipalities out of 76 surveyed in which the leakage at this stage was over 10 percent, with one reaching 15.5 percent. Thus, although the national averages are low, these isolated cases in which the leakage at this point exceeds 10 percent of the total transfer amount are serious. Without taking into consideration any of the leakages at subsequent transfer stages, the beneficiaries-mainly children aged 0 to 6-already are receiving less than 90 cents on the dollar. About one-tenth of all municipalities surveyed were found out to have leaks higher than 5 percent. In addition to this, one would have to consider the possibility of overpricing reflected in two facts: the high price variability found amongst districts for purchasing similar products, and the premium paid when comparing those prices to leading retail supermarket prices, even when adjusting them for quality and transportation costs. For instance, (i) the price of generic Enriquecido L6cteo, a milk substitute, distributed in 32 out of 100 districts visited, varies from NS/. 1-15 per kilogram; (ii) and the 71 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION registered price of cans of milk are in some cases outside Lima, twice higher than in a Lima supermarket! 4.28 Private gains are not the only possible reason of these leaks. One explanation for the leakages at this stage could be a diversion of VdL funds to cover the program's operating expenses (personnel, bookkeeping materials, transportation costs, and warehousing costs). Although prohibited by law, this kind of leak is not a result of a corrupt act. Indeed, the leakages at this stage are found more significant in small, rural, and less accessible districts. In many cases, it was found that in small rural districts, there are severe budget as well as personnel limitations that make the operating costs of the program prohibitive. Moreover, given the large and organized network of Vaso de Leche mothers representing a unified and powerful faction of the constituency that exerts considerable pressure on the mayor, it is no surprise that there may exist many cases in which the municipality supplements the CG transfer with municipal resources. Indeed, we find that oftentimes, leak 2 turns out to be negative, i.e., the municipality spent more in December 2001 than the amount allocated to it by the MEF. "Milk" Leakage #3: Transfer from the Municipality to the Local Committees 4.29 Leakages found from the municipality to local Table 4.10 Leakage #3: Municipality committees were more significant. In Lima, they averaged over to Local Committees 10 percent, but were far lower-only 2.6 percent-in the rest of Leak 3 Peru (Table 4.10). However, it is obvious from the results that Lima 10.06 the poorer, more remote areas have far higher leakages at this Urban 6.83 Rural 18.77 level. Every municipality has an allocation formula, based almost No. ofObservations 37.0 entirely on the size of the target population, that each Vaso de Leche committee services. Thus, criteria of relative poverty do Rest Pr 0.54 not play a role here, but only the number of poor counts. The Poor 5.67 roster of beneficiaries is centralized at the municipal level and More Poor 5.22 provides detailed information on the quantities distributed to each Urban 1.26 Committee within the district. This roster was used to randomly Rural 4.52 select four VdL.committees in order to verify the veracity of the municipal distribution roster. This information was compared to Small 2.83 the quantities that the committees visited declared to have ae 2.23 received from the municipality in the same period for every product distributed. This allowed us to calculate leakages More Accessible 2.31 associated with the transfer from the municipality to each of four Less Accessible 3.70 randomly selected committees. This leak was defined as the Non-Provincial capital 3.10 percentage of the amount listed in the municipal not accounted Provincial capital 1.97 for by the Vaso .de Leche committee and estimated using No. ofObservations 320 municipal and committee data computed at the committee level. 4.30 A clearer picture of the magnitude of the leakage problem that occurs in the transfer from local government to civil society is obtained by examining the worst offenders. The national averages do indeed hide very important information (Table 4.11). There are 27 districts/Vaso de Leche committee pairs (about a tenth of the total surveyed) with leakages in excess of 20 percent and 10 pairs that exceed 40 percent. In the case of such top-ranked worst offenders, the beneficiaries receive 36 cents of every dollar without taking into consideration all 72 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE the leakages in prior segments of the chain of distribution! 13 A possible explanation of these very high leakages is that in some cases municipalities may make changes to. the allocations to every committee, keeping one product already assigned for later distribution, but such informal arrangement significantly diminish transparency of the program and should be prohibited. "Milk" Leakage # 4: Committee to Beneficiary/Household 4.31 A fourth leakage occurs between committees and Table 4.11 Worst Offenders, beneficiaries. Estimation at this step became very difficult Leaent3 because when evaluating the situation inside the committee, we Rank ak #3 found that it is very difficult to quantify what happens to the Lima products distributed to beneficiaries. This is the case because the 2 57.4 committee representatives do not follow the criteria established by 3 48.2 the program regulation. Instead, they make decisions at their 4. 5 43.8 discretion as to how to proceed regarding the distribution of the 6 24.4 product. In most cases, the committee representatives have been est of Peru democratically elected and mostly rely on the approval of the 1 63.7 population of their communities. So, our methodology originally 3 53.2 contemplated the comparison of per-beneficiary rations at the 4 49.6 5 47.4 household level with the total per-beneficiary rations at the 6 47.2 committee level, but this was complicated due to the fact that 7 41.7 multiple products get distributed to beneficiaries and the only way 8 40.1 9 40.0 to aggregate them was to use a common measurable indicator. To 10 40.0 further complicate matters, in the cases of distribution of "prepared" 11 38.9 products, there was no way to gauge whether the servings-per- 12 34.8 13 34.0 container directive was followed and therefore there was no way to 14 34.0 measure the amount of raw product a household was actually 15 31.8 16 29.4 receiving, so we eliminated from the sample the cases in which the 17 28.6 product was not distributed in raw form. 18 27.8 19 27.2 20 26.7 4.32 Estimation of this leakage was done by calculating the 21 25.4 monetary values of each product (using municipal price figures) 22 24.3 and adding these up. This allowed a comparison of the monetary 23 23.5 24 23.0 value of the amount of all the products received by the Vaso de 25 22.6 Leche committee per beneficiary with the monetary value of the 26 22.3 amount received by the individual households per beneficiary 27 20.7 (excluding the committees that distribute prepare d products). The first variable would be obtained from the quantities declared by the mothers' committee representative in the VdL committee survey (in the four committees surveyed in each municipality). The second variable would be obtained from-.the quantities declared by the beneficiaries' household representative in beneficiary household survey (in 'the four households surveyed for each VdL committee). Although the implementation of the proposed formula faced several operational problems, it served to provide very important insights as to the distribution process to the individual households. Some of the problems that made it impossible to quantify the rations received by the beneficiaries are due to the very large variation of types, units, and frequencies of the distribution and products-which makes the program less transparent and thus more difficult to evaluate and supervise, and to products distributed already prepared (approximately 40 percent of the 13 It is important to note that this leakage was computed at the committee level with 320 observations. A lot of committees had a zero leakage and therefore much quite lower than that of the worst offenders. A complete list of worst offenders is found in Instituto Apoyo 2002. 73 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION committee cases). Although the municipality reported (in most cases) the number of rations that could be obtained by each package of product, the VdL committee did not necessarily follow the recommended recipe. Many committee representatives expressed that their objective is to try to service the largest possible number of recipients. Furthermore, the households receive a ration in a large variety of ways-as a cup, a glass, a handful, or just a ration. So, we standardized frequencies, units and products, and eliminated all cases in which the products were distributed in prepared form by the committee or in committees with unclear target beneficiaries. 4.33 The leakage at this level is quite high On average, over a quarter of the product is lost at this stagehousehold level, n percent) in Peru outside of the Lima area (for which we have no F__ I Leak 4 data as shown in Table 4.12). Leaks are markedly more I Lima NA serious in urban districts (34 percent), in provincial Urban NA capitals (40 percent) and in large districts (29 percent). Rural NA To further understand the subtleties of the program, No. ofObservations NA which made this leakage difficult to quantify, one must look at the law itself, which provides for an unnecessarily Rest of Peru 26.70 broad definition of its target beneficiaries. According to Less Poor 26.67 the law, indirect beneficiaries need not exclusively be Poor 321 young children, but may include children from 7 to 13 years, the elderly, and others in need-but only if there Urban 34.53 are enough resources left. The result is that this open- Rul 25.01 ended definition allows such a broad interpretation of eligibility that leakages seem bound to occur, which Small 24.41 causes confusion in the committees and in the population Medium 22.83 in general as to who are the intended beneficiaries. This Large 29.63 problem is further complicated by the indivisibility of the More Accessible 25.71 formula chosen and the ad hoc decisions made at the Less Accessible 28.32 discretion of committee representatives as to the criteria of distribution. These include number of household Non-Provincial capital 22.72 members; number of children; equal quantity for each Provincial capital 40.31 household; or other criteria that the study was unable to No. ofObservations 488 identify, many of which alter the originally estimated quantities per beneficiary in each household of the same committee and municipality. Leak #5: Within the household (dilution of the ration) 4.34 A fifth leakage happens at the household level. This leakage was estimated using household-level data. As a final stage, the fieldwork team visited four households per committee in order to quantify the amounts of the in-kind Vaso de Leche transfers that actually reach the intended beneficiaries. Because of the complications concerning the "target population" mentioned in the previous section, the analysis is restricted to direct beneficiaries only: children aged 0 to 6, pregnant women, and breastfeeding mothers. The leak attributed to "beneficiary dilution"' is defined at the household level as one minus the percentage of household members who consume Vaso de Leche products, who are official direct beneficiaries (Table 4.13). 4.35 Results make clear that, upon reaching the households, there is considerable dilution. On average, target beneficiaries only receive 41 percent of the ration that arrives at the 74 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE household (not taking into account all the losses associated with earlier Table 4.13: Leak 5 leakages)! This dilution effect is (Aithebeneficiaryhoseholdleve) possible, because in most cases the Leak 5 beneficiaries do not receive their rations Lima NA directly from the committee, but Urban NA because the children receive the rations Rural NA filtered through their mothers (and in some cases the father), who pick up the Rest of Peru 58.89% total rations allocated to her/his Not Poor 59.93% household for later distribution. Poor 57.89% Consistent with evidence in studies of Extreme Poor 59.15% other nutritional assistance programs worldwide, the official distribution Urban 59.26% criteria are very difficult if not Rural 58.70% impossible to enforce at this level. In Small 59.01% most cases, it is de facto impossible to 61.46% exclude non-targeted members of the Meium household. Furthermore, in about 60 percent of the committees visited, the More accessible 60.75% products are distributed in unprepared Less accessible 56.11% forms, which as noted above, is understandable since the transactions Non-Provincial capital 58.69% costs in receiving daily prepare rations Provincial capital 59.32% could be too high, but non-prepared frequently result in mixing the nutrition o. Srvaon 985 ration with the families overall food intake. In these cases, considerable variation appears in their final use. 4.36 In sum, the PETS survey l ea (reveals that targeted beneficiaries Fu 4 the ia s leV de LechLProgra get on average 29 cents of each 100% dollar initially transferred by the Central Government! The survey surprisingly indicates the leak is much 60 higher in the bottom (VdL committees 100.00% 99.98% 99N92% and households -leaks 4-5) levels rather 40% than in the top (CG and municipalities- 20 eaks 1-3) levels of the ladder. This o.2e5r% not~No Poor 59.93%ts igiicnt 0 not ony em nsraes siniicnt 0% Initilamount CGto Within Municipalities Committees Within improvements in the official financial MunicipaliiesMunicipalities to toHouseholds Households management of resources by the Committees SIAF/MEF, but amo challenges the predominant view that local private organizations are more accountable in managing resources than official organizations (Figure 4.2). Transfers appear also compounded by the generalized lack of audit controls, poor transparency, and volatility. Finally, leaks clearly affect the poorest, urban, and provincial municipalities more than others, but their level appears similar among districts of different sizes and distances to the province (Table 4.14). 75 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 4.37 Recommendations based Table 4.14 VdL Leakages on the PETS findings can be (in percent) grouped under two areas: those Leak I Leak 2 Leak 3 Leak 4 Leak 5 Combined referring to accountability of Lima 0.06 3.03 10.06 NA NA NA transfers and municipal Urban 0.03 2.73 6.83 NA NA NA Rural 0.11 3.58 18.77 NA NA NA management issues; and those No. referring to the VDL. Observations 20 14 37 NA NA NA Rest of Peru 0.02 0.63 2.59 26.70 58.89 70.84 Recommendations: NotPoor 0.00 0.13 0.54 26.67 59.93 70.81 Poor 0.00 1.36 5.67 19.21 57.89 68.34 Extreme Poor 0.12 1.30 5.22 32.91 59.15 74.39 > On financial management accountability. There is a Urban 0.00 0.42 1.26 34.53 59.26 73.77 need to (i) increase the Rural 0.05 0.85 4.52 25.01 58.70 70.70 transparency of transfers to Small 0.11 0.05 2.83 24.41 59.01 69.94 municipalities through a Medium 0.00 0.59 4.23 22.83 61.46 71.68 monthly report by SIAF, Large 0.00 0.84 2.25 29.63 57.90 71.29 accompanied by thorough Accessible 0.00 0.54 2.31 25.71 60.75 71.67 dissemination of their Remote 0.09 0.82 3.70 28.32 56.11 69.98 redistribution criteria; (ii) build local capacity to Municipal 0.04 0.87 3.10 22.72 58.69 69.35 Provincial manage them; (iii) reduce capital 0.00 0.21 1.97 40.31 59.32 76.25 their volatility, particularly No. among poorest recipient Observations 95 76 320 488 985 N/A municipalities; (iv) upgrade auditing procedures, as a priority, of municipal governments not only by the Comptroller's Office, but also by means of internal audits of the municipal administrations themselves; and (v) overhaul the registration of the canon/sobrecanon petrolero. > On the Vaso de Leche program. A major lesson to learn from the VDL experience is that a poorly designed social program, with a presumed high degree of participation of community leaders grouped in a committee, can be inefficient-voluntarily or involuntarily-and unaccountable to both its agents (constituent beneficiaries) and to its principal (municipal authorities), thereby missing the original purpose or intention of the program. And although there is a need to recognize that the dilution effect allows households to reach indirect beneficiaries and thus overcome rigidities in Central Government guidelines, the nutritional impact on direct beneficiaries is however diminished. Suggested actions can be grouped as follows: In the short term, suggested priority actions should focus on amending regulations to enforce accountability to the municipalities and committees: (i) review VdL regulations, particularly regarding products to be distributed and the form of distribution, so as to make the list of selected milk derivatives shorter and more homogeneous, and thus raise chances of improving the nutritional impact of the program; (ii) establish a proper registry of Vaso de Leche beneficiaries, if possible supported by SLAF; (iii) undertake information campaigns and training sessions to VDL committees and individual beneficiaries, to raise their awareness of new information available and the rules; (iv) undertake surprise audits of worst offenders (municipalities and committees) in the near term, so as to eliminate major deviations; and (v) establish a policy of no-budget increase in real terms of the annual budget for the VDL program. The implementation of the above actions requires a significant overhaul of the system, and should be accompanied by the redesign of a new comprehensive framework for food supplementary programs in Peru (discussed next). 76 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE In the medium term, the question is, given its failures and defective design, should the Vaso de Leche program be transformed to a cash-transfer program? The answer could be affirmative. Mexico's PROGRESA, which successfully moved its food supplement program to an integrated and better-targeted model of social assistance, is a good example. However, as this proposal goes beyond the sole issue of leakages and also addresses targeting, we will return to this subject at the end-of this chapter (Box 4.1). C. ENHANCING TARGETING OF SOCIAL PROGRAMS14 4.38 The GOP's proactive policy of targeted programs on the poor has been fundamental in the poverty reduction achievements of the past years. As a percent of GDP, total social spending has increased from 3.6 percent in 1993 to 6.9 percent of GDP in 2002. Commitment by the Authorities to a set of comprehensive safety net programs has been reflected in a multiplicity of safety net programs specializing in development and relief components. The development component provides permanent access to improving human capital accumulation of the poor in the form of health, education, and basic infrastructure services. The relief component provides a consumption floor for the poor through two types of mechanisms: temporary employment or cash programs, like A Trabajar-Urban and Rural-and direct transfer programs usually in the form of food, or basic services. 4.39 As a point of departure, the new authorities are not increasing resources to existing safety net programs, but rather attempting to improve the efficiency in their administration and targeting. The budget share assigned to extreme poverty reduction programs as a percentage of GDP increased from 1.1 percent in 1994 to about 1.7 percent in 1998, and has. remained constant up to 2002. This level is considered adequate to existing needs (The World Bank, 2000). Striking the right balance between the two types of safety nets, however, requires further assessment of their performance. Indeed, up to the creation of A Trabajar in 2002, direct transfers dominated the landscape of relief and safety net programs in Peru. At present, however, such a temporary employment program competes for scarce resources. Furthermore, ongoing unification of rural infrastructure projects under one administration-PRONAMACHS, INADE-is an initial step to reduce administrative costs and overlaps/gaps in coverage.'" This leaves targeting as a major issue requiring efficiency improvements. Fortunately, Peru already has a well-developed poverty map. 4.40 Broad geographic regional targeting remains regressive in Peru. While adequate targeting in social programs in Peru is a critical component to optimize scarce resources and deal with prevailing heterogeneity in program resources to reach the poor, it is also needed to offset broad mistargeting in regional public expenditure. Simple correlation coefficients between per- capita spending by department (and by CTAR) and poverty and extreme poverty rates produce negative results; whereas they should be positive, had they been allocated to the department with the highest poverty rates. Furthermore, correlation coefficients are positive between per-capita spending by department and poverty rankings, ordered from poorest to richest, whereas they 14 This section partly draws from the work done by Diaz (2001) "s Initial for other infrastructure projects remain, like COOPOP, Caminos Rurales and others from the Ministry of the Presidency. 16 Peru has an excellent FGT2-based poverty measure, which takes into account both the number of people below the poverty line, and the severity of their poverty (the gap between their incomes and the poverty line). The MEF map combined information from the 1993 census with a household survey conducted y INEI in 2000 to "impute" the consumption of households in the census. This in turn was aggregated up to the district level, and combined with information on the poverty line to estimate the number of households in each district below the poverty line (the headcount index), the poverty gap, and the FGT2 (square poverty gap). 77 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION should be negative (Table 4.15). Such severe misallocation is explained by the rigid components of the budget, particularly in current expenditure. Recommendation. > Broad geographic targeting of public expenditure should carefully be reviewed, particularly in light of the incoming process or regional decentralization, giving greatest weight to departments with the poorest populations. Table 4.15 Correlation Expenditure and Poverty by Department Expected direction if Correlation poverty targeted Dept. Expenditure per Capita to Poverty, by definition Poverty rate -0.522 positive Extreme poverty rate -0.463 positive MEF Poverty ranking 0.265 negative INEI Poverty ranking 0.458 negative INEI Extre me poverty ranking 0.507 negative CTAR Expenditure per Capita to Poverty, by definition Poverty rate -0.207 positive Extreme poverty rate -0.130 positive MEF Poverty ranking 0.397 negative INEI Poverty ranking 0.202 negative INEI Extreme poverty ranking 0.135 negative Source: WB estimates. 4.41 Access by the poor to social programs is in general progressive, but important differences prevail among individual programs. Poor household 1m access to social programs is high. A total of 82 percent of poor household have access to social programs (84 percent for those in extreme poverty) and 69 percent 40 access to at least two social programs (73 percent for extreme poverty) (Table 4.17). Poor rural households have proportionally 0 been more adequately reached by social 3O 405W000aty programs than urban households. Among FONCOMS --VASOMLECW- major social programs, FONCODES (excluding A Trabajar PESP-Urbano), joined by Caminos Rurales, remains with the best record in progressive spending, adequately reaching poor beneficiaries (Figure 4.3). PRONAA, however, has lost ground and joined Vaso de Leche in the mild progressivity of its programs. No doubt, their targeting has worsened. Finally, the program PESP-A Trabajar Urbano has started on the wrong foot. Its overall progressivity, particularly at the first four poorest levels of the populations (which goes beyond the population in extreme poverty), reveals that self-targeting of the most needy beneficiaries is failing. The 78 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE reason for this is to be found in the seemingly too high wage late being paid to Figure 4.4 Lorenz Curves for health attract workers from only the poorest expenditure deciles.17 100 Recommendation: 80 > A Trabajar should offer a wage rate that makes it only attractive to poor 40 households. To do this, the GOP should undertake an evaluation study, including 20 assessment of household targeting outcomes, and lower the wage should the 3 study indicate poor urban targeting J-Minuu --V-ESSALUD -Prfect eit eTotaIhealth outcomes. A similar study should be done regarding A Trabajar Rural. The study should be widely disseminated to Figure 4.5 Lorenz Curves for education support the GOP's decisions. expenditure 4.42 The main social sector programs 100 _______________ also reflect a mixed picture in terms of the 80. progressivity/regressivity of their 6 resources allocated nationwide. On the 4 one hand, health expenditure appears with 20- 901 significant regressivity, much exceeded by 2 ESSALUD whose services clearly focus on 0 10 20 30 40 50 60 70 80 90 100 the population located in the non-poor M Perfect equality PTota education deciles, but offset by MWNSA programs, I -Pimar education --Seonda education which show mild progressivity thanks to a few local service delivery programs that clearly target most needy beneficiaries Figure 4.6 Access to Social Programs, 2000 (Figure 4.4). On the cther hand, education (% with respect to total population) expenditure shows some degree of School text & materials progressivity, certainly reflecting a similar Family planning feature in primary education, barely offset by School insurance the mild regressivity of expenditure in secondary education (Figure 4.5). Checking childs growth Comeduor Popular 4.43 In general, safety-net programs Vaso de Leche reach less than 40 percent of their intended extreme poverty beneficiaries. This is the case of the Vaso de Leche and 0 5 1D Comedores Populares programs (Table 17 The current wage rate appears even more generous in the A Trabajar Rural, but this is less alarming because of the sharing income mechanism often being employed de facto by the rural communities, with members being assigned to a project, but actually sharing the working time, or the monthly wage or both. Therefore, the wage rate does not work as a tool for individual self-targeting the way it does in urban areas. Notice that the success of the program is not measured only by the share of beneficiaries found in the poorest income quintiles, but also by the income gains received as a result of participating in the program and the coverage of the program with a proportion of unemployed poor having access to the program, and the efficiency with which it operates (ratio of wages to administrative costs and material inputs) (Schady 2002). 79 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 4.16). Desayuno Escolar and School Text and Materials, however, are the exception, with an acceptable 51 percent and 44 percent of their beneficiaries belonging to the three poorest deciles. Seguro Escolar (School Health Insurance) ranks relatively better among health programs, whereas Planificaci6n Familiar (Family Planning) is found among the worst targeted programs. This ranking also broadly corresponds to the level of access-measured by the ratio of beneficiaries to the total population-that these programs have, with Comedores Populares and Planificacidn Familiar, having the lowest targeting scores (Figure 4.6). Table 4.16: Targeting by Food Assistance, Health, and Education Programs, 2000 No. of Extreme Non- Total Non Social Program Beneficiaries Total Poverty Extreme Poverty Poor Poverty Food Assistance Desayuno Escolar 2,972,859 100 51.1 18.7 69.8 30.2 Vaso de Leche 2,283,919 100 40.1 24.1 64.2 35.8 Comedor Popular 746,134 100 40.2 14.5 54.7 45.3 Health Regulating child growth 1,729,899 100 33.7 18.4 52.1 47.9 School health insurance 1,157,912 100 36.8 20.6 57.4 42.6 Family planning 870,942 100 22.1 24.8 46.8 53.2 Education School text and materials 2,970,567 100 44.4 19.7 64.1 35.9 Table 4.17 Household Access to Social Programs by Poverty Level, 2000 Extreme Non-Extreme Total Poverty Non Poor Geographic area Total Poverty Poverty Total 5,632,815 1,161,588 845,355 2,006,943 3,625,872 Beneficiaries 59.1 83.9 79.7 82.1 46.4 One program 13.4 10.5 17.2 13.3 13.5 More than one 45.7 73.4 62.5 68.8 32.9 Non-beneficiaries 40.9 16.1 20.3 17.9 53.6 Urban 3,607,764 266,243 607,221 873,464 2,734,300 Beneficiaries 50.8 81.9 78.8 79.7 41.6 One program 14.9 19.1 19.1 19.1 13.5 More than one 36.0 62.7 59.7 60.6 28.1 Non-beneficiaries 49.2 18.1 21.2 20.3 58.4 Rural 2,025,051 895,345 238,134 1,133,478 891,573 Beneficiaries 73.9 84.5 81.9 84.0 61.1 One program 10.8 7.9 12.3 8.9 13.3 More than one 63.1 76.6 69.6 75.1 47.8 Non-beneficiaries 26.1 15.5 18.1 16.0 38.9 80 Chapter IV. IMPROVING THE EFFICIENCY OF PUBLIC EXPENDITURE Recommendations: Box 4.1 MEXICO: Integrating Social Programs for Improved Efficiency. The Case > In the short term, application of a common of PROGRESA poverty targeting methodology and In 1996, the Mexican Government designed and criteria-based on the FGTZ poverty map successfully implemented PROGRESA, a single updated every 3 years-in all social and integrated social assistance intervention programs should be made more explicit, replacing a series of disparate food subsidies, transparent and focused on populations that education, health, and other social programs. are extremely poor, and be accompanied by PROGRESA's unified approach provides complementary training, information immediate financial transfers to the rural poor campaigns, communityn while at the same time promoting investment in campigns comuniy moiliatio or children's human capital (future earnings) via nutrition education. increased schooling and improved health and nutrition status. The program provides cash > Programs with substantial mistargeting transfers to selected poor rural families, selected should be redesigned, merged, or closed with household surveys updated every three To evaluate this regularly, benchmarks in years, conditional on keeping children in school terms of the share of expenditure reaching and providing them with basic preventive health extreme poor beneficiaries (lowest 4 care and nutrition. A key feature of the program quintiles) should be set explicitly for at least is the provision of the cash-transfer to registered qui tntin rgrm, n rors mothers, a mechanism designed to ensure that the the ten main social programs, and progressand as an toward meeting. such benchmarks should incentive to empower women in rural also be monitored on an annual basis with communities. In 2001, the program covered 3.2 SIAF and INEI support. As their evaluation million rural families (well over half of the rural indicates that programs do not meet their poor), at a cost of 2.3 percent of the government's objectives or overlap, their immediate social expenditures, or 0.2 percent of GDP. The restructuring will be necessary. Some of program is highly efficient with administrative them, like Vaso de Leche, already showing costs of about 4 percent; therefore, over 95 significant leakages and not meeting percent of its expenditure is transferred as cash nutritional objectives require immediate directly to poor households. majorSix years later, an independent evaluation majo resrucurin. .confirmed PROGRESA's impressive results, i.e., (i) increased primary enrollment rates by about 1 > In the medium term, a comprehensive percent, and increasing secondary enrollment restructuring of the main social programs rates by 8 percent for girls and 5 percent for boys; dealing with extreme poverty is necessary, (ii) children's educational achievement is perhaps following the example of estimated to have increased by about 10 percent, PROGRESA (Box 4.1) in Mexico. Given which would represent an increase in their future very limited fiscal resources, budget earnings of over 8 percent; (iii) increased prenatal rigidities, and low levels of efficiency for a care in the first trimester of pregnancy by 8 substantial amount of resources already percent; and. (iv) decreased incidence of disease devoted to extreme poverty programs, this among children under 5 by 12 percent, as did the seems rather an urgent task. The probability of malnutrition among children restructuring of social rprograms in Peru should aim at several goals: (i) improve the quality of service delivery; (ii) expand the coverage of social programs, particularly among most vulnerable rural groups; (iii) generate fiscal savings that could be used to increase coverage of most effective programs at the local level. 18 A recent study has found that the educational impact of Desayunos Escolares in terms of rate of attendance, rate of repetition and drop -out rates is small, but not negligible: about 10 percentage points in attendance, between a 1-3 percentage point drop in repetition and a 0.3-1.6 percentage reduction in dropout. 81 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION > Restructuring Peru's feeding programs into a unified intervention along the lines of PROGRESA (or, more ambitiously, folding other programs into this unified structure) is a very viable option that Peru might consider. This offers a broad coverage of the poor under a single program structure, with proper targeting and very powerful positive impacts on poor families. Rough calculations based on PROGRESA's costs per beneficiary ($80 per year) indicate that to cover the entire rural poor population of Peru (4.5 million people) with a similar unified intervention (primary education, basic preventive health, and food supplement) would entail a maximum cost of $360 million-or 1,200 million soles (0.6 percent of GDP). Interestingly, this amount represents about a third of total budget expenditure devoted to extreme poverty programs, just over three times the current cost of the Vaso de Leche program, or the combined sum of the Vaso de Leche, Comedores Populares, Desayunos Escolares and Seguro Materno-Infantil programs. Obviously, if the selected target population would be the extreme poor, then the cost would be considerably lower. 82 Chapter V ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION 5.1 In the past decade, Peru made considerable progress towards redefining the role of the State but the reform process remains unfinished In the early to mid-nineties, the size and scope of the public sector was reduced to more manageable proportions; the nature of government shifted from ownership of productive firms and excessive intervention to market regulation and provision of social services; and many state enterprises were privatized or closed. However, in the mid- to late nineties, the State reform process lost steam and some key tasks of State reform were not even initiated. Instead, authoritarian tendencies and open hostility to any democratization or deepening of State modernization gained the upper hand. Not surprisingly, public service delivery, governance, and corruption indicators deteriorated towards the end of the decade. Prominent among the reforms that were set aside by the authoritarian leadership were the decentralization of the State and public administration, the restructuring of the civil service, the implementation of pro-governance and anti-corruption policies, and the design of sound mining fiscal and environment policy agendas. These pending reforms are the objects of the next four chapters of the Report. This chapter is devoted to the issue of decentralization. It takes as its point ofdeparture the fact that, in Peru, there is a consensus among both the political leadership and civil society about the need to decentralize political power and the management of public resources. It goes on to point out both the promising potentialities and the serious risks entailed in the decision to decentralize, drawing for this purpose from the lessons learned in other Latin American experiences in decentralization. It concludes with an examination of the issues raised by the decentralization of education and health as key social services. The main message of the chapter is that the Peruvian resolve to decentralize faces both promises and risks Ifappropriately carried out, centralization can increase both the efficiency ofgovernment-service delivery and the responsiveness ofgovernment to citizens' needs. However, decentralization also poses formidable risks to fiscal discipline, and thereby to macroeconomic stability, in the form of recurring central government deficits, an over-expanded public sector, or the inability to use fiscal policy to adjust to economic shocks. A. THE DECISION TO DECENTRALIZE: PROMISES AND RISKS 5.2 Following almost a decade -long centralized authoritarianism, Per6 is a relative latecomer to decentralization in Latin America. The decision to decentralize initially found its motivation more in the rejection of the-political and social-consequences of centralization, than in an explicit message as to the purposes, content, or goals of decentralization. A number of recent decisions by the new Administration are evidence that there is a growing consciousness about the scope of, and the potential advantages and serious risks entailed in, the decision to decentralize. In the meantime, the decision to decentralize has been taking shape in a number of initiatives, such as: " The passage of a framework Law for State Modernization and the introduction of a framework Law for Decentralization in January 2002. " The decision to hold regional elections next November; 83 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 3 The recent creation of the Public Management Secretariat at the Presidency of the Council of Ministers, charged with formulating a strategy for both central government modernization and state decentralization; u The Ministry of Economy and Finance's decision to implement regional and participatory budgeting; L The Executive decision to create the Regional Coordination Councils, integrating CTARs with majors and social sector ministries representatives. O The legal acknowledgment by the Legislature of the departmental Mesas de Concertacion (Dialogue Roundtables) as consultative bodies of the regional governments, and u The passage in late June 2002, of the Law to Lay the Bases of Decentralization. 5.3 The dominant force behind decentralization is political. The current government's decision to decentralize is part of a broader, deeply positive, trend toward democracy in the LAC region (Burki et al., 1999). However, there is no single regional paradigm for decentralization, and experiences are highly varied (Box 5.1). In the particular case of Peru, the decentralization process is seen by the political leadership as an integral component of State reform, and there is a basic convergence of views as between Congress and the Administration. In the view of one of its leading congressional proponents, the process aims at devolving power to the regions and the local citizenry; promoting autonomous and sustainable regional development; fostering citizens' participation in public affairs and citizens' control of elected officials; and improving efficiency and efficacy of the local public administration. For its part, the Executive branch's view emphasizes seven principles, namely: efficiency in service-provision, macroeconomic stability, gradualism, heterogeneity, equity, participation of civil society, and political independence. 5.4 The decision to decentralize contains big promises and risks. If appropriately carried out, decentralization can, indeed, increase both the efficiency of government-service delivery and the responsiveness of government to citizens' needs. This is because devolving resource- allocation decisions to locally elected leaders can improve the match between the mix of services produced by the public sector and the preferences of the local population. Local officials have better knowledge of local conditions and are more accessible to their constituents. This gives them both the means and the incentive to be responsive. Decentralization may also improve public -resource management since, through sheer proximity, local officials can be held more accountable for their performance. Where the population is mobile and citizens can "vote with their feet," decentralization may also result in local governments competing with each other better to satisfy the wishes of citizens. Decentralization may, however, pose serious risks in the political sphere, in the area of service provision, and for macroeconomic stability. The risks in the political sphere have to do with the possibility that the local administration can be captured by local power elites and be put to the service of special interests. The risks in the area of public service provision have to do with the possibility that, because of the small size of some communities and hence insufficient scale in service provision, or due to institutional weakness of the local governments, the services rendered may turn out to be of lesser quality or more inefficiently provided. The macroeconomic risks are especially troublesome. They may take the form of recurring central government deficits, an over-expanded public sector, or the inability to use fiscal policy to adjust to economic shocks. First, central government deficits can worsen if sub-national governments are unable to reduce expenditures or increase revenues to finance the cost of assuming new responsibilities. Secondly, governments may also find themselves decentralizing both revenues and expenditures, but being unable to.reduce existing levels of central-government spending in those functions that have been decentralized. Third, decentralization can also hamper a government's ability to respond to economic shocks if it reduces central control over aggregate public -sector revenues and expenditures. This is generally the case in countries with high budget rigidity, like Peru. In addition, subnational government 84 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION pro-debt policies and indebtedness behavior can also undermine overall fiscal discipline. Subnational governments may end up borrowing excessively in the expectation that the central government will bail them out. Box 5.1 Lessons from decentralization experiences in Latin America There is no single paradigm for decentralization in LAC. Decentralization experiences in LAC are highly varied in their political, functional, and fiscal dimensions; be it for multi-tier countries (Argentina, Brazil, Colombia, Venezuela and Mexico) or single-tier countries (Chile, Bolivia and Guatemala) (Table A46). At the political level, some countries introduced elected authorities at the regional or municipal level; others did not. At the functional level, some countries were explicit in reallocating functions at the level of regional governments, others did it at the level of municipal authorities, and others did not do it explicitly, but in an incremental way. At the fiscal level, increased revenue sharing or earmarked transfers accompanied new functions by states or municipalities (Table A47). Why decentralize? In theory, three main reasons have been argued in LAC: (i) to promote territorial deconcentration of productive activities and generate employment; (ii) to increase efficiency in public service delivery; and (iii) to raise local participation in fiscal-revenue generation. Outcomes in LAC, however, have mostly been disappointing (CEPAL, 2001). 0 The reduction of territorial concentration of productive activities does not appear positively correlated with decentralization processes. Two highly decentralized countries, Argentina and Brazil, showed higher territorial disparities after decentralization than before. In the case of Colombia, territorial disparities increased as decentralization advanced. Chile, a country with a low level of decentralization is the exception with regional disparities decreasing only temporarily as a result of its decentralization process. O On the expected increased efficiency in service delivery, the experience is mixed. Decentralization appears to be efficiency- decreasing in education and primary health services in seven LAC countries: Argentina, Bolivia, Brazil, Chile, Colombia, Mexico, and Nicaragua (CEPAL, 1998). Exceptions in education are Colombia at improving the allocation of resources, and Chile and Nicaragua, due to the increased participation of the parents in the schools' financing and management. In Bolivia and Colombia, however, the cost per student has increased and quality has decreased following decentralization. Exceptions in health are Colombia and Argentina, and this is explained by the existence of enhanced demand-subsidy mechanisms that have also slightly reduced the acceleration in sector spending levels that usually accompanies the decentralization of health services. 0 On the expected increased participation of local government in generating additional fiscal revenue, LAC empirical evidence has been disappointing. It rather shows a persistent trend in local expenditure to increase faster than local revenue as decentralization proceeds. Being mostly based on increased transfers from the center, decentralization tends to increase the size of the total government sector (Burki, Perry and Dillinger, 1999). Such widening subnational deficits not only had increased national expenditure, but threatened macroeconomic stability in the short term, and aggravated the sustainability of subnational debt in the medium term, thus adding new pressures to already fragile fiscal stances. This was the case in Argentina, Brazil, Mexico, Bolivia, and Chile. Where does the greatest risk lie? The Latin American experience clearly shows that it lies in the potential that decentralization harbors for fiscal instability. Brazil's 1988 Constitution, for example, required a major increase in federal tax sharing by the subnational governments without providing for any devolution of federal spending responsibilities. This threatened to provoke recurrent deficits at the federal level. Colombia decentralized both revenue and expenditures but ended up maintaining and even increasing central government spending on education and health several years after these functions were transferred to the subnational governments. For its part, Mexico substantially increased transfers to municipalities without a reciprocal reduction of spending at the federal level. As a result, both countries have been forced to maintain pre-existing levels of spending while funding and expanding volume of intergovernmental transfers. This has contributed to significant central government deficits. Attempts to eliminate such deficits by raising taxes may lead to over-expansion of the public sector. Brazil's mismatch between revenues and expenditures, for example, was ultimately resolved not by reducing federal government expenditures, but by increasing federal taxes. Finally, decentralization can reduce a country's ability to respond to adverse economic shocks. Governments in Brazil and Colombia, for instance, which must share nearly half of their tax revenues with subnational governments, find it difficult to raise taxes sufficiently to compensate for revenue shortfalls. 5.5 In the case of Peru, the decentralization challenge is compounded by several factors: a vulnerable fiscal stance, at both the aggregate and local level; multiple governance issues afflicting local governments; and the incredible institutional fragmentation in the social sectors in the delivery of services. A variety of vertical programs resulting from years of gradual "add-ons" respond to particular needs with little internal coordination, multiple and overlapping levels, and partial coverage through a variety of institutions. In the health and education sectors, the absence of a unified regulatory structure or regular evaluation and 1 An important caveat is that following cross-section analysis for several decentralized countries, the steady state level of subnational borrowing does not appear associated with higher central government spending or deficits. However, when subnational borrowing increases, the central government seems to have to spend and borrow more in the subsequent periods, which points out to an inter-temporal macroeconomic management problem that not many countries, certainly no one in LAC, have adequately prevented (Burki, Perry and Dillinger, 1999) 85 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION monitoring systems based upon results compounds the problem of either quality control or the assurance of productivity. B. THE D ECONCENTRATION OF CENTRAL-GOVERNMENT SPENDING AND TRANSFERS TO THE MUNICIPALITIES: THE INTERGOVERNMENTAL FISCAL RELATIONS 5.6 The current system of expenditure allocation to the regional and local levels lies on two pillars, namely, deconcentrated spending by the central government and budget transfers to the municipalities. Peru has twenty-five states (called departamentos, or departments). Unlike most countries in Latin America, in Peru there are, however, no intermediate-level governments, as the departamentos are not administrative entities but merely geographical divisions of the national territory toward which the population has historical and cultural attachments. In contrast, the local level of government is fully established. There are two types of municipalities, namely, the provincial municipalities and the districts. Size-wise the relationship between the two types of municipalities is hierarchical in that each provincial municipality consists of a (variable) number of districts. Provincial municipalities are hence bigger than the districts nested by them. In the absence of an intermediate level of government, the central government (CG) executes its department-level spending through the deconcentrated organs known as Transitory Councils for Regional Administration (CTARs) and through the regional directorates of the central-government ministries. The other leg in the current system of expenditure allocation is based on the transfer of resources from the central government to the provincial and district governments. 5.7 Region-level expenditures by the central Table 5.1: Budgeted Expenditure for the government are a significant share of the national Public Sector for FY02 budget. More than half of the central-government Sector Percent of GDP budget is executed through the deconcentrated organs National government 11.4 at the department level. This amounts to a substantial Regional level 6.0 6.0 percent of GDP (Table 5.1). The role of CTARs is Local government 1.8 Public entities 1.2 quite important as they execute above 42 percent of Firs (Fonate) 6.7 the total expenditure outside Lima (Table 5.2). The TOTAL 27.5 relative significance of region-level spending (outside US $ mlns 15,686 Lima) is much higher if we exclude debt service and Source: MEF. pensions from the budget: in 2002, regional allocations account for about half of CG current expenditure and 81 percent of CG capital investments (Table A39), which means that approximately 4 out of every 5 dollars of public investment in Peru is executed outside Lima. The downside of this is that regional and CTAR expenditure are particularly punished when, as a result of an economic downturn, there are cuts in public investment outlays. For instance, in 2000, while total expenditure outside Lima was cut down to 88.4 percent of its budgeted amount, the executed budget for capital investment was much less reduced to 67.5 percent of its respective approved amount (Table 5.3). Hence, if budget autonomy is conceded to future regional governments, the central government would have much less capacity to invest nationwide and degrees of freedom (due to budget rigidity) to reduce expenditure when needed. 5.8 Region-level expenditure in health and education represent a very high share of total regional budgets. In 2002, education and health expenditures represent 50 and 27 percent, respectively, of total budgeted regional expenditure by the central government (Tables A39-40). These ratios show that a future decentralization.of both sectors could be tantamount to a defacto transfer to the future regional governments of up to about three-fourths of current region- level 86 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION budget resources. In fact, if we deduct current expenditure from total regional expenditure, education and health budgeted capital outlays represent a very high 95 percent of the total regional budget devoted to investment! (Tables A39-A40). This share underscores that the burden of budget cuts over regional investment is borne by these two sectors. T ble 5.2 Expenditure Coeffcients by CTAR, 2001 CTAR Expd./ Budgeted Actual Coefficient of Dept No. of Expenditure Expenditure expenditure Differ. Actual CTAR Expenditure Departments UEs Population (MIns of NS) by CTAR (i) by CTAR (ii) (i) - (ii) Expd. by Pop. (%) Amazonas 12 420,606 215.3 89.6 100.4 (10.8) 23.9 46.6 Ancash 27 1,092,662 591.4 277.5 284.1 (6.6) 26.0 48.0 Apurimac 14 455,637 299.0 125.5 129.2 (3.7) 28.3 43.2 Arequipa 19 1,067,469 913.8 261.7 275.0 (13.3) 25.8 30.1 Ayacucho 20 541,427 465.1 163.5 187.3 (23.8) 34.6 40.3 Cajamarca 20 1,480,690 547.0 250.9 264.9 (13.9) 17.9 48.4 Callao 13 774,604 547.0 133.4 71.0 62.5 9.2 13.0 Cusco 17 1,194,275 716.6 264.0 275.2 (11.3) 23.0 384 Huancavelica 11 435,596 215.6 105.3 108.0 (2.7) 24.8 50.1 Hubnuco 12 800,543 322.0 148.1 151.2 (3.1) 18.9 47.0 Ica 14 676,249 464.6 181.6 184.6 (3.0) 27.3 39.7 Junin 20 1,232,343 660.2 278.9 286.0 (7.1) 23.2 43.3 La Libertad 32 1,483,681 741.1 277.6 287.3 (9.7) 19.4 38.8 Lambayeque 14 1,110,129 553.5 173.6 185.2 (11.6) 16.7 . 33.5 Lima 194 7,617,193 23,747.2 9.3 9.5 (0.2) 0.1 0.0 Loreto 18 894,307 511.7 314.8 331.4 (16.6) 37.1 64.8 Madre de Dios 8 96,703 80.0 42.5 42.7 (0.2) 44.2 53.4 Moquegua 6 153,383 147.7 61.5 66.9 (5.4) 43.6 45.3 Pasco 8 259,137 172.7 72.1 75.4 (3.4) 29.1 43.7 Piura 16 1,611,573 788.1 350.8 376.6 (25.8) 23.4 47.8 Puno 33 1,247,494 651.2 304.9 323.7 (18.7) 25.9 49.7 San Martin 23 746,202 445.1 177.3 176.1 1.2 23.6 39.6 Tacna 9 286,539 222.2 93,0 99.8 (6.8) 34.8 44.9 Tumbes 8 197,605 179.1 81.1 90.3 (9.2) 45.7 50.4 Ucayali 16 450,693 225.6 149.0 158.5 (9.5) 35.2 70.2 Total 584 26,326,740 34,422.69 4,387.54 4,540.29 (152.75) 17.2 13.2 w/o Lima 390 18,709,547 10,675.50 4,378.25 4,530.79 (152.54) 24.2 42.4 Source: MEF. Table 5.3 Expendi ure by Department, 2000 (Min NS) Total Expenditures Capital Investments Goods and Services Execute Executed Executed Departments Budgeted Executed Budgete Budgeted Executed Budgete Budgetea Executed Budetea Amazonas 338.5 206.2 60.9 154.9 77.9 50.3 70.8 26.8 37.9 Ancash 641.2 598.2 93.3 203.5 158.6 77.9 70.4 57.5 81.8 Apurimac 367.2 292.7 79.7 182.4 124.1 68.0 48.1 35.8 74.3 Arequipa 917.0 881.2 96.1 300.9 231.8 77.0 104.4 93.7 89.8 Ayacucho 553.0 481.3 87.0 259.9 208.8 80.3 70.0 56.5 80.7 Cajamarca 640.7 494.3 77.2 263.1 116.3 44.2 61.9 52.7 85.2 Callao 448.1 407.7 91.0 148.5 106.5 71.7 108.8 122.8 112.8 Cusco 819.1 698.0 85.2 342.3 237.8 69.5 116.6 103.7 88.9 Huancavelica 278.1 240.5 86.5 109.9 87.3 79.4 41.0 34.8 84.9 Huanuco 346.2 313.2 90.4 117.6 79.3 67.4 42.9 38.9 90.7 Ica 471.7 453.0 96.0 68.2 57.0 83.6 55.2 47.4 86.0 Junin 695.8 613.3 88.2 162.9 83.0 51.0 86.6 72.8 84.0 LaLibertad 782.2 766.6 98.0 193.8 138.2 71.3 79.7 69.4 87.0 Lambayeque 583.5 574.3 98.4 136.1 96.5 70.9 65.9 59.7 90.6 Lima 24,297.4 17,889.1 73.6 2,427.6 1,718.0 70.8 4,341.8 2,759.7 63.6 Loreto 538.4 505.9 94.0 175.0 125.8 71.9 80.2 69.9 87.2 MadredeDios 106.2 79.4 74.8 59.6 33.2 55.6 14.4 12.0 83.2 Moquegua 138.9 124.2 89.4 48.3 31.1 64.3 16.5 13.7 83.1 Pasco 206.8 165.7 80.1 42.7 20.0 46.7 27.3 24.5 89.6 Piura 919.3 781.8 85.0 379.7 221.5 58.3 92.0 97.2 105.7 Puno 815.8 731.6 89.7 323.0 244.3 75.6 89.6 79.0 88.2 San Martin 469.6 408.2 86.9 206.8 141.5 68.4 52.1 45.3 87.0 Tacna 253.4 231.6 91.4 . 74.9 63.7 85.0 33.4 33.7 101.1 Tumbes 195.9 152.4 77.8 69.1 38.2 55.2 19.3 15.9 82.3 Ucayali 260.4 219.0 84.1 81.2 46.8 57.7 47.2 39.2 83.2 Total 36,084.4 28,309.1 78.5 6,532.0 4,487.0 68.7 5,836.0 4,062.7 69.6 w/oLima 11,787.0 10,420.0 88.4 4,104.5 2,768.9 67.5 1,494.2 1,303.0 87.2 Source: MEF. 5.9 Transfers to the local governments are small when compared to region-level expenditures, but constitute the bulk of municipal-government revenues. The transfers to local governments amount to 15.8 percent of the central-government budget (1.8 percent of 87 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION GDP). However, in the case of municipalities other than Lima, they are the most important source of revenue. As a percentage of total municipal revenue, CG transfers represented 48 percent in 2001, and a high 73 percent when Lima is excluded (Figure 5.1). CG transfers are concentrated in the four most important ones, namely: FONCOMUN, Vaso de Leche, petroleum canon, and mining canon, with 68, 17, 7, and 4 percent, respectively, of total budgeted transfers in 2002 (Figure 5.2). When it comes to the locally based sources of revenue, the main revenue items are fees (tasas)-which in 2002 represented 45 percent; taxes-which barely represented a low 25 percent; fines, with 11 percent; public -service fees, with 6 percent; and real estate income-which barely represented a low 5 percent (Table A40). No doubt, these sources of revenue will have to be strengthened through a municipal tax reform, if the municipalities' autonomy is to be strengthened. Figure 5.1 Municipal Revenues, 2001 Figure 52 Central Govt Transfers to Municipals, (exduding Uma) 2001 Own Canon and Canon Minero Other revenues Sobrecanon 4 5% . ...27% Pebid Transfers from central V Lede govt Foncomun 67% 5.10 Local-government finance is additionally constrained by severe liability problems. The municipal debt is not a serious threat for macroeconomic stability, but that is not an indication that the problem is not serious or generally significant, and even explosive, for some municipalities that face severe liquidity and solvency problems. Municipal debt explains a third of the corresponding national registered debt: short-term current liabilities, but this varies widely per municipality. Solvency at the local level can be a serious concern. A World Bank survey among nine municipalities2 shows that four of them had a current liability/revenue ratio above 1, meaning that they require more than a one-year income to cover their present debts (in some cases, such ratio is above 3). Debts have been growing at a rapid annual rate (13 percent on average for the 1997-2000 period) and the debt situation appears to be deteriorating faster in large provincial municipalities (Arequipa) and urban districts (La Victoria in Lima). Debt dynamics is explained by essentially two factors: arrears (cuentasporpagar) with suppliers and public -utility companies (about 55 percent), and non-registered pensions with former municipal employees (21 percent). The municipalities are also becoming increasingly illiquid as all but two had a current-revenue/current-expenditure ratio below one, reflecting that their present current revenues are hardly enough to cover their present current expenses. The problem is becoming explosive, since contingent liabilities are threatening to materialize, mainly because of resolution of labor disputes, whose amount is projected to grow 25 percent in the next few years (Fernandez, 2002).- Municipalities like Arequipa and La Victoria, in particular, are not going to be able to assume additional responsibilities under decentralization if they do not solve their debt problems previously. The GOP seems to have neither an adequate regulatory framework nor an appropriate institutional arrangement for assessing and controlling growth of municipal debt yet. 2 The sample consisted of three provincial municipalities (Lima, Arequipa and Trujillo) and six district municipalities (La Victoria, San Juan del Lurigancho, Jacobo Hunter, Characato, La Esperanza and Poroto). 88 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION C. THE EMERGING LEGAL FRAMEWORK FOR D ECENTRALIZATION IN PERU 5.11 Both the government of president Toledo and the congressional le adership are committed to laying the political and legal basis for the decentralization process. As a result, in late January, the legislature approved the Framework Law for the Modernization of State Management (Law 27658). The Law is an effort to provide the legal base and principles for initiating the process of implementing modernization reforms. It promotes six objectives: a service-provider state; effective access for participation; decentralization and deconcentration; transparency; a new civil service-well qualified and remunerated; and fiscal balance. The Law focuses strongly on the need to redimension the State both in its role and in its size. The State is then perceived less as an actor than as a regulator of actions, while retaining basic human-need- attention functions in health and education. To accomplish this paradigm shift, the Law offers three principal tools: (i) the fusion of state entities to reduce duplication; (ii) the introduction of Management Contracts (Convenios de Gestion) in conjunction with Multi-Annual Strategic Sector Plans (PESEM) in the context of the Multi-Annual Macroeconomic Framework (MMM) objectives and goals; and (iii) pilot programs for modernization through the gradual introduction of convenios. The initial pilot programs will be carried out at the Ministry of Transportation and the Presidency of the Council of Ministers. The Law also defines three lines of action for the central government in terms of the need to (i) clarify functions and structures, (ii) introduce a system of management contracts based upon results, and (iii) modernize administrative systems. As a framework document, the new Law was a step in the right direction. It provided objectives to aim at while remaining flexible. It also opened the door for sector ministries to pursue different avenues of action. In virtue of this Law, the newly created Public Management Secretariat at the PCM was rightly given a leading role. 5.12 In late June 2002, the Peruvian Congress approved by overwhelming majority the Ley de Bases de Decentralization. The Law defines the purpose, principles and objectives, norms and transition measures guiding the decentralization process. The Law distinguishes two sets of principles on which it is based, namely, general principles and specific principles for fiscal decentralization. o Among the general principles, the following are included: (i) decentralization is a permanent State policy, of a mandatory character for all the constitutional branches of the State, the autonomous constitutional entities, and the government as a whole; (ii) the process of decentralization is defined as dynamic and requires a constant monitoring and evaluation of its ends and objectives; (iii) it is irreversible; (iv) it is democratic in that it promotes equality of opportunity; (v) it is integral in that it encompasses the whole State in the national territory need to carry out the process in a gradual manner, so that a sequence of steps must be followed; and (vi) the process is premised on the pinciple of subsidiarity, implying that functions and competencies must be assigned to the level of government closest to the citizens that can efficiently perform a given public service. c The specific principles are next: (i) a clear definition of competencies; (ii) transparency and predictability; (iii) fiscal neutrality so as to avoid any transfers of resources that are not accompanied by a transfer of responsibilities; (iv) public -sector indebtedness is an exclusive competency of the national government and regional and local governments can only contract debt with central government guarantee; (v) fiscal responsibility to be materialized in fiscal rules for the subnational governments, which must include rules on indebtedness and limits to annual expenditures of the subnational governments so that the decentralization process can be sustainable. Prominent among the objectives defined by the Law are the following: (i) institutionalization of the regional and local governments; (ii) economic development and 89 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION competitiveness of the different circumscriptions, according to their productive potential; (iii) modernization and simplification of administrative systems and processes; and (iv) citizen participation in all forms of organization and social control The Law creates intermediate-level governments-the regional governments, and mandates that the regional governments' authorities be directly elected by the population. u The normative foundations are also considered. On this account, the Law defines (i) three types of competencies; namely, exclusive competencies, shared competencies, and delegatory competencies; (ii) a Constitutional Tribunal is called for resolving conflicts of jurisdiction; (iii) the creation of a Regional Compensation Fund (FONCOR), which will finance regional investment projects to be executed by the regional governments; and (iv) new sources of revenue in the national budget and as a percentage of privatization proceeds. o The Law also creates the National Council for Decentralization as the entity that will lead the decentralization process. The Council will be composed of representatives from the central government, the regional governments, the provincial municipalities, and the district-level municipalities. The Council's main responsibility is to oversee the process of transfer of competencies from the central government to the subnational governments, and manage an Intergovernmental Decentralization Fund (FIDE) to promote co-financing of projects by subnational governments. The transfer process will be carried out gradually over several years, in successive stages that initially affect sectors like transport and housing, and at a later stage education and health. The Council will also discharge functions of coordination, technical support, and training for the subnational governments. c Last but not least, the Law includes transition measures of utmost importance, including the elimination of the Ministry of the Presidency and the Transitory Councils of Regional Administration (CTARs) by December 31, 2002. 5.13 This emerging legal framework will be completed with forthcoming laws. These include the Organic Law for the Executive Power, the Law on Regional Governments, the Law of Municipalities, the Law on Fiscal Decentralization, the Law to Create the Regional Compensation Fund, and the Law of Economic and Productive Decentralization. D. ENTERING INTO A DECENTRALIZATION PATH 5.14 Peru's recent efforts toward modernization and decentralization offer the general principles that will guide the transition process, while complementary laws will apply them to specific components of such a process. As decentralization is the result of political will and consensus-not simply of technical expertise, a broad and participatory dialogue (and patient persuasion), supported by the National Dialogue and the Concertation Roundtables, have taken place in the past eight months. As decentralization should be demand-driven, the binding thread for the process must be the idea that its ultimate purpose is to increase the coverage and improve the quality of public services so that the citizenry assumes ownership of both the process and its results. 5.15 The latent risk that fiscal discipline may turn out to be a casualty of the decentralization process will continue to be high until a clear framework for inter- governmental public finances is finally enacted and enforced The guiding criterion for such a framework has to be the establishment of hard-budget constraints for the expenditure and indebtedness decisions of subnational governments. The road ahead must then be carefully trodden if the appropriate framework for fiscal discipline of the subnational governments is to be attained. Some critical actions that will help to secure that outcome seem to be in order. 90 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION 5.16 The issues of substance relate to the rules of the game in the process of decentralization. Five issues deserve special consideration, namely: (i) the degree of autonomy to be given to the subnational governments in their future provision of social services; (ii) the kind and extent of expenditure responsibilities to be transferred to the subnational governments; (iii) the rules and institutional mechanisms for transferring resources from the central government to the subnational governments; (iv) the allocation of taxing powers to the subnational governments; and (v) the rules the subnational governments must follow in contracting debt. 5.17 A large degree of autonomy is desirable. The higher the autonomy the subnational governments can enjoy, the higher the likelihood that the potential benefits of decentralization will be realized. Both the ability of the local governments to respond to the demands of the local population and the participation of the citizenry in local political life will be enhanced if a high degree of autonomy is allowed. However, for the decentralization not to be a travesty of local democracy, higher degrees of autonomy must be accompanied buy higher accountability on the part of subnational governments vis-a-vis the local population. But autonomy has its risks as well. Autonomous subnational governments may end up spending irresponsibly and accumulating unsustainable debts if they are not subject to stringent budget and indebtedness rules. 5.18 There are no hard and fast rules on the kind and extent of expenditure responsibilities to be assigned to subnational governments. The extent to which expenditure and service-provision responsibilities are to be transferred depends mainly on the institutional strength and administrative capability of regional and local governments. In this respect, two broad facts in the Peruvian situation are of compelling importance. First, regional governments are yet to be constituted after the November 2002 elections. There is no reason to expect that the soon-to-be-born regional governments will become capable and efficient administrative apparatuses in a short period of time. Quite to the contrary, there is every reason to predict that the constitution of truly capable regional governments will take some time. Secondly, with the exception of a few provincial municipalities and an even fewer number of district municipalities, local governments are weak. In the light of these two facts, it stands to reason that the transfer of responsibilities must be gradual. In designing the transfer of responsibilities, due account must be taken of the markedly different degrees of development of administrative capability in the subnational governments. In addition, in the Peruvian case, the difficulties of any transfer process are compounded by the sheer existence of the two levels of municipalities, the provincial municipalities, and the district municipalities. This implies that, unlike most countries, after the implementation of the regional government Peru will hIve four levels of government rather than three. Now, most district municipalities are very small to assume responsibility for the more demanding public services and can discharge only the relatively less complex tasks and, in any case, only those where economies of scale are not crucial. Assignment of responsibilities to the two levels of municipalities must be carefully designed having in mind size and human-resource and administrative-capacity constraints at the two levels. Finally, the one mistake that by all means should be avoided is the lack of clarity and precision in the distribution of functions as between the central government, the regional governments, the provincial municipalities, and the district municipalities. Legal and institutional confusion about the distribution of functions is likely to lead to either situations of effort duplication and overlapping or to some public services being left unattended and underinvested. 5.19 The GOP intends to change the way decisions are made about the amount of resources to be transferred from the central government to the subnational governments. In the past, these transfers (1) have been tied to variable sources; (2) have been, to a certain extent, subject to discretional decisions at the central level, sometimes under the guise of 91 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION formulas that have not been very transparent, and (iii) have been pro-cyclical. It is expected that, in the new legal framework, procedures that are more transparent and decision-making criteria will be established. From the standpoint of transfer policy, the two key institutional pieces in the new framework are going to be the Compensation Funds and the Intergovernmental Fund for Decentralization (FIDE, in its Spanish acronym). Defining their long term financing mechanisms, as well as criteria for accessing their resources is a short-term priority. Both types of funds will follow explicit formulae in the calculation of the transfers to a given regional or to local government. As it is customary, these formulae will take hto account the size of the population as well as income and poverty indicators. There is however an important difference between the two types of funds. Whereas transfers from the Compensation Funds will be applied in an unconditional manner, i.e., regardless of the particular local government's fiscal and service-delivery performance, FIDE's transfers will be conditional on each subnational government's meeting pre-defined targets of performance in public service provision and on its adhering to explici standards of fiscally disciplined performance. 5.20 There must be a close connection between the transfer of responsibilities and the transfer of resources. The double golden rule to be followed is that (1) the transfer of resources must be accompanied by the transfer of responsibilities, and (2) responsibilities must not be transferred unless an adequate level of resources to discharge the responsibilities is available at the subnational level (in most cases as a result of a combination of resource transfers from the central government and the subnational government's own tax revenue-see the next paragraph). 5.21 Subnational governments must be given powers of taxation within appropriate boundaries. It would be a monumental mistake were Peru to decide that the subnational governments are to rely only on the resource transfers from the central government. The experience from other countries teaches that an institutional scheme where local governments do not have the power to collect taxes weakens accountability at the local level; reduces incentives to pay any kind of taxes, including taxes to the national government; and weakens the central government's willingness and commitment to refrain from bailing out the subnational governments when and if these incur in insurmountable financial difficulties stemming from lack of fiscal discipline. It is therefore fundamentally healthy that subnational governments have their own sources of tax revenue and make an effort of their own to optimize their tax revenue, taking into account both issues of incentives to entrepreneurial efforts and issues of re-distributive justice as well as the need for local public goods. In this respect, the Law already defines a procedure by assigning to the Executive the sole responsibility to submit regional tax proposals- whose management will be implemented by regional governments-to Congress. The menu of taxes that can be imposed on the populations of the subnational entities by their governments must be unambiguously defined by law, and any tax that is outside that menu or list will be illegal. The types of taxes that are appropriate at the subnational levels are a matter of considerable consensus both in the literature and in practical experience. Real estate taxes and taxes on industry and commerce are the most common at the municipal level. Excise taxes on specific consumption goods such as tobacco products, liquors, and beer are the most frequent at the intermediate, regional level. Within the confines of the tax menu defined by national legislation, subnational governments must be given a fair degree of autonomy to define the bases and rates for their tax systems. 5.22 Subnational-government indebtedness must be subject to stringent rules and procedures. There are good reasons to allow subnational governments to use debt as source of finance. As a way of spending on account of future income, subnational governments' resort to We showed the degree of progressivity of FONCOMUN and saw little reason for modifying its redistribution criteria. 92 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION debt makes as much sense as it does for any other economic agent. The issue then is not whether or not debt should be allowed. The issue is rather to which constraints must subnational government indebtedness be subject, so as to prevent these constraints from becoming a source of national instability and local ruin. A related issue is how to avoid the moral hazard that becomes rampant when and if subnational governments are given (implicit or explicit) assurance that they will be bailed out. Since neither the market nor the spontaneous workings of the political system enforces debt discipline, it is necessary to impose limits to subnational indebtedness. A positive initial step is that the Law already specifies that external debt is an exclusive competence of the National Government and its amount must be consistent with the overall public sector debt ceilings. It also defines that sub-national governments can only contract new debt supported by a state guarantee. However, domestic and contingent-registered and unregistered-debts are not mentioned directly, and should be addressed in the incoming reform of the Law of Fiscal Prudence and Transparency. Recommendations: > There is an immediate agenda of urgent issues that must be addressed. The agenda includes issues of paramount importance such as designing alternatives for intergovernmental fiscal relations, hard budget and indebtedness constraints on subnational governments, the specific functions and competencies to be decentralized by sector (see Annex F), the characteristics and funding sources of the future Regional Compensation Fund, and any possible changes in the transfers scheme for municipalities. As the Law of Fiscal Prudence and Transparency gets revamped, adding a chapter on subnational fiscal responsibility would be a positive development (Chapter II). > A critical goal of decentralization is to develop an increased local revenue -generating capacity, based on local governments' own tax systems. Measures to be taken may include the following: (i) increasing the real estate tax in the short term and upgrade it with a new system (a territorial-based tax on property extension) in the medium term with the support of a fiscal cadastre, initially piloted in a few municipalities; (ii) transferring the proceedings of gasoline taxes to subnational governments; (iii) abolishing obsolete taxes; and (iv) promoting horizontal technical assistance on the basis of best-practice municipalities. > The issue of local governments' already existing-domestic and external, registered and non-registered-debt must be urgently tackled. A curative approach would complete a full diagnosis of all municipalities and design a transparent program for the restructuring of the existing debt. Its implementation would be another essential pre-condition for assuming new functions and administrative responsibilities. A preventive approach would establish prior conditions for contracting further debt and limits to its size.4 Ditto for future regional governments. > The building of institutional capacity both in the current local governments and in the future regional governments is critical for the success of decentralization, especially regarding the development of integrated financial management systems at the regional and local levels in advance of, and as a prior condition for, decisions about transfers of expenditure responsibilities. The resulting local and regional systems must be linked to the national SIAF so as to have consolidated accounts for all levels of the government. > The use of performance contracts has proved useful and should be promoted. Pert's experience with FONAFE in developing management contracts with public enterprises is a 4 Two interesting examples of good sub-national risk monitoring are Brazil and Colombia. In the former, complementary limits are established in the Brazilian Fiscal Responsibility Law. In the latter, rules are embedded in the Colombian "traffic light system" that links each subnational government's debt to its payment capacity and use indicators as traffic lights to alert the central government about potentially excessive subnational debt (Ma and Polackova, 2002) 93 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION success story for the LAC region (Annex Box 5.2 The Decentralization Drive in Education in E). The experience can be applied, under LAC: Some Cases a pilot-case approach, to deconcentrated In the case of the ParanA state in Brazil, the decentralization ministerial units and subnational effort has aimed at the municipalization of basic education. governments.Recent efforts in Ecuador have focused on the extension of govenmets. Specal ncetive ca be networks, both for pedagogical and administrative reasons and on given to such units and governments to a model of voluntary municipalization of education. That entice them to enter into management municipalization process is incipient and not subject to any evaluative criteria (Moore and Rosales, 2001). In rural Mexico, contracts. The latter's success would parents have been authorized to certify teacher attendance under provide important lessons and examples national jurisdiction, while transfers to state education boards because, in the end, decentralization and have been extended since 1998. Municipalization combined'with direct decisional control has been a successful model in reform place a premium on accountability. Nicaragua. School autonomy in Nicaragua is based upon three This is not simply probity, transparency, elements: fiscal transfers to schools based upon both technical and internal efficiency in the use of and equity criteria (formula based); parental control of fiscal resources through Parent Councils; and complete local authority resources. It is also an emphasis on over hiring and firing decisions. While the effort has been results, performance, and social successful in the transfer of decisional control and in the encouragement of community and parent involvement in the accountability. Success breeds success, as educational process, educational quality has not been noticeably the demonstrative effect of carefully affected (Arcia and Belli 1999). In the Dominican Republic, monitored successful pilot experiences recent reforms and training efforts aim at increasing the role of parent associations in monitoring teacher attendance, in receiving can prove fruitful and critical in the the transfer f some funds for local operating costs, and in public's support to the process. exercising a more participatory role in local schools. In pp Venezuela, a current decentralization effort focuses on the transfer of management authority to the school level ("the school E. D CEN RALZATIN ~THEas the center of basic education") and introducing and certifying E.(vocational) educational training at the level of the workplace EDUCATION SECTOR' (cada empresa una escuela). In Chile, a varety of efforts have been attempted including where parents have been given the option to select among publicly financed schools (Winkler and 5.23 Specialists have noted that public Moore, 1996). In El Salvador, rural parents associations have education in Per( remains among the most been reinforced and have been empowered to enter into contracts centralized in Latin America. (Box 5.2) to successfully manage schools. EDUCO schools lie outside the purview of the Ministry of Education's regular school system. During the decade of the 1990s, the effort to They are located in remote rural areas principally, and were/are decentralize public education was limited to a intended as a means of extending coverage. The management of series of "on-again, off-again" initiativesresources is the responsibility of community seris o-agan, ff-aain iniiatves educational associations (ACEs) based upon annual plans. The driven more by political interests, than transfer of funds to the ACEs is based upon these annual plans technical reasons (Ortiz de Zevallos et al., and these are received monthly as earmarked funds (salaries, operations). Personnel decisions are the responsibility of the 1999). These initiatives were also short-lived. ACE, but pay scales are Ministry responsibilities. The ACE has Opposition of municipalities and labor unions, clear governance responsibilities, although the Minis names andthe director. Pedagogical concerns and standards remain and he ear f te los o fre pulic ministerial responsibilities. The evidence to date suggests that, education doomed these efforts from the as a means for increasing community participation and outset. As a result, few efforts had any major involvement ofparents, the model has been successful, but as a means for improving educational quality, there is no clear impact toward either a deconcentration or evidence that EDUCO has accomplished much. Finally, in decentralization of the educational system. Guatemala, four different models of decentralization have been Further, even as recent (2001) policy pursued simultaneously: (local management autonomy and 'control - PRONADE); limited decisional authority at the school directives with strategic objectives have been level (Juntas Escolares); the decentralization of a competitive formulated there is little consensus as to the system of teacher recruitment and selection; and the mechanisms or organizational units that deconcentration of finance and supervision to departmental should5be the agents of the process.. What isDgoven not clear to date is whether the agent of decentralization should be the future regional governments, the existing Educational Service Units (Unidades de Servicios Educativos, or The next two sections partly draw from the mimeograph prepared by Richard Moore (2002). 6 These directives provide an ambitious set of measures and goals. Three objectives are noted: (a) quality basic education for all; (b) strengthening at the school level with increased autonomy, participation and quality learning; (c) dramatic improvements in the condition and qualifications ofteachers (MED, 2001). 94 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION USEs), the municipalities, or the schools, and what is the appropriate role of the community in the process. Worst of all, the institutional framework for the delivery and administration of educational services has become sort of a labyrinth of fragmented units. 5.24 The first element of this fragmentation stems from the creation of CTARs. They are responsible directly to the Ministry of the Presidency and charged with responsibilities over the bulk of the regional recurrent budget in both health and education. Under the current scheme, the Ministry of Education has little control over the allocation of resources within its own sector at the regional level. In addition, even at the central level, the confusion of authority between the Ministry of the Presidency and sector ministries remains. 5.25 The second element of this fragmentation is the myriad of institutions exercising some administrative and/or decisional responsibility at the regional level. The result is a high degree of overlap and duplication of responsibilities and authority. The responsibility for providing educational services resides in Regional Directorates of Education (DREs), yet these entities depend fully on transfers from the central government. The DREs exercise their functions (normative, financial, regulatory and supervisory) in a very partial manner and contingent upon the relationship of any individual Director with MED. At the regional and subregional level, institutional fragmentation is even more pronounced due to the existence of a variety of institutions exercising some functions, yet whose roles in the process of deconcentration or decentralization are unclear. No fewer than five organizational units play a role at this level: DREs, CTARs, AEs, USEs, and ADEs. Under the DREs, there exists no less than three units or levels of action. Areas of Execution (AEs) and Educational Service Units (USEs) are administrative units charged with purely administrative functions. The USEs are line units with responsibilities for exercising budget functions, yet they receive large portions of that budget from the CTAR. They are responsible for the management of 58,000 schools and additional 18,000 non-formal educational programs. In some cases, there are regions with sub- DREs with essentially the same administrative functions as USEs. In addition, under the DREs, there exist educational supervisory bodies, the Areas of Educational Development (ADEs). The sector fragmentation is further evidenced in the extreme degree of confusing, poorly coordinated and overlapping jurisdictions, particularly in the case of the 82 USEs, and 189 AEs. 5.26 Historically, relations between Regional Directorates and both the central level of MED and the local level have been conflictive, mainly due to MED limited budgetary control, and poor regulatory and managerial capacity. While theoretically MED exercises regulatory authority over DREs, in fact appropriate control is not exercised. This regulatory weakness on the part of MED over its own bureaucratic apparatus is in part a function of the sheer size of the sector, and in part a function of limited budgetary control. There are more than 300,000 teachers in the public sector, and budgetary control over the labor force rests with the CTAR. The weak managerial capacity, in turn, is a function of several factors, including the low quality of functionaries of MED, high rotation among ministers and regional directors, limited communication between those charged with educational planning and execution. Finally, the triplication of authority between the Ministry of the Presidency, MED and the CTARs affects capacity and coordination. The first-through transfers to the CTARs-transfers funds to executing units and exercises control over payroll. MED has, in theory, educational quality as its principal concern, but also exercises control over hiring of teachers. Finally, CTARs have a discretional focus on non-personnel budget austerity and control. Perhaps the most fundamental manifestation of the lack of coordination resides in the area of payroll and personnel policy, also reflected by the absence of a generalized information system or, even an inventory of personnel 95 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION (Chapter VI).7 This historical multiplicity of functions also affects the quality and supervision of education. At the same time, inadequate regional supervisory capacity and quality results in a supervisory function that is virtually bereft of a concern for pedagogical quality in the face of administrative and bureaucratic demands. 5.27 The legal framework for the sector is not particularly propitious for decentralization. Both the Teachers' Law (Ley Magisterial) and the General Education Law are rigid in terms of preventing the conditions for flexible response at either the school level or in terms of local functions that might be exercised. 5.28 However, a promising beginning towards a redefinition of the legal framework for decentralization is evident in the Executive Decree promulgated in February 2001. This decree (i) expanded and fortified the role and functions of the school director; (ii) created Consejos Escolares Consultivos (CEC) at the local level; (iii) strengthened the role of Parents Associations, intensifying relations among MED, teachers, and the community; and (iv) mandated the introduction of planning and participation at the local level for the formulation of Institutional Development Projects (PDI) at the school level and the latter through the (optional) creation of the CEC. 5.29 A 2001 pilot exercise in the creation of the CEC in some 467 schools has been positive, but.still is far from expectations. In November 2001, the incoming director of the program made a very cursory evaluation of the experience as follows: (i) the experience has been positive and should be expanded; (ii) there is a need to create a baseline for monitoring and evaluation of the program; (iii) the principal limitation of the effort to date is its voluntary character and the fact that the functions of the CEC are limited to advise and consult; (iv) the CEC has no resources and its leveraging capabilities are weak; (v) support from USEs, DREs or MED in the formation, training, and capacity to evaluate on the part of he CEC has been extremely limited; and (vi) little promotion of the program has occurred. As a result of these factors, there is little incentive to participate or even create the CEC at the school level. 5.30 Several actions are needed. Recommendations: > Both the Teachers' Law and the General Education Law need to be revised in this context. They are too rigid for most forms of decentralization. > Decentralization should follow a sequential process. Central government reorganization prior to decentralization, but with an eye to decentralization, needs to be first. The suggested model for defining functions and structure in Annex F should be carried out within the central ministry and with the participation of key personnel from the various structural units. > The current institutional morass of jurisdictional levels needs to be integrated. The decision to transfer funds and functions from CTARs to DREs is only a partial move. The proposed Law of Decentralization offers broad guidelines for jurisdictional levels (regions, municipalities) but the non-exclusive nature of which level promotes, develops, and regulates 7 At present there is an effort to create this information system. However, to date only Lima-Callao is included in the system and there are pressures not to implement it at the regional level for fear of loss of regional discretionary authority. In the effort to link the information system to a system of payroll, a problem that needs to be addressed in the near term is the inconsistency in the definitions of benefits (bonificaciones). 96 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION activities of education (and health) leaves gaps Box 5.3 Decentralization of the Health Sector for defining these functional differences.8 What in LAC is clear is that the creation of multiple leve ls of A review of decentralization efforts in the health intermediation may not be an appropriate sector in six countries of Latin America shows that the strategy and may well inhibit deeper forms of e have been very uneven (Burki et al., 1999). straegy an ma wel inibi deper orm of Effrtsto reform the health sector invariably aim at the decentralization to the provider level. So, as separation of financing from service delivery, a split the process of redefinition of roles and between purchaser and provider, and the introduction often of management or performance contracts. functions progresses, there must be a broad Decentralization involves the transfer of parts of the simplification of administrative units and a public-sector health institutions from the central to clarification of responsibilities at the subnational governments. Central governments retain policy-making, overall financing and highly specialized intermediary level. operations and subnational governments have been given > The introduction of management or primary (municipal often), secondary and tertiary hospital performance contracts should wait for a more are (states and provinces). More often than not, efforts tdecentralize have generally failed at either improving advanced stage of the decentralization of the coverage or efficiency and when successful have often education sector. Once the institutional morass included some torm of contractual agreement, such as is simplified and monitoring indicators are agreed, they could serve as an agent for culture In Nicaragua, for instance, there is an ongoing effort to decentralize a number of basic institutional characteristics change, for the transformation from an (provision and acquisitions, finance, personnel organizational culture based upon norms, management) to both hospitals and to regional/local units bureaucracy, procedures, and inputs to an (SILAIS) through the introduction of management contracts. The experience of two years with management organizational culture founded in transparency, contracts in hospitals has not been successful, although individual and institutional responsibility and the learning curve has improved the quality of these contracts, particularly in terms of moving from input - authority, and results. Performance agreements based criteria to output and outcome-based criteria for must be accompanied by individual contractual agreement. performance agreements evaluating teachers regularly. > The new legislation should clarify the role of service providers and local councils. Much of the current move to decentralize in education focuses on direct transfers authority, finance and accountability mechanisms to educational providers: schools and their communities. > The CEC model should be expanded, but with strengthened fuvrctions. Examples from LAG countries would encourage greater school autonomy and the use of tools of social/legal accountability, such as institutional development plans and management agreements. > Authorities need to create political "space" for the introduction of modalities of decentralization with a reasonable level of consensus. Such consensus does not exist in the educational sector today. The National Consultation on Education is a good beginning. F. DECENTRALIZATION IN THE HEALTH SECTOR 5.31 The public provision of health services is characterized by the ministerial absorption of multiple vertical programs, fragmentation, inefficiency, limited programmatic integrity, driven by the supply of programs for health care rather than the demand for health services. (Box 5.3) In addition, there is limited focus on preventive health care and the promotion of health care, as the supply of curative health services dominates the terrain. Today, there exists no fewer than 23 vertically controlled programs administered from the center, and each with its own disconnected set of norms, procedures, supervision system, evaluation, data bases, training and mechanisms for accessing local level participation. The result d During the PER main mission a brief seminar was held with the technical leadership of MED. It practiced a methodology for initiating the dialogue on functions and structures, based on technical workshops led by key technical personnel at the central level (central restructuring) and participation from regional directors and their technical staff. 97 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION is an incredible. fragmentation of policy and administration in the national system.' Further, at the regional level many of these vertical programs maintain their own coordinators for administration, monitoring, data collection, and training, virtually independent of the regional directorate. Here, too, the result is fragmentation in the financing of health services and little coordinated control over policy direction or planning. The regional directorates for health (DISA) suffer from a minimum of authority, the inability to program in a coordinated fashion and a high degree of fragmentation as well. As is true in the case of education, the financing functions that reside in the CTAR further erode any authority that the DISA might have. DISAs have very little programming authority and receive mostly pre-programmed funds, including earmarked entrusted funds (Altobelli, 2001). Finally, barely about 20 percent of the health expenditure budget is devoted to preventive health care, which reflects its lower priority vis-a-vis curative health. 5.32 Ironically, while the health care system has become increasingly centralized, the central line ministry has minimal control over any attempt to integrate the health care system. There are rather three areas or initiatives that favor decentralization in the health care system, and these should be promoted: (i) the Performance-Contracts Administration Program (Programa de Administracidn de Acuerdos de Gesti6n - PAAG); (ii) the Regional Directorates of Health (DISA); and (iii) the Unidades de Seguro Integral de Salud (USIS) recently created. 5.33 The PAAG-CLAS Program is a very useful model for health decentralization that should be strengthened PAAG already executes some 17 percent of the MINSA budget under financial management rules of its own. There are approximately 550 CLAS and some 1,200 health establishments under the system (approximately 20 percent of total health establishments) under the system. It was created in 1994 as a means to improve the quality of health care expenditures, to bring transparency to the system, to improve the cost effectiveness of service delivery and to promote (private) community participation in health care management. The CLAS model builds upon the existing local and community organization, and essentially was designed to increase coverage in rural and periurban communities, in areas where existing health services were weak or absent. Similar to Convenios de Gesti6n Institucional, the CLAS model provides for a management contract. However, one fundamentally different characteristic of the contract is that the model applies private law to the use of public resources, and by doing so proffers greater flexibility in the management and utilization of public resources. CLAS not only combines community and public health services in the planning process, but also gives CLAS direct control over hiring and firing of contracted health personnel and assures a unique degree of social accountability. In its initial stages, the formation of CLAS was very localized: each CLAS managed a single establishment or health center. It became apparent that this approach was less than ideal in terms of the potential coverage of the system and gradually the CLAS became a management contract over a group of establishments in a network. While this change in scope and structure seems to be more effective in terms of economies of scale, there is a danger that increasing scale reduces "local" social control at the level of the community, in favor of control at the site of the center of the network. CLAS has shown to be an effective model of co- management in the financing and provision of services. It -has increased the "capture" of local resources (because fee for services remain at the local level) and the sense of local accountability. However, CLAS should not be seen as a means for reducing government expenditures on local health care. Further, CLAS has not proven to be as effective in the supervision of the quality of services. By its very nature, members of CLAS are not competent to supervise quality. 9 Recent documents provide excellent diagrammatic representations of the institutional disorder in the health sector. They offer dismaying evidence of the chaos in the sector. See World Bank (2001e) and Cortez (2001). 98 Chapter V. ADDRESSING THE PROMISES AND RISKS OF DECENTRALIZATION Recommendations: > expanding the number of CLAS, possibly to attain some 25 percent of health establishments, but without increasing their scale of the CLAS for to maintain the sense of local social accountability, the ratio of CLAS to establishments should not be increased. > the need to strengthen other entities within the health care system, more capable of quality assurance and supervision 5.34 For CLAS to function well requires the reinforcement of Regional Directorates of Health (DISA). The DISA suffers from fragmented authority, an administrative system that further fragments the coordination of financing in the sector and in the supervision of the quality of health provision. In addition, execution.by the DISA is driven as much or more by the supply of services rather than demand for services, and introduces certain rigidity in the use of funds. However, despite norms and regulations; DISAs have played an important role in the health sector, since they are able to manage financial accounts to their own benefit and to the advantage of the CLAS model if they enjoy a larger role in planning, technical assistance, administrative control supervision, and evaluation of management agreements with various- service providers (Altobelli, 2001). A positive relationship between DISA and CLAS would allow the former to assure that the appropriate transfers of funds to CLAS takes place. The proposed Law of Decentralization correctly provides a legislative framework for achieving DISA strengthening. Recommendations: > The relationship between CTARs and DISAs should change by removing the former intermediary from sector driven purposes. This implies a need for the immediate strengthening of the institutional and human resource capacity within DISAs. > DISA's ability and authority to supervise and control the quality of service delivery should improve. The current nature of the finance mechanism and the lack of control over financing at the DISA level have affected it. > the renewed efforts to consolidate and integrate vertical and virtually autonomous programs implies the need to rethink the planning, administration, supervision and financing roles for DISA, and to restructure these units in an integrative fashion. Further, DISAs should be increasingly demand driven, relying on local diagnostics of needs. 5.35 . A third initiative that needs strengthening is the radical transformation of health insurance with the creation of the Integral Health Insurance Unit, or USIS in 2001. The USIS lies outside MINSA and is an ambitious effort to integrate Maternal Infant Insurance (SMI) and the General School Insurance (SEG) programs under one roof. While at present the two programs-funds remain separate, the current drive is to consolidate them and to universalize coverage, to separate financing from service provision and to strengthen the decentralization process. The model incorporates a variety of service providers, ranging from public health centers to management contracts with CLAS or private local service providers. There are several concerns. First and foremost is the legal status of USIS. Because of the absence of legal autonomous status, USIS cannot act as an executing unit to provide direct financing for services. Rather, most financing by USIS is channeled through the DISA, creating yet another administrative instance in financing. Further, the use of DISA as direct administrator of USIS funds generates delays in transfers to providers (and CLAS) and'has fragmented the capacity to supervise the cost effectiveness of service provision. 1o The control and evaluation functions over 1o Not only have there been complaints from CLAS over delays in transfers, but also that in some cases the DISA simply has not made the transfers to CLAS (Moore, 2002). - 99 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION what is financed and who is providing it rests within the DISA and not in USIS. A second concern is the potential for increased fragmentation in financing local services. Various financing entities provide resources to a local health establishment and the potential for duplication of payments for the same service is real. A third concern is the limited institutional capacity of the USIS. At present, USIS is incapable of providing the necessary technical capacity to supervise or evaluate. In addition, quality control over providers (accreditation) is weak or non-existent. Recommendations: > The solution to the dilemma of indirect financier can be resolved through the granting of legal status to USIS as a Public Decentralized Organism (OPD). It would remove one level of unnecessary administration in financing services, provide strong impetus to the strengthening and extension of the CLAS model, and allow for the separation of finance (USIS) from provision (CLAS, public health centers, private providers) and finance from supervision (DISA). > Fragmentation should be reduced through the consolidation of existing vertical programs within MINSA (para 5.37). > The strengthening of the supervisory role of DISA, as well as of USIS itself would improve the monitoring, evaluation and supervision of both the quality of service and of financial accountability. 5.36 In recent months, there has been a strong focus on attempting to bring some order to the public health system. This focus can be summarized as an effort to consolidate some of the vertical and somewhat isolated programs within the DGSP; a simplification of the central government structure through the reduction of Executive Directorates to three, and renewed interest in decentralization. Recommendations: > It is equally imperative that while pursuing decentralization policies and initiatives, reform of the core of the Ministry also takes place. Up to now, the principal focus of efforts is on restructuring of the central level, and less on decentralization, but both avenues should be pursued simultaneously. There would be a notable improvement in internal communication, whereas previously each vertical program was much of an island unto itself. > Central ministerial restructuring has been initiated, but must continue with a greater eye towards decentralization. Again, the functional matrix model offered in Annex F is an example of how such an identification of structures and functions can be a useful applied tool. > The work of the Commission on a Decentralized Health Care System (Sistema Coordinado y Descentralizado de Salud) is of paramount importance. Its main responsibility is to (i) clarify the roles of the various entities involved in the health care system---MINSA at the central level, DISA, PARSALUD, municipalities, CLAS, and local health centers; and (ii) integrate programs and projects, ending the multiplication of vertical programs and the creation of parallel bureaucracies virtually removed from the core of the MINSA. > Prior review, strengthening the regulatory functions of the central and intermediate levels is increasingly imperative. CLAS is the answer for decentralization of service delivery to locally accountable providers in Per, and provides the key for longer-term models of decentralization. A program of expansion over the next several years without altering the ratio of CLAS to individual health establishments should be pursued, perhaps at a rate of some 5-10 percent per year. 100 Chapter VI UPGRADING THE CIVIL SERVICE 6.1 Civil service reform is the missing link of an outcome- and poverty reducing-oriented modernization of the state in Peru. The past decade witnessed a severe deterioration of Peru's civil service with deep implications for the fiscal accounts and the costs of the pensions regime. Hyperinflation of the late eighties and the beginning of the nineties sharply reduced the real earnings of civil servants. A poorly conceived downsizing of the central government work force in the early nineties, which, by the end of the decade was reversed, led, however, to the departure of the most capable civil servants. The fiscal burden of the surviving public labor force is very high and unsustainable. There is an entrenched disregard for the provisions of the existing civil service law and its associated regulations. Civil servants are employed under a variety of legal regimes with a wide dispersion in salaries for similar tasks. The existing formal salary structure has not been revised in over a decade, so base salaries are increased through resort to a variety of supplements and benefits in cash and in kind. With the proclamation of a ban on hiring into permanent positions ten years ago, institutions have resorted to hiring on a contractual basis, a process that is not subject to any required procedures for recruitment and selection and has severe fiscal implications. All these factors have contributed to the existing legacy of a work force that is poorly prepared, poorly trained, poorly supervised, and consequently of very low productivity. Consequently, there is an urgent need for a thorough reform to re-establish some semblance of control over the wage bill and to promote the creation of a more efficient government Any initiative designed to improve the performance of government must include a major overhaul in the institutional framework that governs public sector employment and in the management of its human resources. The current civil service law has major weaknesses that argue for its replacement by a new legal framework designed to promote a higher level of efficiency in the delivery of government services. A sequencing path of reforms is suggested, starting by a modest, but valuable, pilot effort to rationalize HR management on a ministry by ministry basis and supported by a SIAF-controlled payroll, to the ultimate creation of an autonomous institution or Directorate with the responsibility to administer the civil service system, with normative and enforcement functions. A. THE SIZE AND EMPLOYMENT CONDITIONS OF THE PUBLIC SECTOR 6.2 There is no unambiguous estimate of the size of public sector employment in Peru. The range varies between 616,200 and 890,728, the latest number including members of the armed forces and police, which represents 7.9 percent of the employed labor force.2 If we accept the most conservative estimate of 616,200, it represents approximately 2.4 percent of the estimated population for the year 2000. This ratio closely resembles the average proportion for middle -income countries of 2.5 percent. If total government employment, including police and the armed services, is considered, the ratio to population for Peru is 3.5 percent, which compares favorably to the group of middle-income countries whose ratio is 4.3 percent (Figure 6.1).3 Similarly, as a percentage to the GDP, the wage bill4 is lower than the average in LAC or middle ' This mostly draws from the mimeograph prepared by Gregory (2002). 2 Encuesta Nacional de Hogares, op. cit., p.11 The information for international comparisons is drawn from the World Bank database on Public Sector Employment and Wages. 4 Such a wage bill is significantly underestimated as it is explained below. 101 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION income countries, 6.2 percent. While Peru would compare Figure 6.1 International Government favorably with other countries, it Employment Comparison should be kept in mind that these are poor indicators of whether 30 government employment or 20 remunerations are excessive or inadequate. Such a judgment is 10 66 made only after taking into account 0 the range of obligations assumed by Peru Latin America Middle-Income Countries different governments and the 0 Total Govt Empt (% of pop) 0 Wage bill (% of GDP) efficiency with which they are U Wage bill (% of govt exp.) carried out. 6.3 The fiscal costs of the public servants and the associated pension regimes are very high, bringing rigidity to the budget and absorbing a significant share of it. On the one hand, Peru's wage bill, equivalent to 39 percent of total expenditure and 6.6 percent of GDP, is second highest in LAG, and this ratio does not include public servants hired under other categories. This suggests overstaffing (Figure 6. 1). On the other hand-and this is very relevant for a country suffering fiscal retrenchment, the budgetary cost of the public pension system, which has three components-recognition bonds issued to workers that switched to the private system, unfuinded liabilities under the Oficina de Normalizaci6n Previsional of both current and fuiture workers, and special regimes-amounted to a minimum of about 0.6 percent of GDP in 2002 and is projected to peak up to at least 1 percent of GDP by 2010. 6.4 Furthermore, a negative public perception prevails on the overall performance of the public work force. A WBI survey on governance and corruption in Peru (Chapter VI1) reveals that public servants in Peru rank unfavorably when compared to Honduras and Colombia, in terms of bribes paid for obtaining public contracts and services, bad administration of public funds, and the practice of buying positions in public entities (which favor capture of state by determined groups). Peru ranked better, however, in frequent reporting of corruption in the public administration (Figure 6.2) 6.5 Public employment conditions in Peru's public sector are governed by a legal framework that has rarely been fully applied The 1979 Constitution mandated that a law be enacted that ought to establish merit as the governing principle in definiing conditions of employment, and set up a single structure of remunerations and fringe benefits for all covered employments. Hence, the Civil Service Law and the Ley de Bases de la Carrera,Administrativa y de Remuneraciones del Sector Puiblico were enacted in 1984. Both Laws are basically a statement of principles underlying the employment relationship. The specific norms and procedures regulating that relationship were left to be defined in the Laws' implementing regulations (or Reglamento). The initial Reglamento was published in 1985 and was superseded by a far more detailed Reglamento in 1990. However, for all practical purposes, both Laws have been a dead letter for the better part of their life. Concomitant with a partial reduction in employment between 1991 and 1993, a ban on hiring for permanent pDsitions was decreed. Presumably, all hiring since then has been on a contractual basis and has not been subject to any statutory procedures. Other provisions of the Law, such as the creation of a tribunal to hear appeals by public servants fiom disciplinary actions, have never been fulfilled. Notice that the Constitution of 1993 omits any reference to a civil service law or to the principle of merit in public employment. 102 Chapter VI. UPGRADING THE CIVIL SERVICE 6.6 The present legal framework is weak and the status Figure 6.2 Corruption in Public Services of civil service is chaotic. The Laws did not create a centralized agency with normative and enforcement % of firms reporting capture by the state I functions. Nor did it provide for the creation and maintenance of an administration of public funds employees' database. Therefore, there exists no source within the oitins in the uyin Government of Peru that can provide exact and systematic information about the number of government employees, their employment status % of firms reporting bribes demanded to or histories, and their remuneration. The existing formal salary structure % of users reporting bribes demanded to has not been revised in years and is receive service hopelessly inadequate. In short, there % of firms reporting bribes for obtaining is no reliable way one can trace the public contracts evolution of government employment 0 10 20 30 40 50 60 70 80 and remuneration over the past C Peru Inubli 0 Colombia decade. Because of this vacuum, a Multisectoral Commission to Study the Situation of the Central Public Administration Personnel (henceforth the Commission) was created in 2001 and was charged with examining the issues related public sector employment. The Commission's Final Report provides preliminary information that is available on the state of government's employment as of the first half of 2001. Public Sector Employment and Earnings 6.7 There are four classes of public servants: Er Those who entered under the legal provisions of the existing Civil Service Law (Regime 276). L% Since the hiring ban instituted a decade ago, no appointments to permanent positions have supposedly been made, but ministries have instead resorted to hiring employees under contracts that are also governed by provisions of Regime 276. 13 Another class of contract employees is subject to no regime and is financed from diverse sources, but largely from the budgetary category of goods and non-personal services (NPS). c Because of the restrictive provisions of Regime 276, some decentralized agencies have opted for coverage under the labor code that applies to the private sector (Regime 728). 6.8 The Commission did preliminary major findings concerning employment and salaries in the public sector. o If the two first classes of public employees are aggregated into a single category referred to as Regirne-276 public employees, the monthly average salaries under the three regimes are 1,4396 1,796, and 1,775 soles respectively for an overall average of S/1,500. Earnings of employees under Regime 728 would exceed on average those of statutory employees by 25 percent, while those under NPS would enjoy a 23 percent premium. lz n The high concentration of employees in Regime 276 (82.2 percent of the public work force) is heavily influenced by the inclusion of public servants that do not form part of the 103 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION administrative ranks of the government (Table 6.1). In fact, if we were to deduct from the total employment appearing in the table, teachers, health professionals, judges, and attorneys in the Attorney General's office, the remaining employees would, for the most part, represent the staff of line ministries. When these deductions are carried out, it turns out that only 45.3 percent of the work force within the line ministries is employed under Regimen 276, while almost as many, 40.2 percent, represent contract employees under NPS. So, NPS introduces major distortions to the wage bill and overall budgetary flexibility (para 3.14). Table 6.1 Number of Employees and the Average Monthly Wage Bill Regimen Number of Employees Monthly Wage Bill Thousands Percent Millions of Soles Percent Regimen 276 506.8 82.2 729.4 78.9 Regimen 728 49.5 8.0 88.9 9.6 NPS 59.9 9.7 106.3 11.5 Total 616.2 100.0 924.6 100.0 Source: Comisi6n Multisectorial Encargada de Estudiar la Situaci6n del Personal de la Administraci6n Pililica Central. 6.9 Contrary to common belief, the current employment level is higher than in the early-nineties. The number of public servants accounted for by the Comisi6n, 616,200, is slightly higher than that in 1990. The number of employees ascribed to the national government was 591,498 in 1990 (The World Bank, 1994). An alternative measure of urban employment in public administration, obtained from third-quarter household surveys and data available for 1997- 2001 reports a total public employment of 707,000 (Table 6.2) .6 Table 6.2 Urban Public and Private Sector Employment and Average Monthly Earnings 1997-2001 (nuevos soles) . % Change 1997 1998 1999 2000 2001 1997-2001 No. Average No. Averagc No. Average No. Average No. Average No. Average '000s Earning '000s Earning '000s Earning '000s Earning '000s Earning '000s Earning Public Total 561.6 771 619.7 820 662.5 941 648.6 1030 707.0 1078 25.9 39.8 Permanent 339.8 903 397.1 851 399.4 1042 389.1 1120 433.9 1175 27.7 30.1 Fixed Term Contract 131.8 667 127.4 881 156.8 888 156.6 950 145.0 1134 10.0 70 Other 90.0. 422 95.2 609 106.3 638 102.9 812 128.1 685 42.3 62.3 Private Informal 1001.9 482 1149.6 521 1147.9 521 1089 523 1223.6 443 22.1 -8.1 Private Formal 1217.1 1042 1373.6 1288 1279.9 1369 1305.8 1246 1255.3 1285 3.1 23.3 Source: Gregory (2002). 6.10 Household surveys confirm that the ban on new hiring of public employment proved ineffective as a very significant increase in government employment took place over 1997-2001. All but one category increased their number of public servants. In total, public 6 It should be kept in mind that the survey's measure of public employment is not limited to only the central government, but includes municipalities. However, it would seem that the large difference between the two measures of employment is to be attributed to the inflated number of educators reported by the household survey. Whereas government sources estimate the number of teachers to be in the order of 280,000, the survey reports some 377,000 in 2001. No other source conies close to reporting numbers as large as these. Another reference point is provided by a census of schools undertaken in 1999. It reported the existence of 309 thousand teachers, of whom 235 thousand were public school employees. This compares with the survey's measure of almost 340 thousand public school teachers in that same year. Furthermore, the variations in this number over the interval 1997-2001 are larger than one would expect, and the increase of about 45,000 reported for 2001 over 2000 certainly is unlikely to reflect reality (Saavedra et al, 2001) 104 Chapter VI. UPGRADING THE CIVIL SERVICE employment is believed to have significantly increased by over 25 percent; but if we correct this number by the likely over-estimation of the number of teachers in 2001 (reducing them to approximately the level of 2000), growth in total public employment would still be significant at about 20 percent. 6.11 There is also a considerable divergence between the Commission's estimate of earnings and those reported by the household survey for 2001. The former are significantly higher, ranging from 33 to 66 percent greater than the a-erage public sector earnings reported by the household survey for 2001. This is an uncomfortably large difference that cannot easily be explained, except by the frequent tendency of survey respondents to understate their real revenues. However, another difference appears on reported average earnings for all employees, which advanced by almost 40 percent between 1997-2001. Those in central administration reported an increase of almost 52 percent, while educators reported a more modest increase of almost 27 percent. Since the consumer price index rose by only 15.5 percent between the two end points of the series, public employees realized real percentage increases in double digits on average earnings in such period. Comparison ofPublic and Private Sector Earnings 6.12 The earnings of civil servants in the central Figure 6.3 Employment and Average administration gained Earnings of the Private and Public Sector significantly in relation to 1600 1600 those of private sector W employees7 over the period O 1200 4 1997-2001. According to I household surveys, in 1997, E 800 800 average earnings in the central _ administration were 87 percent 4 400 of those in the private formal sector. They reached virtual 00 parity in 2000 and surpassed 97 98 99 00 01 them by about 7 percent in Public Emp Informal Emp Formal Emp. 2001. Average earnings for --*Public, RHS Informal, RHS -W-Formal, RHS permanent teachers also surpassed average earnings for the private sector in 2001. Average earnings in the total government, also increased, but remained below average earnings for formal private employment and above average earnings for informal employment in the same period (Figure 6.3). These findings differ from the Comisin's estimates; public sector average earnings, including those of educators, appear to range between 30 and 66 percent above those in the private sector. The Commission's number, however, has a larger margin than one would find to exist in most other countries of similar or higher incomes' 6.13 A third more accurate, but conflicting, measure of private sector earnings is derived from establishment surveys that extract data from payrolls. In June of 2001 average monthly 7The data for the private sector employees is for firms with ten or more employees. 8A note of caution is needed in evaluating these 'findings. Household surveys of income and expenditures typically fail to capture all income actually received. Aggregate incomes reported generally fall short of those recorded in the aggregate national income accounts by 15 to 20 percent, and in the case of Peru may be understated by as much as 30 percent. It is reasonable to believe that private sector incomes are more likely to be understated than those of salaried public servants. If so, the reported margin by which public earnings exceed private may be exaggerated. 105 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION earnings were reported to be NS/1,854 as compared to the NS/1,285 reported for private formal sector employees by the household surveys.9 Relative to this value, the average monthly earnings of public employees reported in Table 6.2 are only six-tenths. 6.14 In view of the conflicting data, it is very risky to make definitive pronouncements regarding the actual earnings of civil servants and their relationship to those from the private sector. The safest statement that can be made is that public sector earnings appear to have gained relative to those in the private over the interval 1997-2001. Since the establishment surveys are a more accurate source of information regarding private Rctor earnings, it can be concluded that while the gap between the two sectors has narrowed, public sector earnings are probably still somewhat below their private sector counterparts0. Earnings by Occupational Category ofPublic Employees 6.15 The salary payroll is highly inconsistent and fragmented, as the official payroll barely covers less than half the real one. The formal payroll of the central administration for employees covered by Regimen 276 is called the Unified Payroll (Planilla Unica de Pagos, or PUP). The Comisi6n's study reports that only 41 percent of all salary payments are charged to the PUP. All the rest of the payments are made by resort to over 40 different salary supplements in cash and in kind! For example: incentive payment, vacation bonus, birthday bonus, consumption subsidy, family food basket, monthly bonus, transportation, schooling subsidy, uniforms, etc. The result of these ad hoc arrangements is a total loss of any transparency and coherence in the formal salary structure. Employees in the same salary grade or performing the same tasks receive widely disparate salaries and the salary relationships existing among the different levels of the salary structure obey no consistent or transparent relationship to the skills required 10 perform the assigned tasks as reflected by the minimum and maximum salaries in each salary grade that are financed out of the PUP (Table 6.3). o The ratios of the highest to the lowest salary within grades vary greatly. For example, the maximum salary h grade Fl is 5.55 times greater than the minimum while in F7, the ratio stands at 1.36. o There is no perceptible order in which this ratio moves along salary grades from the minimum to the maximum and the overlap of salaries is extreme. For example, the maximum salary in the bottom grade (SA-E) is greater than the maximum salary in ten of the next 11 higher grades." Indeed, it is greater than the maximum recorded for vice-ministers and secretary general! 9 Ministerio de Trabajo y Promocidn Social, Encuesta Nacional de Sueldos y Salarios as reported in hforme Estadistico Mensual, Agosto 2001, p.39. 0 The importance of the level of public sector remunerations relative to the private sector should not be underestimated. According to World Bank's 1997 World Development Report (p. 8), there is a clear and strong inverse relationship between public sector wages relative to those in manufacturing and the incidence of corruption. The higher public sector salaries are relative to those in manufacturing, the lower is the incidence of corruption. Regardless of whether one refers to ENAHO's average monthly earnings in the central administration in 2001, S/1,379, or the Comisi6n's average, S/1,500, both fall far short of the average monthly wage in urban manufacturing in June, 2001 of S/2,337. " However, the ComisiOn report notes that this particular anomaly is not normally observed within institutions. 106 Chapter VI. UPGRADING THE CIVIL SERVICE Table 6.3 Dispersion of Salaries Charged to the PUP by Occupational Group and Salary Grade (Nuevos Soles) Occupational Salary Maximum Minimum Ratio of Group Grade Value Value Max/Min F8 2,594 1,326 1.96 UPPER MANAGEMENT F7 1,574 1,154 1.36 F6 1,651 1,059 1.56 F5 1,928 1,005 1.92 F4 1,458 964 1.51 MIDDLE MANAGEMENT F3 2,695 906 2.97 F2 2,934 669 4.39 FI 3,805 686 5.55 SP A 2,670 651 4.10 SP B 1,608 620 2.59 SP C 1,798 600 3.00 PROFESSIONALS SP D 2,469 576 4.29 SP E 1,382 551 2.51 SP F 1,158 598 1.94 STA 1,960 526 3.73 ST B 1,725 518 3.30 TECHNICAL STC 816 520 1.57 ST D 788 508 1.55 STE 778 507 1.53 SA A 1,765 502 3.52 SA B 1,555 500 3.11 AUXtLiARY SA C 1,748 491 3.56 SA D 1,052 498 2.11 SA E 1,803 480 3.76 Notes: Grade F8 does not include the President of the Council of Ministers. Some Secretaries General (F6) have a rank of Vice-Minister (F7). Source: ComisionMultisectora4 op.cit., p.2. 6.16 On average, contracted personnel earn more than permanent, except for auxiliary ones, but salary dispersion is high (Table 6.4 and Figure 6.4). Peru's civil service regime appears to offer higher-than-market rates of pay for employees at the bottom of the occupational hierarchy while paying less-than-market salaries to highly skilled employees. Managerial personnel with permanent status under 276 appear to earn less on average than do those contracted under the same regime, but the Comisidn's report comments that more disaggregated data by salary grade reveals that the permanent managers actually earn more. Table 6.4 Maximum, Average, and Minimum Earnings by Occupational Group and Regimen (Nuevos Soles) Permanent Contract Min Avg Max Min Avg Max Upper Mgmt 13634 23360 43083 Middle Mgmt 700 3377 12438 932 4047 5016 Professionals 684 1943 5554 764 2246 6770 Technical 679 1373 3816 579 1499 3488 Auxiliary 498 1209 3779 578 1106 2416 Source: Gregory (2002). 107 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Figure 6.4 Permanent and Contractual Average Earnings (Nuevo Soles) Auxiliary Technical Professionals Middle Mgmt Upper Mgmt 0 5000 10000 15000 20000 25000 9 Permanent 0 Contract 6.17 A wholesale reform of employment and remuneration practices in the public sector is in order. Public employment is growing rapidly and uncontrolled. Salary levels are being assigned without reference to any benchmarks. There is no mechanism for assuring that individuals performing similar tasks at similar levels of productivity receive similar pay. B. REFORMING THE LEGAL AND INSTITUTIONAL FRAMEWORK 6.18 Neither the Laws, nor the Reglamento establish procedures and offer benefits to employees that are more likely to discourage rather than encourage efficiency: L A glaring omission of the Laws is the absence of any entity responsible for ensuring its enforcement. The 1985 Reglamento leaves enforcement to the human resources offices at the ministry- level.'2 This is a weak reed on which to place such a responsibility. Typically, the human resources offices are poorly staffed and exercise little authority. To the extent that their Directors serve at the pleasure of the higher authorities, there would be little disposition to oppose the latter should they wish to by-pass procedures outlined in the law or regulations. O Upon appointment to a post, a civil servant immediately acquires tenure since the 1990 Reglamento expressly rejects the use of probationary periods. O The Reglamento seems to reflect a determination to move employees upwards through the occupational and salary structure as quickly as possible. It calls for setting annual quotas for promotions based on either existing vacancies or by the reclassification of a servant's existing post. Rather than restricting access to promotion to maintain a strong incentive to excellent performance, it creates a mechanism that risks converting the promotion into a pro forma event. As the Reglamento seeks to fill all vacancies above the entry level to an occupational group from within the existing personnel pool, this limits the flexibility required to ensure that the best possible available candidate is selected. 6.19 The procedures outlined for the recruitment and selection of new employees conform to standard practice in theory, but prove ineffective. Vacancies at the entry level 12 Reglamento Inicial del Decreto Legislativo No. 276, Decreto Supremo No. 018-85 PCM, Disposiciones Generales, 2. 108 Chapter VI. UPGRADING THE CIVIL SERVICE must be advertised and selection follows a competitive process in which antecedents are evaluated and examinations may be required. But there are loopholes. For example, individuals hired on a contractual basis are not required to undergo a competitive selection' process and after one year they may be granted permanent status if a vacancy exists. Then, after three years on contracts, an individual acquires a right to become permanent if the institution decides it needs to retain the position (Art. 40). Clearly these provisions are inconsistent with procedures to ensure that the best candidates are hired into tenured positions. 6.20 Monetary rewards are merely a function of tenure. Article 51 grants a five percent increase in the basic salary after each five years of service. Artic le 52 provides for bonuses that are a function of family size. Both provision violates the principle of equal pay for equal work and certainly have no link to performance. Article 53a provides for a bonus to be paid to those occupying positions of responsibility or of supervision. No definition of "responsibility" appears either in the Laws or in the Reglamento, so that it obviously lies within an area of discretion of the institution's authorities". Moreover, articles 140-46 of the Reglamento oblige all government institutions to provide a variety of welfare benefits extending to the employee and his/her family. Among the benefits to be provided are meals, transportation, in-house medical services, housing, nurseries and educational subsidies, cultural, social and recreational activities, subsidized credit, and others. The Need for a Civil Service Directorate 6.21 Peru intends to enact a new Civil Service Law and Reglamento designed not only to protect the legitimate rights of employees, but also to create the incentives for a higher level of efficiency and an improved quality of services to the public. Recommendations. It is advisable that a new Civil Service Law establishes a central entity-a Directorate-with authority for normative and enforcement functions. The Directorate must ideally be autonomous from the executive power. The person who heads the Directorate should be viewed as apolitical as possible and have had professional experience in the area of human resource management. The qualifications for the position should be clearly set forth in the law'4 The Directorate could be assigned a variety of functions designed to ensure a measure of consistency in the administration of human resources across the various institutions of government. Among these could be: > Prepare the Reglamento of the new civil service law; > Prepare a model Reglamento (subject to the approval of the Council of Ministers), which would serve as a basis for institution-level Reglamentos; 1 Why a salary supplement should be necessary is not clear. After all, the job evaluation process is supposed to take account of the duties and responsibility that characterize each post and assign it to the appropriate salary category. No further supplement should be required. But even less understandable is the granting of a permanent bonus in recognition of past supervisory services after one moves to a non-supervisory position. The Reglamento provides for this salary supplement for those serving in a managerial position for at least 3 years though it does not specify what proportion of the bonus becomes permanent (Art. 124). 14 The appointment process that is to be followed is of the utmost importance. One possible appointment mechanism is the constitution by law of a nominating panel with representation from Congress, civil society organizations and the business sector, which would forward a short list of approved candidates to the President from which he or she would make the appointment. The Director (Rector) should enjoy a rank equivalent to that of a minister and hold such position for a term longer than that of the president. Removal from the position would require a demonstration of incompetence. In the absence of such a demonstration, renewal of the director's term would be expected. 109 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION > Create and maintain a database containing personnel information of all government employees; > Design of a single system of job description and evaluation that could be applied across all government agencies; > Design and implement the training program of the personnel assigned to the human resources offices in the public entities; > Provide technical assistance to the institutional human resource offices; > Assist ministries in assessing training needs and training programs, maintaining a listing of qualified sources of training; > Design an appropriate wage structure that permits government to recruit and retain employees of the desired quality; > Perform periodic surveys of the labor market to identify salary trends on the basis of which it could offer policy recommendations to the executive designed to urintain public salaries in the desired relation to the market; > Perform technical analyses of public sector employment. Personnel management 6.22 The procedures outlined in the existing Reglamento are, for the most part, adequate, but could be improved They prevent opening all vacancies to competitive qualification by both, internal and external candidates. Some greater degree of regulation of the process of contracting would appear to be desirable, since it is an area that could give rise to abuse. Currently, an institution can contract anyone it wishes without resort to a competitive process of qualification. Recommendations: > Contracts with a duration of less than a year, the current practice be followed, but contracts should be non-renewable to prevent avoidance of the competitive process. > Contracts for longer periods of time should be filled through a competitive process. Holders of longer-term contracts should also be entitled to all the other benefits, e.g., paid vacations, participation in a pension scheme, etc. > Consideration must be given to employing occupants of all professional and managerial positions on a contractual basis rather than on a permanent basis. To compensate for the increased risk of separation, a premium could be offered over the position's salary grade, e.g., ten percent, to those opting for the contract option. The increase in flexibility afforded by the contractual arrangement could well be worth such a premium. 6.23 Remunerations and Other Incentives to Perform. The Directorate should be made responsible for elaborating a single market-oriented salary structure designed to accommodate the full range of services employed in government from the unskilled service worker to the minister. Recommendations: > Budgetary provisions should be made for payment of bonuses to employees as rewards for superior performance. 110 Chapter VI. UPGRADING THE CIVIL SERVICE > The budget should allocate to each institution some proportion of its wage bill for the distribution of such bonuses. Each institution would be responsible for administration of the bonus. > Permanent payments that are unrelated to performance, e.g., length-of-service supplements, should be avoided. If it is desired to grant some recognition for length of service it would be preferable to grant a lump sum award when recognition is due rather than make a permanent addition to the basic salary as the current law provides. > The new Civil Service Law should provide that salary increases mandated for all employees be made in proportional form, i.e., an equal percentage increase for all. The common practice of awarding the same absolute increase to all employees has the unfortunate consequence of reducing relative differentials among salary grades, leading to a compression of the salary structure and to future difficulties in the recruitment of high-level skills. It also reduces the incentives to perform in order to better qualify for promotion. > All new employees should be subject to a probationary period during which they may be discharged without payment of an indemnity. 6.24 The Implementation of a New Civil Service Regime. The new Law and salary structure should not be extended to the whole central government immediately upon its enactment, except for the general provisions governing recruitment and selection, but start with modest human resource (HR) management reform pilot exercises, developed on a ministry by ministry basis. The government ought to use accession to the law and the new salary structure as an incentive for each institution to undergo a careful review of its mission- and structures, and their adequacy for meeting the future institutional needs. Recommendations: > A basic HR restructuring process would include (i) the discontinuation of activities that do not fit within the newly defined mission of the institution; (ii) the elimination of duplications and consolidation of units; (iii) the re-engineering of processes and procedures to simplify the delivery of services, completion of new job titles, descriptions and evaluations; and (iv) a commitment to the professionalization of management. > Pilot institutions should be required to prepare a detailed plan of action and a calendar for its completion to qualify for coverage under the Law. > An appeals procedure is needed for any disciplinary system. The Ley de la Carrera Administrativa provided for the creation of a body to hear appeals (Tribunal del Servicio Civil). However, it was never established, so it is a priority. Such an entity should lie outside the institution. 6.25 The Scope of a Civil Service Law and decentralization. A final note concerns the desirable scope of a new civil service law. The central government is committed to processes of de-concentration of administrative functions of ministries to regional or local political entities. At the same time there is a potential devolution of functions and resources to regional or municipal governments. Recommendation: > It is imperative that the staffing of subnational units of government be subject to the same rigorous standards and procedures advocated for the central government, if decentralization is to result in better economic and social services to the public. 111 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION C. ADDITIONAL MEASURES FOR IMPROVING HUMAN RESOURCE ADMINISTRATION 6.26 There are additional features that an overhauled human resource administration would require: > A virtual database itegrated into the SIAF system would support the main urgent task of developing a unified payroll system that allows to know the exact number of public workers and to eliminate ghost workers. Any high performing system of personnel administration requires it also containing a full profile of each employee including: a public sector employment history that records data of entry, positions held, disciplinary actions, classification, salary, supplements, etc. > Gradual simplification of the remunerations package to cash payments as every effort should be made to avoid in-kind supplements. The resort to a wide assortment of supplements to the basic wage gives rise to difficulties in accounting for the full cost of the remunerations package. Typically, individual welfare or satisfaction is maximized when an individual receives all of his income in cash and can allocate it according to his consumption preferences. > The current variety of payment practices should gradually be replaced by a wage structure that provides for a single basic salary for grade and assures that pay differences among employees performing the same tasks are minimized. > Pension regimes require an ad hoc solution. If a new salary structure were to result in increases in salaries the government would be faced with an immediate increase in its liabilities to the pensioned population as well as to its employed work force. There are currently some 225,000 beneficiaries of the pension system established by DL-20530 whose pensions are automatically adjusted by the current salary of the positions they occupied prior to retirement, the cedula viva arrangement. > A true system of performance evaluation has yet to be established. The current Ley de la Carrera Administrativa adopts merit as a criterion for promotions and for granting performance bonuses, but it has been ignored. Depending on the nature of the tasks, a performance bonus is desirable. > A prior condition to the introduction of performance contracts is the professionalization of management. Indeed, the introduction of trained managers can lead to improvements in productivity even in the absence of performance bonuses. Able managers can raise productivity by improving the organization of their units, modifying and simplifying work procedures, and by motivating employees to perform. It is desirable to make a clear distinction between the policy-making and the executive functions of the ministerial hierarchy. The General Director and all lower-level directors and supervisors would constitute the backbone of the permanent professional management structure D. THE EDUCATION SECTOR 6.27 The education sector is a major destination of government resources and public school teachers constitute, by far, the largest single group of public employees. The education sector accounts for about 15 percent of total expenditures and deserves special attention because teachers are not covered by the civil service law that applies to administrative employees but are governed by their own legal regimen. The institutional framework is defined by two basic documents, the Ley del Profesorado, enacted in 1984, and the Reglamento de la Ley del Profesorado, which dates back to 1990. 112 Chapter VI. UPGRADING THE CIVIL SERVICE 6.28 There are various categories of teachers defined by the nature and amount of education and/or training that they possess. Those who hold diplomas in pedagogy from one of the teacher training institutes, (institutos superior pedagdgicos, or ISPs) occupy the highest category. In addition, there are five categories of teachers who do not hold a teaching title." In 1997, 62 percent of the approximately 248,000 primary and secondary school teachers held pedagogic degrees. The distribution among schools of teachers with different qualifications is hardly random. The more rural and distant from urban areas a school is the greater the likelihood that it will employ a larger proportion of untitled and poorly prepared teachers. 6.29 Teachers' salaries have shown huge variations, in real terms over the past few decades. They attained their peak during the decade of the '70s, declined precipitously during the '80s, and recovered during the past decade. By the end of the last decade, teachers' earnings, in real terms had recovered to a significant level. Their salary structure had also been severely compressed. 6.30 Teacher's present assignment mechanisms provokes Misallocations. The assignment of a teacher to a post vests an ownership right to the post. A consequences of this is a serious misallocation of teachers. As enrollments shift from one school district to another as a result of population changes, schools in districts losing students will find themselves with a surplus of teachers, but the surplus cannot be reassigned and the teacher may retain his/her post as long as he/she wishes. 6.31 Promotion procedures are inoperative, since they were suspended a decade ago. The result is that most teachers are found in the bottom two levels of the career structure.'7 6.32 The current salary structure does not provide adequate incentives that can elicit a more efficient performance from teachers. Consider, for instance, one of the categories that is likely to encompass a substantial number of teachers, namely, the category that includes those under the pension regimen 19990 and who are appointed for 30 hours of classroom time. The difference between the salary in Level I and that in Level V is only S/67.75, or barely ten percent of the starting salary in Level I.8 The differentials for the other categories are of a similar magnitude. 19 '5 The first of these consists of individuals who have completed their pedagogic studies but have not completed the degree requirements. A second category recognizes those who hold degrees in non-pedagogic fields from institutions of higher learning other than the ISPs. The third is for those who have not completed pedagogic studies at a higher level of education. Fourth are those with incomplete university training. Finally are the teachers with only secondary school education. 16 The World Bank, Peruvian Education at a Crossroads: Challenges and Opportunities for the 21' Century (Washington, D.C. 2001), p.47. 17 Teachers with pedagogy degrees have in theory a defined career path consisting of five levels, each of which has a salary associated with it. A teacher must spend at least five years in a level prior to becoming eligible for promotion to the next level. The passage from level I to level II is automatic at the end of five years. From then on, promotion depends on the evaluation of the teacher. The evaluation is based on three categories of factors, namely, (i) education and degrees obtained, the number of training courses attended, papers presented at professional meetings, length of service, and posts held; (ii) an evaluation of services that is based on attendance and punctuality and participation in joint faculty activities; and (iii) meritorious recognition as evidenced by opinions from the state, parents' organizations, or by intellectual production. 8 The discussion of salary differentials is based on the salary structure in effect in 1999 and does not reflect the 16 percent increases instituted since then (The World Bank, 2001). 19 Last year, a 50-sol increase was extended to all teachers, the first of similar increases promised over three years. This practice of granting equal absolute increases across the board to all teachers leads to an even greater compression of the salary structure. 113 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 6.33 The adequacy of teachers' salaries is a controversial issue. Comparing the remunerations received by teachers and professionals in the private sector using several different measures of pay for 1997 (Table 6.5) shows that if no allowance is made for hours worked at home by teachers, the hourly earnings of teachers fall short of those of private sector professionals but exceeds those of non-professional workers with a university education. The disadvantaged position of teachers relative to the labor market is mainly due to the higher earnings reported in Lima. In the rest of urban areas, teacher's earnings exceed those of other workers with advanced education. If it is assumed that teachers spend an extra 40 percent of classroom hours in school-related work at home, the disadvantaged position of teachers is 20 accentuated in Lima, though in other urban areas, the differences are narrow. Table 6.5 Hourly Earnings by Occupational Groups and Areas, Third Quarter, 1997 (Nuevos Soles) Teachers' Earnings Earnings Teachers' Weekly Adjusted for (Assuming 20 Earnings (Assuming Area and Hours Hourly Weeks Hours Worked at 40 Hours Worked at Occupation Worked Earnings Worked Home) Home) Urban Peru Teachers 26.7 5.7 7.2 6.0 5.2 Professionals 42.2 8.1 8.2 8.2 8.2 Univ. Educated 42.3 6.8 7.0 7.0 7.0 Lima Teachers 25.4 5.8 7.3 6.1 5.2 Professionals 42.1 10.0 10.2 10.2 10.2 Univ. Educated 42.3 8.3 8.5 8.5 8.5 Other Urban Teachers 27.2 5.6 7.2 6.0 5.1 Professionals 42.4 5.8 5.9 5.9 5.9 Univ. Educated 42.3 5.1 5.2 5.2 5.2 Source: Saavedra, Jaime (2001). 6.34 Whether teachers are earning salaries that approximate or exceed their opportunity cost, is an inadequate guide to the salary policy to be pursued in the educational sector. If the goal is to improve the quality of education one would think that the first objective should be to improve the quality of the teachers entering the profession. To the extent that a higher quality teacher has a higher opportunity cost than the current cadres, it will be necessary to offer higher salaries in order to attract him/her. However, it makes little sense to raise all existing teachers' salaries to the levels required to attract a more promising teacher. What this suggests is that the remunerations policy of the Ministry should be tailored to fit more closely the opportunity costs of those entering the teaching profession. A New Institutional Framework for Human Resources in the Education Sector 6.35 Taking a closer look at other provisions of the Ley del Profesorado and the Reglamento allows to offer suggestions that might be incorporated in a new legal framework. 20 The Reglamento provides for a monthly bonus equal to 30 percent of total remuneration for class preparation and it is assumed that the earnings figures reported include this bonus. 114 Chapter VI. UPGRADING THE CIVIL SERVICE Recommendations: > The Ministry of Economy and Finance needs to establish a central employment registry integrated into the SLAF, similar to the one advocated for the civil service sector. This would establish a general single payroll management system, supported by banking transfers. > The remunerations package should be simplified as it would be more transparent and equitable for the sums expended on fringe benefits to be distributed in the form of salary and allow employees to purchase the benefits or services they would wish to enjoy. The current Ley and Reglamento are replete with a wide variety of fringe benefits and bonuses. There are ten provisions for leave with pay; discounts available on public transport and for state- sponsored cultural events; 20 percent of all publicly financed housing is reserved for teachers; preferential access to mortgage credit; automatic annual two percent increases in basic pay' family allowances; bonuses for length of service; payments upon the death of a teacher or his/her spouse, parent, or offspring; funeral benefits; upon retirement, a bonus in recognition of years of service; a 30 percent monthly bonus for class preparation; a five percent bonus for directors and university professors for performing their tasks; funds for the construction of facilities for the delivery of social services; and for cultural and entertainment purposes, among others. However, the value of these benefits to employees is positive only to the extent that they receive them. Not all employees receive them in equal amounts and, therefore, the compensation package has a different value for different individuals. . > A new Law should also place a heavy emphasis on merit as a condition for entry to the profession and as a determinant of compensation. If committees are viewed as desirable to review the salary recommendations of principals or supervisors, these should be composed of persons of similar rank, perhaps with the participation of parent associations. Union officers tend to be not appropriate members of such committees. > A more intensive use of contracting of teachers' services is bound to provide greater flexibility in the allocation of posts. Contracting would permit the reallocation of positions from schools with declining enrollments to those districts with expanding student bodies. It would also facilitate the process of relieving ineffective teachers of their duties. If, however, contracting of teachers were not to become the standard form of employment, then the procedures for hiring teachers and granting tenure ought to be reformed. Because the removal of ineffective teachers is more difficult under a system that grants tenure, it becomes imperative to place greater emphasis on the way teachers are selected and evaluated. The now practice of holding qualifying exams for entry into the field should be made a legal requirement. Given the large number of potential applicants, the Minitry can afford to establish high standards for entry to the field. We believe it is important that tenure not be granted upon entry to the profession, but rather every new teacher should be subject to a probationary period during which performance can be evaluated. > A new Law should redefine the rationale of a new salary structure for the profession and avoid bonuses or other payments that are unrelated to performance. For example, the current provision of automatic yearly increases rewards the good and the poor teachers equally and, therefore, provides no incentive to improve performance. Effective attendance to school should also be taken into account, so as to prevent teachers' absenteeism. > A suggested sequencing of reform is depicted below (Box 6.1). 115 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Box 6.1 The Sequencing of Civil Service Reform Measures Stage Institutional Actions Civil Service Actions I. Creation of a Reform Commission. II. Issuance of invitations to pilot individual Preparation of new civil service law. institutions to prepare and submit plans for human resource management reform. III. Preparation of institutional reform plans. Submission of law to Congress. IV. Review and approval of reform plans. Enactment of civil service law. V. Creation of management structures in to- Creation and staffing of.Civil Service Directorate. be-reformed institutions. VI. Management training. Start integrating Preparation of internal Reglamento payroll into SIAF by pilot institutions VII. Begin the institutional restructuring Prepare model Reglamento for adaptation to process.. institutions. VIII. .Identify the positions and job content Upgrading directors and staff of institutional required to meet programmatic objective human resource offices. Training. TX. Prepare internal Reglamento Select methodology forjob description to be applied throughout the government and train institutional human resources department personnel in application of methodology. X. Perform job descriptions and evaluations; Prepare model salary structure that reflects market specify qualifications for filling each rates of pay. post; formalize in manuals. XI. Staff decisions followed by retraining or Apply civil service law provisions to reformed recruitment of new personnel. Separation institutions. of redundant personnel with severance payments. XII. Personnel training. Design a personnel information system with standard personnel forms and the networking of institutional information systems with a Directorate's centralized database. XIII. Fully integration ofipayroll into SIAF. 116 Chapter VII IMPROVING GOVERNANCE AND REDUCING CORRUPTION 7.1 The collapse of the Fujimori Administration in the middle of corruption scandals raised awareness in the country about the need for a long-term pro-governance and anti-corruption strategy. Mr. Fujimori's third term was short-lived: mounting evidence of corruption by close associates forced Fujimori to resign. A Constitutional transition took place. Mr. Paniagua was converted from President of Congress into Interim President until July 2001, when elected President Toledo took over the Presidency. In April 2001, the GOP established the "Iniciativa Nacional Anti-Corrupcion - INA, " a committee with the task of promoting a national workshop with representatives of major agencies in charge of the fight cgainst corruption and developing an anti-corruption strategy. With the support of the World Bank and other international donors, the GOP requested a diagnostic study ofcorruption in Peru, whose results follow below' 7.2 Peru's most recent scores on governance are mixed There is strong evidence of a causal relationship between good governance and better development outcomes, including higher per capita incomes, lower income mortality, and higher literacy. World Bank researchers recently compiled a massive cross-country database of some 300 governance indicators, yielding six clusters of composite measures: voice and accountability, political Figure 7.1 Perus Governance Indicators instability, government efficiency, 1998/2001 regulatory burden, rule of law, and Global Ranking - lower is better corruption for two years, 1998 and 2001. Peru's scores in Government effectiveness and regulatory quality have dropped significantly as the country moves from an authoritarian- U. - like regime to one with a more vocal and active opposition (Figure 7.1). Its overall score in corruption remains 0 m Voice and Politicat Government Regulatory Rule of Law Corruption approaching the LAC average, but Accountability Instability Effectiveness Quality some scores are in the upper end, 01998 I2001 appearing particularly high-close to 60 percent-in procurement contracts of public works. To offset such a gloomy picture, Peru's scores in voice and accountability, political stability and rule d law-have improved, thus reflecting the political opening that is taking place, but significant challenges remain in these areas too. This section summarizes the World Bank Institute's study on governance, rule of law and corruption for Peru (The World Bank, 2001c). This study was carried out between February and March 2001 and, although present Authorities have indeed taken actions to address some of the issues below identified, their persistent status allows change only to take place slowly. Is main objective is to support the anticorruption effort and GOP's commitment in Peru. The rich survey data collected facilitates the process of unbundling the many faces of corruption across governmental agencies and regions, from administrative corruption associated to laws, rules and regulations, to state capture by vested interests affecting their actual design and implementation. Overall, Peruvian citizens (1696 public service users), 401 enterprises and 1123 public officials were forthcoming in responding. The refusal rates for each of the three surveys were low and, particular among public officials, very low: barely 11.3% of interviewees did not complete the survey. 117 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 7.3 The evidence provided by the Box 7.1 Regulation of Infrastructure in Peru survey indicates that Peru faces, however, Any evaluation of the effectiveness of a regulatory framework would need to serius orrptin c alle ges(Fiur ~'~ be based more on input that output measures since the latter, although usually seriouses. being the more appropriate, their measurement and interpretation in and 7.3). More than 85 percent managers of regulation is non4rivial. On the input side, the evaluation is based on the enterprises interviewed rated corruption as legal, institutional, organizational, and procedural elements. T e In Peru, the legal and instituuional elements of the regulatory framework are the most serious problem they face. Thegood and above average in the Region. Regulation is grounded on sector- uncertainty surrounding the decisions over specific laws and complemented by concession contracts. The institutional component is good, with separate sector agencies; a single regulator but an the judicial system, the low quality and advisory board, staffed by mostly professionals and not politicians; a budget inefficiency of public administration, financed by sector firms contributions-fees; pay scales comparables to private a mis lon, sector; and all the tudes of transport regulated by a single agency, a most unstable economic policies and tributary desirable structure and a rarity in the Region. On the weak side is security of employment for the regulator, which could be stronger, and the appeal instability are alo perceived as a very process that relays on the judiciary, but that is also the case practically serious problem by more than one-half everywhere else. On the organizational element, the agencies also score well. S They are well structured by functions and sectors, and staffed by competent enterprises. Comparing various dimensions professionals. Perhaps their major weakness is on the procedural ide, in the of governance and quality of service sense that the framework does not give enough guidance on the process so as to inspire confidence and transparency. For example, there is little to compel delivery across 4 regions (Lima, Selva, the agency to explain or justify its timing and process to reach decisions, and Sierra ad schecks and balances are scarce. Siera an Reto .de osta, g vernnce The sector is evolving so as to improve those components. A number of appears uneven across Peru, with the Selva deficient structural elements have been corrected such as eliminating region ranking systematically higher than operators from the regulatory board at OSIPTEL and merging the institutional and the tariff agencies on the energy sector. Clarifying the roles of the the other three regions. How costly bribery Antitrust agency and the regulatory agency, at least on energy issues, is is for the competitive enterprise sector isneeded. One weakness identified has been the managerial instability of the is fr te co pettiv entrprse scto is transport regulatory agency, OSITRAN, which has had five Presidents in four suggested by the fact that most enterprises years. On implementation, or outcome measures, the agencies score rather to well. They used level of conflicts, delays, price levels, quality of service, are prepared to materially contribute t compliance with contract and so on, as evaluators. The agencies controlling corruption. Enterprises report ineased their administrative capacity, trained their personnel, and developed that they are willing to offer on average 6.8regulatory instruments. thatthe ar wilin to ffe onaveage6.8 Regarding their financial and operative autonomy, Peru's regulatory agencies percent of their gross revenues in order to have still to be improved. The institutional base of their autonomy is fragile. eradicate corruption. This figure is higher This is, fundamentally, because of the absence of effective checks on possible eradcat corupion Thi fi ureis ighr dcisions coming from the Executive branch, which undermine their effective only to improve the administration of the independence. In previous governments, when presidential objectives shifted tx away from efficient, fair, and predictable regulation and toward other fiscal system. This large potential ta fisal ystm. his lare p te t a x priorities, the funding of these agencies declined and their scope narrowed. contribution is not only an indicator of how This has to do with two flaws in their institutional design. First, they enjoy only limited budget independence. Secondlly, these agencies' chief executive costly corruption is financially for many officers and board members are vulnerable so being removed by the President competitive businesses, but also it is or by members of the ministerial cabinet for essentally any reason. They have potential powers so he very effective, if the regulator has that attitude, indicative of the fiscal losses to the Treasury and they have shown in general that they have it. An indication of their autonomy has been the continuous claims and complains by the Peruvian if a more transparent system were to exist. Executive and Congress, since 1996, that regulatory agencies have too much Performance by regulatory agencies is power, too many resources, and are accountable so no one. The limited autonomy these agencies have is fragile. The current government was 1 initially inclined so reduce it. Ye isis tbe commended for having adopted and open and competitive process for the selection of the heads of the bregulatory agencies. eAll in all, the perception about the performance of regulations in Peru remains administration of justice are particularly above average in the region. The rankings for the World Economic Forum's 2002 Global Competitiveness Report placed Peru among the top four afflicted, other government institutions countries in the LA region. This framework, however, remains fragile to appear to exhibit good governance and eventual institutional and legal changes, and can be improved, but future changes need so be done with transparency and with predictability, honoring are highly regarded. In all three types of procedures established in the contracts, so as not to increase investors surveys and respondents, it is reported that concerns about changing the rules of the game. Congress and rules-of-elaw institutions rate among the most corrupt state agencies, as well as core public security institutions (such as the Police, the Army, and the Judiciary (Figure 7.3). About 90 percent of enterprises and users of public services believe that the judicial power is neither independent from the government or political groups, nor do they believe that justice is administered in a fair, just, or transparent manner. In addition, enterprise managers report that bribes are used predominantly to facilitate judicial proceedings. In contrast, the Ombudsman, National Bank, CONASEV, and provincial authorities are not regarded as misgovemed in general-with the notable exception of municipal fiscal/tax collection services, which were not viewed positively. 118 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION Figure 7.2 Obstacles to Doing Business Corruption Insecurity - delinquency Trafficking Uncertainty generated by the Judicial Power Low level of efficiency and quality of bureaucracy Unstable economic politics Frequent changes in public politics Tributary instability Low levels of technological development Financial access and cost Monopolies Inflation Tax administration Difficulties in 'competitiveness" Cost of regulations and transactions Terrorism Inadequate infrastructure Requirements to process exterior commercial operations Legal procedures to establish an enterprise Regulations for forex transactions 0 20 40 60 80 100 Share of respondents considering it a very serious obstacle Figure 7.3 Perceived Level of Corruption in Selected Agencies (as ranked by users) Congreso de Ia Republica Poder Judicial Ministerio de Justicia Policia Nacional del Peru Fuerzas Armadas JNE Consejo Nacional de Ia Magistratura Ministerio de la Presidencia ONPE Prefectura Local authority (Municipal) Ministerio de Salud RENIEC Registros Publicos Defensoria del pueblo Local authority (Province) Banco de Materiales Conasey Universidades Publicas Banco de Ia Nacion a 10 20 30 40 50 60 70 80 % reporting that the agency is very corrupt M Officials N Users I0 Firms 7.5 Bribes are used by firms predominantly to obtain public contracts, deal with the Judiciary, and obtain licenses and permits. To facilitate judicial procedures and obtain licenses, enterprises pay bribes at least twice more often than for import/export documents and 119 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION other procedures (Figure 7.4). About half of all enterprises always pays bribes to win public contracts (Figure 7.5). This number rises to around 67 percent among foreign enterprises. Not only does this estimate of public contracts tainted by this problem reflect widespread and large- scale corruption, but also the reported percentage bribe 'fee' expected to secure such contracts is also rather large: enterprises report that on average, regardless of their ownership structure, 17 percent of the value of the contract is paid as a bribe. Further, in terms of time spent dealing with the public administration, foreign enterprises appear to spend more time than domestic firms. The poor performance of the Judiciary leads to the use of alternative mechanisms for conflict resolution. In addition, both managers and public officials report that bribes are paid to judicial authorities significantly more frequently than to guarantee contracts or obtain public services. Figure 7.4 Bribes by Type of Activity (Based on enterprise survey) To obtain public contracts r 4 To expedite judiciary proceedings To obtain licenses and permits To obtain contract with state companies To update tributary statu - Foreign To obtain public services a Domestic To obtain import and export permits 0 20 40 60 80 100 % reporting that bribes are very frequent The thin lines represent margins of errors (or the 95% confidence intervals) for each value. They show the range where the true (population) proportion would lie with probability 0.95 if we had drawn a random sample from this population. Figure 7.5 Corruption In Public Service (as reported by public officials, users, and managers) Average % of users that were asked to pay a bribe while 15% trying to obtain a public service Average % of managers that were asked to pay a bribe 20.10% while trying to obtain a public service Bribes paid to obtain public contracts (public officials reported the estimated % of cases in which bribes are 45% necessary, average) Mis-use and mis-management of public funds (% public 15% officials reporting it is very frequent) Purchasing of positions (% of public officials reporting it is a 13.40% very frequent practice) Corruption in public administration (% of public officials reporting it is very frequent) 0% 10% 20% 30% 40% 50% 60% % of respondents who reported the presence of corruption / or average across all responses 120 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION 7.6 State Capture. Some key 'institutions' outside of the public sector appear to also fuel misgovernance by exercising undue influence over the state (hint at modicum of 'state capture'). l Both drug cartels and large financial and economic groups are perceived as wielding enormous influence in shaping laws, policies and regulations, in contrast with associations of professionals, trade unions, etc. which are not seen as wielding excessive or undue influence. o Bribes are used to shape key government's functions (high level judiciary, legislative and executive) much more often than to obtain public contracts or services. 3 Bribes are also linked to political funding. Nearly 30 percent of firms report that very often firms in Peru make political contributions to affect the political process. Though the majority of managers deny providing campaign contributions during the last elections, one- third of firms report that bribe revenues are very often used to finance political campaigns, suggesting the existence of a tight, non-transparent link between government officials and private sector. o Ilicit payments are made to secure public administration positions. 13 percent of public officials reports that this practice exists for some higher level jobs as a 'private investment', while this number (and amount) declines significantly for lower level positions. A. THE COSTS OF WEAK GOVERNANCE AND CORRUPTION 7.7 Corruption increases inequality. Bribery is a significant and regressive Figure 7.6 Average Percent of Income 'tax' on public service users. Users report Paid in Bribes major bribe expenditures extorted from institutions in charge of driving licenses 2.9 3.02 and passports, construction permits, taxes, school enrollment, etc. Furthermore, the poorer groups are disproportionately 0 affected by corruption, with lower income 4 1.5 - users spending on average in bribe 1.0 - 0.6 payments a percentage of their income - 0.5 - thrice as higher than wealthier households. 0.0 (Figure 7.6) Low Medium High 7.8 Businesses are also affected Income FNote: Average percentage of income paid in bribes was spend more on bribes than on security, computed for all respondents that used any public service and and this bribery 'tax' is regressive. On reported that they were or were not asked to pay a bribe. average, 5.2 percent of enterprises' monthly gross revenue goes to bribes. For smaller enterprises, the bribery tax is particularly onerous: it accounts for 8.9 percent of their revenue, whereas large ones pay less than 2.5 percent. And there is evidence that much of the onerous bribery tax on enterprise is indeed extorted, in the sense than on balance the firms do not benefit from it. 7.9 Bribes contribute to the lack of effective access.to public services by the poor. Users, especially the poorest, are often discouraged by bureaucratic inefficiencies and choose not to seek a public service. Users point out at Government agencies like ESSALUD (health services), Empresa de Agua y Desague, and SUNAT as the ones where they are more likely to not attempt 121 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION to obtain a service or complete a procedure given their absence of an effective service orientation, if any. B. IMPACTON PUBLIC SERVICE DELIVERY 7.10 Bribery payments do not translate in higher quality Figure 7.7 Indices of Service Quality and Bribery of public services for the users (based on public officials responses in 57 government and firms. Regression analysis agencies, 2001) show that on average, paying more bribes does not result in 70 higher quality of service delivery 60 and that the opposite might be (D ~0 true, with poor quality and ~ 4 bribery going hand in hand (Figure 7.7). Users report the 20 highest levels of service quality 10 in the offices devoted to issue identity cards and passports, the 0 50 60 70 80 90 100 real estate registry and public Index of service quality schools. Users report the lowest levels of service quality in the 7.8a Quality of Publc Services offices handling national taxation as ranked by users, 2001) (SUNAT), construction permits, real estate duties, and public Identity cards health and social security (Figure Passports 7.8a). Enterprises report highest Realeseregistry levels of service quality by utility Educton companies and the lowest levels Pennits ofthe National Bank by the municipal authorities. Acs oelectricity_______ 71 HelhadEuainVehicle registration - 71i re ciioatndcsoSevcQultanBrby Services. Except Lima, the Jdca etfctos______ quality of education by public Drining licenses schools is fair (more than 50 Health care ______ percent of users rate such service Social security 0 Selva high), somewhat more Wee estate dlutles 8 Lima homogeneous across the country Construction permtits ____ OResto than that of other social services, Natwilnion cs Sierra and bribes are not very frequent 0 10 20 3 8 0 9 60 70 s0 (7 percent) (Figure 7.8b). % reporting high quality of service However, users' survey data also suggest that the most pressing issues of the education system are the quality of the personnel, especially since teachers' promotion is very significantly influenced by political power, the high access costs to school and the poor quality of the school equipment. Conversely, the delivery and quality of health services is quite poor and very uneven across regions. While respondents report that they do not pay bribes for health care quite as often as for other services, they are very often discouraged to seek health care due to its internal inefficiencies. 122 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION 7.12 There is wide geographical heterogeneity in the quality of public service delivery. The quality of public services appears on average to be higher in the Selva region and Lima than in the rest of the country, and significantly greater in the case of offices handling driving licenses, judicial documents, basic health services, and schools. Figure 7.8b Quality of Health and Education Services (as reported by users) 6)0 50 Health Social Education Services Security a Resto de Costa * Sierra o Selva n rn 7.13 There is also an extremely varied service delivery performance across public institutions in Peru. Such variance permits a rigorous analysis to distill the key governance elements that influence service delivery. This analysis between governance structures (as rated by users, enterprises, and public officials) suggests that the following factors may be important for improving governance: Li Voice. Simple and partial correlation analyses suggest that corruption reporting mechanisms are associated with lower levels of corruption. Complaint mechanisms deter corruption. oi Transparency. The relationship between levels of corruption and transparency is statistically significant: agencies with more transparent procedures are less likely to display corruption o Quality of Rules. When rules and regulations over personnel are clearly and well specified there are less incentives for corruption ai Social Inclusion and Collective Action. Agencies with employees supporting civil society participation reforms in the public sector display lower levels of corruption. Li Meritocracy. Agencies in which personnel decisions are based on merits and performance are not associated with lower frequency of corruption. 7.14 Local governments perform much worse than national agencies in service delivery. The survey findings support the extent and seriousness of corruption and governance problems within municipal governments, when compared to national agencies. Overall, the degree of satisfaction in the provision of public services by local governments appear much below other government agencies in term of quality, low cost, degree of satisfaction with the service expected 123 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION and accessibility to the poorest (Figure 7.9). Corruption levels in local governments also appear much higher than those in national agencies (except with respect to bribes paid to obtain a public service (Figure 7.10). Figure 7.9 Provision of Public Services - Local Governments vs Other Government Agencies %o (based on public officials' responses) 90 80 70- 60 50 - 40- 30 - I 20 -- 10 -- 0 . . Is of high quality Is provided at Satisfies completely Is accessible to the relatively low costs the demand of the poor users M Local Government W Other Figure 7.10 National and Municipal Agencies are ridden by different types of corruption (based on public officials' responses) 35 30 25 20 15 - 10 Irregular use of Bribes to obtain Bribes to change a Bribes to obtain public resources public contracts legal decision public services I Municipal Agencies I National Agencies C. A POLICY AGENDA FOR IMPROVING GOVERNANCE 7.15 Reducing corruption and improving governance is more than just a matter of law enforcement-it requires a society-wide set of institutional reforms. A multi-pronged strategy that addresses the various forms of corruption-from state capture to administrative corruption-must address political accountability and transparency, promote a competitive private sector, strengthen institutional restraints, improve public sector management, and embrace civil society participation. It is important to emphasize that while the comprehensive approach described in Box 7.2 applies generally, the detailed components described within each box can lead to the determination of reform priorities in Peru. 124 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION 7.16 The strategy needs to center on fundamental public sector reforms, while at the same time strengthening civil society participation and engaging constructively the competitive segments of the private business sector-within a coalition-building approach of shared responsibility. The evidence suggests that public institutions characterized by effective voice and accountability mechanisms, efficient corruption reporting mechanisms, clear and well-defined rules, and a reform-minded staff, perform better and are able to control corruption more effectively. The evidence also identified public officials support for civil society oversight and meritocracy reform as anticorruption mechanisms. The survey evidence underscores the importance of implementing reforms to enhance transparency and public oversight. 7.17 Design of an anticorruption strategy is facilitated when there is a good understanding of the levels of resistance that are likely to be encountered. Public officials were asked about their levels of support to various reforms. Public officials favor strengthening public administration, in terms of both merit-based promotions and simpler administrative procedures. They are also in favor of the introduction of a system of checks and balances and the oversight of public sector's activities by the civil society (Figure 7.11). Box 7.2 A Multi-pronged Strategy for Addressing Corruption and Improving Governance Institutional Restraints: Political Accountability: Independent and effective judiciary * Political competition, credible political parties SLegislative oversight * Transparency in party financing * Independent prosecution, enforcement * Disclosure of parliamentary votes * Asset declaration, conflict of interest niles Competitive Private Sector: Civil Society Participation: * Economic policy reform Frie,Ndom * Competitive restructuring of monopolies * Freedom of information k* Regulatory simplification for entry * Public hearings of draft laws * Transparency in corporate govemance * Role for media/NGOs * Collective business associations Public Sector Management: * Meritocratic civil service with monetized, adequate pay * Budget management (coverage, treasury, procurement, audit) * Tax and customs * Sectoral service delivery (health, education, energy) * Decentralization with accountability Source: World Bank, 2000, Anticorruption in Transition-A Contribution to the Policy Debate. 125 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Figure 7.11 Public Officials' View on Reforms in Public Sector (% of respondents) Simplification of administration Clearer procedural guidelines Introduction of meritocratic system in the public sector Introduction of a system of checks and balance for public sector Better supervision of political party financing Free and better access to state information Greater supervision of civil society over public sector activities Decentralization of state administration Reduction in number of public officials Privatization of public services 0% 20% 40% 60% 80% 100% 1 In favor M Indifferent 0 Against 7.18 The Government needs to build credibility. There is a significant discrepancy between users and public officials on the official commitment to fight corruption. Moore than half of users, versus only 20 percent of public officials, reported that the government is not committed at all in the fight against corruption. This lack of trust toward government's commitment to eradicate corruption is especially felt among low-income users of public services. Conversely, the fact that a high proportion of public officials report preparedness of the public sector to fight corruption is an encouraging sign that not only the challenge is recognized within the public sector, but that there would be support among civil servant for an anticorruption and governance improvement program. Recommendations: > The government needs to make the fight against corruption a top priority. The majority of citizens (users, public officials and firms) would support it. > The church and the media should be integrated as potential contributors in the fight against corruption: all respondents indicate they are key players in developing an anticorruption strategy. Low-income users exhibit high confidence in the media. 7.19 The Government needs to continue enhancing transparency. Transparency in government is synonymous with openness. Introducing transparency requires that officials provide the public-civil society organizations, the media, or anyone else who is interested- sufficient information about the budget and its activities in order for the public to serve as an effective check on abuses. (Chapter III). 126 Chapter VII. IMPROVING GOVERNANCE AND REDUCING CORRUPTION Recommendations: > The Government should consider issuing a freedom of information law that require governments, national and sub national, to provide information to the public unless there is a valid reason (such as threat to national security) for the information to remain secret. > Sufficient investment in training civil servants producing and handling potential public information is essential to prevent misunderstandings. > A proactive approach that invites open oversight by Congress, civil society, and the media, for example with respect to large privatizations or important tenders, is usually effective. 7.20 The Government needs to be accountable. Separation of political and economic interests is a challenge facing governments everywhere. Building accountability into political life helps reduce the potential gains for "captor" firms and politicians alike. Recommendations: > The GOP should improve its civil service recruiting system and migrate toward a merit- oriented promotion system, while guaranteeing insulation of the civil service from political changes (Chapter VI). > The GOP should issue an Ethic Code Law for public service with clear prohibitions of conflict of interests, and an effective enforcement for both political and civil service positions. This is an essential corruption-deterrent tool. As income and assets of high- ranking politicians in Peru are already publicly disclosed, the media and the general public may act as allies in uncovering conflicts of interest and questioning politicians who cannot explain their wealth. Such measure should also reach Congressmen, incoming regional presidents, and municipal authorities. > The GOP should demand transparency of the financing of political parties. Campaigns should make clear the links, both explicit and implicit, between politicians and the interests that they support. The GOP should mandate and publicize detailed reports on the finances of all political organizations, identifying contributors and beneficiaries, and providing the civil society, the media, and the general public with tools they can use to identify the sources of state capture. > Other possible measures by the GOP include banning the use of state resources for political campaigns, limiting the amount that can be spent on politial campaigns, providing public funding, and prohibiting certain types of entities from contributing to political campaigns. 7.21 The Government needs to promote a sound business environment. Since every bureaucratic and regulatory intervention creates an opportunity for corruption, reducing the regulatory burden on firms is a key element of the strategy. Recommendation: > The GOP should examine and consider reducing the number of regulations and inspections. Clear rules defining when regulation is appropriate, sound analysis of a regulation's impact on firms and a constant, mandated, dialogue with the business community would increasingly be required to make regulatory reform sustainable. 127 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 7.22 In sum, the Government needs a comprehensive and inclusive approach to deal with governance. Peru is at a crossroad in governance matters and it is crucial to implement a series of institutional reforms in key areas in order to improve transparency and accountability of government agencies. By addressing the systematic institutional weaknesses within key public sectors and agencies in Peru, that progress would be possible. Survey results show that reducing corruption is not just a matter of strong enforcement, but also hinges directly on the incentives facing public offichls, enterprises, and households. Many reforms that are important in their own right-state and regulatory reform, judicial system reform, civil service development, and openness in government to name a few-should now be viewed as part of a pro-governance agenda. Although there is no single recipe for success, several countries have followed a three pronged criteria focusing on enforcement of anticorruption statutes, education of the population about their rights when dealing with the public sector and the harm corruption causes, and prevention of corruption by improving public sector governance. A sustainable strategy also endorses an inclusive approach with governmental and non-governmental leaders. Recommendation: > Creating a high-level steering committee, supported by a professional secretariat, can help to bring together representatives of governmental bodies, both central and local, to develop the specific action plan that would implement the governance strategy. By explicitly including representatives from outside government, the steering committees can further build credibility while mobilizing an important ally for reducing corruption. 128 Chapter VIII FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES 8.1 While the Peruvian mining sector reform in the early 1990s resulted in the attraction of substantial exploration and development investment and the subsequent increase in mineral production and exports, since the late nineties, key fiscal and environmental issues have negatively affected the competitiveness and quality of insertion of mining with foreign investors and local society, and particularly with the local communities. The mining reform brought about the effective opening of the sector to direct foreign investment and the shift of the role of the state from owner-operator to lessor-regulator. This was accomplished by reforming the legal and institutional framework, privatizing the mining state owned operations, and starting up an environmental management program. While the three reforms have been widely considered successful, three issues remain controversial and are considered in need of adjustment. First, the taxation system might be producing a certain loss of competitiveness. To the extent that it affects rates of return on investment, the effective tax rate in a given country is a major factor influencing foreign investors' decision as to where to invest their capital. Nations whose mining fiscal systems impose high levels of overall taxation should expect to see lower levels of foreign investment in their mining sectors than countries with lower tax levels. Thus, as the world moves forward in the new century, there is a clear trend for mining taxation systems to become increasingly similar. Second, poor transparency is eroding the business climate. The highly centralized and non-transparent nature of the Peruvian governments of the 1990s created an issue of credibility with the local populations around the mining transfers from the Central Government to municipalities. Although the Mining Law establishes that 40 percent of the income tax generated by the mining operations-the canon minero-should go back to the municipalities (20 percent) and the regions (20 percent), and Bank findings show that its overall redistribution has de facto been progressive, the management of the canon has not been transparent and there is a widespread perception that the local communities are not benefiting from it. Third, the perception of inadequate environmental control and of widespread conflict between mining companies and local populations is endangering the contribution and the stability of a key sector of the Peruvian economy. Furthermore, the environmental information that the Ministry of Energy and Mines (MEM) has received from the mining companies has not been adequately released to the public, resulting in a feeling of mistrust against mining activities. This chapter shows empirical evidence that supports the view that the current Peruvian mining tax regime is among the most globally competitive systems worldwide and provides an appropriate balance between investors' needs and government needs. However, room for improvement is constrained by the need of additional fiscal resources. Finally, it overviews other structural issues and provides detailed recommendations. A. BACKGROUND1 8.2 The mining sector is essential for growth and development in Peru. The sector generates more than 45 percent of the country's export earnings and contributes about 57 The sections on taxation mainly draw on Otto (2002). 129 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION percent of GDP.2 The mining industry has long been Peru's most prosperous sector, with growth rates between 6.6 percent to 15.2 percent, between 1994 and 1999, attracting a total of US$6.3 billion in investments between 1992 and 2001, and peaking in 2001 at US$1.3 billion. Potential tax revenues from the sector would be highly relevant to keep up its macroeconomic -financial projections. Tax expenditure associated to mining benefits is estimated at 0.4 percent of GDP (IMF 2000). The canon minero accounts for about 4 percent of total transfers to municipalities. Over one quarter million people depend on formal mining for their livelihood, and it is estimated that this could double given the sector's potential. Mining is particularly relevant in some of the mountainous areas of the country, which count amongst the poorest regions. Attracting and sustaining mining investment in these areas would contribute to reduce the widening inequalities between the coastal areas and the lagging interior. However, investors' interest in the sector has recently dropped dramatically, resulting in moderate investment projections of about $500 million p.a. between 2003 and 2005. The key question for Peru is now whether it can institute reforms that will return the sector to its role as an engine of growth, much like demonstrated by the Chilean example. 8.3 Past reforms were left incomplete. The Peruvian General Mining Law provided a sound basis for the modernization of the sector. Mining results of reforms in the early 1990s were impressive: Between 1990 and 1997, while global exploration investment went up 90 percent and grew four-fold in Latin America, in Peru it grew twenty-fold. Since 1997, as a consequence of the drastically lower metal commodity prices, the share of exploration investment of most developing countries was reduced substantially-as most international mining companies retreated to their traditional exploration areas-yet Peru was able to keep is share of the total. This increased investment during the 90s doubled the mineral production and mineral exports value of Peru. However, structural--taxation and environment-and revenue-transfer transparency issues have negatively affected the competitiveness and credibility of the mining regime, expanded the perception of inadequate environment control, and widespread conflict between the government and the local populations, thus endangering the contribution and required stability of this critical source of growth. B. THE PERUVIAN TAXATION REGIME 8.4 The existing Peruvian mining tax system is complex (Table 8.1). The term "tax" is defined as any levy imposed by the government on a productive agent, regardless of whether it takes the form of an explicit tax, a fee, or any other payment that is paid to the government or to another party because the taxpayer is required to do so by government. C. COMPARISON OF PERU'S TAX SYSTEM TO MINERALS TAX SYSTEMS IN SELECTED COUNTRIES 8.5 Many of the major tax types of the mine ral sector are already applied by Peru. Comparison of mining tax systems between different countries endowed with mineral resources is relevant from the standpoint of determining how attractive is a particular country to foreign investors in world where investment in mining depends, inter alia, upon decision criteria such as after-tax profitability, the investor's ability to predetermine the applicable tax liabilities, the stability of the tax regime, and the level and method of tax levies (Table 8.2). 2 Peru's main mineral exports are iron, silver, zinc, copper, lead, platinum and gold. 130 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES Table 8.1 PERU: Description of the Existing Mining Fiscal System Income tax: * Distribution: 50 percent of income tax paid by a mine to the Central Government is to be remitted back to the Canon where the mine is located. * Rate. 27 percent under the general tax regime. If the taxpayer has elected to enter into a Stability Agreement or Mining Contract, the rate is 29 percent. An additional 4.1 percent tax is applied to net profits remitted (see withholding tax description below). Deductions for computing taxable income: * Feasibility studies: there are two possible interpretations. (1) treated as a development cost; may either be expensed in the year costs were incurred, or costs may be amortized over a period of three years from the year the minimum production is achieved; or (2) treated as a preoperative cost; may either be expensed in the year the cost is incurred, or costs may be amortized over a period of ten years from the year in which the minimum production is achieved. * Pre-production exploration costs: costs may either be expensed in the year they were incurred or amortized as from the year the minimum production is achieved, over a period determined based on the life of the mine. * Development expenses: costs may either be expensed in the year they were incurred, or may be amortized over a period of three years from the year the minimum production is achieved. * Capital costs: taxpayer can select the rate of straight line depreciation up to the allowed maximum; most mining, processing and power equipment has a maximum of 20 percent per year; roads and buildings have a maximum of 3 percent unless a stability agreement (15 years) is in effect in which case a 5 percent maximum applies; costs incurred for govemment approved infrastructure such as a school, hospital or recreational facility can be expensed as incurred. * Costs qualifying for depreciation or amortization may be adjusted for inflation using the whole price index; however, adjustment is not allowed if a) there is a stability agreement and b) the taxpayer has elected to keep its books in U.S. dollars. The following types of costs may be deducted for computing net taxable income: pre-production exploration expenses, mine site development costs, feasibility study cost, operating costs, capital costs, qualifying loan interest, withholding tax on interest,. property tax, fee based on land area, payroll taxes, workers profit share, value added tax (when IGV is not used as a credit). Excess profits type tax: none. Royalties: there are no royalties or similar taxes. Withholding tax on loan interest paid to foreign lenders: 4.99 percent is applied to loans from abroad provided monies are sent into Peru, lender is a financial institution person and the maximum interest is less than prime + 6 points or LIBOR + 7 points. 30 percent is applied for the excess of interest exceeding the maximum limit; in cases where loans are entered into between related partners; and if the lender is a resident of a tax haven. Withholding tax on dividends remitted abroad: none y that name. However, an additional rate of 4.1 percent on income tax is applicable when dividends are distributed. Withholding tax on salaries and fees paid to foreign consultants: 30 percent tax rate is applied on 80 percent of gross income (thus, in practice, the effective rate is 24%) if technical services are rendered in Peru. If technical services are rendered partly abroad and partly in Peru, a 30 percent tax rate is applied on 40 percent gross income (thus, in practice, the effective rate is 12%). Services totally rendered abroad are not subject to withholding tax. Import duty on foreign equipment rates are generally 12 - 20 percent the representative rate for mining equipment is 12 percent import duties are not immediately deductible, rather, they are added to the asset acquisition cost and depreciated at the asset's depreciation rate. Export duties on minerals: none. Sales tax on equipment: none Value added tax (IGV) on purchased goods or services: 18 percent IGV is levied on sales of goods and services, imports of goods, construction contracts, sales of real property by construction concerns, and services rendered abroad, but used in Peru; IGV may be reimbursed by means of assignable credit notes based on export sales; can offset IGV against income tax, or obtain a refund check. If minerals are sold locally, IGV applies to sales except for gold. The amount of IGV that may be credited is restricted to 18 percent of export sales, and any amount exceeding this may be carried forward as a credit against future export sales. IGV paid during exploration is refundable. Time to claim back value added tax: 3-6 months. Education tax: none. Property tax: only applies to operations located in urban areas. Local development requirement none. If a taxpayer voluntarily contributes to local development and that local development project is approved by the proper government ministry, the costs may be expensed as incurred. In practice, approval is difficult or time consuming. Land use fees: a validity tax, also called a good standing fee, is calculated based on the area in mining concession from the moment the claim is filed. The fee is US$3/ha/yr and is deductible. 75 percent is distributed by the central government to the Canon where the mine is located. Stamp tax: none. Payroll taxes paid by employer: extraordinary solidarity tax: 2 percent of salaries paid; health service (ESSALUD): 9 percent of salaries paid; national technical industrial training service, if more than 20 workers: 0.75 percent of salaries paid. (Tahle 8 1 continued on next paPe} 131 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Thl R 1 (enntinuiod) National Service for Construction Industry Training: 0.2 percent of total incomes comprised of materials, workmanship, general expenditures, technical direction, profits or any other item that might be invoiced to the company's clients. Workers' profit sharing: it is obligatory to pay a workers' participation of 8 percent on the net profits; of this amount up to 18 times monthly salary goes to the worker, with the remainder going to a special educational, social and recreational fund. Disbursements are decided by a Board comprised of representatives for mining companies, government, and workers. The amount paid is allowed as a tax deduction. Not all foreign governments recognize this as a creditable tax and double taxation can thus occur. Tax incentives: * Loss carry-forward: 4 years. * Loss carry-back: none. * Tax credits: income tax paid abroad in respect to income of foreign source taxable in Peru, may be deducted against Peruvian income tax within certain limits; qualifying IGV may be credited (see above). * Tax stability: title-holders of mining activities may enter into several types of tax stability agreements. Two types are defined under the Foreign Investment Law and two others under the General Mining Law. They are not mutually exclusive and a company can have both (one under the Foreign Investment Law and another under the General Mining Law). If an agreement is in place under the Mining Law, the income tax is increased by 2 percent. Under the Foreign Investment Law, a Stability Agreement (as an investee-the company which received the investment) granted by the National Commission for Foreign Investment and Technology guarantees for 10 years stability concerning: income tax regime, currency exchange regime, free availability of foreign currency and non- discrimination). To qualify, the investor must invest a minimum of US$10,000,000 within two years of the Stability Agreement. Under the General Mining Law, an investor can enter into a Mining Contract. The agreements can be for a period of 10 or 15 years: 10-year - the investment must equal US$2 million and be destined to either start up an operation with a production capacity of 350 to 5,00OMt/day and 15-year - this agreement targets production of at least 5000Mt/day and requires an investment of US$20 million for a start-up operation, or US$50 million to capitalize an existing operation. A Mining Contract guarantees the following: free marketing of mineral products for export or domestic sale; free disposal within the country and abroad of foreign currency generated by exports; free convertibility into foreign exchange of local currency generated by mineral sales; non-discrimination in exchange matters; depreciation rates of 20 percent for mining and processing equipment and 5 percent for buildings (15 year agreements); the electable option to keep books in U.S. dollars (15-year agreements); general tax stability including income tax, regime, compensation and/or tax refunds, customs duties, municipal taxes, and validity fee; non- discrimination in exchange matters; and freedom to remit profits, dividends, financial resources. Tax reduction in Selva regions: special tax rules and rates apply to IGV, income tax and excise Tax. Requirement to use local goods and services: none. Local equity requirement: none. Government equity requirement: none. No ring fencing principles apply (may consolidate books). Foreign external accounts: allowed for receipt of revenues. Exchanee controls* no significant restrictions 8.6 Comparisons of mine taxation in different taxing jurisdictions are a complex matter (Table 8.3). An isolated comparison of any one type of taxing mechanism may lead to certain insights, but taken alone may not p-ovide a useful indication of how mine taxation in one jurisdiction compares to that in another. To gain a broader understanding of how overall tax systems compare, it is necessary to analyze them in a holistic manner. Both to facilitate and formalize such an analysis, it is a common practice to define a hypothetical pre-feasibility-type model mine and then apply different taxation systems to that mine and compare the results. In the background work for this chapter, model mines were specified and various measures of taxation and profitability calculated to allow comparison between Peru and other countries. Yet, this is not the only possible approach. Another familiar one is to analyze competitiveness by level of investments in new deposits and expansion in already existing ones. The model mines used in the background work for this chapter were based on the standard base metal (copper) mine model and gold mine model developed at the Institute for Global Resources Policy and Management at the Colorado school of Mines. Description of the model, its 132 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES Table 8.2 M neral Taxes in Peru and Other Countries Does Peru impose this levy? Comments Tax type Yes No Income tax X Excess profits tax X Royalties X Withholding tax on remitted X An additional rate of 4.1 percent (income type tax) is applicable when dividends dividends are distributed (similar to a dividend withholding tax, but it is not technically termed a withholding tax). Withholding tax on remitted X loan interest Import duties on equipment X Equipment imported for temporary usage, such as for exploration work, may be imported without payment of import duties if such equipment is re-exported within a year. Export duties on minerals X Sales tax on purchased X equipment Sales tax on minerals paid by X mine Value added tax on services X Fully refundable during exploration (as of the date of this report, refund is not (IGV) applied because the procedural rules are pending). Value added tax on X Fully refundable during exploration (as of the date of this report, refund is not equipment (IGV) applied because the procedural rules are pending). Value added tax on mineral X IGV does not apply if the mineral is sold for export. sales (IGV) Property tax/fee X A property tax applies only to mines located within urban areas. Education tax/fee X An education tax is not applied, however an 8 percent profits based profit sharing requirement is applied, and a portion may gD to a special educational, social and recreational fund. Local development tax/fee X Fees based on Land Area X Called a "validity tax" or "good standing fee." Stamp tax X Payroll taxes X Excise tax on fuel X Exempted during exploration (the tyemption is not applied because the Strocedual rul a x - ). 8.7 The current tax system in Peru is globally competitive for mining. When countries are ranked from the standpoint of the foreign investors' internal rates of return (IRR), Peru ranked 11mh out of 23 (base metal) and 1h out of 23 (gold) and in the second lowest quartile (Tables 8.3 and 8.4). 8.8 Sensitivity of the Peruvian Tax System to Prices and Costs. Using a standard copper mine model, the Peru's mining tax system is quite neutral to changes in prices, operating costs, and capital costs, except when prices become low (Table 8.5). The impact of a tax system on a mine can vary according to its profitability. If the overall effective tax rate (ETR) goes up as profitability goes up, the system is said to be progressive. If the ETR decreases as profitability goes up, it is said to be regressive. Progressive tax systems tax more profitable mines at a higher effective rate than lower-profit mines. Most economists agree that neutral or slightly progressive tax systems are better than regressive systems. methodology, attributes and limitations can be found in Otto, James M. (2002). The model's estimates are particularly sensitive to price assumptions. 133 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table 8.3 Comparative Measures of Profitability and Effective Table 8.4 Comparative Economic Measures for a Model I Gold Metal T-r Rat, far M,dl RFt- MetAl Min- in qpl,,td urisdiction Miin- in Rpleetod TI .di..cti.on. s .. Country Foreign Investor's Total Effective Country Foreign Investor's Internal Total Effective Tax Countr R R05. T-. R,= ~ Cunr R't69 of Rturn (U) Pt Lowest taxing quartile Lowest taxing quartile Sweden 15.7 28.6 Sweden 19.2 29.1 W. Australia 12.7 36.4 South Africa 18.8 32.6 Chile 15.0 36.6 Chile 18.3 36.8 Zimbabwe 13.5 39.8 Philippines 18.4 38.2 Argentina 13.9 40.0 Argentina 16.6 42.5 China 12.7 41.7 W. Australia 15.2 43.1 Second lowest taxing quartile Second lowest taxing quartile Bolivia 11.4 43.1 Peru 14.7 43.3 South Africa 13.5 45.0 Zimbabwe 15.7 45.9 Philippines 13.5 45.3 USA (Nevada) 15.1 49.3 Kazakstan 12.9 46.1 Bolivia 12.2 52.4 Peru 11.7 46.5 Kazakstan 13.5 54.4 Tanzania 12.4 47.8 Greenland 14.7 54.9 Second highest taxing quartile Second highest taxing quartile Indonesia 12.2 48.6 Ghana 13.6 56.7 Poland 11.0 49.6 Tanzania 12.7 57.9 USA (Arizona) 12.6 49.9 Indonesia 11.4 60.4 Mexico 11.3 49.9 UzbekistAn 11.2 62.0 Greenland 13.0 50.2 Mdxico 10.4 62.9 Ghana 11.9 54.4 Ontario Canada 10.7 68.3 Highest taxing quartile Highest taxing quartile Papua NG 10.8 57.8 Ivory Coast 9.1 69.1 Uzbekistan 9.3 62.9 Papua New Guinea 8.7 72.3 Ivory Coast 8.9 62.4 China 7.1 73.9 Ontario Canada 10.1 63.8 Poland 3.0 90.2 Riuki,l Fia 3 5R 9 Rjriim, F-a -1 6 106 0 Source: Otto (2002). Source: Otto (2002) 8.9 Comparing Tax Stability Eet ie T P. Across Countries. In a global E c Tax survey of mining companies, over 50 Price Sensitivity: percent of the respondents listed tax US$1.10/lb (base case) 47 stability as a "very important" factor US$1.50/lb 39 * . . .Operating Cost Sensitivity: in investment decision-making. Out US$0.40/lb 45 Fairly Neutral tx US$0.45/lb (base case) 47 of a list of 60 investor criteria, tax US$0.50/lb 49 system stability ranked 10th in Capital Cost Sensitivity: US$550,000,000 43 Fairly Neutral importance. In the mining business the investment horizon is long and (base case) I I lqwnnn nnn Al companies are usually reassured by Note: a rate higher than 100 percent indicates that he mine is paying taxes in excess of tax systems that reduce tax payment its net income. uncertainty and vulnerability, particularly during the loan and project payback periods. However, many governments are hesitant to resort to tax stability agreements. This is because there is a basic tenet that one generation of lawmakers should not be able to tie the hands of future lawmakers. Another relevant consideration is that tax stability is desired by all productive sectors as it reduces uncertainty. Then, if stability is offered to one sector, other sectors will also seek it. 8.10 The government faces a trade off. On the one hand, stability agreements enhance the potential for mineral sector investment, but, on the other, they complicate the tax system and raise administrative costs. Besides, if taxes are stabilized for various mines, then an administrative challenge can arise over time. As the underlying tax laws change, each stabilized mine will have a tax regime dating to the time the stability agreement was entered into. This means that, over time, there will be multiple tax regimes and the government agency charged with tax administration will increasingly face a more complicated situation monitoring and enforcing each. This entails costs. 134 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES Table 8.6 Availahility of Tax Stshility in Selected urigdictifons Country Some form of tax Description stability availablee Argentina Yes 30 yrs, provincial & municipal taxes, import duties, exchange rules. Bolivia No Burkina Faso Yes During the term of the contract; except mining taxes and fees. Canada (Ont) No Yes 10 yrs, if mine elects a higher income Chile tax rate (42%). China No Ghana No Greenland No Indonesia Yes Tax stabilized for life of mining agreement or a shorter period (Contract of Work). Ivory Coast No Yes Taxes stabilized for life of mining Kazakhstan agreement. Mexico No Peru Yes Two systems of tax stability: mining contracts (10-15 yrs tax stability), and 10-yr Legal Stability Agreements that fix the income tax regime and certain other fiscal imposts; 2% additional rate Philippines No Poland No - South Africa No - Sweden No - Tanzania No - USA-Arizona No - Uzbekistan Yes Most major taxes may be frozen for 10 yrs from date of establishment; tax experts warn that there may be difficulties with the practical implementation W. Australia No Zimbabwe No- Source: Otto (2002). 8.11 At the present time, in Peru, there are hundreds of mining concessions with stability agreements. However, as these agreements stabilize taxes for only around 10-15 years, over time, the affected mines will revert to the general tax system in the future. Stability also carries a price for companies-an additional 2 percent rate is applied to the 27 percent income tax. Stability is, indeed, important to investors, witness the number of mining concession holders who have agreed to pay the premium. One option open to the Peruvian government is to raise the amount of this premium. The effect of raising the stability premium is shown in Table 8.7. As can be seen in the table, for the two model mines a premium of 5 percent on the regular income tax rate of 27 percent reduces the investor's rate of return by about 1 percent, and the effective tax rate remains at about 50 percent or lower. Recommendation: Companies find tax stability very attractive and Peru's stability agreements are an incentive to investors. 135 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION > It is recommended that tax stability Tahle R-7 Tay System Sengtoyfty to Income Tax Rate agreements be retained but that the Govt Revenue: premium for all future such Effective Tax Investor IRR Fees (US$ agreements be raised to 5 percent. C T Mo - de -10) mill-na The disadvantages of such 25% 45 11.9 451 agreements (sector discrimination 27% (current) 47 11.7 471 and administrative burden) are 27+2=29%(stabilized) 48 11.5 491 rate) more than offset by the increased 27+3=30% 49 11.4 501 tax base that may be created in the 27+532% 51 11.1 521 future by higher levels of Gold Model: investment, and the higher levels of 25% 42 15.0 73 tax paid based on the 2 percent 27% (current) 43 14.7 76 27+2=29%(stabilized) 45 14.4 79 premium. Investors would rate) probably be willing to pay a higher 27+3=30% 46 14.3 80 premium for tax stability. 27+5=32%/ 41 13 9 113 8.12 The income tax rate of 27 Table 8.8 Income Tax Rates Applied to Mining Projects in percent in Peru is somewhat lower Selected Iuri-.dict*onn. than in some nations (Table 8.8). In rn"*rt Cnrpnrat inarnv tesr I F Argentina 35% addition, a 4.1 percent additional income Bolivia 25% (a surtax may also apply in some tax is applied to remitted dividends; the cases) net effect for many foreign companies Burkina Faso 35% (0.5% of previous year turnover is the is a rate of around 30 percent. If a minimum tax) taxpayer desires a stability agreement, Canada (Newfoundland) Federal: effectively 29.12% including 4% and.many miners probably would, the surtax); provincial: 14% base rate is 29 percent. Chile 15% (two elective regimes are available) China 33% (30% to central gov't., 3% to Recom endaion:provincial gov't.) > The current income tax rate should Grn n 35% be retained. Indonesia 30% (previous COWS range from 229 - 48%) 8.13 Peru uses a form of Ivory Coast 35% accelerated depreciation for Most Kazakhstan 30% (excess profits tax may apply mine equipment (20 peent Mexico 35% straight-line method). The most P.N.G. 35% for large (SML) mines, 25% for most common form of tax-base incentive for other mines mining is accelerated depreciation Peru 27% (29% for taxpayer with stability Most nations provide the mining agreement) industry with some sort of accelerated Poland 2000, 30%; 2000-2001, 28%; 2003, 24%; depreciation (Table 8.9). Until recently 2004+,22 Per aso roide a for o South Afica 30% for other than gold; formula > 30% 8.13pvd a fr ofor gold mines accelerated depreciation on mine Sweden 28% buildings at a rate of 20 percent Tanzania 30% (straight-line). After tax reform, this USA-Arizona progressive based on income (profit level) was lengthened to a 3 percent Uzbekistan 33% depreciation rate, which is in-line with W. Australia 2000/2001, 34; 2001+, 30% depreciation periods allowed by many zimb we 35% other nations. Source: Otto (2002). 136 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES T.hblR ft TprorintionAppliedtinTypirol Mining R ulpment in RltdInidtnu Accelerated method Country available for some capital Example rqupmrnt (rornn) Argentina Yes 3 yr straight-line Bolivia Yes 8 yr straight-line Burkina Faso Yes useful life minus one year Canada Yes up to 100% in yr incurred for new mine or 25% declining pool Chile Yes 3 yr straight-line China Yes 10 years Ghana Yes 75% in 1 yr, then 50% declining balance Greenland Yes the company may decide the rate and period Indonesia Yes 10 yr straight-line or 20% declining balance Ivory Coast Yes method of acceleration depends on life of equipment Kazakhstan Yes 25% declining balance method Mexico No Peru Yes 5 yrs straight-line (20%),except for mine buildings Philippines Yes twice the normal straight-line rate Poland Yes 5 yrs straight-line (20%) South Africa Yes expensed in I year of production Sweden Yes 5 yrs straight-line (20%) Tanzania Yes 12.5% straight-line USA-Arizona No Uzbekistan No 8% straight-line W. Australia Yes prime cost or diminishing value methods (less than effective life) Source: Otto (2002). 8.14 Unlike most industries, when a mine closes its buildings usually have no future function, i.e., no market value. Thus, mines whose duration is less than the depreciable life of their buildings will never be able to fully depreciate these buildings. For this reason, some nations allow a shorter depreciation period, or an alternative depreciation method (such as life of the mine). Table 8.10 shows the impact of a 3 percent versus a 20 percent depreciation rate. The effect is considerable for the short-lived gold model mine but there is less of an impact on the longer-lived copper model mine. Recommendation: Table 8.10 Tax Sensitivity to Building Depreciation Rates > A longer period of depreciation (8- Efectiv Government 10 years) should be considered and e Tax Investo Revenue: All the previous system of 20 percent Loss Carry Forward Rate r R Taxes & Fees depreciation should be reinstated Timp Limit millins) for buildings; a 3 percent rate Copper Model: forbuldng; 3 eren rte 3% rate (current system) 47 11.7 471 yields a depreciation period of over 20% rate 46 11.9 461 33 years (longer than the life of Gold Model: most mines). 3% rate (current system) 43 14.7 76 20% rate 41 15.2 72 8.15 Peru has an exceptionally Souce8: Otto (2002). short loss carry forward time limit of 4 yearse One of the most common tax incentives is to allow taxpayers the ability to carry forward losses from one year to offset taxable income in the succeeding years. For capital- intensive industries and for industries exceptionally prone to commodity price fluctuation, criteria that are no doubt met by the mining, loss carry forward is an important issue. For a sample of countries, a loss carry forward time limits as of year 2000 shows that, of the 22 nations, all but one had a longer time limit than Peru (Table 8.11). Table 8.12 shows the effect of lengthening the loss carry forward time limit 137 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Recommendation: Table8.11 Loss Carry Forward/Back Policy > The loss carry forward time limit Loss carry Loss carry-back should be extended to at least 10 G W rw-FA 3:11 "MU (+FQ M-gyn- years, or preferably, that no time Aeia Yes N No limit be imposed. While a short Burkina Faso Yes 5 No period is adequate for most small, Canada (Ont) Yes 7 Yes 3 Chile Yes None Yes none short-lived mines, it is not China Yes 5 No conducive to attracting Ghana Yes None No - investment for larger mines. Gdena Yes N Ye - Ivory Coast Yes 5 No- Kazakhstan Yes 7 No- 8.16 Peru does not have a Mexico Yes 10 Yes none provision for mining closure. Of Peru Yes 4 No increasing concern to governments is Philippines Yes 5 No Poland Yes 5 No- the issue of mine reclamation and South Africa Yes None No closure. These costs are primarily Sweden Yea None No Tanzania Yes None No incurred by the miner late in the USA-Arizona Yea 15 Yes 3 project at a time when production is Uzbekistan No No - falling off or close to nil. Thus, W. Australia Yes None No Zimbabwe Yea None No companies cannot recover the costs involved with closure unless the tax system is adjusted to take these into Table 8.12 Tax Sensitivity to Loss Carry account while cash flows are still Forward Time Lomot Loss Carry Effective Investor Govt Revenue: being generated. It is in the Forward Time Tax Rate IRR All Taxes & Fees government's interest to see that the Ioit (0/4 company doesplan and set aside funds Copper Model: 4 years (current 47 11.7 471 for this activity, for at the end of the system) mine life, funds will not be generated. 5 years 45 12.1 451 One way to encourage companies to 7 years No time limit 45 12.1 450 fund closure is to require an annual Gold Model: set-aside over the life of the mine and 4 years (current 43 14.7 76 to allow this set aside as a tax system) 5 years 43 14.7 76 deduction. Table 8.13 indicates the 7 years 43 14.7 76 effect of allowing a deduction for such No time limit 43 14-7 76 costs spread equally over the life of Source: Otto (2002.) the mine. Recommendation: Table 8.13 Tax Sensitivity to Annual Allowed Closure > The tax system should be modified Deduction Over the M pi Life Effective Tax Investor Government Revenue: so that money set aside for this Closure Deduction Rate IRR All Taxes & Fees purpose may be deducted straight cary () (o) (cror miliba) line as amortization over the No deduction (current 47 11.7 471 productive life of the mine, system) Annual life ofmine 46 11.8 463 deduction 8.17 The current Peru's tax system Gold Model: No deduction (current 43 14.7 76 recognizes t importance of mine system) contributions to local communities and Annual life of mine 42 15.1 73 infrastructure and allows a deduction for such expenditures. There is intense interest by many stakeholders in furthering the concept of sustainable development. One way to foster this is to invest in communities impacted by mining so that when the mine closes, the affected communities will be able to carry-on with social and 138 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES alternative economic activities. Tahle 8.14. Tax Sensitivity to Reinvestment Allowance Deductions are only allowed if the oEffective Investor Rvernment Deduction areci: neso Revenue: All expenditure is. approved by the Scenario Tax Rate IRR Taxes & Fees government and companies have found it Coper(odel -il0i)(us extremely difficult to get such approval xpar Mo 46 12.1 533 20% expansion, no461153 from the relevant ministries. reinvestment allowance (current system) Recommendation: 20% expansion, 80% 44 12.5 506 reinvestment allowance > The current practice of allowing government-approved, miner-paid-for investment in communities and infrastructure to be tax deductible should continue. Deductions should apply not only to hard investment (transport, power), but also to soft investment in local infrastructure/capacity building by community groups (training, education). > However, the current system of approval does not work well, if at all, and needs reform. Such approval authority should vest in a single ministry, not with the ministry responsible for the particular type of infrastructure. > The relevant law should also state that if the ministry does not give its approval for such tax status on the proposed investment within a reasonable time period, say 60 days, the request will be deemed to have been approved. 8.18 Until recently, Peru offered mining investors an incentive to increase the company's production with a reinvestment provision, but this benefit was eliminated in a recent tax reform. Before reform, the company was not subject to pay income tax on the profits it reinvested, provided these sums were in accordance with an investment program approved by the mining authority. The period of validity was four years, but could be extended for another three. However, the benefit was capped at 80 percent of the total profits, with the balance being subject to income tax. A-reinvestment incentive such as this is very, very uncommon globally. Most government's use the "penalty" approach to encourage reinvestment rather than the "incentive" approach, i.e., they impose a foreign-dividend withholding tax. To assess the impact of such an incentive, the copper mine model was modified (Table 8.14).4 8.19 A critical policy question when assessing a proposed tax incentive is whether or not the incentive (which, presumably, will lead to a short-term tax decrease) will lead to increased tax revenues over the long run. There are two ways to look at this: (i) will government returns from an individual project be increased over the life of the project? and (ii) will government revenue increase because other potential investors (who would not have invested had there been no incentive) will invest? With regard to (i), for the particular mine model used here, the table shows that the government will end up with less revenue. Even though there will be more metal sold over the life of the mine and various taxes will be applied to those sales, the additional revenue does not equal the amount of revenue lost in the four tax-incentive years. With regard to (ii), it is doubtful that such an incentive would attract new investors. In practice, the incentive might further reduce government revenue if companies purposefully underbuild initial capacity in order to take advantage of the liberal capacity increase allowance in later years. 4 It allows for a four-year, phased 20% mine capacity expansion commencing in the 8th year after production began. The cost was assumed to be $80 million for the expansion. It was assumed that an 80 percent profit exemption cap is applicable. The. economic measures for the expanded project were assessed with and without the tax incentive to determine its effect on government revenue and company rate of return. 139 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Recommendation: T,hlg A1 P 4 I)t hi Mineral royalty type tax on most > The decision to eliminate the G mlgeva-" reinvestment tax allowance should not Argentina Yes be reversed.Yes bervre.Burkina Faso Yes Canada (Ontario) No 8.20 Peru is one of the mining Chile No 8.20China Yea countries in the world that do not Ghana Yes impose a traditional royalty type tax. Greenland No Th t o 1taaton of Indonesia Yes The theory of optimal Ivory Coast Yes nonrenewable natural resources states Kazakhstan Yes that, under suitable assumptions, a MdxiGo Ye government can tax away the pure rents Philippines Yes generated in the extraction of those Peru No Poland Yes resources with no loss of allocation South Africa No efficiency. This justifies the imposition Sweden No of special taxes on nonrenewable - USA Ye resource producers over and beyond the Uzbekistan Yes taxes levied on firms operating in the A other sectors of the economy. It is Source: Otto (2002). widely acknowledged that the simplest special tax to administer is the gross royalty and this goes a long to explaining the reason why 60 percent of the countries in the reference sample impose royalties on mining producers (Table 8.15). 8.21 In the Peruvian context, a Tnhle- R.16. Tax System Sen%otilAty to a Royalty Tax Government decision on royalty taxes must be Royalty Tax on Effective Tax Rvne l ae Roii eve :Al ae made taking into account the Gross Sales Rate Investor IRR & Fees mining taxation system as a whole. Fa(Uso Yes RS) One of the simpler forms of royalty is Copper Model: CanadarOntaro No117 7 a tax-deductible gross sales revenue 0Ce ad valorem tax. The impact of such 1%*/ 49 11.3 493 a tax was assessed on the model 2Y% e 11.0 516 mines (Table 8.16). Taking into 2.5% 52 10.8 527 account the rest of the system, it is 3% 53 10.6 538 relevant to point out that Peru already Gold Model: imposes substantial input taxes (import 0% (current) 43 14.7 76 duty, IGV). On the other hand, o% 47 13.9 82 whereas a royalty tax is assessed only 2% 5I 13.0 88 in years during which there are commercial revenues, an input s 25% such as import duty, entails collection 1 3 54 12.1 95 during the project startup period, before there is any mine income. Companies would much prefer to pay taxes during years when production has begun rather than during construction. Recommendation: > A reasonable royalty should be imposed, provided that the import duty is eliminated by exemption or zero-rating, or that workers' participation is modified in a process affecting* all 140 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES public sector.5 The royalty tax should (a) be based on gross mineral sales revenues; (b) not exceed a rate of 3 percent; (c) have a uniform rate should for all mineral types; and (d) be offset by the elimination of the import duty. 8.22 Peru does impose Table 8.17 Tax System Sensitivity to Import Duty substantial import duties and Representative Effective Tax Government Revenue: All Mining Equipment Rate Investor IRR Taxes & Fees the rates vary according to the IMFnJt n&5 Ua* (LSS milions) item. However, import duty is Copper Model: allowed as a depreciable deduction 15% 48 11.2 484 against income subject to income 12% (current) 47 11.7 471 tax Mining is capital intensive and 5% 44 12.9 438 utilizes specialized equipment that is 0% 41 13.9 413 usually imported. This means that Gold Model: an import duty on equipment has a 15% 45 14.1 78 direct impact on project economics 12% (current) 43 14.7 76 in the project's early years. 5% 40 16.3 69 Project feasibility studies calculate o% 37 17.5 65 various projections of profitability, such as discTunted rate of return, and such measures are very sensitive to large costs in the early years of a project. Even modest levels of equipment import duties can sink a marginal project. Competition for mineral sector investment worldwide is fierce, and many countries have either eliminated import duties on mine equipment or have found ways to exempt projects or their equipment from such duties. The effect of eliminating the import duty is shown in Table 8.17. The impact of import duty is high on both government and taxpayer. Recommendations: SPeru should follow the lead of most other nations and eliminate import duty through either an exemption or if duty. category lists are sufficiently detailed to isolate most mining equipment, by zero rating such categories > However, if this recommendation- is adopted, a royalty tax of 2-3 percent should h- imposed on gross mineral sales revenues. The combined effect of eliminating import duty and adding a royalty tax is shown in Table 8.18. 8.23 Peru levies an 18 general sales tax (lGV in its Spanish acronym) on goods and services purchased. It must be paid (n goods and services, but in the case where a mine exports its output, it can be credited against the Income Tax up to a limit of 18 percent of such sales of exported minerals. Because the VAT is a "consumer" tax and export minerals must compete globally, almost all mineral-exporting nations have chosen to eliminate its impact on both export mineral sales and equipment purchases. Peru eliminates that impact through allowing the possibility that the IGV on export sales be reimbursed. The amount of IGV that may be credited is restricted to 18 percent of export sales, and any amount exceeding this may be carried forward as a credit against future export sales. Moreover, the IGV paid during exploration is refundable. Recommendation: > The current system of IGV as applied to mining should remain unchanged. 5There is a negative impact on mining competitiveness of the 8 percent workers' participation since it increases the real income tax rate burden to firms from around 30 to about 38 percent (para.8.13). Authorities informed that this participation could be diminished and unified across all similar activities in an incoming initiative. 141 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 8.24 Additional recommendations Table 8.18 Tax Sensitivity to Eliminating Import Duty and for improving the use of Impoing Royalty Government intergovernmental transfers are next: Effective Revenue: Al Tax Rate Investor IRR Taxes & Fees > Transparent managing of mining Copper Model: tax revenue. The administration of 12% import duty, 0% 47 11.7 471 royalty (current the channeling and use of canon system) minero transfers need to be better 0% import duty, 1% 43 13.5 436 known. An important step royalty forward is the GOP's recent 0% import duty, 2% 46 13.1 460 decision to publish the monthly royalty revenue transfers to municipalities 0% import duty, 2.5% 47 12.8 472 in the Portal de Transparencia royalty in hePota d Tanpaenia 0% import duty, 3% 48 12.6 484 Econ6mica. This is not enough. royalty In the PETS developed (see Gold Model: Chapter IV), about 6 out of ten 12% import duty, 43 14.7 76 municipalities outside Lima 0% royalty (current claiming that they knew how their system) transfers were estimated, actually 0% import duty, 1% 41 16.6 71 royalty did not know, and in the poor and 0% import duty, 2% 45 15.6 78 extreme poor, such number royalty decreased to 3 out of ten. 0% import duty, 46 15.2 86 Transfers also show high volatility, 2.5% royalty unpredictability and are subject to 0% import duty, 3% 48 14.7 84 limited auditing capacity by the royalty Central Government (Apoyo Institute, 2002). The emphasis of new regulations and its subsequent administration should be on ensuring that the local communities are well trained and informed on how and when these transfers are estimated. >' Ways to stabilize the annual flows Box 8.1 Creating Mining Stabilization Funds? The Ancash of mining transfers (Stabilization Model? 6 Fund) should be explored, as the Given the unpredictability of fiscal transfers, a Mining Stabilization cyclical behavior of international Fund is desirable. It could be created with private sector participation affet icom taxrevnue and have a long-term view to finance: local infrastructure projects prices afeticm a .eeu prepared by municipalities under participatory plans and capacity and, therefore, canon minero building activities. The Fund could be supervised by a Mesa de flows (Box 8. 1). Concertacit5n and be subject to strengthened auditing procedures by the official authorities. > Local adminrstration capacity A pilot case, though not based on government transfers, has already needs to be strengthened. been created with the Ancash Fund. This Fund contains about US$1 10 Municipalities should be supported million of payments, which the firm Antamina still had to pay in the in learning how to do optimal use context of its privatization. Rather than paying those straight to the Treasury, the company and the involved mayors have proposed a of these resources. They should "Fund model": The Treasury, upon collecting from Antamnina, puts the develop capacity for managing entire revenue into this Fund, which then finance infrastructure them: taking investment decisions, projects in the region (13 projects making up about 60 percent of te available funds are already identified, including roads, assessments, and and eventually managing te electrification). Resources of the Fund could also regularly be assets created on their basis. relilenisbed with part of ranon minero transfers- Local administration officials and local community representatives would need to be enabled to participate in local economic development programs that might be initiated together with or by mining companies. 6Once the annual amount is defined, the monthly transfers of the canon minero are constant. 142 Chapter VIII. FINE-TUNING MINING FISCAL AND ENVIRONMENTAL POLICIES D. IMPROVING MINING ENVIRONMENT MANAGEMENT 8.25 Mining environment is cause for heated social and political concerns in Peru. Complex political debates have emerged around existing and newly planned mining operations, typically in the context of criticisms that local communities and regions do not gain much from mining operations, the marginal employment these mines create for unskilled indigenous workers, the often negative social consequences n the communities living close to the mines (import of prostitutes and alcoholism are often mentioned as major social concerns) and the lack of genuine interest of some mining companies to promote regional development and respect environment provisions. 8.26 Three weaknesses in Peru's environmental management are conflict of interests, credibility, and capacity concerns. The mining sector, in general, has complied better with the environmental institutional assessment requirements, as well as with territorial environmental assessments. However, there are several instances of potential conflict of interests. O A conflicting case occurs between core technical groups within the Ministry of Mines and Energy (MEM), who wish to promote mining, and the environmental uit within the Ministry, whose mandate is to prevent environmental damage. O Another conflicting case is the inspection of the Environmental Compliance and Management Programs in the energy and mining sectors. Aside from potentially presenting technical deficiencies, such inspection is under the responsibility of the General Directorates (electricity, hydrocarbon and mining) with the help of independent experts contracted by the respective enterprises being evaluated, again in a clear conflict of interest. o The reform of the early 1990s had approached the environmental issue from a sectoral perspective, i.e. one where each sectoral ministry has a General Directorate of the Environment (in the case of the MEM it would be the GDE-MEM), which in effect regulates and monitors environmental compliance. The only entity resembling a national environmental authority-the National Environmental Council (CONAM)-is limited to a very weak, coordinating intersectoral role. This has resulted in a perception by local communities of a conflict of interest at the MEM, which plays a two-fold role of promoting mining investment and implementing environmental control, and the fear that environmental control is too lax at the expense of the health of the locals. This situation is aggravated by the fact that GDE- MEM does not release environmental information to the public. o The institutional structure is geographically centralized at the Lima-based GDE-MEM as the only responsible entity (the regional offices of the GDE-MEM are quite weak). Local populations only participate in consultations and do not form part of the institutional structure of the process. 8.27 The mining industry's environmental record has begun to impede investments. In practice, even if better than in some other sectors, the environmental record of the mining industry has begun to impede private sector development. Even when controls existed on paper, they are seldom implemented in practice. Social and political conflicts that were mainly triggered by accidents impacting the environment or the social fabric of the local community, such as spills or resettlement issues, have threatened the ability of companies to pursue mining permits or to continue to run their already existing operations. For example, Tambogrande in Piura, a medium- sized gold mining project awarded to Canadian Manhatten Minerals, is unlikely to go ahead, because the "Tambogrande defence committee" has successfully mobilized public opinion, 143 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION including the Catholic church, staging demonstrations and boycotts serious enough to cause the company to abandon completion of its environmental impact analysis. Other examples of such conflicts causing interruptions of ongoing project development or operations are the mines of Yanacocha and Antamina. Both were involved in massive community protests, one in the context of a mercury spill, the other in the context of a resettlement program that halted their operations temporarily and threatened to stop further expansion. 8.28 The current enforcement by the state entities responsible for the environment is inadequate. The National Environmental Council is a small organization lacking political power. In practice, it is the Ministry of Energy and Mines that has most influence over investment decisions. However, the team at the ministry responsible for environmental controls lacks the personnel or budget to carry out its functions properly. Naturally, given its mandate, rather than impose burdensome environmental controls, the ministry's main priority is to maximize investment and eventual output. 8.29 The role of the private sector: From an environmental perspective, the worst offenders tend not to be the large multinational mining companies, but rather the-Peruvian- owners of older and smaller mines, less sensitive than multinationals to allegations of disregarding environmental norms. Most multinationals have made serious efforts to comply with the environmental code and invest in technologies to reduce pollution and to create a good working relationship with surrounding communities. Only legal obligations, enforced by competent and credible institutions, will be able to influence the large number of Peruvian owned mining firms to control water pollution and spills, to control air pollution, and to appropriately solve conflicts over land rights. Recommendations: > A new national and independent environmental authority that is structured geographically and supported by local monitoring needs to be created. Such authority should regulate all sectors and be integrated with future regional authorities, which will be responsible for the monitoring and control of all environmental aspects. > As part of the decentralization process, an agreement on roles and responsibilities in the environment provisions of the mining sector is also needed to foster further private investment. This agreement should define a certain bvel of inter- and intra-provincial coordination to organize environment controls smoothly and to promote cooperation. > New roles and responsibilities will require the buildup of institutional capacity, especially at the subnational level. > The regulatory framework of environmental management also needs to be reviewed, so as to make sure stricter environment regulations are well designed and introduced. > Local interest groups, supported by non-governmental organizations (NGOs), should be integrated in environmental monitoring and consultation. These groups will certainly continue to pressure the government to protect community interests from mining projects. 144 Bibliography Altobelli, L. (2001), "Decentralization to Improve Health Care for the Poor," Background Paper for The World Bank (2001?). Apel, M. J. Hansen and H. Lindberg (1996), "Potential Output and Output Gap," Quarterly Review of the Bank of Sweden, Vol. 3, pp. 24-35. Apoyo Institute (2002a), "A Public Expenditure Tracking Survey (PETS): Central Government Transfers to Municipalities in Perii," Background Paper for the Public Expenditure Review. Apoyo Institute (2002b), "PETS-The Education Sector in Peri," Background Paper for the Public Expenditure Review. Arias L. et. Al. (2000), "Ajuste Fiscal Ex-Ante o Ex-OPost: D6ficit Ciclicos en Perd," on Talvi E. and C. Vega, Como Armar el Rompecabezas Fiscal, IDB, Washington DC. Backus, D. K. and P. J. Kehoe (1992), "International Evidence on the Historical Properties of Business Cycles," The American Economic Review, Vol. 82, pp 864-888. Banco Central de la Repfiblica de -Peru (2002), "Sostenibilidad de la Politica Fiscal," Jorge Estrella, Seminar on Fiscal Rules, Lima. Brown, E. Cary (1956), "Fiscal Policies in the Thirties: A Reappraisal," The American Economic Review, Vol. 46, December, pp. 857-879. Burki, S., G. Perry and W. Dillinger (1999), Beyond the Center: Decentralizing the State, The World Bank (Washington, D.C. 1999). CEPAL (1998), "La Descentralizaci6n de la Educaci6n y la Salud," CEPAL, Santiago de Chile. CEPAL (2001), Decentralizaci6n en Amirica Latina: Teoria y Practica, IvAn Finot (CEPAL LC/L 1521), Santiago de Chile. Cortez, R. (1998). Equidad y Calidad de los servicios de Salud: El Caso de los CLAS," Universidad del Pacifico/Ministerio de Salud, Lima. Cortez, R. (2001) "Capitulo de Salud," Universidad del Pacifico, Lima. De Leeuw, Frank, Thomas M. Holloway, Darwin G. Johnson, David S. McClain, and Charles A. Waite (1980), "The High Employment Budget: New Estimates, 1955-80," Survey of Current Business, Vol. 60, November, pp. 13-43. Deutsche Bank (2001), "Public debt Sustainability in Selected Emerging Markets." Diaz, J.,(2001), "Grado de Focalizaci6n de Algunos Programas Sociales, 1998-2000," Ministerio de Economia y Finanzas, Direcci6n de Asuntos Econ6micos y Financieros, Lima. 145 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Dolado, J. J., M. Sebastidn and J. Vall6s (1993), "Cyclical Patterns of the Spanish Economy," Investigaciones Econdmicas, Vol. XVII, pp. 445-473. Dombusch, R. and S. Fischer (1990), Macroeconomics, New York: McGraw-Hill. Dornbusch, R. and S. Edwards (1991), The Macroeconomics of Populism, NBER, The University of Chicago Press, Chicago. European Central Bank (2001), "Potential Output Growth and Output Gaps: Concept, Uses and Estimates," ECB Monthly Bulletin, October, pp. 37-47. Fellner, William (1982), "The High Employment Budget and Potential Output: A Critique," Survey of Current Business, Vol. 62, November, pp. 25-33. Femndez, E. (2002), "La Deuda Corriente de los Municipios Peruanos," Background Paper for the PER. Washington DC. Finot, Ivan (2001), "Descentralizaci6n en Amdrica Latina: teoria y prdctica," Instituto Latinoamericano y del Caribe de Planificaci6n Econ6mica y Social - ILPES. Giorno, C., P. Richardson, D. Roseveare and P. van den Noord (1995), "Estimating Potential Output, Ouput, Output Gaps, and Structural Budget Balances," OECD Economics Department, Working Papers, No. 152. Gonzalez C., D. Rosenblatt and S. Webb (2002), "Stabilizing Intergovernmental Transfers in Latin America: A Complement to National/Subnational Fiscal Rules," Paper for the Seminar on Fiscal Rules, Lima. Gregory P. (2002), "Public Sector Employment in Peru," Background Paper for the PER, Washington DC. Gupta, Sanjeev, Marijn Venhoeven and Erwin Tiongson (1999), "Does Higher Governance Spending Buy Better Results in Education and Health Care?" IMF Working Papers WP/99/21. Fiscal Affairs Department. Hagemann, Robert (1999), "The Structural Budget Balance: The IMF's Methodology," IMF Working Papers, July, Washington, DC. IMF (2000), "Peru: Racionalizaci6n del Sistema Tributario," G. Kopits, J.P. Ley y J. C6rdova, IMF departamento de Finanzas P-6blicas. IMF/WB (2001), "Tracking of Poverty Reducing Spending Heavily Indebted Poor Countries (HIPC)," IMF, Washington DC. IMF/WB (2002), "Actions to Strengthen the Tracking of Poverty Reducing Spending Heavily Indebted Poor Countries (HIPC)," IMF, Washington DC. Instituto Apoyo (2002), "Central Government Transfers to Municipalities in Per6: A Detailed Look at the Vaso de Leche Program," Public Expenditure Tracking Survey. Report to the Public Expenditure Review of Peru. 146 BIBLIOGRAPHY Jayasuriya R. and Q. Wodon (2001), "Explaining Country Efficiency in Improving Health and Education Indicators: The Role of Urbanization," Background Paper to World Development Report 2003. The World Bank. Jim6nez F. (2002), "Peril: Sostenibilidad, Balance Estructural y Propuesta de Regla Fiscal," Seminar on Fiscal Rules, Lima Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobat6n (2002), "Governance Matters II: Updated Indicators for 2000/01," The World Bank, Washington D.C. Kopits, George; Juan Pablo C6rdoba, Eduardo Ley and Juan Toro (2000), "Peri: Racionalizaci6n del Sistema Tributario," Fondo Monetario Intemacional. Kopits, George"(2002); "Una Vista Panorimica de las Reglas Macro Fiscales," Paper of the Seminar on Fiscal Rules, Lima. Larrain F. and J. Sachs (1991), Macroeconomics in the Global Economy, Prentice Hall, New Jersey. L6pez-Cilix, Jos6 R. (2002), "Guatemala: La Macroeconomia del Ciclo Electoral," Banca Central 40, Enero-Abril, Guatemala City. Ma J. (2002), "Monitoring Fiscal Risks of Subnational Governments," paper in H. Polackova and A. Schick, Government at Risk, The World Bank, Washington DC. Matsuda and Shepherd (2000), "Peru's Public Administration and the Delivery of Public Services," Background Paper to the Peru IGR, Washington DC. MED (2001), "Principales Lineamientos de Politica Educativa," Lima. Moore, Richard (2002), "Decentralization in the Social Sectors: Prospects for the Decentralization of Health and Education in Peril:." Background Paper to the Public Expenditure Review of Peru. Mor6n, Eduardo (2001), "El Debate de la Transparencia," Universidad del Pacifico. Lima. Mostajo, Rossana (2002a), L"Protecei6n Presupuestaria? - Los Programas Sociales Protegidos en el Peri," Background Paper to the Public Expenditure Review of Peru. Mostajo, Rossana (2002b), "El Sistema Presupuestario en el Peri," Serie: Gesti6n Pblica, Instituto Latinoamericano y del Caribe de Planificaci6n Econ6mica y Social - ILPES. Oliva C. (2002), "Fiscal Trends and the Fiscal Sustainability Isuue," Background Paper to the Public Expenditure Review of Peru. Ortiz de Zevallos, G. et al. (1999), "La Economia Politica de las reformas Institucionales en Perd: los Casos de Educaci6n, Salud y Pensiones," prepared for the IDB, Dopcumento de Trabajo R-348, Washington DC. 147 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Otto, M. James (2002), "Position of the Peruvian Taxation System as Compared to Mining Taxation Systems in Other Nations," Ministry of Economy and Finance. Report to the Public Expenditure Review of Peru. PCM (2002), "The Strategy of Decentralization," Mimeo, Lima. Perry G. (2002), "Can Fiscal Rules Help Reduce Macroeconomic Volatility in LAC," Paper presented at a Seminar in Oaxaca, Mexico Perry et al. (2000), Securing Our Future in a Global Economy, The World Bank, Washington DC. Reinikka, R. and J. Svensson (2002), "Assessing Frontline Service Delivery," Draft, World Bank. Rose-Ackerman (1999). Corruption and Government: Causes, Consequences and Reform. Cambridge: Cambridge University Press. Saavedra, J. et al. (2001), "El Financiamiento de la Educaci6n Pblica en el Pei: El Rol de las Familias," GRADE, Lima. Schenone, Osvaldo H. (2001), "Exoneraciones y Regimenes Tributarios Especiales en Peri," Background Paper to the Public Expenditure Review of Peru. Seligson, M. (2002), "The Impact of Corruption on Regime Legitimacy: A Comparison Study of Four Latin American Countries," Journal ofPolitics 64: 408-33. Schady, N. (2002), "Notes on A Trabajar Program," Mimeo, The World Bank, Washington DC. Treisman, D. (2000), "The Causes of Corruption: A Cross National Study," Journal of Public Economics 3: 399-458. Valderrama J. (2002), "Peri: Propuesta de Reforma a las Reglas Fiscales," paper for the Seminar on Fiscal Rules in Lima. Wodon, Q. et al. (2001). "SIMSIP: a Simulation Tool," Draft, The World Bank. Washington DC. Wodon, Quentin (2002), " The Efficiency of Public Spending in Peru," Background Paper to the Public Expenditure Review of Peru. World Bank (1994), "Peru: Public Expenditure Review," Washington DC. World Bank (1999), "Peru: Improving Health Care for the Poor," Washington DC. World Bank (2000), "Peru: Policy Notes," Washington DC. World Bank (2001a), "Peru: Institutional and Governance Review." Poverty Reduction and Economic Management Unit, Latin America and the Caribbean. World Bank (2001b). "Peruvian Education at a Crossroads: Challenges and Opportunities for the 21" Century," Washington DC. 148 BIBLIOGRAPHY World Bank (2001c), "Voices of the Misgoverned and Misruled: An Empirical Diagnostic Study on Governance, Rule of Law and Corruption for Peru. A Survey," WBI, Washington DC. World Bank and Inter-American Development Bank (2001a), "Peru: Country Financial Accountability Assessment." Draft, Washington DC. World Bank and Inter-American Developiment Bank (2001b), "Peru: Country Procurement Report," Draft, Washington DC. 149  ANNEX A A PUBLIC DEBT SUSTAINABLITY ANALYSIS FOR PERU' Introduction We estimate the level of Peru's public debt that would improve its sustainability in the medium term. Authorities have shown their intention to develop a prudent debt management strategy. The 2000 debt exchange with private creditors slightly worsened the maturity profile of debt payments,2 but authorities believe that the present value of public debt fell. Our baseline medium-term scenario indicates some external vulnerability over the medium term, conditional on low inflation, moderate real GDP growth (average 4 percent) per year, which in turn would depend on recovery of private sector and foreign investment, increased productivity and sound macroeconomic policies. Our debt sustainability model assumes the absence of major terms of trade shocks that would negatively affect the current account deficit, and no private capital outflows that would require a tighter monetary policy to avoid currency depreciation. We examine the sensitivity of key public debt ratios to alternative assumptions on financing terms and growth rates. Our central premise is that both foreign financing and adjustment to fiscal fundamentals are necessary to ensure more sustainable public debt ratios. The path towards sustainability is estimated, with annual target primary balances consistent with non-financial public sector debt and debt-servicing requirements. The analysis concludes that the Government could reach a public debt to GDP sustainable ratio of 40 percent in 2004 and 36.2 in 2010, provided it builds primary balances close to 0.9 percent of GDP in 2004 and 1.3 percent of GDP in 2010. Nonetheless, the Government would face a sizable debt-service burden that would peak at about 36 percent of exports and 44 percent of tax revenue in 2004 and decrease thereafter, the latter result being a direct reflect of the fiscal deficits and low tax ratios that Peru features in the past years. Faster growth accompanied by tight control of expenditure and additional foreign financing would be essential to bring the GOP to a high case scenario, so as to channel more resources toward a poverty reduction strategy, while reducing its debt-servicing burden. A comprehensive tax reform is needed to improve such prospects. The Annex is divided into two sections. The first one is a conventional debt sustainability analysis as described above. The second one considers the impact of two hypothetical shocks. The first shock consists of a slowing down of capital inflows (sudden stop) resulting in real exchange rate depreciation. The second shock reflects the materialization of the government's contingent liabilities, arising from an eventual failure of the private sector to fulfill its short-term external debt obligations. Peru's Macroeconomic Background While economic activity remains weak, the macroeconomic situation has been stabilizing since 2000, with inflation decreasing from 7.3% in 1998 to 2% in 2001. The combined fiscal deficit reversed its upward trend, but still was significant at 2.5% of GDP. The current account as a 'This Annex was prepared with thorough research assistance of Rashmi Shankar, numerous comments to early drafts and data support from IMF staff, and a very timely suggestion from Sara Calvo, who motivated its extension to shocks and provided enthusiastic advise to its development. 2 New government 10-year bonds with bullet payment at maturity exchanged for front-loaded interest reduction Brady bonds or FLIRBs, which were selling at a discount of 28.5% at, end 2001. These carry low interest rates, and the principal is to be amortized gradually from 2005 to 2017. 151 PERU: RESTORING FISCAL DISCIPLINE FOR POvERTY REDUCTION percent of GDP has improved significantly between 1998 and 2001, from -6.4% to -2.1%. Measures to strengthen the banking system are reflected in improvements in some indicators of bank soundness. Loan quality has improved, as measured by the ratio of non-performing to total loans (excluding restructured and refinanced loans), and liquidity and capital adequacy ratios have risen. While total (public and private, short, medium and long-term) debt to GDP ratio remains high at 52.3%, adverse developments in the region have had little effect on Peru's financial indicators. Yet, the external public debt represents around 70 percent of total debt and virtually all is long term. Around 60 percent of this debt is with multilateral and bilateral creditors. Because of the high degree of dollarization of domestic debt, it is vulnerable to currency risk, making fiscal and monetary policy lose some degree of freedom. For illustrative purposes, a comparison of Argentina's (before the crisis) and Pern's fiscal deficit,. public debt-GDP ratio, public debt-servicing-exports ratio and current account balance is contained in Table 1. Peru has a lower non-financial public sector deficit as well as a lower current account deficit. Argentina's public debt and public debt servicing burden is higher than Peru's, but the rising trend in the public debt service/exports ratio in the latter is worrisome. Strengthening of the export performance, the banking sector and particularly the fiscal position will play a significant role in keeping Peru's public debt manageable. In a base case, inflation remains low at approximately 2% which would allow the government to relax somehow monetary policy in the face of weak economic activity. A managed floating exchange rate regime is preserved and the international reserves position would remain comfortable. Table 1: Current Macroeconomic and Financial Trends (% of GDP) 1998 1999 2000 2001* Argentina Peru Argentina Peru Argentina Peru Argentina Peru Current Account -4.80 -6.37 -4.30 -3.50 -3.20 -3.00 -2.80 -2.10 Public Debt Service 56.80 27.46 73.50 32.65 70.50 42.80 78.20 43.30 (% of Exports) Public Debt Service 31.40 25.26 36.80 31.89 38.90 33.53 48.30 28.89 (% of Revenue) Public Debt 37.60 42.70 43.00 48.00 .44.90 45.90 51.0 46.50 NFPS Budget Deficit -2.40 -0.70 -4.40 -3.10 -3.80 -3.20 -3.50 -2.50 *Pre-swap estimate for Argentina The Target Fiscal Adjustment under a Baseline and Alternative Scenarios with no External Shocks We present alternative scenarios of growth and interest rates, multilateral financing and the sensitivity of key public debt ratios to the assumptions made. The dynamic is detailed in Table 2. In each case, the target primary balance needed in order to sustain the public debt outcome is estimated. 152 ANNEX A General Assumptions * Receipts from privatization are US$600 million for 2002 and 2003, and fall to less than $50 million per year thereafter. * Inflation in the GDP deflator (US$ terms) increases from 0.9 percent in 2002, to approximately 2 percent over the period 2003-2010. * Continued rollover of short-term debt over the entire period 2002-2010. * The terms of new multilateral financing depends on Peru's classification as a category IV country. The loan period is on average 17 years, with a five-year grace period, and the interest is LIBOR (World Bank forecast of 4 percent) plus a 70 basis point spread, on average. * The non-financial public sectors elasticities of revenues and expenditures with respect to GDP remain constant at 0.98 and 0.95 respectively. Elasticities are estimated over the 1994- 2002 period. Scenario A. This is the baseline scenario. It incorporates the preliminary figures for the Brady buy-back under which all-outstanding front-loaded interest reduced bonds have been exchanged. The baseline assumptions are: * moderate growth: 3.7 percent in 2002, followed by 4 percent over the period 2003-2010; * moderate average international interest rates: 10 percent throughout the period; * moderate cut in fiscal deficits: tapering off from -1.9 percent of GDP to -1.4 percent of GDP over the period; Results: Under the baseline assumptions, the public debt-GDP ratio would come down to 36.2 percent in 2010, from 44.9 percent in 2002. The primary balance required to cover interest liabilities increases from 0.2 percent to 1.3 percent over the period. The public debt servicing ratio falls from 27.4 percent to 23.7 percent as a proportion of exports, rises from 3.8 percent to 4.6 percent as a proportion of GDP, and rises from 32.1 percent to 42.4 percent as a proportion of central government tax revenues, over the period. The assumption on mixed concessional- multilateral (reduced balance of payments support from 2003, partly offset by sector investment disbursements) and limited commercial borrowing results in an implicit average interest rate on the total stock of NFPS debt between 6.7 percent and 7.9 percent over the period 2004-2010. Scenario B. This is the best-case scenario, which presents a sensitivity analysis of the public debt-GDP ratio to higher growth and strengthened fiscal discipline. The assumptions are: * high growth: 3.7 percent in 2002 followed by 5 percent over the period 2003-2010; * low average international interest rates: The interest rate on new public debt is 7.5 percent over the period; * low fiscal deficits: tapering off, as a proportion of GDP, from -1.9 percent in 2002, to -1.8 percent in 2003, -1.3 percent in 2007 and -1 percent by 2010. This is based on the assumption of higher revenue, tighter expenditure and faster growth. The revenue elasticity remains .98, but the expenditure elasticity is constrained to remain at .88 so as to cap expenditures at 18.3% of GDP in this scenario, in contrast to 18.7% of GDP in the baseline. Results: The public debt-GDP ratio would fall lower to 32.2 percent in 2010 under this scenario, from 44.9 percent in 2002. The primary balance required for reaching sustainability would also be lower than in the baseline, due to faster growth and better financial conditions. To cover debt service liabilities it increases from -0.5 percent in 2000 to 1 percent in 2005, and remains stable at that proportion of GDP till the end of the forecast horizon. The public debt-servicing ratio falls from 27.4 percent in 2002 to 20.5 percent as a proportion of exports, remains constant at 3.7 153 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION percent as a proportion of GDP, and increases slightly from 32.1 percent to 34.3 percent as a proportion of central government tax revenue, over the same period. The assumption on mixed concessional-multilateral (full balance of payments support) and commercial borrowing results in an implicit average interest rate on total NFPS debt around 6.4 percent over 2004-2010, which is lower than in the baseline due to a higher share of multilateral financing and better placement conditions on international markets. Scenario C. This is the worst-case scenario, reflecting the sensitivity of the public debt-GDP ratio to low growth and poor fiscal discipline. The assumptions are: * Low growth: Growth rate of 2.5 percent is assumed over the entire period; * High fiscal deficit: The fiscal deficit falls from -2.4 percent to only -1.9 percent of GDP over the entire period; * High average international interest rate: The interest rate on public debt is 12 percent over the period. Results: The public debt-GDP ratio falls slightly to 42.7 percent in 2010 from 45.4 percent in 2002, under this scenario. However, the primary balance required to cover debt service liabilities increases to 2.2 percent in 2010 from -0.5 percent in 2000. The public debt-servicing ratio increases from 27.4 percent to 28.7 percent as a proportion of exports, rises from 3.8 percent to 6.2 percent as a proportion of GDP, and rises from 32.1 percent to 56.7 percent as a proportion of central government tax revenue, over the 2002-2010 period. The implicit average interest rate on total NFPS debt increases to between 7.3 percent and 9.9 percent over 2004-2010 because of the increased reliance on commercial borrowing following from the previous assumption of reduced multilateral financing and worst bond placement conditions in international markets. Thus external financing is more expensive in this scenario. It is assumed that there is reduced multilateral financing and no fast disbursing loans. Table 2: Debt Dynamics Non-Financial Public Sector: Scenario With No External Shocks (in percent of GDP) 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 201 1. Baseline: Scenario A Primary balance -0.5 -0.6 0.2 0.4 0.9 1.1 1.1 1.2 1.3 1.3 1. Public Debt to GDP 45.9 46.5 44.9 40.9 40.0 39.3 38.7 38.2 37.5 36.9 36. Public External Debt Service/Exports 25.0 23.3 27.4 31.5 36.1 38.1 30.9 28.7 26.7 25.6 23. Public Debt Service/GDP 3.8 3.9 3.8 4.5 5.3 6.0 5.2 5.0 4.9 4.8 4. Public Debt Servicefrax Revenue 27.4 26.1 32.1 36.1 43.5 50.0 43.7 43.3 43.1 43.6 42. Memorandum items Fiscal balance (percent ofGDP) -3.2 -2.5 -1.9 -1.8 -1.8 -1.7 -1.7 -1.6 -1.5 -1.5 -1.A GDP (millions of USS) 53513 53983 56816 60068 63033 66800 70861 75170 79740 84671 89815 Exports (millions ofUS$) 7028 7132 7915 8525 9325 10548 11833 13107 14568 15984 1740Y L Scenario B Primary balance -0.5 -0.6 0.2 0.4 0.8 1.0 1.0 1.0 1.0 1.0 1. Public Debt to GDP 45.9 46.5 44.9 40.8 39.4 38.2 37.1 36.0 34.8 33.6 32.' Public External Debt Service/Exports 25.0 23.3 27.4 31.2 35.2 36.7 29.0 26.4 24.2 22.7 20. Public Debt Service/GDP 3.8 3.9 3.8 4.4 5.2 5.7 4.7 4.4 4.2 4.0 3. Public Debt Service/Tax Revenue 27.4 26.1 32.1 35.7 42.0 47.3 40.0 38.4 37.2 36.5 34. Memorandum items Fiscal balance (percent ofGDP) -3.2 -2.5 -1.9 -1.8 -1.7 -1.5 -1.4 -1.3 -1.2 -1.1 -1. GDP (millions of USS) 53513 53983 56816 60068 63639 68091 72925 78103 83648 89675 9604 Exports (millions ofUSS) 7028 7132 7915 8525 9325 10548 11833 13107 14568 15984 174 3. Scenario C Primary balance -0.5 -0.6 -0.3 0.0 0.7 1.1 1.3 1.6 1.9 2.1 2. Public Debt to GDP 45.9 46.5 45.4 41.3 41.4 41.7 42.0 42.4 42.6 42.7 42.7 Public External Debt Service/Exports 25.0 23.3 27.4 32.2 37.9 40.8 34.3 32.6 31.1 30.3 28.' Public Debt Service/GDP 3.8 3.9 3.8 4.6 5.7 6.6 6.0 6.0 6.1 6.2 6.: Public Debt Servicefrax Revenue 27.4 26.1 32.1 36.8 46.3 55.0 50.7 52.1 53.8 56.2 56. Memorandum items Fiscal balance (percent of GDP) -3.2 -2.5 -2.4 -2.3 -23 -2.2 -2.1 -2.1 -2.0 -2.0 -1. GDP (millions of USS) 53513 53983 56816 60068 62124 64887 67839 70926 74153 77603 811 Exports (millions ofUSS) 7028 7132 7915 8525 9325 10548 11833 13107 14568 15984 17 154 ANNEX A Summary: Table 2 and the charts (Figure Al) provided below present details of the dynamics of public debt financing toward medium-term sustainability. It is important to note here that in the worst-case scenario, the implication of a required government primary surplus of 2.2 percent of GDP merely reflects that a bigger fiscal effort would be required in a deteriorated scenario for Peru to honor its payments (see tables Al-A3). The Target Fiscal Adjustment under a "Sudden Stop" Scenario: Accounting for a Real Fxchange Rate and Contingent Liability Shocks Motivation The recent literature on crises identifies a combination of fiscal sustainability and real exchange rate misalignment as a likely cause of rapidly growing public debt and a loss of access to international credit markets. One perspective stresses that unexpected stops in capital flows of a permanent nature can in itself generate substantial swings in the real exchange rate. Our purpose here is to identify the real exchange rate adjustment necessary to close the current account in the event of Peru being the victim of such a shock, i.e. a sudden stop in capital inflows, and to assess its impact on debt sustainability. Computing the Real Exchange Rate Adjustment A sudden stop in capital inflow is typically accompanied by large contractions in international reserves and a decline in the relative price of non-tradables with respect to tradables (or a depreciation in the real exchange rate). This follows from the current account identity: a relative price adjustment is necessary to absorb the demand for tradables in order for the current account deficit to be reduced in line with the fall in capital flows. Calvo et al (2002)3 demonstrate that the necessary real exchange rate adjustment is given by a percentage fall in the resource imbalance necessary to reduce the current account deficit to the required extent (current account deficit/imports), divided by the elasticity of demand for non-tradables with respect to p (the relative price of non-tradables to that of tradables). Based on their model, we compute the following RER adjustments for Peru, Bolivia and Ecuador: Table 3: Real exchange rate adjustment in the event of a "sudden sto ": Perd9 Bolivia Ecuador Elasticity of demand for non-tradables with 0.4 0.4 0.4 respect to p (x) w =1-(current account deficit/imports) 0.79 0.72 0.68 Real exchange rate depreciation 34% 41% 45% Source: World Bank staff calculations based on Calvo et al (2002). It must be emphasized that these numbers are extreme values based on the assumption that capital flows will cease completely, and the current account deficit will have to be closed. A more reasonable assumption, i.e. a 10 percent depreciation in the real exchange rate would capture the idea that in the event of a shock, capital inflows may slow down, but not cease completely. Figure A2 contains a graphical representation of the implication of 10 percent real exchange rate adjustments for public debt sustainability in Peru. In the sensitivity exercises (based on the same assumptions as in the previous section), we are therefore accounting for a shock that leads to a slowing down, but not complete cessation of capital inflows. 3Calvo, Guillermo, Alejandro Izquierdo and Ernesto Talvi, "Sudden Stops, the Real Exchange Rate and Fiscal Sustainability: Argentina's Lessons", IADB Research Department, 2002. 155 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Additionally, we account for the impact on public debt dynamics of the contingent liabilities arising from a possible failure of the private sector to meet its short-term contractual debt obligations. We compute contingent liabilities as the share of foreign credit to the private sector that is allocated to non-tradable activities. In Peru's case, the non-tradables share is roughly two- thirds of the economy (i.e., 1-(trade/GDP)=.67). Short-term private sector liabilities are roughly US$3.9 billions, i.e., nearly 45 percent of the total private sector debt, reported to be US$8.7 billions. Sensitivity Analysis and the Target Fiscal Adjustment (if the RER adjusts by 10% and contingent liabilities are accounted for) Scenario A: Under the baseline scenario, incorporating a real exchange rate adjustment of 10% (assumed to occur in 2002), results in a higher public debt to GDP ratio of 37.8% in 2010, compared to the unadjusted figure of 36.2%. Similarly, the required primary balance rises to 1.5% of GDP from 1.3%. Once we add the contingent liabilities, public debt to GDP ratio increases to 40.8%, in 2010, while the required primary surplus increases to 1.8% of GDP. Scenario B: Under the best-case scenario, incorporating a real exchange rate adjustment of 10% (assumed to occur in 2002), results in a higher public debt to GDP ratio of 33.5% in 2010, compared to the unadjusted figure of 32.2%. Similarly, the required primary balance rises to 1.3% of GDP from 1%. Once we add the contingent liabilities, public debt to GDP ratio increases to 36.4%, in 2010, while the required primary surplus increases to 1.5% of GDP. Scenario C: Under the worst case scenario, incorporating a real exchange rate adjustment of 10% (assumed to occur in 2003), results in a higher public debt to GDP ratio of 44.9% in 2010, compared to the unadjusted figure of 42.7%. Similarly, the required primary balance rises to 2.6% of GDP from 2.2%. Once we add the contingent liabilities, public debt to GDP ratio increases to 48.4%, in 2010, while the required primary surplus increases to 3.0% of GDP. Summary: Figure A3 presents a comparison of public debt to GDP ratios and required primary surplus as a percent of GDP, under the three scenarios with and with no shocks. Table 4 presents a summary of these results under alternative assumptions (see Tables A4-A6). Table 4: Public Sector Debt Sustainability With and With No Shocks: in percent of GDP) 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 201 1. Baseline: Scenario A Primary balance (unadjusted) -0.5 -0.6 0.2 0.4 0.9 1.1 1.1 1.2 1.3 1.3 1. Primarybalance(tealexchangerateadj.) -0.5 -0.6 0.3 0.5 1.0 1.3 1.3 1.4 1.4 1.5 1. Primary balance incl contingent liabilities -0.5 -0.6 0.8 1.0 1.5 1.7 1.7 1.7 1.8 1.8 1. Public Debt to GDP (unadjusted) 45.9 46.5 44.9 40.9 40.0 39.3 38.7 38.2 37.5 36.9 36. Public Debt to GDP (real exchange rate adj.) 45.9 46.5 47.3 43.3 42.2 41.4 40.7 40.1 39.3 38.6 37. Public Debt to GDP (contingent liab.adi.) 45.9 46.5 52.1 47.9 46.6 45.5 44.6 43.8 42.8 41.8 40. Z. Scenario B Primary balance (unadjusted) -0.5 -0.6 0.2 0.4 0.8 1.0 1.0 1.0 1.0 1.0 1.( Primarybalance(realexchangerateadj.) -0.5 -0.6 0.3 0.5 1.0 1.2 1.1 1.2 1.2 1.3 1. Primary balance incl contingent liabilities -0.5 -0.6 0.6 0.8 1.3 1.5 1.4 1.5 1.5 1.5 1. Public Debt to GDP (unadjusted) 45.9 46.5 44.9 40.8 39.4 38.2 37.1 36.0 34.8 33.6 32. Public Debt to GDP (real exchange rate adj.) 45.9 46.5 47.3 42.9 41.3 40.0 38.8 37.6 36.3 34.9 33. Public Debt to GDP (contingent liab.adi.) 45.9 46.5 52.1 47.5 45.6 44.0 42.5 41.1 39.5 38.0 36.4 3. Scenario C Primary balance (unadjusted) -0.5 -0.6 -0.3 0.0 0.7 1.1 1.3 1.6 1.9 2.1 2., Primary balance (real exchange rate adj.) -0.5 -0.6 -0.2 0.2 0.9 1.4 1.6 1.9 2.2 2.4 2.( Primary balance incl. contingent liabilities -0.5 -0.6 0.4 0.7 1.5 1.9 2.1 2.4 2.6 2.8 3. Public Debt to GDP (unadjusted) 45.9 46.5 45.4 41.3 41.4 41.7 42.0 42.4 42.6 42.7 42.7 Public Debt to GDP (real exchange rate adj.) 45.9 46.5 47.8 44.4 44.3 44.4 44.7 44.9 45.0 45.0 44. Public Debt to GDP (contingent liab.adi.) 45.9 46.5 52.6 49.1 48.9 48.8 48.8 48.9 48.8 48.6 48.4 156 ANNEX A FIGURE Al: SCENARIO WITH NO SHOCKS Primary Balance (% of GDP) Public Debt-GDP ratlos (% of GDP) 2.5 50.0 2.0 0 35.0 1.5 30.0 25.0 1.0 2000 5.5.0 10.0 5.0 50I 1= = W m 21 .0 202 03 00 00 20 207 208 209 20020 2003 2004 200 200 2=W 20M 2M9 2010 F*cenr~ A --seen.o B - sad= C -B A ~=2r B ± car C Public debt serviclexports Public debt seice/GDP 45.0.0 a0. 6.0 a 5.0 .0 30.0 25.0 :.0 20.0 3.0 2 5.0 2.0 0.0 .01.0 5..0 0.0' 0.0 = 2M I 23 00 2M02W 20 2O 2010 32=2 00 204 2W 2M 2M 2010 Public debt ser~tax revenues Public Se r Financing Need/GDP 60.0 6.00 50.0 - 5.00 40.04.0 *» . 30.0 3.00 2.00 20.0/ 1.00 / 10.0. 0.01 O~~.000 = 2 21 0.0 2 = 2 M 20= = 2WI 2W ffl 2010 .00304 2 7 200 01 I~S~IA -s~Bl4~ bl s~gg,C -aa A - ~Buo flé sr F"sa bal~neGDP Bmsete Co~ftsion of p~uar surpius 0.04 2= 3 4 20 2M6 20 2 2010 -0.52 0 20025 2003 2W4 2M206 2» 2M2010 -2.0 -é.''2 k ~ ~ é ~ .. -3 j--*S~0i A --S~~an Bl 4 S~w C] 1~ ~~Iy£II0 - lu~ pea~l å F~o M »=. 157 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION FIGURE A2: SCENARIO WITH SHOCKS Primary Balance (RER adjusted) Public Debt (real exch. rate adj.)-GDP (% of GDP) ratios (% of GDP) 3.0 60.0 2.5 A 50.0 2.0 40.0 1.5 .30.0 1.0 20.0 0.5 10.0 -0.522 2003 2004 2005 2006 2007 200 2009 2010 2002 2003 2004 2005 2006 2007 2008 2009 2010 -4.-scenario A - scenario B ; scenaio C -scenario A - scenario B A scenario C Primary Balance (RER adj. and incl. Public Debt (RER adj. and incl. contingent contingent liabilities, % of GDP liabilities, % of GDP 3.5 60 3.0 2.040 30. 1.5 1.0 20- 0.5 10 0.0 0 .. 2002 2003 2004 2005 2006 2007 2008 2009 2010 2 03 2 0 M 20 2 0 20 2010 -*--scenario A -*-scenario a -- scenrio C 1-scenario A -- scenrio a *--scenario C 158 ANNEX A FIGURE A3: PERU DEBT DYNAMICS: COMPARISON OF SCENARIOS WITH AND WITH NO SHOCKS Baseline: Public Debt-GDP Ratios Baseline: Primary balance (percent of GDP) 60 40 1 3011 10 0.5 20002001200220032004200520062007200820092010 -0.5 2 20022003'200420052006200720082009201( -10 -*-Public Debt to GDP(unadjusted) primary balance (unadjusted) - Public Debt to GDP(real exchange rate adj.) * -primary balance (real exchange rate adj.) - Public Debt to GDP (contingent liab.adj.) ---primary balance incl. contingent liabilities Best case: Public Debt-GDP Ratios Best case: Primary balance (percent of GDP) 60 20 2 40 ____________ 20 1 . 10 0.5 0 0.0 2 01 20002001200220032004200520062007200820092010 -0.5 2002200320042005200620072008209201( --4-Public Debt to GDP(unadjusted) -4- pdimary balance (unadjusted) Public Debt to GDP(real exchange rate adj.) - -0 primary balance (real exchange rate adj.) -+-Public Debt to GDP (contingent liab.adj.) - primary balance incl. contingent liabilities Worse case: Public Debt-GDP Worse case: Primary balance Ratios (percent of GDP) 60 50-4 40 30 2 20 10 200020012002200320042002006200120080092010 * 2200320042005200620072008200920 --4-Public Debt to ODP(unadjusted) ---primary balance (unadjusted) -- - - -Public Debt to ODP(realexchange rate adj.) --.* -primary balance (realexchange rate adj.) ---Public Debt to GDP (contingent IHab.adj.) ----primary balance incL contingent liabilities 159 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table Al: Scenario without shocks 2002 2003 2004 2005 2006 2007 2008 2009 2010 Scenario A: Baseline (in million of US$) Proj. Proj. Proj. Proi. Proi. Proj. Proj. Proj. Proj. Nominal GDP 56,816 60,068 63,033 66,800 70,861 75,170 79,740 84,671 89,819 real growth 3.7 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 deflator 0.9 1.9 2.0 2.0 2.0 2.1 2.0 average Interest paid on new debt 10 10 10 10 10 10 10 10 Primary balance 94 225 551 719 773 892 1007 1106 1208 Primary balance as % ofGDP 0.16 0.37 0.87 1.08 1.09 1.19 1.26 1.31 1.35 interest payments 1160 1318 1672 1868 1949 2098 2242 2370 2503 existing debt 1160 1233 1439 1468 1357 1321 1267 1179 1080 domestic 194 200 176 156 54 41 26 17 9 foreign (incL new multi- and bilateral disbursements) 966 1033 1263 1312 1303 1280 1241 1162 1071 newly issued commercial debt 0 85 233 400 592 777 975 1191 1423 interest payments as % of GDP 2.0 2.2 2.7 2.8 2.8 2.8 2.8 2.8 2.8 overall fiscal balance -1066 -1093 -1121 -1148 -1177 -1206 -1235 -1265 -1295 Revenues 11376 11822 12286 12767 13268 13788 14328 14890 15474 Revenues as a % of GDP 20.0 19.7 19.5 19.1 18.7 18.3 18.0 17.6 17.2 Tax revenues (IMF projections for 2002-3) 6761 7448 7740 8044 8359 8687 9027 9381 9749 Expenditures 12443 12916 13406 13916 14445 14993 15563 16155 16768 Expenditures asa % of GDP 21.9 21.5 21.3 20.8 20.4 19.9 19.5 19.1 18.7 overall fiscal balance as % of GDP -1.9 -1.8 -1.8 -1.7 -1.7 -1.6 -1.5 -1.5 -1.4 privatization receipts 600 600 400 47 45 45 40 40 40 privatization as % of GDP 1.1 1.0 0.6 0.1 0.1 0.1 0.1 0.0 0.0 anortization of existing debt 1009 1371 1693 2153 1706 1662 1653 1716 1629 domestic debt 148 316 549 863 305 188 132 128 54 fbreign debt 873 1078 1179 1334 1445 1528 1575 1676 1696 multilateral 410 419 445 515 589 600 601 650 651 on new additional multilateral disbursements 0 0 0 0 0 0 0 29 54 Paris Club 429 592 633 686 723 762 808 815 793 Brady Bonds 34 67 101 133 133 166 166 182 198 PDI 34 67 101 101 101 101 101 101 101 FLIRB 0 0 0 32 32 65 65 81 97 new foreign disbursements 1250 1096 1156 1174 1063 905 809 660 524 Multilateral creditors 1071 740 808 771 713 635 586 514 422 Bilateral creditors 229 356 348 403 350 270 223 146 102 financing need (issuance of new commercial debt) -275 268 1258 2080 1775 1917 2039 2281 2360 domestic 0 foreign (residual) -275 total debt service 2169 2689 3365 4021 3655 3760 3895 4087 4132 debt service/tax revenue 3208 36.10 43.47 49.98 43.73 43.28 43.14 43.56 42.39 total debt service as % of GDP 3.8 4.5 5.3 6.0 5.2 5.0 4.9 4.8 4.6 Implicit average interest rate 5.39 5.26 6.72 7.27 7.27 7.48 7.65 7.75 7.86 domestic 6.20 foreign 5.14 Total debt as % of GDP 44.92 40.88 39.95 39.25 38.70 38.16 37.52 36.86 36.16 Total debt 25524 24557 25183 26222 27421 28685 29922 31208 32478 Stockof new commercialdebt 1155 1698 2956 5036 6811 8728 10767 13048 15408 Total existing debt 24369 22859 22227 21186 20610 19957 19155 18160 17069 Existing stock MLT External Debt of the Non-Financial Public Sector (A+D) 18576 17382 17299 17121 16850 16385 15715 14848 13811 BUY BACK OPERATION Eurobond issue 1430 FLIRB Brady Outstanding principal (nominal) 1212 Discount rate 77 Market Value in percent of nominal 930 Additional financing 500 Debt Servicing 2.17 2.69 3.36 4.02 3.66 3.76 3.89 4.09 4.13 Exports 7.92 8.53 9.33 10.55 11.83 13.11 14.57 15.98 17.40 TDS/Export 0.27 0.32 0.36 0.38 0.31 0.29 0.27 0.26 0.24 TDS/GDP 0.04 0.04 0.05 0.06 0.05 0.05 0.05 0.05 0.05 Financing Need/ GDP 0.00 0.00 0.02 0.03 0.03 0.03 0.03 0.03 0.03 Debt Servicing/ Revenue 19.07 22.74 27.39 31.49 27.55 27.27 27.18 27.44 26.70 160 ANNEX A Table A2: Scenario without shocks 2002 2003 2004 2005 2006 2007 2008 2009 2010 Scenario B (in million of US$) Proj. Proj. Proj. ProJ. Proj. Proj. Proj. Proj. Proj. Nominal GDP 56,816 60,068 63,639 68,091 72,925 78,103 83,648 89,675 96,042 real growth 3.7 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 deflator 0.9 1.9 2.0 2.0 2.0 2.1 2.0 average Interest paid on new debt 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 Primary balance 94 230 535 679 701 793 877 941 1005 Primary balance as % of GDP 0.2 0.4 0.8 1.0 1.0 1.0 1.0 1.0 1.0 interest payments 1160 1291 1589 1721 1730 1804 1867 1905 1939 existing debt 1160 1240 1453 1489 1385 1357 1310 1229 1136 domestic 194 200 176 156 54 41 26 17 9 foreign (incl. new multi- and bilateral disbursements) 966 1040 1277 1333 1331 1316 1284 1212 1127 newly issued comercial debt 0 51 136 232 345 448 557 676 803 interest payments as % ofGDP 2.0 2.1 2.5 2.5 2.4 2.3 2.2 2.1 2.0 overallfiscalbalance -1066 -1061 -1054 -1043 -1029 -1011 -989 -964 -933 Revenues 11376 11934 12519 13132 13775 14450 15158 15901 16680 Revenues as a % of GDP 20.0 19.9 19.7 19.3 18.9 18.5 18.1 17.7 17.4 Taxrevenues(IMFprojectionsfor20023) 6761 7448 7813 8196 8598 9019 9461 9925 10411 Expenditures 12443 12995 13572 14175 14804 . 15461 16148 16865 17614 Expenditures as a % of GDP 21.9 21.6 21.3 20.8 20.3 19.8 19.3 18.8 18.3 overall fiscal balance as % of GDP -1.9 1.8 -1.7 -1.5 -1.4 -1.3 -12 .1.1 -0.97 privatization receipts 600 600 400 47 45 45 40 40 40 privatization as % of GDP 1.1 1.0 0.6 0.1 0.1 0.1 0.0 0.0 0.0 amortization of existing debt 1009 1371 1693 2153 1706 1662 1653 1716 1629 domestic debt 148 316 549 863 305 188 132 128 54 foreign debt 873 1078 1179 1334 1445 1528 1575 1694 1732 multilateral 410 419 445 515 589 600 601 650 651 on new additional multilateral disbursements 0 0 0 0 0 0 0 47 90 Paris Club 429 592 633 686 723 762 808 815 793 Brady Bonds 34 67 101 133 133 166 166 182 198 PDI 34 67 101 101 101 101 101 101 101 FLRB 0 0 0 32 32 65 65 81 97 new foreign disbursments 1250 1396 1456 1474 1363 1205 1109 960 824 Multilateral creditors 1071 1040 1108 1071 1013 935 886 814 722 Bilateral creditors 229 356 348 403 350 270 223 146 102 financing need (issuance of new commercial debt) -275 -64 891 1675 1327 1423 1493 1680 1699 domestic 0 foreign (residual) -275 total debt service 2169 2662 3282 3874 3436 3466 3520 3621 3568 debt service/tax revenue 32.08 35.74 42.00 47.27 39.96 38.43 37.20 36.49 34.27 total debt service as % of GDP 3.8 4.4 5.2 5.7 4.7 4.4 4.2 4.0 3.7 Implicit average interest rate 5.39 5.16 6.41 6.74 6.52 6.54 6.52 6.43 6.35 domestic 6.20 foreign 5.14 Total debt as % of GDP 44.92 40.83 39.42 38.21 37.12 36.03 34.82 3336 32.25 Total debt 25524 24525 25084 26018 27068 28138 29130 30097 30970 Stock of new commercial debt 1155 1366 2257 3932 5258 6681 8175 9855 11553 Totalexisting debt 24369 23159 22827 22086 21810 21457 20955 20242 19417 Existing domestic debt stock (inc. BR) 5793 5477 4928 4065 3760 3572 3440 3312 3258 Existing stock MLT External Debt of the Non-Financial Public Sector (A+D) 18576 17682 17899 18021 18050 17885 17515 16930 16159 BUY BACK OPERATION Eurobond issue 1430 FLIRB Brady Outstanding principal (nominal) 1212 Discount rate 77 Market Value in percent of nominal 930 Additional financing 500 Debt Servicing 2.17 2.66 3.28 3.87 3.44 3.47 3.52 3.62 3.57 Exports 7.92 8.53 9.33 10.55 11.83 .13.11 14.57 15.98 17.4 TDS/Export 0.27 0.31 0.35 0.37 0.29 0.26 0.24 0.23 0.21 TDS/GDP 0.04 0.04 0.05 0.06 0.05 0.04 0.04 0.04 0.04 Financing Need/GDP 0.00 0.00 0.01 0.02 0.02 0.02 0.02 0.02 0.02 Debi Servicing/ Revenue 19.1 22.3 26.2 29.5 24.9 24.0 23.2 22.8 21.4 161 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A3: Scenario without shocks Scenario C (in million of US$) 2002 2003 2004 2005 2006 2007 2008 2009 2010 Proj. Proj. Prol. Proj. Proj. Proj. Prol. Prol. Proj. Nominal GDP 56,816 60,068 62,124 64,887 67,839 70,926 74,153 77,603 81,134 real growth 3.7 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 deflator 0.9 1.9 2.0 2.0 2.0 2.1 2.0 Average Interest paid on new debt 12 12 12 12 12 12 12 12 12 Primary balance -190 -1 441 728 904 1143 1377 1597 1822 Primary balance as % of GDP -0.3 0.0 0.7 1.1 1.3 1.6 1.9 2.1 2.2 interest payments 1160 1373 1839 2151 2351 2614 2874 3120 3371 existing debt 1160 1229 1434 1461 1347 1306 1246 1151 1042 domestic 194 200 176 156 54 41 26 17 9 foreign (incl. new multi- and bilateral disbursements) 966 1029 1258 1305 1293 1265 1220 1134 1033 newly issued commercial debt 0 144 406 689 1004 1308 1628 1969 2329 interest payments as % of GDP 2.0 2.3 3.0 3.3 3.5 3.7 3.9 4.0 4.2 overall fiscal balance -1350 -1374 -1398 -1422 -1447 -1472 -1497 -1523 -1549 Revenues 11376 11934 12519 13132 13775 14450 15158 15901 16680 Revenues as % ofGDP 20.0 19.9 19.7 19.3 18.9 18.5 18.1 17.7 17.4 Tax revenues (IMF projections 2002-3) 6761 7448 7813 8196 8598 9019 9461 9925 10411 Expenditures 12443 12995 13572 14175 14804 15461 16148 16865 17614 Expenditures as % ofGDP 21.9 . 21.6 21.3 20.8 20.3 19.8 19.3 18.8 18.3 overall fiscal balance as % of GDP -2.4 -2.3 -2.3 -2.2 -2.1 -2.1 -2.0 -2.0 -1.9 privatization receipts 600 600 400 47 45 45 40 40 40 privatization as % ofGDP 1.1 1.0 0.6 0.1 0.1 0.1 0.1 0.1 0.0 amortization of existing debt 1009 1371 1693 2153 1706 1662 1653 1716 1629 domestic debt 148 316 549 863 305 188 132 128 54 foreign debt 873 1078 1179 1334 1445 1528 1575 1666 1684 multilateral 410 419 445 515 589 600 601 650 651 on new additional multilateral disbursements 0 0 0 0 0 0 0 19 41 Paris Club 429 592 633 686 723 762 808 815 793 Brady Bonds 34 67 101 133 133 166 166 182 198 PDI 34 67 101 101 101 101 101 101 101 FLRS 0 0 0 32 32 65 65 81 97 new foreign disbursments 1250 674 734 752 641 483 387 238 102 Multilateral creditors 1071 318 386 349 291 213 164 92 0 Bilateral creditors 229 356 348 403 350 270 223 146 102 financing need (issuance of new commercial debt) 9 971 1957 2776 2467 2606 2723 2961 3036 domestic 0 foreign (residual) 9 total debt service 2169 2744 3532 4304 4057 4276 4527 4836 5000 debt service/tax revenue 32.08 36.83 46.29 55.05 50.65 52.11 53.85 56.15 56.67 total debt service as % of GDP 3.8 4.6 5.7 6.6 6.0 6.0 6.1 6.2 6.2 Implicit average Interest rate 5.39 5.42 7.27 8.15 8.46 8.93 9.33 9.65 9.95 domestic 6.20 foreign 5.14 Total debt as % of GDP 45.42 41.35 41.43 41.69 42.04 42.37 42.55 42.66 42.70 Total debt 25808 24838 25741 27054 28522 30053 31552 33107 34643 Stock of new commercial debt 1439 2401 4358 7134 9601 12206 14929 17890 20926 Total existing debt 24369 22437 21383 19920 18922 17847 16623 15217 13717 Existing domestic debt stock (Incl. BR) 5793 5477 4928 4065 3760 3572 3440 3312 3258 Existing stock MLT External Debt of the Non-Financial Public Sector (A+D) 18576 16960 16455 15855 15162 14275 13183 11905 10459 BUY BACK OPERATION Eurobond issue 1430 FLIRB Brady Outstanding principal (nominal) 1212 Discount rate 77 Market Value in percent of nominal 930 Additional financing 500 Debt Servicing 2.17 2.74 3.53 4.30 4.06 4.28 4.53 4.83 5.00 Exports 7.92 8.53 9.33 10.55 11.83 13.11 14.57 15.98 17.40 TDS/Export 0.27 0.32 0.38 0.41 0.34 0.33 0.31 0.30 0.29 TDS/GDP 0.04 0.05 0.06 0.07 0.06 0.06 0.06 0.06 0.06 Financing Need/ GDP 0.00 0.02 0.03 0.04 0.04 0.04 0.04 0.04 0.04 Debt Servicing/ Revenue 19.07 23.54 29.58 35.18 32.37 33.30 34.41 35.89 36.21 162 ANNEX A Table A4: Scenarios with shocks (10% depreciation in real exchange rate and adjustment for contingent liabilities) Scenario A: Baseline (in million of Icu) 2002 2003 2004 2005 2006 2007 2008 2009 2010 Pmj. Pro Pro. Proj. Proj. Proj. Proj. Proj. Proi. Icu$ 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 Nominal GDP 203,280 213,314 223,843 237,220 251,643 266,943 283,173 300,684 318,966 real growth 3.7 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 Deflator 0.9 0.9 0.9 1.9 2.0 2.0 2.0 2.1 2.0 average Interest paid on new debt 10 10 10 10 10 10 10 10 10 primary balance incLeontliab.% of GDP 0.8 1.0 1.5 1.7 1.7 1.7 1.8 1.8 1.8 primary balance as % ofGDP 0.28 0.51 1.04 1.25 1.26 1.36 1.44 1.48 1.52 Interest payments including cont.iab 5362.56 5957.89 7292.80 8032.01 8340.58 8901.37 9444.04 9928.1410429.15 newly issued commercial debt 0 320 878 1508 2235 2931 3678 4493 5368 interest payments as % of GDP incLcontliab. 2.6 2.8 3.3 3.4 3.3 3.3 3.3 3.3 3.3 overall fiscal balance as % of GDP -1.9 -1.8 -1.8 -1.7 -1.7 -1.6 -1.5 -1.5 -1.4 privatization receipts 2264 2264 1509 177 170 170 151 151 151 amortization of existing debt 3807 5173 6387 8123 6437 6270 6236 6476 6148 domestic debt 558 1192 2071 3256 1151 709 498 483 204 foreign debt 3292 4068 4447 5034 5453 5764 5941 6325 6401 multilateral 1547 1581 1679 1943 2222 2264 2268 2452 2456 on new additional multilateral disbursements 0 0 0 0 0 0 0 111 205 Paris Club 1619 2234 2388 2588 2728 2875 3049 3075 2992 Brady Bonds 127 253 380 503 503 625 625 686 748 PDI 127 253 380 380 380 380 380 380 380 FURB 0. 0 0 123 123 245 245 306 368 new foreign disbursements 4716 4135 4362 4430 4011 3415 3052 2490 1977 Multilateral creditors 4041 2792 3049 2909 2690 2396 2211 1939 1592 Bilateral creditors 864 1343 1313 1521 1321 1019 841 551 385 financing need (issuance of new commercial debt) -1244 769 4496 7595 6435 6967 7418 8326 8617 Domestic 0 foreign (residual) -1244 total debt service 8184 10145 12694 15169 13792 14186 14695 15418 15591 totaldebtserviceas%ofGDP 4.0 4.8 5.7 6.4 5.5 5.3 5.2 5.1 4.9 Implicit average Interest rate 5.39 5.28 6.75 7.31 7.33 7.56 7.75 7.87 7.99 domestic 6.20 foreign 5.14 Total debt as % of GDP (rer adjusted) 47.27 43.32 42.23 41.39 40.71 40.07 3932 38.55 37.75 Total debt as % of GDP (rer adjusted+contingent liab.) 52.12 47.94 46.63 45.5 44.63 43.76 42.80 41.83 40.84 Total debt 96095 92411 94524 98190 102451 106956 111349 115919 120422 Stock of new commercial debt 4151 6165 10661 18256 24691 31658 39076 47402 56019 Total existing debt 91944 86246 83863 79934 77760 75298 72273 68517 64403 Existing domestic debt stock (incl. BR) 21857 20665 18593 15337 14186 13477 12979 12496 12292 Existing contingent liab.(year 1998) 9859 9859 9859 9859 9859 9859 9859 9859 9859 Existing stock MLT External Debt of the Non-Financial Public Sector (A+D) 70087 65582 65270 64597 63574 61820 59294 56021 52111 163 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table AS: Scenarios with shocks (10% depreciation in real exchange rate and adjustment for contingent liabilities) Scenario B: (in million of Icu) 2002 2003 2004 2005 2006 2007 2008 2009 2010 Proi. Proj. Pr. Proj. Prol. Pro. Pro. Proj. Proj. IcU$ 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 Nominal GDP 203,280 215,365 228,168 244,129 261,462 280,026 299,908 321,516 344,344 real growth 3.7 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 deflator 0.9 0.9 0.9 1.9 2.0 2.0 2.0 2.1 2.0 average Interest paid on new debt 7.5 7.5 7. 7.5 7.5 7.5 7.5 7.5 7.5 primary balance% of GDP incl.contliab. 0.6 0.8 13 1.5 1.4 1.5 1.5 1.5 1.5 primary balance as % of GDP 0.28 0.50 0.98 1.15 1.12 1.19 1.24 1.26 1.28 interest payments incLcont.liab. 5116.09 5616.71 6757.11 7285.79 7361.52 7699.34 8008.82 8243.33 8480.38 interest payments as % of GDP 2.2 2.3 2.6 2.7 2.5 2.5 2.4 2.3 2.2 interest payments as % of GDP incl.contliab. 2.5 2.6 3.0 3.0 2.8 2.7 2.7 2.6 2.5 overall fiscal balance as % of GDP .1.9 -1.8 -1.7 -1.5 -1.4 -1.3 -1.2 -1.1 -1.0 privatization receipts 2264 2264 1509 177 170 170 151 151 151 amortization of existing debt 3807 5173 6387 8123 6437 6270 6236 6476 6148 domestic debt 558 1192 2071 3256 1151 709 498 483 204 foreign debt 3292 4068 4447 5034 5453 5764 5941 6391 6534 multilateral 1547 1581 1679 1943 2222 2264 2268 2452 2456 on new additional multilateral disbursements 0 0 0 0 0 0 0 178 338 Paris Club 1619 2234 2388 2588 2728 2875 3049 3075 2992 Brady Bonds 127 253 380 503 503 625 625 686 748 PDI 127 253 380 380 380 380 380 380 380 FLlRB 0 0 0 123 123 245 245 306 368 new foreign disbursements 4716 5267 5493 5561 5143 4546 4184 3622 3109 Multilateral creditors 4041 3924 4180 4041 3822 3528 3343 3071 2724 Bilateral creditors 864 1343 1313 1521 1321 1019 841 551 385 financing need (issuance of new commercial debt) -1244 -439 3162 6123 4813 5179 5449 6158 6234 domestic 0 foreign (residual) -1244 total debt service 8184 10050 12405 14670 13059 13230 13506 13980 13889 total debt service as % of GDP 4.0 4.7 5.4 6.0 5.0 4.7 4.5 4.3 4.0 implicit average interest rate 539 5.18 6.45 6.83 6.66 6.74 6.80 6.79 6.80 domestic 6.20 foreign 5.14 Total debt as % of GDP 47.27 42.87 41.30 39.97 38.76 37.56 36.26 34.90 33.49 Total debt as % of GDP (rer adjusted+contingent liab.) 52.12 47.45 45.63 44.01 42.53 41.09 39.55 37.97 36.36 Total debt 96095 92334 94245 97572 101342 105190 108746 112214 115332 Stock of new commercial debt 4151 4956 8118 14242 19054 24233 29682 35840 42074 Total existing debt 91944 87378 86127 83330 82288 80957 79064 76374 73259 Existing domestic debt stock (Incl. BR) 21857 20665 18593 15337 14186 13477 12979 12496 12292 Existing contingent liab.(year 1998) 9859 9859 9859 9859 9859 9859 9859 9859 9859 Existing stock MLT External Debt of the Non-Financial Public Sector (A+D) 70087 66714 67533 67993 68101 67480 66085 63878 60966 164 ANNEX A Table A6: Scenarios with shocks (10% depreciation in real exchange rate and adjustment for contingent liabilities) Scenario C: (in million of cu) 2002 2003 2004 2005 2006 2007 2008 2009 2010 Proj. Proj. Proj. Proj Proj. Proj. Proj. Proj. Proj. IcU$ 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 3.77 Nominal GDP 203,280 210,237 217,433 227,103 237,436 248,239 259,534 271,609 283,967 real growth 3.7 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 deflator 0.9 0.9 0.9 1.9 2.0 2.0 2.0 2.1 2.0 average interest paid on new debt 12 12 12 12 12 12 12 12 12 primary balance % of GDP incl. Cont.1iab. 0.4 0.7 1.5 1.9 2.1 2.4 2.6 2.8 3.0 primary balance as % ofGDP -0.22 0.18 0.94 1.38 1.60 1.90 2.16 2.37 2.57 interest payments incl. Cont. liab. 5559.74 6361.91 8122.52 9296.92 10053.35 11047.14 12026.63 12953.96 13900.42 interest payments as % of GDP incLconLiab. 2.7 3.0 3.7 4.1 4.2 4.5 4.6 4.8 4.9 overall fiscal balance as % of GDP -2.4 -23 -2.3 .2.2 -2.1 .2.1 -2.0 -2.0 -1.9 privatization receipts 2264 2264 1509 177 170 170 151 151 151 amortization of existing debt 3807 5173 6387 8123 6437 6270 6236 6476 6148 domestic debt 558 1192 2071 3256 1151 709 498 483 204 foreign debt 3292 4068. 4447 5034 5453 5764 5941 6284 6352 multilateral 1547 1581 1679 1943 2222 2264 2268 2452 2456 on new additional multilateral disbursements 0 0 0 0 0 0 0 71 156 Paris Club 1619 2234 2388 2588 2728 2875 3049 3075 2992 Brady Bonds . 127 253 380 503 503 625 625 686 748 PDI 127 253 380 380 380 380 380 380 380 FL.RB 0 0 0 123 123 245 245 306 368 new foreign disbursements 4716 2543 2769 2837 2418 1822 1460 898 385 Multilateral creditors 4041 1200 1456 1317 1098 804 619 347 0 Bilateral creditors 864 1343 1313 1521 1321 1019 841 551 385 financing need (issuance of new commercial debt) -228 3288 7001 10086 8912 9429 9865 10757 11032 domestic 0 foreign (residual) -228 total debt service 8184 10351 13327 16237 15307 16134 17080 18247 18865 total debt service as % ofGDP 4.0 4.9 6.1 7.1 6.4 6.5 6.6 6.7 6.6 implicit average Interest rate 5.39 5.44 732 8.23 8.57 9.07 9.51 9.86 10.19 domestic 6.20 foreign 5.14 Total debt as % of GDP 47.77 44.40 4432 44.44 44.67 44.90 44.96 44.97 44.91 Total debt as % of GDP (rer adjusted+contingent Ilab.) 52.62 49.09 48.85 48.78 48.83 48.87 48.76 48.60 48.38 Total debt 97112 93338 96363 100929 106075 111449 116697 122146 127520 Stock of new commercial debt 5167 8684 15685 25771 34683 44112 53977 64734 75766 Total existing debt 91944 84654 80678 75158 71391 67337 62720 57412 51754 Exsting domestic debt stock (Incl. BR) 21857 20665 18593 15337 14186 13477 12979 12496 12292 Existing contingent liab.(year 1998) 9859 9859 9859 9859 9859 9859 9859 9859 9859 Existing stock MLT External Debt of the Non-FinancIal Public Sector (A+D) 70087 63989 62085 59820 57205 53859 49740 44916 39462 165  ANNEX B THE FISCAL EFFORT REQUIRED FOR A SUSTAINED STRUCTURAL DEFICIT' We estimate the required fiscal adjustment that is consistent with the path for more sustainable debt-to-GDP ratios calculated in Annex A. Fears of excessive fiscal deficits stem from the likelihood that they may be unsustainable in the long run. The fiscal effort in question is calculated in terms of the path that the structural deficit (as opposed to the actual deficit that incorporates a cyclically adjusted component) must follow to be consistent with debt and debt- service sustainability. So, first, a structural primary balance is calculated, consistent with maintaining and even reducing the overall NFPS debt as a ratio of GDP, both in terms of principal and servicing requirements. Second, we estimate the fiscal adjustment required to achieve sustainability on the assumption that the cyclical component of the primary balance will approach zero over the 10-year forecast period. The fiscal effort is estimated under three alternative scenarios and its sensitivity analyzed according to alternative interest rate, financing, and output growth assumptions. The underlying assumptions, target debt-to-GDP and primary- surplus-to-GDP ratios, as well as the implicit average interest rates on total debt follow from the analysis in Annex A. Rationale for Using the Structural Balance and Methodology to Calculate It It has long been recognized2 that the nominal budget balance is an imperfect indicator of the government's true fiscal stance. The problem is that, in general, the fiscal balance depends not only on the authorities' tax and expenditure decisions but also on the extent to which autonomous spending decisions by the private sector reflect themselves on the country's output and national income levels and thereby on tax collection. Thus, if the economy enters a recession, because, say, a reduction in private investment, tax revenue automatically declines, irrespective of the authorities' fiscal stance, and the fiscal balance worsens. Conversely, if an autonomous increase in private spending leads to and 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, as its stands, the nominal fiscal balance is a distorted measure of whether government fiscal policy is 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. To address the issue of the need to have a measure of fiscal policy that is 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 literature3. The basic idea is to carry out a simple decomposition of the nominal budget balance (B) into two unobservable components, namely, 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) (B.1) 1 This Annex was also prepared with technical assistance from Rashmi Shankar. 2 See, for instance, Dornbusch and Fischer (1990). 3 See, for instance, Brown (1956), de Leeuw et al. (1980), and Hagemann (1999). For a criticism of this approach, see Fellner (1982). 167 PERU: RESTORING FIScAL DISCIPLINE FOR POVERTY REDUCTION Under the admittedly inaccurate, simplifying assumption that government spending is not a function of the state of the business cycle, the structural component is defined by: SB(t) = SR(t) - G(t) (B.2) Where SR(t), the structural revenue, is the fiscal revenue that would be collected if actual output in t were the level of output determined by the long-run trend, which will be called trend output. G(t) is total government spending. SR is estimated by means of the following equation: SR(t) = R(t)[TY(t)/Y(t)] (B.3) Where R(t) is the observed value for the government revenue; TY(t) is the economy's trend output, Y(t) the actual level of output, and 8 the income elasticity of total government revenue. Completion of the calculations implied by equations (B2) and (B3) requires that estimates for both trend output and the income elasticity of revenue be generated. These two issues are addressed in the forthcoming paragraphs. The revenue elasticity of.the cyclical component over the 1970-2001 period was estimated. It was found to be 1.144. For the purpose of estimation trend output, the estimation method proposed by Hodrick and Prescott (1997) is used. In terms of this method, if the logarithm of Y(t) is denoted y(t), the time series for y(t) can be decomposed in its trend component ty(t) and its cyclical component cy(t): y(t) = ty(t) + cy(t) ; for t = 1, ...,T (B.4) then the Hodrick-Prescott (HP) trend output, hpy (t), is the series that minimizes the expression: E Cy(t)2 + X { [ty(t) - ty(t-1)] - [(ty(t-1) - ty(t-2)]}2 (B.5) Where the sums are carried out from t = 1 to t = T and X is a parameter that determines how smooth the trend line will be. The method's idea is to minimize the sum of two terms where the first term is the sum of squares of the cyclical component and the second term is the sum of squares of the trend component's second differences. While the first term in (B.5) penalizes large residuals (i.e. poor fit)5, the second term penalizes lack of smoothness in the trend. If X is zero, the trend will simply equal the original series for all t; but, if X is very large, changes in the slope of the trend are avoided, and, in the limit, the trend will simply be a straight line. For quarterly data there is a consensus among practitioners around the value of X = 1600, originally proposed by Hodrick and Prescott (1980)6. Although there is no consensus, however, for data of other frequencies and originally Dolado et al. (1993) used X = 400, Backus and Kehoe (1992), Giorno et al. (1995), and the European Central Bank (2000) use )= 100 for the structural deficit of the European Union countries. Econometric software like Eviews also uses the latter value. Hence, a value of X = 100 is assumed. 4 This is different from the revenue elasticity of 0.98 assumed for the main debt sustainability analysis for a shorter period, since what is considered here is the elasticity of cyclical tax revenues with respect to cyclical output. The cy(t) values are deviations from trend and the conceptual framework implies that, for a good fit to exist, their average over long time periods must be close to zero. 6This part of the discussion relies heavily on Maravall and del Rfo (2001) 168 ANNEX B Figure 2.12 (see Chapter II) above shows actual output and estimated structural or cyclically adjusted output for Peru in the period 1990-2002. Figure 2.13 shows actual and estimated fiscal balance. Calculation of the structural balance allows us to carry out the next steps in the simulation exercise, namely: 1. Estimation of the required primary surplus consistent with the target debt-GDP ratio in period "t+n" where n is the number of forecast periods. This requires a forecast of future real interest and growth rates. 2. Estimation of the structural primary surplus in "t+n ", whose present value must equal the present value of the primary surpluses over the n periods in the forecast horizon. 3. Estimation of the fiscal effort required: Here a time horizon of 9 years is assumed since the purpose of the exercise is to capture the transition dynamics in terms of the permanent, or long-run fiscal adjustment required. The issue is then explored of how the debt-GDP ratio would evolve under alternative assumptions on interest rates, growth, and financing. The purpose here is to capture the impact of changes in the assumptions regarding the values taken by key parameters on the target permanent primary surplus that is consistent with sustainable debt and debt servicing. Scenarios and assumptions. Following the above indicated methodological steps, three simulation scenarios for the period 2002-2010 are run. The assumptions defining each scenario are similar to those in Annex A and are summarized in what it follows7. The baseline scenario assumes a moderate growth rate of 4.0 percent that is slightly higher than the average growth rate of the Peru's economy in the nineties (3.6 percent). This scenario incorporates the preliminary figures for the Brady buy-back under which all outstanding Front- Loaded Interest-Reduced Bonds have been exchanged. It is assumed that maintain moderate levels. The average interest rate on new debt is fixed at 10% throughout the period. It is also assumed that there will be reduced balance-of-payment support from the multilaterals (being phased out in 2003) but that this reduction will be partly offset by sector-investment. There is limited resort to commercial borrowing. The financing mix is such that the implicit average interest rate on the total stock of NFPS debt grows gradually from 5.4% and 7.9% over the period 2002-2010. A second, optimistic best-case scenario assumes a higher average growth rate of GDP (5.0 percent). The interest rate on new debt will be 7.5% over the simulation period. The multilaterals are assumed to provide full balance of payments support. The mix of multilateral financing and (limited) commercial borrowing is such that the implicit average interest rate on total NFPS debt stays at lower levels than in the other two scenarios, growing slowly from 5.4% to 6.1% over 2002-2010. 7 There is a set of assumptions common to all scenarios. They are as follows: (i) by 2010, automatic stabilizers will ensure that the cyclical component of the primary balance is close to zero. (ii) The permanent primary balance is - 0.43% in 2001. (iii) Receipts from privatization will be US$600 million for 2002 and 2003, and will fall to less than $50 million per year thereafter. (iv) Inflation in the GDP deflator (US$ terms) will increase from 0.9% in 2002 to approximately 2% over the period 2006-2010. (v) Continued rollover of short-term debt over the entire period 2002- 2010. (vi) The permanent primary surplus grows "smoothly" to the required level. (vii) The non-financial public sector elasticities of actual unadjusted revenues and expenditures with respect to GDP remain constant at 0.98 and 0.95 respectively. Elasticities are estimated over the 1994-2002 period. 169 PERU: RESTORING FiscAL DISCIPLINE FOR POVERTY REDUCTION Finally, a pessimistic worst-case scenario assumes both a low average growth rate (2.5 percent) and an even higher implicit interest rate on debt. The latter is the result of: (i) reduced multi- lateral financing (no fast-disbursing loans and reduced new commitments from 2003 onwards), and (ii) the interest rate on new commercial debt is assumed to be 12.0% and the country has to rely more on this type of debt. Results.For the baseline scenario a substantial fiscal effort is required. The primary balance will have to be steadily increased from the 2001 deficit of -0.4 percent of GDP to a surplus of 1.2 percent of GDP by 2010, for a net fiscal correction of 1.6 percent of GDP (Table B.1). In the optimistic scenario a lighter but still considerable fiscal effort in its own right will be required. The goal would be to reach a surplus of 0.3 percent of GDP by 2009-2010, for a net fiscal correction of 0.7 percent of GDP. This is consistent with higher growth and fiscal discipline. Not surprisingly, the pessimistic scenario would entail an extremely demanding fiscal effort. The (structural) primary balance would have to be reduced to zero immediately. With no delay, increasingly high primary surpluses would have to be attained and the net fiscal correction in the nine-year period through 2010 would amount to 3.4 percent of GDP. Table B.1: Required Fiscal Effort for Debt Sustainability (diagonal elements corresponding to main Est. Projections DSA) (as a percentage of GDP) 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Primary surplus -0.6 Permanent primary surplus -0.4 Cyclical component of primary surplus -0.3 Scenario A: Baseline Interest rate (implicit average on NFPS debt) 6.6 5.4 5.3 6.7 7.3 7.3 7.5 7.7 7.8 7.9 Rate of growth of potential output (annual, %) 3.7 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 4.0 Target Debt stock (% of GDP) 36.2 Target permanent primary surplus (% of GDP) -0.4 -0.2 0.0 0.2 0.4 0.5 0.7 0.9 1.1 1.2 Fiscal Correction (% of GDP) 1.6 Scenario B: Best Case Interest rate (implicit average on NFPS debt) 6.6 5.4 5.1 6.3 6.6 6.4 6.4 6.3 6.2 6.1 Rate of growth of potential output (annual, %) 3.7 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 5.0 Target Debt stock (% of GDP) 1 34.0 Target permanent primary surplus (% of GDP) -0.4 -0.3 -0.2 -0.1 0.0 0.0 0.1 0.2 0.3 0.3 Fiscal Correction (% of GDP) 0.7 Scenario C: Worst Case Interest rate (implicit average on NFPS debt) 6.6 5.4 5.4 7.3 8.2 8.5 9.0 9.4 9.7 10.1 Rate of growth of potential output (annual, %) 3.7 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 Target Debt stock (% of GDP) 42.7 Target permanent primary surplus (% of GDP) -0.4 0.0 0.4 0.7 1.1 1.4 1.8 2.2 2.5 3.0 Fiscal Correction (% of GDP) 3.3 170 ANNEX C QUESTIONNAIRE ON PUBLIC FINANCIAL MANAGEMENT' AGGREGATE FISCAL DISCIPLINE A. Budget preparation and approval la. Are there formal constraints (constitutional or legislatively mandated) on aggregate spending and/or deficits? Yes. The objective of the Law of Fiscal Prudence and Transparency is to establish the guidelines for a better management of the public finances in order to contribute in the economic stability, essential condition to obtain rates of a sustained economical growth and social welfare. At the same time, the Law establishes macro fiscal rules for the General Government, the Consolidate Public Sector, and it creates the Fiscal Stabilization Fund. Besides discipline on the expenditure side, the support of the fiscal equilibrium needs to generate fiscal savings in booms in order to face negative internal or external situations. It is.necessary to establish fiscal goals for the medium-term. In this way, the law contributes to the formation of expectations for the economic agents, in a period of three years. For that reason, the Law establishes ceilings to the levels of expenditures and debt. lb. Is the Government required to publish actual figures relative to these constraints? Yes. MEF is in charge of the elaboration of the Multi annual Macro-Economic Framework (MMM) and the evaluation of its execution and publication. The MMM includes the Declaration of Principles of the Fiscal Policy, the goals of fiscal policy, in the next three years, and the targets for macro economic indicators in the next three years. It also contains the projection of fiscal revenue and expenditure, the sum of public investment and the level of public debt. 1c Are these constraints imposed and monitored by donors? No, they are not, but all economic agents observe the fiscal performance. 2a Are there formal constraints (constitutional or legislatively mandated) on public debt and domestic/external borrowing by (i) central government (ii) sub national governments; and (iii) public enterprises? In the Law, it is established that the public debt in the medium term should be consistent with the principle of fiscal equilibrium. It is necessary to establish limits to the levels of expenditures and debt. There are not formal restrictions in the inner of the components of public debt but limits from the Annual Law of Public Debt. 2b. Is this monitored by the Central Bank? Yes. Debt ceilings are established in the projections of the MMM and are coordinated with the General Direction of Public Credit at the Ministry of Economy and Finances(MEF). 2c. Are these constraints imposed and monitored by donors? No, but all-economic agents observe fiscal performance. 2d. Is the government required to publish actual figures relative to these borrowing constraints? Yes. They are in the MMM 3a. Is there a medium-term expenditure framework which projects and aggregate expenditure ceiling over a three to five year horizon, consistent with the macroeconomic targets? The MMM has a horizon of three years and the Sector Strategy Plans (SEPs) a horizon of 5 years. 'Nelson Shack, Budget General Director, and Oscar Pajuelo, Accountant General of Peru responded this questionnaire. 171 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 3b.- Is this published? Yes, the MMM and the SEPs are published in " El Diario Oficial " and in the MEF's "Portal de Transparencia" Website. 3c. What is the percent difference between the aggregate spending in the projection and the annual budget? Very small, because the annual budget is elaborated as a function of the levels of expenses and incomes, consistent with the revised fiscal projections that the MMM realizes every year. 4a. What is the percent deviation between the aggregate spending in the budget as proposed by the central agencies (i.e., Minister of Finance in the Budget Call Circular) and that approved by cabinet at the end of budget discussions? No response. 4b. What is the percent deviation between aggregate spending proposed by the cabinet and the legislature? There is no deviation. Congress cannot alter the level of aggregate expenditure proposed by the Executive. B. Budget execution and monitoring la. Are there formal rules that guard against overspending by agencies relative to budgeted amounts (e. g., central agencies, chief accountants or banks having the authority to refuse expenditures if there are insufficient funds in the ministerial account?) Yes. No agency can spend more than the budget authorized by the National Office of the Public Budget (DNPP). Moreover, the SIAF contributes to respect those rules. lb. Is there a published reconciliation of actual expenditures versus budgeted amounts? Yes. Monthly Treasury report , the quarterly and annual one from the DNPP and the annual one from the DGAES (General Office of Economical and Social Affairs). All are available in the web site. Ic. Is there punitive action taken against overspending agencies? Yes, but it depends on a case - by - case basis. There are administrative, civil and penal responsibilities. The system of authorization in the SIAF allows a permanent and continuing evaluation of the authorizations of commitments and payments of checks. In this way, there is an efficient control of expenditure. 2a. Is there a formal or informal requirement to report on aggregate.-fiscal outcomes relative to targets? Yes. The report about the MMM's execution must be published semiannually. MEF goes to Congress annually to inform about the execution of the budget. On the other hand, it is legally consistent with the revised fiscal projections that the MMM realizes every year. 2b. Are these published? Yes, in the website. 2c. If so, with what lags? The access of information in the SLAF is virtual and has information aggregated up to the last month. 2d. What is the percent deviation between the aggregate spending in the annual budget and the total amount actually spent at the end of the fiscal year? No response. A. EXPENDITURE PRIORIZATION AND ALLOCATIVE EFFICIENCY Budget preparation and approval breadth of consultations la.- What percent of expenditures are allocated by the central government (as opposed to sub national governments)? Approximately 90%. lb. Which of these activities do sub national governments have constitutional responsibility for in allocating their budgetary expenditures: (i) primary education; (ii) secondary education; (iii) university education, (iv) hospitals, (v)health clinics ? Check only those which apply. There exist the 172 ANNEX C Transitory Councils from the Regional Administration (CTARs) in which one can find the Regional Offices of Health and Education. The local Government, at the provincial or district level, does not have the responsibility to provide education and health services. 2a. Are there explicit pre-budget consultations about budgetary priorities between government and the following groups in the private sector: (i) business community;(ii) public interest groups (e.g. NGOs), (iii) labor unions; (iv) farmers associations? Check only those which apply. The Budget is formulated internally in the public sector. There is not a framework that involves the participation of other sectors of the society yet. 2b. How large a change vis-h-vis existing priorities in the current budget have emerged from such consultations: negligible, modest or large? It is not applicable but incoming pilot participatory budgeting will address investment expenditure. 2c. Are there post-budget consultations with the same group, which attempt to reconcile pre-budget understandings with actual allocations? No, there are not. 3a. At the start of budget preparation, is there a session in the legislature about budget priorities? No. 3b. How large a change vis-h-vis existing priorities in the current budget have eierged from such a session: negligible, modest or large? It is not applicable 4a. Rank the following in terms of their relative influence of the following in deciding upon broad priorities for the composition of expenditures: (i) Ministry of Finance/Planning; (ii) the Cabinet,(iii) the Legislature; (iv) Donors, (v) private sector-government consultation committees. That ranking is ok. 4b. What is the average percent deviation in the allocation for the major sectors and programs: (i) between the budget as proposed by the central ministries and that by the cabinet; and (ii) between the budget ad proposed by cabinet and that approved by the legislature? Range: negligible (0-10%), modest (10-30%), high (more than 30%). Negligible (0-10%) 5a. Does the government publish expenditure priorities corresponding to the following levels of disagreegation: (I) sector expenditures; (ii) programs; (iii) projects? Check only those that apply. Yes, it does, but at the level of sectors and programs. These priorites are included in the MMM. 5b. If so, are these expressed in terms of outcomes (i.e. impact on beneficiaries e.g., infant mortality) or outputs (i.e., goods' and services produced e.g., number of health clinics or immunizations provided)? No, they are not. 5c. Are actual achievements of sector expenditures published? Yes, they are in the website. 5d. If so, is there a public or published reconciliation with the targets? Yes, there is. 6a. What percentage of public spending is financed by donors? It is less than 2% of the budget. 6b. Is there a prior agreement among donors about the composition of expenditures that are being collectively financed? No. 6c. if so, is this agreement induced by the leadership of a central donor? It is not applicable. 173 PERU: RESIORING FISCAL DISCIPLINE FOR POVERTY REDUCTION B. BUDGET PREPARATION AND APPROVAL ALLOCATION RULES AND CRITERIA la. Are expenditure allocations across ministries and programs increased or decreased in the same proportion across-the-board? The variations, which could happen in the different assignments of expenses, are not proportional. They depend on the level of the expenditure priority. lb. Are there formulae or rules which earmark funds for specific expenditures? What proportion of total expenditures do they constitute? No. 2a. Is there a formal or informal rulel which requires an explicit consideration of whether individual programs or projects that are to be funded by the budget can be undertaken by the private sector? No. 2b. For which sectors is this done? For what percentage of programs/projects is this actually done (100%, 50-99%, 20-49%, less than 20%)? It is not applicable. 3a. Is there a requirement to conduct and ex ante quantitative analysis of costs and benefits before a new program/project is initiated? Yes, in the case of projects, the Law of National System of Public Inversion requires such analysis. 3b. For which sectors is this done? Indicate the percentage of programs/projectsfor which this is actually done (100%, 50-99%, 20-49%, less than 20%)? Since 2001, is is 100%. 4a. Is the distributional impact of public spending explicitly quantified and considered in allocating resources among programs and projects? No. 4b. For which sectors is this done? Indicate the percentage of programs/projects for which this is actually done (100%, 50-99%, 20-49%, less than 20%)? Not applicable. C. BUDGETING PREPARATION AND APPROVAL NORMS la. Is there a system of forward estimates which projects the future cost implications of existing and proposed porgrams and projects? No. lb. Are these automatically roiled over into the next budget, adjusted only for key national parameters such as inflation rate? Not applicable. Ic. Are these forward estimates published? Not applicable. Id. Does the government publish a reconciliationstatement explaining any significant deviations in the composition of expenditures between the original forward estimates and the annual budget? Not applicable. 2a. Are line agencies required to identify cuts in their existing programs to match new spending proposals? No. 2b. Are various new spending proposals and offsetting cuts discussed systematically at a Cabinet or sub-Cabionet level? Yes. D. BUDGETING PREPARATION AND APPROVAL CAPITAL/ RECURRENT BUDGETING la. Are there separate budgets for capital and recurrent expenditures? Yes,there are. lb. Is there a requirement to estimate the recurrent cost implications of new capital investments? Yes, there is. 174 ANNEX C Ic Are there different ministries responsible for preparing capital budgets (e.g., Mistry of Planning) and recurrent budgets (e.g., Ministry of Finance)? The investment office (ODI) at MEF does it. Id. What percent of public investments is donor financed? Non available. E. BUDGETING PREPARATION AND APPROVAL DONOR RULES Ia. Is there a donor conditionality on the overall composition of expenditures? No. lb. Has expenditure composition been changed in accordance with this conditionality? Not applicable. Ic. What percent of donor financed expenditures are earmarked for particual programs and projects? Non available F. BUDGETING EXECUTION AND MONITORING la. What is the average percent deviation between the composition of expenditures as approved in the annual budget and the actual allocation at the end of the budget year? Non available. lb. On what basis was the compositon changed: (I) arbitrary/ad hoc; (ii) related to specific problems? It was changed in relation of specific problems such as internal shocks of natural character (El Niflo phenomenon), political,and external shocks (financial crisis and the world recession). Ic. What was the relative role of the following in inducing these changes: (I) Ministry of Finance/Planning; (ii) the Cabinet; (iii) the Legislature; (iv) private sector-government consultation committees? Rank these in order of importance, with 1 for the least influence and 4 the most. Non available 2a. Is there a requirement for carrying out ex post evaluation of programs/projects ? By whom: central agencies, line agencies, or by independent external agencies? Check all those that apply. They are done by the ODI, Special Programs, MEF 2b. Are the results used in expenditure allocations for the next budget? Yes. 3. Are client surveys routinely carried out as part of these evaluations? No. ACCOUNTABILITY la. Is there a clear specification of the output to be produced by: (I) a ministry; (ii) a department within a ministry; and (iii) a division, program or project unit within a department? Yes there is financially for the ministry ruled by the norms issued by the Nation Public Auditor's Office and they are in agreement with the Principles of the "Generally Accepted Accounting" which includes, according to the case, the development of the International Norms. lb. If so, are these outputs published?. They are published, with limited distribution, to the Executive and the Legislative bodies. They are available in the Nation Public Accountant' Office Web site. 2a. Are performance indicators specifically linked to senior manager (I) tenure, (ii) promotion; and (iii) compensation? No. 2b. Are these performance indicators based on the achievement of outputs (i.e., goods and services produced, e.g., number of immunizations of health clinics) or outcomes (i.e. , impact of beneficiaries e.g, lower infant mortality). They are collected by the National Office of Public Budget with little details. 2c. Have chief executives been fired on account of nonperformance? No. 175 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION 3a. What is the percentage deviation between public and private pay for different grade levels? It is 5 to 1 in proportion with respect to staff. 3b. Is there an explicit link between pay and performance? No. 4a. Is competitive bidding required for the procurement of major expenditure items? Yes, procedures are under standard. The Superior Council of Contractings and Procurements from the State - CONSUCODE -supervises their fulfilment. 4b. Are the rules for bidding made public? Yes. When the bidding amount is larger, the publication is larger too. Sa. When are financial accounts of line agenciesprepared: (I) quarterly during the budget year; (ii) semestral during the budget year; (iii) within six moths from the end of the fiscal year; (iv) more than six months but less than one year; (v) between one and three years; (vi) more than three years. Yes, for the budget. Yes. The budget program is conciliated with the numbers rendered by the entities of accounts, the National Office of Public Budget and the Nation Public Auditor's Office. (iii) (iv) (v) and (vi) No. 5b. Are there punitive actions taken against (I) delays; and (ii) discrepancies? Under the law, there are actions, but in the practice it depends on the actions of the Republic General Auditor's Office which is the entity in charge of that. It is not frequent to see actions by the General Account of the Republic. Sc. Are these accounts tabled before a separate session of the Legislature? The procedure indicates that the General Account of the Republic is forwarded to the Republic General Auditor's Office in order to create the Auditor's report. After that, it is forwarded to the Presidency of the Republic, to the Ministry of Economy and Finances and to the Review Commision of the General Accounts of the Congress. 5d. Are they made public? They are distributed to the Executive and Congress. They are in the Public Auditor's Office Web Site for fiscal year. They are divulged in the training, local, domestic or international competitions. 6a. Are the agency accounts audited? In the mayority of cases, specially those of public enterprises entities of the Central Government, the Republic General Auditor's Office is in charge of the auditing of the General Account from the Republic. 6b. If so, by whom: internal agency auditor, the government auditor within the Executive, independent auditor? No response. 6c. When are audits of agency accounts undertaken: (I) quarterly during the budget year, (ii) semestral during the budget year; (iii) within six months from the end of the fiscal year; (iv) more than six months but less than one year; (v) between one and three years (vi) more than three years. It is once per year. 6d. What percent of programs have been audited in the last five years? As a percentage of the number of entities, approximately 20%. In Budget amounts, it is approximately 80%. 6e. What percent are financial audits as opposed to performance audits? Approximately 35% 6f. Are the result published? No, they are not. 6g. Has there been punitive action or promotion based on these audits? Yes there has been. The Republic General Auditor Office, with the help of the General Attorney' Office, is in charge of that. 7a. Are there client surveys undertaken? Depends on each case. Part of the control of procedures includes actions with the providers and clients. 176 ANNEX C 7b. How frequently? There is not an establish schedule. It depends on the Control Plans. 7c. Are the results published? No, they are not. 7d. Do these surveys measure satisfaction with srevice delivery ( i.e. outputs), or with success of the program (i.e. outcomes), or both?. No. 8a. How many major donors provide projects financing?. There is no data available. 8b. Do these projects specify the amount and type of expenditures on which project resources will be spent? There is no data available. 8c. Does each donor have its own rules about disbursement, procurement, accounting and auditing of projects funds? Yes. We are working in order to adapt those requirements with the Integrated Financial Management System in the area of funtional and operational competence. 8d. Do these rules match those of the government? Yes. 177  ANNEX D A PUBLIC EXPENDITURE TRACKING SURVEY: METHODOLOGICAL ISSUES A. Sample Design Measuring leakages in transfers to subnational governments, local organizations and program beneficiaries is not an easy task since it involves two central problems. First, all parties affected by the leak have no interest in having it revealed. Second, leakages can occur at so many levels that tracing them all requires a complex methodology. Those are the two challenges that this report faced and, to a considerable degree, overcame. The core of our methodology was to collect data at each stage in the transference of public funds from the top of the chain, namely that of the central government, down to the bottom, namely that of the beneficiary. In order to gather data on each of these levels, the study carried out in Peru involved an extensive amount of fieldwork based upon a series of questionnaires. The questionnaires themselves are attached to Apoyo Institute (2002a). In this Annex, we briefly describe how we went about gathering the data. The project began with a pilot study in Lima, Peru. The objectives of the pilot included an assessment of the duration of the fieldwork and a test of its effectiveness for the purposes of the study (for example, to explore whether the information collected in the fieldwork would be sufficient to rigorously estimate leakages). The pilot consisted of 20 districts of the department of Lima (out of a total of 177 districts). Each district included: a survey for the municipality; 3, 4, or 5 surveys' for the VDL mother's committees; and 16 surveys for beneficiary households2 (4 per VDL committee). Additionally, Lima was selected for the pilot because it considerably differs from the rest of the country and needs a separate treatment. The advantage to do it in this way was that we were able to report comparisons between Lima and the rest of the country, which in many cases were quite large. Based on the experience of the Lima sample, and after a thorough process of consultation, and with guidance from the INEI (Peruvian Institute of Statistics and Information), a national sample selection methodology was agreed upon. The following departments were chosen to be representative of Peru (excluding Lima & Callao): Ancash, Arequipa, Cajamarca, Cusco, Loreto, and Piura. These departments gave us the broadest range of geography, population density and poverty distribution, while reducing our field costs to reasonable levels (i.e., we did not have the resources to include municipalities from all of Peru's provinces). We then selected a total of 100 municipalities in which the surveys were carried out. Our method of selecting the municipalities was focused on poverty as a central stratification variable since the Vaso de Leche program was meant to deal directly with poverty. The sample, then, represents Peru (when the Lima sample is included) and the 100 municipalities selected are stratified by level of poverty so that the efficiency of the sample would be maximized (stratification reduces sampling error). The sample is self-weighting, making it easy to work with when complex multivariate programs are employed. Our method of municipality selection followed next steps: For the Pilot the following rule was established: 3 glass of milk committees if there were less than 30 committees total ; 4 if the number of committees was between 30 and 70; and 5 if there were more than 70 committees. 2 When the pilot was carried out, the project had not formally included surveys to household beneficiary yet. A tentative instrument was tested on households to evaluate the importance and viability of including beneficiaries. Therefore, the pilot survey was shorter than and different from the survey applied to beneficiaries during the final fieldwork. 179 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Stratification according to poverty A database consisting of the entire universe of districts in Peru, excluding Lima & Callao (Total of 1,651 districts) was used as a starting point. * The MEF's continuous index of poverty FGT23 was used to calculate poverty population deciles. * The deciles were arranged into three groups such that group 1 consisted of deciles 1-3, group 2 contained deciles 4-7 and group 3 had deciles 8-10. These three groups approximate the categories of "not poor", "poor" and "extreme poor" and were used to stratify the districts of our sub-population (Ancash and Piura) into three strata. * The three strata represent 14 percent, 41 percent, and 45 percent of the districts in Peru (excluding Lima and Callao) respectively. * In order for the sample to be self-weighted 14, 41, and 45 municipalities (total of 100) were chosen from each stratum respectively (from the sub-population of six departments). The selection for each stratum was done using PPS4 relative to district population. Once the above procedure was carried out, individual municipalities were selected according to PPS criteria, using a complete listing of all districts selected that were ordered within the stratums by geographic order to allow a systematic selection that ensured geographic heterogeneity. Within each municipality, the field used the roster of Vaso de Leche committees and used systematic sampling to select four of those, unless there were fewer than four in a given municipality, in which case all were selected. The only restriction was that if travel time to a given committee would have required more than 24 hours, a substitute was used. This means that the sample slightly under represented remote areas within the neighborhoods of the selected committees, the field team selected four households from the beneficiary lists that are maintained by each committee. Recalls were not made, but the next household on the list was used as a substitute when blanks were encountered. The survey was carried out from February 3, 2002 through February 17, 2002. Descriptive statistics for the samples are provided in the full report. Within each municipality we interviewed the mayor, obtained municipal level data from him/her, and also obtained the municipal roster of committees participating in the Vaso de Leche program. The committees were selected from the roster using a systematic selection. Once we met with at least one committee member and interviewed that individual with our survey instrument, we obtained from that member a list of individual beneficiary households, and interviewed four households in each committee catchments area, using the survey instrument intended for households. B. Description of Transfers from Central Government to Municipalities and Downwards Municipal resources come from two primary sources: central government transfers and local revenues. Central government transfers are a very important source of revenue, particularly for small and rural municipalities, which have very little income collection capacity. Central 1 FT=tPL - EXPpc 2 where PL=-Poverty line, EXPpc=per capita household expenditures, Q-Number of FGT, = f F N i1 PL poor, and N=Population 4 Probability Proportional to Size (PPS) is a method used in sample selection whereby the probability that a given element enters the sample is proportional to some quantity (in our particular case, the district's total population). 180 ANNEX D government transfers include FONCOMUN and Vaso de Leche (VDL) for all municipalities and Canon Minero and Canon/Sobrecan6n Petrolero for some of them. Central government transfers represent an important percentage of total district-level income (for the districts outside of Lima in our sample, transfers, on average, represent 72% of total income and among the districts of the extreme poor stratum they can represent in excess of 90% of total income). In 2001, these four major central government transfers totaled at the national level 1.9 billion Nuevos Soles (roughly $560 million) (Table DI).5 For these reasons the leakages, delays, volatilities, and inefficiencies associated with the execution process of these transfers have considerable impact on municipal-level finances and their understanding is, consequently, of paramount importance. Of these four transfers, only the Canon Minero is not variable month-to-month. In theory, the VDL transfer should not be variable month to month, but as the following table shows this is not true in practice. Foncomun and the Canon/SobreCanon Petrolero are percentages of a variable quantity (primarily the national sales tax in the case of Foncomun and the ad-valorem petroleum production in the case of the Canon/Sobrecanon Petrolero). The implicit volatility in these two transfers is a source of hardship for municipalities, whose budget depends heavily on them. Table D1. Total Transfrs to Municipalities in 2001' (in U.S. dollars) Sobre"anouItj Le Peru $400,023,180 $23,765,654 $37,461,817 $97,148,245 $558,398,895 Lima $72,466,783 $714,003 NA $33,753,411 $106,934,197 Urban $69,482,021 $691,650 NA $33,359,639 $103,533,311 Rural $2,984,762 $22,353 NA $393,772 $3,400,887 No. Observations 177 171 NA 177 N.A. Rest of Peru $327,556,397 $23,051,651 $37,461,817 $63,394,833 $451,464,698 Not Poor $76,066,085 $5,692,301 $9,924,653 $15,642,446 $107,325,483 Poor $132,062,598 $8,603,523 $17,670,749 $24,705,129 $183,041,999 Extreme Poor $119,427,714 $8,755,827 $9,866,416 $23,047,259 $161,097,216 Urban $152,245,705 $9,629,748 $20,840,016 $30,939,959 $213,655,428 Rural $175,310,692 $13,421,902 $16,621,801 $32,454,874 $237,809,270 Small $65,758,919 $3,206,004 $3,916,796 $8,990,266 $81,871,986 Medium $50,188,784 $3,515,896 $5,744,644 $10,333,395 $69,782,718 Large $211,608,694 $16,329,751 $27,800,377 $44,071,173 $299,809,994 More accessible $239,125,681 $17,287,765 $22,227,123 $46,727,882 $325,368,452 Less Accessible $88,430,716 $5,763,885 $15,234,694 $16,666,951 $126,096,246 Non-Provincial capital $188,468,161 $11,999,547 $18,319,805 $41,422,875 $260,210,387 Provincial capital $139,088,236 $11,052,104 $19,142,013 $21,971.958 $191,254,311 No. Observations 1641 1296 142 1641 N.A. 1/Information based on national official statistics Source: MEF 5 Using MEF statistics at the national level. 181 ANNEX D Month-to-month percentage changes in amounts for each of the four major transfers were calculated. The volatility of the transfers was defined as the standard deviation of these month-to- month changes. The following Table details the volatilities of Foncomun, Canon/Sobrecanon Petrolero, and Vaso de Leche in the year 2001.6 The measure of dispersion is the degree to which the volatilities vary within each of these sub classified categories. The volatility of the Canon/Sobrecanon Petrolero transfers is the combination of the variations in world petroleum prices as well as the variability of local production and is, on average, twice as high as that of the Foncomun transfer. Foncomun's volatility is not considerably different between urban/rural districts or between poverty strata but is twice as high in larger districts (as measured by their projected 2001 populations). The volatility of the Canon/Sobrecanon Petrolero, on the other hand, is quite a bit higher in poor, rural, remote districts of Peru. The former finding is quite interesting given that the assignment formula does not vary within the year. Therefore, a percentage change in the sales tax revenue (and/or any other component of the Foncomun) should imply an equal percentage change in the amounts allocated to the districts across the board. However, we see that the standard deviations of the monthly percentage changes are not equivalent. This is partly due to the fact that these transfers have a built-in floor (15,600 Nuevos Soles) and therefore their variations are asymmetrical. The VDL transfer, surprisingly enough, also suffered from considerable volatility in the year 2001 in the districts outside the department of Lima. The poorest districts were the most affected with a standard deviation of percentage changes on the order of 15 percentage points for the.year. This figure is in sharp contrast to the average volatility of 0.3 percent experienced by the 177 districts of Lima and Callao (Table D2). Table D2. Volatility of Transfers to Municipalities' Lima 4.9% 4.1% NA 2/. NA 0.3% 1.6% Urban 5.5% 4.2% NA NA 0.2% 1.2% Rural 2.9% 3.2% NA NA 0.5% 2.6% No. Observations 177 NA 177 Rest of Peru 6.3% 3.4% 13.4% 6.9% 11.6% 30.3% Not Poor 6.1% 3.8% 9.2% 3.1% 10.0% 56.8% Poor 6.0% 3.6% 11.6% 4.7% 8.1% 22.2% Extreme Poor 6.5% 3.2% 18.8% 8.4% 15.4% 23.9% Urban 5.7% 3.8% 10.4% 4.1% 5.2% 19.5% Rural 6.5% 3.2% 15.7% 7.7% 14.7% 33.9% Small 4.0% 3.1% 13.2% 5.0% 10.7% 34.9% Medium 8.2% 1.7% 15.5% 9.4% 15.3% 26.3% Large 9.4% 0.8% 12.4% 5.7% 10.5% 21.7% More accessible 6.8% 3.3% 10.3% 5.0% 9.5% 27.5% Less Accessible 5.5% 3.5% 16.4% 7.2% 14.2% 33.3% Non-Provincial capital 5.8% 3.4% 13.4% 6.8% 12.1% 28.9% Provincial capital 9.6% 0.5% 13.8% 7.4% 7.7% 39.7% No. Observations 1642 142 1642 Source: National Official Statistics, MEF IZ Standard Deviation of the percent changes month-over-month 2001 2. The department of Lima does not receive Canon/Sobrecanon Petrolero 6 The volatilities are estimated by calculating the standard deviation of the rate of change of the monthly amounts for each of the three transfers. 182 ANNEX D FONCOMUN The Fondo de Compensaci6n Municipal (FONCOMUN) is mentioned in subsection 4 of article 193 of the Constitution of the Republic of Peru. It is comprised of two percent of the federal sales tax (know as IGV), 8% of gasoline sales, 5% of recreational embarkations, and 25% of the net income generated by casinos and other gambling establishments. The contribution from the sales tax accounts for approximately 93% of FONCOMUN revenues and therefore constitutes its main source of variability. The law which permits the redirection of a portion of the federal sales tax under the name of "Impuesto de Promoci6n Municipal" is dictated by article 86 of Legislative Decree 776 of the Law of Municipal Taxes. The Fund is calculated on a monthly basis and is distributed with monthly frequency to all the municipalities, based on a predefined allocation formula of the Ministry of Economy (MEF). The process for disbursing the transfer from the central government to the municipalities is the following. The collection of funds is centralized at the Banco de la Naci6n (BN). Within the first five days of each month, the BN informs the National Direction of Public Budget (DNPP) of the MEF regarding the amount collected. This office in turn informs the General Direction of Fiscal Policy (DGPF) of the MEF, which is in charge of estimating the amount to be distributed to each municipality. Once the distribution is determined, DGFP informs DNPP, which issues the transfer order to the BN. Finally, the BN deposits the money in the account of each municipality on the 13t to 15th day of the month (see diagram). According to MEF personnel our team interviewed, the process is automatic and should not take any longer than the terms established. The distribution criteria do not change frequently although there are certain minimums established which require a complicated re-allocation if there is a downturn in resources for a given month. Nonetheless, as the results of the fieldwork indicate, the municipalities do face some delays. In principle, the transfer process funds from the central government to the municipalities is straight forward and does not suggest ex ante that there will exist a leakage in this segment of the chain. At this stage, the process is quite systematic and the Treasury (BN) serves a limited function: electronically transferring funds from one account to another and issuing arrival notifications. The DGFP office of the MEF has a formula used to calculate the way it will distribute the total amount of funds the each municipality. For the districts outside of Lima and Callao, the formula allocates amounts to each province based on the infant mortality rate and the total population of the province. The formula further assigns amounts to the districts based on district-level populations with a bias towards rural populations (rural inhabitants count twice as much as urban inhabitants do). For Lima and Callao the objective is to prioritize urban marginal districts. The allocation criteria considered are the following: population; illiteracy rate for those older than 15 years old; households with at least one child attending school; households without a water connection in the home; households without sewerage; households without electricity; households that occupy makeshift homes; households with three or more inhabitants per room. All these indicators are provided by the national institute of statistics (INEI) and are based on the 1993 census. Until December of 2001 (December 29d), the law established that only up to 30% of FONCOMUN could be used to for current expenditures. At least 70% of the fund had to go towards capital expenditures. Since January of 2002, the Foncomun is not subject to the 70-30 ratio any longer and can be freely utilized as the municipalities best see fit. 7 FONCOMUN transfer, as well as the other municipality transfers and finances in general are still not incorporated in SIAF. At the time this report was under elaboration, SIAF's office was working with a pilot sample of at least 25 municipalities to integrate them (see Chapter 111). 183 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION FONCOMUN TRANSFER PROCESS Municipality Bank of the Nation __ Receives the (BN) transfer on the Centralizes tax 13th or 15th of the collection month Within first 5 days of DNPP requests BN to the month BN deposit the transfer in informs DNPP of the municipality account amount collected ' National Direction of Public Budget (DNPP) DNPP informs DGPF DGFP informs the of the total amount transfer amount that available for corresponds to each distribution to the " municipality municipalities General Direction of Fiscal Policies (DGPF) Canon Minero Dictated by supreme decrees 88-95-EF and 041-97-EF, the Canon Minero is a Central Government transfer, which distributes a portion of the mining sales to the localities in which the mineral ore was extracted. A 20% of the income tax paid by the mining companies is distributed among the municipalities in the following manner: 40% for Provinces and districts in the region or regions8 in which the ore was extracted or in which the mining company's economic headquarters are located; and 60% for Provinces and districts in the department or departments in which the ore was extracted or in which the mining company's economic headquarters are located (in addition to the 40%). The total amount allocated to a set of municipalities is then distributed by the MEF according to a formula similar to the one used for FONCOMUN. Within a region, the allocation to provinces is based on the infant mortality rate and the total population of the province. Of that amount, 20% remains at the provincial municipality (provincial capital) and the remaining 80% is distributed among the municipalities of the Province based on their respective populations (once again giving a two-to-one preference to rural populations). For Lima and Callao, where the objective is to give priority to the urban-marginal districts, the additional criteria used in the distribution of the FONCOMUN are also used for the Canon Minero. The process of disbursement of the Canon Minero is simple. The tax agency, SUNAT, deposits the taxes collected in a special account at the BN. The DNPP then distributes the total amount among municipalities (following the criteria mentioned above) in twelve equal monthly installments made by the BN by way of electronic deposits. There is a restriction in place governing the usage of the Canon Minero funds: 8 Under the administration of Alan Garcia regions (several adjacent departments) were created. 184 ANNEX D Municipalities are only authorized to use the resources obtained through this transfer to cover capital expenditures. Canon v SobreCanon Petrolero These transfers were established by Law Decree 21678 in November of 1976 and are comprised of a percentage of the ad-valorem petroleum production. The Canon is 10% of the ad-valorem petroleum production and is distributed among the districts within the department where the petroleum was extracted. The SobreCanon is 5% of ad-valorem petroleum production and is distributed among the districts that are not within a department that produces petroleum but are within the same region. The Canon/SobreCanon is equally distributed among the provinces? in the department and then distributed to the district-level based on criteria such as'o: population; geographic area; whether it is a Provincial or departmental capital; and whether the population is greater than 100,000 inhabitants. Glass of Milk (Vaso de Leche - VDL) The Glass of Milk program was started by the Mayor of Lima in 1984 for the Metropolitan area and as its name reflects it originally consisted of distributing glasses of milk to school aged children. Since then, the program has grown to national coverage (in 1985), not only restricted to milk products, and has become one of the most significant transfers from the central to local governments (only second to FONCOMUN). It came to its current form under Law 24059 and Article 7 of Law 27470 (2001). Although the spirit of the program remains the same, its organization has undergone considerable changes due to an increased coverage. By Law, the intended primary beneficiaries are: Children 6 years old or less and pregnant and breastfeeding mothers. Priority is given to those showing clear signs of malnutrition or tuberculosis. The Law also establishes that if there are resources left over after attending the needs of the primary beneficiaries then attention may be provided to secondary beneficiaries: children between 7 and 13 years old; the elderly and those suffering from tuberculosis. The main goal of the program is to improve the nutritional level of infants, small children and pregnant or breastfeeding mothers, and improve the quality of life of the poorest segments of the population. Given empirical evidence, which identifies milk as an important nutritional source, a decision was made to distribute milk and milk-related products. Unfortunately, organizational hurdles, inefficiencies, leakages, and low nutritional value of the products chosen for distribution are aspects that .limit the effectiveness of the VDL program to accomplish its original goals. Because of the program's flawed design organization and the way in which it functions, there are no effective mechanisms that enforce its priorities. Currently, the MEF makes monthly transfers to municipalities for use in the Vaso de Leche Program. The transfer amounts are calculated by the MEF using distribution indices that are based primarily on poverty and demographic indicators, supplied by the National Institute of Statistics. Local governments are solely responsible for the operation of the VDL program in their jurisdiction and have autonomy, except for the definition of the beneficiaries, from the Central Government as far as the product they wish to distribute and other operational details. However, three additional. restrictions are placed upon the municipalities by the Central Government. First 100% of the Vaso de Leche funds must go towards the purchase of the product(s) to be distributed. Second, the product(s) should have a minimum of 90% of national inputs (up to 100% in the areas in which the local supply can cover the entire demand). The 9 The departments of Loreto and Ucayali, the distribution at the province-level is in proportion to their share of production. In Piura and Tumbes the distribution makes no distinction between the provinces. 10 In the case of the province of Puerto Inca, Hu6nuco, the distribution among districts is simply in equal parts. 185 PERU: RESTORING FIsCAL DISCIPLINE FOR POVERTY REDUCTION products distributed can be milk in any form and/or milk substitutes, and/or other products such as soybean, oatmeal, quinua, kiwicha or others." Third, the Central Government requires that each municipality creates an Administrative Committee (VDLC) with a certain membership composition: the mayor; a municipal employee; a representative from the Ministry of Health; three representatives of the Mother's Associations (elected by the mothers following the rules established in their own statutes); and a Representative of the local agriculture/farming association duly accredited by the Ministry of Agriculture. By Law, the municipalities or committees should not charge anything to the beneficiaries for the products distributed. However, some municipalities and many committees charge some money (or require that in-kind contributions be made) to cover some of the expenses that the program implies. Requests for contributions are more common in poor municipalities that need to cover distribution expenses, and in committees that prepare the product on-site and need to pay for things such as fuel and other inputs. In general, larger districts have an entire office dedicated to the Vaso de Leche, including a director and other full-time employees. In smaller districts, often times, the mayor himself or one of his regidores12 administers and directs the program. It is quite noteworthy to mention that beneficiaries (the mothers) organize themselves into Comit6s/Clubs de Madres (mother's associations), which have noticeably increased the social capital of the communities. These mothers' associations span the entire country, including the most remote areas, and together form a huge national network. The social network that results from the mother's associations has become a way for these rural and impoverished communities to gain a voice at the local level even on matters unrelated to the Glass of Milk program. The transfer process takes on the following scheme: * The municipality receives the transfer from the Central Government. A monthly transfer is deposited in the Banco de la Naci6n (Treasury) local branch office and the Municipality is notified. The monthly amount does not vary in a given year except in very particular situations, and the transfer usually arrives around the same day each month. * The Product is chosen and, in most cases, a public bidding process is used to select a provider, normally for a year and for the whole municipality jurisdiction. The selection of the product or products is done differently in different districts. In some cases, mothers have a direct say in the choice of product via surveys or direct vote. In other cases, the Administrative Committee makes the decision. Finally, in some cases it is the Mayor's decision. * A roster of all direct beneficiaries in each committee by category (children form 0 to 6 years, pregnant or breastfeeding mothers) is elaborated. The mothers' committees normally do this process, which is supposedly closely supervised by the municipality. This roster should also be frequently revised. However, in most cases these rosters are not updated and the municipality does not supervise the process. Using these registers, the municipality determines the product allocation to each of the mothers' committees of its jurisdiction. * The municipality (or the provider) distributes the product to the mothers' associations. These usually cover a given neighborhood and can have a variable number of beneficiaries. A president, a vice-president, treasurer and secretary usually lead them. The officers are voted into office for extended periods of time and they are rarely voted out of office unless problems arise. Depending normally on the size of the municipality, the products are directly distributed to the committees or to intermediate instances: groups of committees or "centros " Recently, a new law has been passed that requires that distributed products have a minimum of 207 calories. Mothers, however, are not pleased with this requirement because it implies that milk substitutes will be distributed and they do not like them. Mothers made street protests demanding evaporated or fresh milk. 12 Municipal council member 186 ANNEX D de acopio" (special locations where the products are collected by areas and from where the committees pick up their allocations). The mothers' committees should follow the rosters of beneficiaries to determine the allocation per household or beneficiary, but in most cases, rosters are only "referential" and follow other criteria. The mothers (or the intended beneficiary) pick up their ration(s) at the mothers' association headquarters. Usually the headquarters is the house of the president of the mothers association and there is a certain time of the day/week/month that the distribution is made depending on the frequency with which the product is distributed. The committees either distribute a product already prepared to the beneficiaries or in its raw form. The decision may depend on the type of product, on the resources availability, or on the geographical dispersion of the beneficiaries. In the cases in which the products are not prepared by the committee, mothers use them at their home at their total discretion (usually distributing the products among all children or household members). C. Technical Definitions of Leakages in the Vaso de Leche Program Leak 1. From the Central Government to the municipalities. Leak 1 is defined as the percentage of the transfer reported by the MEF that is unaccounted for by the municipality. We compare the amount the Ministry of Finance reports as outgoing with the amount the municipality reports to have received. This leakage is estimated with municipal-level data of December 2001. = 1 Amt. Municipality Reported Amt. MEF Reported Leak 2. Within Municipality. This leak is defined as the percentage of the amount transferred to the municipality "i" from the Central Government for the month of December 2001 that is unaccounted for by the total expenses of the municipality for that month (in terms of products purchased for the VDL program). Leak 2 is zero if the municipality spends the entirety of the resources available in December 2001 under the Glass of Milk program on products to be distributed by the program. This leakage is estimated based on municipal-level data. [ (Quantity, xPrice) Leak = 1- Transfer Amount J Leak 3. From Municipality to VDL Committees. This leak is defined as the percentage of the amount listed in the municipality not accounted for by the VDL committee. This leakage is estimated using municipal and committee data and is computed at the committee-level. This leakage indicates how much is lost in this segment of distribution, but does not allow one to attribute it to one of the two parties involved at this stage. In other words, we estimate the leakage from the municipality to the individual VDL committees, but do not know if the leakage is a result of misappropriation or inefficiencies of the municipality, the VDLC, or both. LeakeouM =1- Amount Receivedco. Amount listed in Municipal RosterOM Leak 4: From VDL committees to Beneficiaries/Households. Leak 4 is the loss due to the difference between what VDL committees receive according to beneficiaries registered and what they actually distribute to households. The estimation of the leakage at this level is done by 187 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION calculating the monetary value of each of the products (using municipal price figures) and adding these up. This allows a comparison of the monetary value of the amount of all the products received by the VDL committee per beneficiary with the monetary value of the amount received by the individual households per beneficiary (excluding the committees that distribute prepared products). The first variable is obtained from the quantities declared by the mothers' committee representative in the VDL committee survey (in the four committees surveyed in each municipality). The second variable is obtained from the quantities declared by the beneficiaries' household representative in a beneficiary household survey (on four households surveyed for each VDL committee). (Quantity, x Price,) Lek1 Beneficiary H Leak4 = 1- Z (Quantityj xPrice,) Beneficiary Leak 5: Inside the Household. This leak is attributed to beneficiary dilution at the household level. It is defined as one minus the percentage of household members who consume Vaso de Leche products, which are direct beneficiaries. This leakage is estimated using household-level data. L I [Beneficiaries HH HH Consumers HH 188 ANNEX E PUBLIC ENTERPRISE REFORM IN PERU: INTRODUCING MANAGEMENT CONTRACTS AT FONAFE1 In 1999, as part of some isolated efforts of state modernization in Peru, FONAFE, an institution of no more than 30 employees became a useful instrument to increase public savings and improve the efficiency and effectiveness of public enterprises. FONAFE is the public entity assigned to the Ministry of Economy and Finance, which is in charge of directing the entrepreneurial activity of the state. Temporarily, it has extended its management instruments to regulatory and revenue collecting agencies in Peru's public sector. Enterprises under FONAFE's control include those in which the state owns either the total of the social capital or the majority of shares. In those cases in which the state is a minor partner, FONAFE only exercises stockholder options corresponding to its shares. Its main functions are: (i) to approve the consolidated budget of the enterprises in which it holds the majority of shares; (ii) to approve management norms for those enterprises; (iii) to exercise stockholder options and administer resources resulting from its shareholder status; and (iv) to name representatives to the board of those enterprises in which it is a shareholder (in those cases in which the state holds the total of shares, FONAFE names a board of five members). Upon introducing reform, FONAFE developed and implemented four basic management tools: * An Integrated Expenditure on Personnel, which sets a global amount that enterprises and other public entities under its supervision are authorized to use for this purpose, while allowing management flexibility within that limit. * A Maximum Salary Limit, which establishes the maximum annual salary any employee may receive, excluding the productivity bonus. * A Productivity Bonus, an annual monetary incentive to personnel which can sum up to 6% of the payroll and that is only granted if 90% or more of the annual objectives are achieved. * A Management Contract that establishes the specific terms of the agreement between FONAFE and the enterprise or public entity under its supervision. These instruments have shifted management from being focused on inputs and procedures to outcomes and results. A key element of the system is the capability to develop adequate indicators in the areas of cost, quality, coverage and internal efficiency. FONAFE's approach has been to let each enterprise to develop its own indicators based on its strategic planning processes, which are then analyzed and validated by FONAFE. According to its key officials, this allows for greater legitimacy and applicability. Indicators proposed by each entity need to fulfill usually recommended requirements: * Appropriateness.- They should be the expression of the processes and products of the institution in order to properly reflect the fulfillment of the institutional objectives. * Independence.- Should be independent of each other and respond to the actions developed by the institution. Indicators depending on external factors should be avoided when possible (e.g. situation of the country, demand variations, third party decisions, tariff changes, etc.). * Measurement.- Availability of data and methodological issues should be considered in the proposal. * Control.- Basic data should be part of a system that can be audited. * Cost.- Information for the management indicators should be collected and processed within reasonable limits of cost and time. This Annex was prepared by Gabriel Ortiz de Zevallos. 189 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Introducing these instruments has not been easy. Preliminary results have been very positive (see Table). However, some difficulties encountered are: * Reactions to implementing change have been uneven and diverse. Enterprises under FONAFE operate in different sectors, can be old or relatively new, may or may not have data easily available, and have different levels of managerial capabilities, as well as their personnel may not have the same willingness to be part of a change of this sort, * Managers have often started with unrealistic goals. In the first year of their implementation, most enterprises overestimated the results they would be able to achieve. After the first year, during which most targets were not met, managers changed to more modest, realistic goals. SChairacteristics ^ ' '. - Ya G. 1999 2000~ Y2001 2002, Enterprise SEDAPAL (Lima's Water and Sewerage Service) Indicator: Service continuity 15.9 17.6 19.3 22.0 (in hours) Enterprise SEAL (South-West Electricity Society) Indicator: Energy loses in the distribution system 21.7% 21.25% 18.12% 16.40% Entity OSIPTEL (Telecommunications Regulation Agency) Indicator: Supervision in the delivery of the 364 524 telecommunications public service (supervision actions) Note: Starting on year 2001, Osiptel is no longer under the scope of FONAFE, but management instruments continue to be negotiated directly with the Ministry of Economy and Finance. Besides political will, leadership and a strong technical team, two factors have been key for success: * Its legal and administrative status has allowed FONAFE not to be burdened by administrative procedures. FONAFE was created as a line unit of the Ministry of Economy and Finance budget, so its budget does not have to be approved by Congress, but by the Council of Ministers. It has sufficient authority to dictate its norms, including budgetary issues, and the legal capacity to control management of the enterprises under its supervision. * Its approach to transition has been incremental and case-by-case, enabling it to adapt to specific difficulties. For instance, FONAFE's 1999 agreement with EGASA, the electric-generating company of Arequipa, set up as its initial performance indicator to develop a strategic plan for the next decade. Subsequent goals and indicators have been based upon the implementation of new systems and plans. FONAFE has extended management contracts to other public agencies-regulatory and revenue collecting. Although desirable, extending them to the central government implies new difficulties: * Productivity bonuses come from savings, but generating surpluses to finance bonuses through the improvement of management is easier in State-owned enterprises than in ministries. The Education Ministry, for example, spends almost 90% of its budget in payroll, making it much harder to maneuver and generate savings to reward thousands of teachers. However, the current administration is barely tackling with the very difficult issue of identifying how many "ghosts" are currently parts of the teacher's payroll. * Different personnel regimes coexist within the public sector. While in public enterprises there are very few cases of workers not registered in the payroll, central government agencies tend to have large numbers of employees contracted as professional service staff outside the payroll. Legally, it is hard to justify a bonus in those cases. * Reliable data are fundamental, but extremely scare. The need of having quality administrative systems that generate and constantly update data is central for the success of management contracts. A recent study of twenty social programs that implied a budget of US$ 770 million in year 2000, found that only three of those programs, accounting for US$17 million, have institutionalized proper monitoring and evaluation systems in manuals. 190 ANNEX F A METHODOLOGY FOR FUNCTIONAL .. REVIEWS AND FOR ANALYZING FUNCTIONAL PROSPECTS The absence of an implementation strategy in the Framework Law for Modernization may result in rather haphazard efforts to apply it to specific ministerial or sectoral contexts. Such an implementation strategy is imperative. Below, Moore (2002) proposes some initial elements for conducting functional reviews as part of this implementation strategy. The basic questions that guides the methodology is how are these functions organized currently in the structure of the ministry/sector? and how can or should they be organized in the future, either by units or levels of the ministry? Its central premise is that in order to be efficient any organization must fulfill seven basic functional subsystems. The strategic management of the organization requires an analysis of how these are structured and how they interact. Each of these functions consists of sets of activities that can be identified, and several levels of the organization may exercise each of these activities either exclusively or concurrently. A matrix for identifying functions, activities, levels and exclusivity can be designed for the application to the organizational structure. These subsystems include the following: * The programmatic function that provides the why of the organization and its vision and mission. This function defines the substance of the institution and explains its purpose and linkages to other sectors or institutions. * The planning function that provides the how for carrying out the institutional vision and combines activities related to policy formulation and planning, and their budgetary expression. At the same time, the function incorporates those activities related to investigation (for strategic planning). * The financial administration function that provides for the identification of the processes related to what resources (financial) are needed and utilized to carry out the how. This function is delimited by financial analysis, the administration of resource generation, budget execution, and the control of the use of financial resources. * The human resource administration function that provides for the management of what resources (human) are necessary to carry out the activities of the institution and relates to the various subsystems of human resource management (recruitment, selection, promotion, development, evaluation, etc.). * The management control, supervision and evaluation function(s) that determine how the institution is regulated, how supervision takes place, how resource execution of projects and plans is carried out to assure accountability in policy execution. * The administrative support function that guarantees the acquisition of goods and services, assures the capacity to carry out institutional actions, and includes those activities related to acquisitions, maintenance, construction and their organization. 191 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION * The information function for producing, processing, distributing and consolidating the information requirements to guarantee opportune and efficient decision-making, management and execution and the appropriate diffusion of information. Each of these consists of activities, and these can be identified (current, future) by units and levels in the form of a matrix of activities. Such an initial matrix for identification has been utilized recently in Ecuador's Ministry of Education.' In real terms, the carrying out of the exercise is organized in a series workshops/seminars of relevant (primarily internal) persons to arrive at the completed matrix. The matrix is then applied to a reorganization strategy. 'See Annex 3 in Moore (2002) for a full description of details of the Ecuadorian experience. 192 STATISTICAL APPENDIX' Page # Table Al Peru: Combined Public Sector Operations (millions of nuevo soles).............................................................................. 195 Table A2 Peru: Combined Public Sector Operations (percent of GDP) ......................................................................................... 195 Table A3 Peru: Combined Public Sector Operations (millions of 1994 soles) ............................................................................... 195 Table A4 Peru: Central Government Operations (millions of nuevo soles).................................................................................... 196 Table AS Peru: Central Government Operations (percent of GDP)................................................................................................ 197 Table A6 Peru: Central Government Operations (millions of 1994 soles)...................................................................................... 198 Table A7a Peru: Structure of Central Government's Fixed Capital Formation (millions of nuevo soles)........................................ 199 Table A7b Peru: Structure of Central Government's Fixed Capital Formation (percent of total).................................................... 199 Table A8a Peru: Operations of the Non-Financial Public Sector (millions of nuevo soles)............................................................. 200 Table A8b Peru: Operations of the Non-Financial Public Sector (percent of GDP)......................................................................... 200 Table A8c Peru: Operations of the Non-Financial Public Sector (millions of 1994 soles)............................................................... 201 Table A9 Peru: Local Government Operations...................................................... ................................................................. 201 Table A10 Peru: Operations of the Non-Financial State Enterprises...............................................................................................202 Table All Peru: Operations of the Rest of the Central Government................................................................................................202 Table Al2a Peru: Functional Classification of Central Government Budget Expenditures (millions of nuevo soles)........................203 Table Al2b Peru: Functional Classification of Central Government Budget Expenditures (percent of GDP) ................................... 203 Table A13 Peru: Central Government Spending by M inistry or Institution...................................................................................... 204 Table A14 Peru: Social Spending by Type....................................................................................................................................... 204 Table A15 Peru: Composition of Social Expenditures ..................................................................................................................... 205 Table A16 Peru: PESEM : Projected Social Spending by Institution, 2002-2006............................................................................206 Table A17 Peru: Pledges at the 2002 Consultative Group for Peru.................................................................................................. 206 Table A18 Peru: Privatizations and Concessions, 2001-04 .............................................................................................................. 207 Table A19 Peru: Current Revenues by Awarded Concessions......................................................................................................... 207 Table A20 Peru: Comparative Performance in Social Indicators .................................................................................................... 208 Table A21 Peru: M ain Tax Breaks..........................................................-.-.......-............................................................................. 209 Table A22 Peru: Alternative M easures for Tax Reform by Sector or Region................................................................................... 209 Table A23a Latin America General Government's Tax Revenues, 1998........................................................................................... 210 Table A23b Latin American Economies: Outline of Fiscal Policy Rules........................................................................................... 210 Table A24 Peru: Elasticities of Social Spending to Total Spending of the Consolidated Public Sector in Peru (1997-2002)........... 211 Table A25 Peru: Degree of Transparency in the Publication of Institutional Information................................................................ 211 Table A26 Peru: Published Information of the Budget in 2002........................................................................................................211 Table A27 Peru: Composite Governance Indicators, International Comparisons, 1998................................................................... 212 Table A28 Volatility in Latin America............................................................................................................................................. 212 Table A29 Economic Distribution of General Government Revenue and Expenditure, Selected Countries .................................... 213 Table A30 Peru: Distribution of Social Investment by Different Social Programs and by Decile of Poverty Severity................... 213 Table A31a Peru: Average M onthly Per Capital Income of Households, 2000.................................................................................. 214 Table A31b Peru: Household Access to Social Programs by Income Per Capita Deciles, 2000......................................................... 214 Table A31c Peru: Household Access to Social Programs by Income Per Capita Deciles, 2000.........................................................214 Table A32 Peru: M is-targeting by Types of Food Assistance, 2000................................................................................................215 Table A33 Peru: M is-targeting by Types of Health Programs, 2000................................................................................................ 215 Table A34 Peru: M is-targeting by Types of Education Programs, 2000 .......................................................................................... 215 Table A35 Peru: Evolution of Household Access to Social Programs, 1998-2000........................................................................... 215 Table A36 Peru: Annual Private Expenditure on Education by Per Capita Income Decile, 2000..................................................... 216 Table A37 Peru: Annual Expenditure on Health by Per Capita Income Decile................................................................................ 216 Table A38 Peru: Definitions of the Consolidated Public Sector (CPS)............................................................................................ 217 Table A39 Peru: Central Government National and Regional Budget Composition, 2002.............................................................. 217 Table A40 Peru: M ain Revenue Sources for M unicipalites,2002..................................................................................................... 218 Table A41 Peru: Departmental Budget - Education and Health by Expenditure Category, 2002 ....................................................218 Table A42 Peru Indices on Corruption........................................................................................................................................... 219 Table A43 Peru: Corruption in the Judiciary, National and M unicipal Agencies............................................................................. 219 Table A44 Peru: Expenditure by Department, 2001......................................................................................................................... 220 Table A45 Peru: Expenditure Coefficients by Department, 2001..................................................................................................... 221 Table A46 Peru: Decentralization in LAC ....................................................................................................................................... 221 Table A47 Peru: Principal Revenue Sources of Subnational Governments...................................................................................... 222 Table A48 Peru: Summary of M ining Taxation Authority for Selected Taxes and Fees................................................................. 222 Table A49 Peru: M ining Fiscal M ethods and their Amenability to Fiscal Decentralization.............................................................223 Table A50 Total Transfers in Soles................................................................................................................................................223 'This detailed Appendix was prepared by Elaine Tinsley. 193  STATISTICAL APPENDIX Table Al. Combined Public Sector Operations Mlons of haevo Soles 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 1/ L Currant Revenues 4115 7485 11474 17025 21470 25288 29712 31261 30653 33077 32382 1 NWn-financlal expenditures 3682 7268 10915 16 824 21335 24136 27 795 29728 32956 34425 33 620 1. Current expenditures 2 932 5475 7903 11671 15 289 17813 20523 22614 25396 27610 27755 2. Capital expenditures 750 1794 3011 5153 6046 6323 7272 7115 7 560 6815 5864 L Others 33 57 -28 357 329 613 224 571 577 554 316 IV. Primary Balance 466 273 531 58 463 1785 2141 2104 -1726 -793 -922 V. interest 1042 2007 2908 3588 4064 3332 2844 3174 3729 4120 4122 VI. Overall Balance -577 -1734 -2377 -3031 -3 601 -1 587 .704 -1070 -5455 -4913 -5044 VilI. Not Financing 577 1734 2377 3031 3601 1567 704 1070 5455 4913 5044 1. Ekternal 954 1426 2141 2453 2878 1081 12 561 -313 2097 1980 2. Domestic -379 239 -64 -4454 -1428 -4792 -800 -255 4449 1389 1930 3. Privatttation 1 69 300 5032 2151 5279 1492 764 1318 1427 1134 1/ PreIrrinary Source: W Banco do Ia Necln, BCFP, Sunat, Aduanas, EsSalud, pubic benefilary socleties. local government and public institutions. Table A2. Combined Public Sector Operations Percent of GDP 1991 1992 1993 1994 1995 1998 1997 1998 1999 2000 2001 11 L Current Revenues 15.4 16.7 18.6 17.3 17.8 18.5 18.9 18.8 17.5 17.7 17.1 L Non-financlaiexpenditurea 13.8 16.2 15.8 17.1 17.7 17.6 17.7 17.9 18.9 18.4 17.7 1. Current expenditures 11.0 12.2 11.4 11.8 12.7 13.0 13.0 13.6 14.5 14.8 14.6 a Ospital expenditures 2.8 4.0 4.3 5.2 5.0 4.6 4.6 4.3 4.3 3.6 3.1 l Others 21 0.1 0.1 0.0 0.4 0.3 0.4 0.1 0.3 0.3 0.3 0.2 IV. Firrary Balance 1.7 0.6 0.8 0.6 0.4 1.3 1.4 1.3 -1.0 -0.4 -0.5 V. iterest 3.9 4.5 4.2 3.6 3.4 2.4 1.8 1.9 2.1 2.2 2.2 Vi. Overall Balance -2.2 -a9 -3.4 -3.1 -30 -1.1 -0.4 -0.6 -3.1 -2.6 -2.7 VU. Not Rnancing 2.2 3.9 3.4 3.1 3.0 1.1 0.4 0.6 3.1 2.6 2.7 1. E5temal 3.8 32 3.1 2.5 2.4 0.8 0.0 0.3 -0.2 1.1 1.0 2. Domestic -A 0.5 -0.1 -4.5 -1.2 -3.5 -0.5 -0.2 2.5 0.7 1.0 3. Privatization 0.0 0.2 0.4 5.1 1.8 3.9 0.9 0.5 0.8 0.8 0.6 GDP(MUlon sales) 26,686 44.953 69.262 98577 120,858 136,929 157,274 166,514 174,719 186,756 189,532 Source: Statistical Appendix, Table Al Table AS. Combined Public Sector Operations. In Mons 1994 Soles 1991 1992 1993 1994 1995 1996 1997 1998 1990 2000 2001 11 L Ourrent Revenues 13127 13761 14197 17025 19319 20399 22081 21661 20527 21348 20495 I I.WNn-fnanclal expenditures 11746 13363 13505 16824 19198 19470 20656 20599 2069 22218 21276 1. Ourrent expenditures 9352 10065 9779 11671 13758 14370 15252 15669 17007 17819 17567 2. Capital expenditures 2394 3298 3726 5153 5441 5101 5404 4930 5062 4398 3712 . Others 2 105 104 -35 357 296 495 167 395 387 358 200 IV. Prary Balance 1486 502 657 558 417 1424 1591 1458 -1156 -512 -583 V. Interest 3326 3689 3598 358 3657 2888 2114 2199 2497 2659 2609 VI. Overall Balance -1839 -3187 -2941 -3031 -3240 -1264 -523 -741 -3653 -3171 -3192 VII. Ht Financing 1839 3187 2941 3031 3240 1264 623 741 3653 3171 3192 1. Bternal 3045 2621 2649 2453 2590 872 9 389 -209 1354 1253 2. Domestic -1210 439 -80 -4454 -1285 -3868 -595 -177 2979 896 1222 3. Privatlzation 5 127 371 5032 1935 4258 1109 529 883 921 717 AverageCPlDeflator 31 54 81 100 111 124 135 144 149 155 158 Source: Statistical Appendix. Table Al 195 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A4. Central Government Operations Imlons of Nuevos Soles 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 11 2002 2( I. Current revenues 3193 6059 9424 14386 18319 21 522 25 001 26174 25334 27515 26733 28341 I. TAX FF.VeNES 2958 5416 8478 12866 16212 19036 22122 22995 21873 22 663 23383 23919 1. Icore tax 250 749 1406 2526 3462 4981 5710 5861 6072 5130 5630 6030 2. Property tax 177 201 254 85 30 3 3 1 11 0 3. Bcport tax 4. hport tax 336 650 1228 1700 2144 2308 2471 2891 2848 2913 2738 3015 5. Value-added tax 765 1702 3534 5954 7648 8578 10344 11040 11029 11996 11806 12462 -Dorrestic 437 911 1992 3563 4217 4723 5 929 6384 6470 7004 6866 7035 -rports 328 791 1543 2391 3429 3856 4415 4655 4559 4 993 4940 5 427 6. Excise tax 1 196 1743 1611 2267 2486 2761 3365 3427 3446 3421 3533 3417 -Fuel 781 995 991 1288 1493 1599 1931 2000 2092 2118 2318 2211 -Others 415 748 620 979 993 1162 1434 1426 1354 1303 1215 1206 7. Other tax revenues 312 442 528 834 1160 1268 1737 1569 1568 1873 2507 2412 8. Tax refund -78 -71 .84 -499 -715 -863 -1509 -1794 -2101 -2669 -2832 -3417 b. NON-TAX EVENUBS 236 643 946 1 520 2106 2485 2879 3179 3460 4851 3350 4342 II. Non-interestexpenditurea 2842 5839 8992 14270 18575 20729 23729 25379 27697 28989 28253 29426 1. Current expenditures 2319 4401 6448 9 929 13545 15680 17934 19756 21797 23757 23857 25004 a. Wages and salarles 911 1770 2659 3866 5098 5433 6397 6979 7774 8180 8389 8643 b. Goods and services 552 966 1413 2570 3928 5129 5319 6022 6210 7068 7068 7236 c. Transfers 856 1665 2375 3493 4518 5118 6217 6756 7813 8508 8400 9045 2. Capftal expenditures 523 1438 2544 4341 5031 5048 5795 5623 5900 5232 4396 4422 a. Capital fomiation 504 1 115 2150 3342 3955 3921 4448 4964 5652 4749 3901 4221 b. Transfers 4 122 45 4 109 16 0 0 0 254 c. Others 15 200 349 995 987 1 111 1347 659 248 229 494 201 Ill. Others 31 23 28 -58 324 274 578 192 533 539 530 291 603 IV. Primary balance 375 248 373 440 18 1371 1465 1327 -1825 -944 -1229 -603 II. Interest 1035 1 998 2888 3560 4023 3289 2789 3119 3674 4074 4062 3819 1. Domestic debt 86 91 61 107 90 76 116 176 238 543 466 3417 2. Ectemal debt 949 1907 2827 3453 3933 3213 2673 2944 3436 3532 3596 603 VI. Overall Balance - 660 -1 750 -2515 -3120 -4005 -1918 -1 324 -1792 -5499 -5 019 -5291 -4422 VIl. Not Rnancing 660 1750 2615 3120 4005 1918 1324 1792 5499 5019 5291 4422 1. Ektoemal 954 1426 2141 2453 2878 1077 12 561 -313 2097 1980 1407 2. Domestic -296 256 74 -4365 -1023 -4438 -180 467 4493 1494 2177 603 3. Privatization 1 69 300 5032 2151 5279 1492 764 1318 1427 1134 2412 1/ Prentnary 2/ lEstirrate 3/ Less payrrints to American itemational Group and Convanlo Per0-Alerrania. Source: MFY, Banco do la Nacidn, BCRP, Sunat, Custom, Ehpresa Naclonal de Conarcialliacidn de Iunas S.A. (Enc), Brpresa Conarcialzadora de Altentos S.A. (Ecasa) and Patropern. Wand World Bank staff estirates for 2002. 196 STATISTICAL APPENDIX Table AS. Central Government Operations Percent of GIP last 1992 1993 1994 1995 1996 1997 1998 1999 200 2001 11 202 21 L Current revenues 12.0 13.5 13.6 14.6 15.2 15.7 15.9 15.7 14.5 14.7 14.1 14.1 la. TAX FEVENIJES 11.1 12.0 12.2 13.1 13.4 13.9 14.1 13.8 12.5 12.1 12.3 11.9 1. hoorre tax 0.9 1.7 2.0 2.6 2.9 3.6 3.6 3.5 2.9 2.7 3.0 3.0 2. Property tax 0.7 0.4 0.4 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3. Export tax 4.iporttax 1.3 1.4 1.8 1.7 1.8 1.7 1.6 1.7 1.6 1.6 1.4 1.5 5. Value-added tax 2.9 3.8 5.1 6.0 6.3 6.3 6.6 6.6 6.3 6.4 6.2 6.2 Donestlc 1.6 2.0 2.9 3.6 3.5 3.4 3.8 3.8 3.7 3. 3.6 3.5 -inports 12 1.8 2.2 2.4 2.8 2.8 2. 2.8 2.6 2.7 2.6 2.7 6. Excise tax 4.5 3.9 2.3 2.3 2.1 2.0 2.1 2.1 2.0 1.8 1.9 1.7 Ral 2.9 2.2 1.4 1.3 1.2 1.2 12 12 1.2 1.1 12 1.1 -Others 1.6 1.7 0.9 1.0 0.8 0.8 0.9 0.9 0.8 0.7 0.6 0.6 7. Ohr tax revenues 12 1.0 0.8 0.8 1.0 0.9 1.1 0.9 0.9 1.0 1.3 1.2 8. Tax refund -0.3 -0.2 -0.1 .0.5 -0.6 -0.6 -1.0 -1.1 -1.2 -1.4 -1.5 -1.7 b. NO?+TAX IEVENUES 0.9 1.4 1.4 1.5 1.7 1.8 1.8 1.9 2.0 2.6 1.8 2.2 I. Non-interestexpenditures 10.6 13.0 13.0 14.5 15.4 15.1 15.1 15.2 15.9 15.5 14.9 14.6 1. Ourrent expenditures 8.7 9.8 9.3 10.1 11.2 11.5 11.4 11.9 12.5 12.7 12.6 12.4 a. Wages and salaries 3.4 3.9 3.8 3.9 4.2 4.0 4.1 4.2 4.4 4.4 4.4 4.3 b. Goods and services 2.1 2.1 2.0 2.6 3.3 3.7 3.4 3.6 3.6 3.8 3.7 3.6 c. Transfers 3.2 3.7 . 3.4 3.5 3.7 3.7 4.0 4.1 4.5 4.6 4.4 4.5 2. Capita expendliures 2.0 3.2 3.7 4.4 4.2 3.7 3.7 3.4 3.4 2.8 2.3 2.2 a. Capitalfonation 1.9 2.5 3.1 3.4 3.3 2.9 2.8 3.0 32 2.5 2.1 2.1 b. Transfers 0.0 0.3 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.1 c. Others 0.1 0.4 0.5 1.0 0.8 0.8 0.9 0.4 0.1 0.1 0.3 0.1 IIL Others 21 0.1 0.1 -0.1 0.3 0.2 0.4 0.1 0.3 0.3 0.3 0.2 0.3 IV. Primary balance 1.4 0.6 0.5 0.4 0.0 1.0 0.9 0.8 -1.0 -0.5 -0.6 -0.3 t Interest 3.9 4.4 42 3.6 3.3 2.4 1.8 1.9 2.1 2.2 2.1 1.9 1. Domestic debt 0.3 0.2 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.3 02 1.7 2.Ektemal debt 3.6 4.2 4.1 3.5 3.3 2.3 1.7 1.8 2.0 1.9 1.9 0.3 VL Overall Balance -2.5 -3.9 -3.6 -3.2 -3.3 -1.4 -0.8 -1.1 -3.1 -2.7 -2.8 -2.2 VIL Nt FnancIng 2.5 3.9 3.6 3.2 3.3 1.4 0.8 1.1 3.1 2.7 2.8 2.2 1. Bkternal 3.6 3.2 3.1 2.5 2.4 0.8 0.0 0.3 -0.2 1.1 1.0 0.7 2. Domestic -1.1 0.6 0.1 -4.4 -0.8 -3.2 -0.1 0.3 2.6 0.8 1.1 0.3 3. rIVatizatin 0.0 0.2 0.4 5.1 1.8 3.9 0.9 0.5 0.8 0.8 0.6 12 GDP (MEon soles) 28.686 44,953 69262 98.577 120.858 136,929 157,274 166,514 174,719 186.756 189,532 201,000 Source: Statistical Annex, Table A4, WB calculations. 197 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A. Central Govemrent Operations In Rllons 1994 Sales 1991 1992 1993 1994 1995 1996 1997 1998 1999 200 201 i 2002 2i 1. Cuwnt reees 10187 11139 11661 14386 18485 17361 18590 18136 16965 17758 16920 17620 la. TAX REVENUES 9435 9957 10491 12866 14588 15356 16440 15933 14648 14627 14799 14871 1. fnce tax 796 1377 1740 2526 3115 4018 4243 4061 3396 3311 3564 3749 2. Propertylax 665 370 315 85 27 2 2 1 7 0 3. Eport tax 4.Importtax 1070 1195 1520 1700 1929 1862 1836 2003 1907 1880 1733 1874 5. Va~e-addedax 2441 3128 4373 5954 6880 6920 7688 7649 736m 7742 7472 7748 - Domelic 1395 1674 2465 3563 3795 3810 4406 4424 4333 4620 4346 .4374 - Imports 1045 1454 1909 2391 3085 3110 3281 3226 3053 32 3127 3374 6. Exmse tax 3816 3205 1994 2267 2237 2228 2501 2374. 230B 2208 2236 2124 -FuGI 2493 1829 1227 1288 1343 1290 1435 1386 1401 1367 1467 1375 -Oems 1323 1375 767 979 893 937 1066 988 907 841 769 750 7. Oher tax revenes 996 813 654 834 1044 1023 1291 1087 1050 1209 1587 1500 . Tax 1efun -249 -131 -104 -499 -643 -696 -1121 -1243 -1407 -1723 -1792 -2124 Ib. NON-TAX REVENUES 752 1182 1170 1520 1897 2005 2140 2203 2317 3131 2120 2699 IL Non4ntrestexpnditu~B 9066 10734 11127 14270 16715 16721 17834 17585 18548 18710 17862 18296 1. Currmnlexpndtums 7398 8092 7978 9929 12188 12649 13328 13689 14597 15333 15100 15546 a. Wages and 8ala 2906 3254 3291 3866 4587 4383 4754 4835 5206 5280 5310 5374 b. Goods and services 1761 1776 1749 2570 3535 4137 3953 4172 4159 4562 4474 4499 c. Trnsers 2731 3061 2939 3493 4066 4129 4620 4681 5232 5491 5316 5623 2. Capitaexpendtures 1667 2643 3146 4341 4527 4072 4307 396 3951 3377 2782 2749 a. Capital formation 1609 2050 2661 3342 3569 3163 3306 3439 3785 3065 2469 2624 b. Trasfe 12 224 55 4 8 13 0 0 0 164 c. Oes 47 369 432 995 870 896 1001 456 166 148 313 125 Il. Ohers 2 75 51 -72 324 246 466 143 369 361 342 184 312 IV. Primry balnc 1196 455 462 440 16 1106 1089 920 -1222 -609 -778 -375 IL Interst 3302 3670 3573 3560 3620 2653 2073 2161 2460 2630 2571 2374 1. Dom~ec debt 275 166 75 107 81 62 86 122 159 350 295 2124 2. Extemal debt 3027 3507 3498 3453 3539 2592 1987 2040 2301 2279 2278 375 VL OveralI Balane -2106 -3218 -3111 -3120 -3604 -1547 -984 -1242 -3682 -3239 -3349 -2749 Vik Net Fiancig 2106 3218 3111 3120 3604 1547 984 1242 362 3239 3349 2749 1. Extemal 3045 2621 2649 2453 2590 869 9 389 -209 1354 1253 875 2. Domes0 c -944 471 91 -4365 -921 -3580 -134 323 009 964 1378 375 3. Pvalizati n 5 127 371 5032 1935 4258 1109 529 883 921 717 1500 CPI DeBator 31 54 81 100 111 124 135 144 149 156 158 181 Footnotes and cource: See Table A4, WB oakCaaom. 198 STATISTICAL APPENDIX Table A7a. Structure of Central Government's Fixed Capital Formation Milions of Nuevo Soles 1991 1992 1993 1994 1995 1996 1997 1996 1999 2000 ECONOMIC SETORS 384 885 1344 1942 2169 2222 2810 2434 3220 2563 1. Agriculture 233 373 544 589 580 510 615 809 1089 911 2. Transport and comnminations 116 380 568 904 1162 1096 1688 1194 1596 1132 3. Energy and rining 2 63 151 380 365 544 483 420 494 483 4. Idustry, comrce, tourdsm 3 3 11 20 20 19 22 11 21 16 6. Fishing 10 66 71 49 52 53 22 0 19 21 SOCIAL SETORS 59 122 489 738 994 751 684 939 1085 910 1. Educalion 34 41 166 368 509 356 377 416 608 496 2. Health 15 25 37 88 186 271 159 323 271 260 3. Housing and Construction 4 6 0 0 0 0 0 0 0 0 4. Employment and others 7 49 285 282 299 125 148 200 187 154 OTHERSECTORS2f 28 90 231 411 371 726 571 902 728 702 MULTISETORALPROGRAMS3 53 38 88 251 421 222 383 690 639 573 TOTAL 504 1115 2150 3342 3955 3921 4448 4964 5662 4749 1/ Prellrrnary 2/ bcludes: Mnistry of the President. Mnistries of Justice, Iterior, Brterlor Relations, Gconomy and Finance, WnisterD Pblco, Juredo Nacinal do Becciones and Instituto Naclonal do Plarnilicaci6n, armng others. 31 icludes projects wIth diverse sectoral scopes. The latest periods incorporate expenditures in the nicroroglons. Source: EF, WB staff estnates Table A7b. Structure of Central Government's Fixed Capital Formation Percent of Total 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 ECONOMIC SETORS 72.1 77.6 62.5 88.1 54.8 56.7 63.2 49.0 57.0 54.0 1. Agriculture 46.1 33.4 25.3 17.6 14.7 13.0 13.8 16.3 19.3 19.2 2. Transport and conrunicatons 23.1 32.3 26.4 27.0 29.1 28.0 37.9 24.1 28.2 23.8 3. Eergy and irtning 0.3 5.7 7.0 11.4 9.2 13.9 10.4 8.5 8.7 10.2 4. Industry, comrrerce. tourism 0.5 0.3 0.5 0.6 0.5 0.5 0.5 0.2 0.4 0.3 5. Fishing 2.0 5.9 3.3 1.5 1.3 1.3 0.5 0.0 0.3 0.4 SOCIAL SECTORS 11.8 10.9 22.7 22.1 25.1 19.2 15A 18.9 18.8 19.2 1. Education 6.7 3.7 7.7 11.0 12.9 9.1 8.5 8.4 10.8 10.4 2. Hlealth 3.0 2.2 1.7 2.6 4.7 6.9 3.6 6.5 4.8 5.5 3. Housing and Construction 0.8 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4. BRployrnont and others 1.4 4.4 13.3 8.4 7.5 3.2 3.3 4.0 3.3 3.2 OTHER SECTORS 21 5.6 8.1 10.7 12.3 9.4 18.5 12.8 18.2 12.9 14.8 MULTISETORALPROGRAMS31 10.5 3.4 4.0 7.5 10.7 5.7 8.6 13.9 113 12.1 TOTAL 100 100 100 100 100 100 100 100 100 100 1/Prerinary 2/ bIcludes: nistry of the President, unistries of Justice, iterior, Exterior Relations. Economy and Finance, Mnisterlo POblco, Jurado Nacional do Becciones and Istituto Naclonal de Planflcacl6n, arng others. 31 ncludes projects w lh diverse sectoral scopes. The latest periods incorporate expenditures In the Iricroreglons. Source: M, WB staff estirmates 199 PERU: RESTORING FISCAL DIsCIPLINE FOR POVERTY REDUCTION Table A8a. Operations of the Non-Financial Public Sector Millons of Nuevo Sales 1991 1992 1993 1994 1995 1196 1997 1998 1999 200 0111 . PrfrnryEaance 480 548 1038 972 461 2113 3210 1948 -1675 -1706 -5983 1. Pimry Balance of te C 375 248 373 440 18 1371 1465 1327 -1825 -944 -1229 a. Curtrevanues 3193 6 059 9424 14388 18319 21522 25001 26174 25334 27515 26733 L Tax rnus 2958 5416 8 478 12 868 18212 19038 22122 22995 21 873 22 83 23383 1L. Non-tax rvanues 236 643 946 1 520 2108 2 485 2879 3179. 3460 4851 3350 b. Non-InancLalexp~ntures 2842 5839 8992 14270 18575 20729 23729 25379 27697 2989 28253 i. Curnent 2319 4401 6448 9 M 13545 15 680 17 934 19756 21797 23 757 23857 ICapma 523 1438 2544 4341 5031 5048 5795 5823 5900 5232 4398 c. Othss2/ 23 28 -58 324 274 578 192 533 539 530 291 2. Primary13nceoftheMst 105 299 865 532 443 743 1745 618 150 -792 8s a. Rest of Cmntral Goavmt 75 57 97 107 371 409 679 659 233 109 143 bL. Ocal Govemmens 16 -32 61 11 74 -15 -3 117 -134 42 154 c. Statenterpdes 14 273 507 414 -2 349 1003 -158 51 -913 377 ILinturest 1225 2302 3159 3727 4261 3507 2987 3275 3107 4235 4231 a.Jtemnaldobt 956 1919 2844 3486 3946 3220 2697 2972 3469 3582 3688 b. Domes det 269 383 314 241 314 267 259 32 338 654 564 l.Overallaence(MI) -745 -1755 -2121 -275B -3800 -1394 253 -1329 -482 -8942 -4623 IV. NtFInancing 745 1755 2121 2756 3800 1394 -253 1329 5462 5942 4823 a. Extema Net FInandng 1108 1447 2107 2383 3048 1092 - 591 642 -111 2 280 1755 (M oof US$ 1372 1052 13 1m 1361 443 29 224 -21 654 495 i VIsbsmnw~ Y 874 320 1388 49 62 382 1566 657 812 1299 1318 . Amord3mc1s3 1032 770 931 941 a2 842 735 799 883 635 778 EL OhrY 1531 1502 Em 1535 1545 904 -10 365 50 -10 -43 b. NetomestcfRnancng -33 239 - 28 -4 660 -1 39B -4976 -1155 -77 4274 2235 135 r. PrvaRlaonFRerenue 1 69 300 5032 2151 5279 1492 784 1318 1427 1134 11 Prelrriary 2/ Less paymnt to Arnerican lntemalanal Gmu and to Convaro P~r~.Alemanla. Y Includs excepöal and short-tem Anan~ing. S~nca NEF, Banco de la Nadön. BCRP, Sut Customn, EsSalud, pulic bensfiday sodelhs, local goemment, state mntpdses and pubo giitulans Table ABb. Operations of the Non-Financial Public Sector Percentof GDP 1991 1992 1993 1994 1996 1998 1997 1998 1999 20W 2012 1. Prlmary Balmnce 1.8 1.2 1.5 1.0 0.4 1.5 2.0 1.2 -1.0 -0.9 -0.3 1. PrdmwyBalance of IM CG 1.4 0.6 0.5 0.4 0.0 1.0 0.9 0.8 -1.0 -0.5 -0. a. Current revenuas 12.0 13.5 13.6 14.6 15.2 15.7 15.9 15.7 14.5 14.7 14.1 L Taxvnum s 11.1 12.0 12.2 131 13.4 13.9 14,1 138 12.5 12.1 12.3 l. Non-tax revnus 0.9 1.4 1.4 1.5 1.7 1.8 1.8 1.9 2.0 2.6 1.8 b. Non-financiapendture 10.6 13.0 13.0 14.5 15.4 15.1 15.1 15.2 15.9 15.5 14.9 L Cunent 8.7 9.8 9.3 10.1 11.2 11.5 11.4 11.9 12.5 12.7 12.6 l]. Capita 2.0 3.2 3.7 4.4 4.2 3.7 3.7 3.4 3.4 2.8 2.3 c. Ohrs 2 0.1 ai .1 0.3 0.2 0.4 0.1 0.3 0.3 0.3 0.2 2. PrmayhBan of the Rest 0.4 0.7 1.0 0.5 0.4 0.5 1.1 0.4 0.1 .0.4 0.3 a. Rs of etra Goemmett. 0.3 a 0.1 0.1 a3 0.3 0.4 0.4 0.1 0.1 0.1 b Local Govennents 0.1 -0.1 0.1 0.0 0.1 0.0 0.0 . 0.1 -0.1 0.0 .1 c. Stata Entstptse 0.1 0.6 0.7 04 0.0 0.3 0.7 -0.1 0.0 -0.5 0.2 [1. n1.299 4.6 5.1 4.6 3.8 3.5 2.6 1.9 2.0 2.2 2.3 2.2 a. Exeral deA 3.6 4.3 4.1 3.5 33 2.4 1.7 1.8 2.0 1.9 1.9 b. Domest dbt 1.0 0.9 0.5 0.2 0.3 0.2 0.2 0.2 0.2 0.4 0.3 iL. OveraI~Baanc (l -2.8 -49 -1 -2.8 -3.1 -1.0 0.2 0.8 -3.1 -3.2 .2.5 fV. Nt FInandng 2.8 3.9 3.1 2.3 3.1 1.0 -0.2 0.8 .1 3.2 2.5 a.ExtemaNetFInancIng 4.1 3.2 3.0 2.4 2.5 0.8 -0.4 0.4 -1 1.2 0.9 . visu mnls y 3.3 0.7 2.0 05 0.5 0.3 1.0 0.4 0.5 0.7 0.7 d. Arnrdaons 3.9 1.7 1.3 1.0 0.7 0.6 0.5 0.5 0.5 0.3 0.4 . Ohrs Sa 5.7 3.3 0.9 1.6 1.3 0.7 -0.7 0.2 0.0 0.0 0.0 b. Ntomesc~ nancng -1.4 0.5 -0.4 -4.7 -1.2 -6 -0.7 0.0 2.4 1.2 1.0 . Prfvatzation ROnes 0.0 0.2 0.4 5.1 1.8 &9 0.9 0.5 0,8 0.8 0.6 GDP (MlHon sales) 26,688 44.953 69,282 9,577 120,858 138,929 157.274 166.514 174.719 18,758 189.532 Footnotes and souroe: See TabUe A8a. 200 STATISTICAL APPENDIX Table A~c. Operations of the Nor-nancial Public Setor In MUlons 1994 Sle 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 21 1. Primay Blanoe 1531 1004 124 972 415 1705 2386 1348 -1122 -1101 -375 1. PrimryBancftheCG 1196 455 462 440 16 1106 1089 20 -1222 -609 -778 a. Cumn rvenue 10187 11139 11661 14386 16485 17361 16660 18136 16966 17758 16920 1. Tax rae 9436 9967 10491 12(366 14668 1536 16440 15933 14648 14627 14799 . N-tx re u 752 1182 1170 1520 197 2005 2140 2203 2317 3131 2120 b. NWniane edtu0s 9066 10734 11127 14270 16715 16721 17834 17585 18548 16710 17882 1. Cuent 7396 802 7978 92 12188 12649 1332B 13689 14597 15333 15100 D. C~*al 1667 2643 3148 4341 4527 4072 4307 3696 3961 3377 2782 cOher2/ 75 51 -72 . 24 246 466 143 389 361 342 184 2.PrimayalacetthePds 335 549 .823 532 399 99 1297 420 100 -492 403 a Rs of Ce Goemmen 238 106 120 107 334 330 04 457 156 70 91 b.Lcä Gomer~s 52 -68 75 11 67 -12 -2 81 -90 27 97 c.StaeE ntpdses 45 502 627 414 -2 281 794. -110 34 -59 239 IL r~erest 3907 4232 3908 3727 3834 29 2197 2269 2549 2734 2678 ELa Jtema delt 3051 352 3520 3486 3%51 2698 2006 2069 2 2312 2321 , Ocmesc~dett 667 704 3e9 241 263 232 193 209 226 422 357 IILOv~elIBaance(I)4 . -2376 .27 -2624 -2756 -3419 -1124 188 -921 -3871 -3835 -3053 IV. Nu Ficig 2376 3227 2824 2786 3419 1124 -188 921 3671 3830 3053 a Exteal Net1 nanng 3530 2661 2607 2363 2742 881 -439 445 -74 1471 1111 L isbu~semres 2787 589 1718 496 66 308 1164 455 544 . 38 834 B. Amaf o y 3292 1416 1152 941 749 679 546 554 691 410 492 rd. rs3/ 4883 2761 749 153 1390 729 -766 253 33 -7 -27 b Net Domesc Fnaning -1158 440 -354 -46 -1258 -4014 -88 -53 262 1443 1225 . Piva~aon Reuejs 5 127 371 5 1936 4258 1109 629 183 21 717 Avg CIDe or . 31 54 81 100 111 124 135 144 149 155 158 Fcolnotes and sowe See Täle A8a. Table A9. Local Govemment Operations Mlons of Nuevo Soles - 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 L Current evenues 495 825 1 338 1713 2170 2539 3144 3454 3279 3549 3754 I. N0n-financial expenditures 488 886 1307 1735 2151 2 588 3175 3374 3451 3784 3 832 1. Current expenditures 317 582 920 1 086 1 285 1 483 1 840 2022 2079 2239 2289 2. Capital expendtures 172 303 387 649 866 1106 1335 1352 1372 1545 1543 K. Others 9 29 30 33 55 35 28 38 38 278 233 IV. Prinnary Balance 16 -32 61 11 74 -15 -3 117 -134 42 154 V. Interest 2 9 19 28 40 42 53 53 54 45 59 VL Overall Balance 14 -40 42 -17 34 -56 -56 64 -188 -2 95 Source: Local governmnts and Contadurfa Påblica de la NacCón. 201 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table AIO. Operations of the Non+Inancial State Enterprises MMons of Nuevo Soles 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 1. Current levenues 4083 6105 6 917 11272 12486 12304 11388 9524 10700 12914 11581 1. Fttropero 1691 2716 3867 5 154 5861 6187 5216 4285 5370 7724 6797 2. BectroperO 238 356 618 869 1108 1390 1386 1450 1265 1214 1221 3. Flagionales de Eectricidad na ne na na 736 825 975 1197 1454 1568 1339 4. Sedapal 66 133 208 356 418 507 621 581 626 643 665 5. Controrrin 311 515 738 1018 1336 1440 1554 603 546 258 202 6.Others 1778 2386 3486 3875 3027 1953 1635 1408 1438 1507 1358 U. Non-financial Current Expenditures 3833 5503 7826 9985 11575 10932 9450 836 9047 12588 10495 1. Flropefo 1694 2685 3754 4878 5774 5828 4782 4212 5344 7823 6343 2. Bectropen 124 254 391 591 755 899 1 019 1126 675 782 858 3. Flegionales de Bactricklad na na na na 608 619 652 893 1122 1509 958 4. Sedapal 55 116 136 248 300 331 391 379 357 447 400 5. Centrorrin 333 456 684 899 1205 1323 1169 606 365 507 203 6. Otros 1628 1992 2861 3371 2934 1932 1436 1092 1184 1520 1732 I. Gastoa de capital 280 459 684 954 1115 1367 1 571 205 1711 1312 785 1. FPtroperd 35 48 124 150 449 182 23 41 29 32 40 2. Bectroperd 95 106 127 157 77 450 101 528 533 453 150 3. Rgionales de ectricidad na na na na 96 114 229 204 428 367 244 4. Sodapal 24 38 91 124 88 170 371 426 453 264 196 . entrotn 21 7 28 23 39 61 65 46 28 42 8 6. Others 105 260 315 500 366 389 783 779 241 154 147 IV. Others 44 129 100 82 202 344 703 650 108 73 75 V. Primary Balance (Q-IAV) 14 273 507 414 -2 349 1069 -156 51 *913 377 VL Interest 182 295 251 139 197 175 112 101 78 115 107 VII. Overall Balance -168 -22 256 275 -199 173 957 -259 - 27 .1028 270 1. Ftropero -87 -59 -53 97 -464 90 385 7 -18 - 154 398 2. Bectroperd - 1 -36 99 102 263 22 241 -220 S6 -21 214 3. Regiorales do Eectriclidad na na na na 26 88 88 182 -96 -333 182 4. Sedapal -8 -9 -20 -19 48 37 -56 -64 -167 -100 27 5. Centrornin -57 35 -7 80 69 41 308 -60 144 -298 -15 6. Others -15 46 236 14 -141 -103 -9 -103 44 -114 -536 na not avalable Source: State Enterprises, Fando Iacional de Financariento dola ActIvidad Brpresaral del Estado (FONAF). Table All. Operations of the Rest of the Central Government I Mions of Nuevo Soles 1991 1992 1993 1994 1995 1996 1997 1996 1999 2000 2001 L Currentrevenues 707 1147 1556 2314 3070 3731 4715 5169 8090 6657 6667 1. ContrbutionstoEsSakudandONP2/ 539 933 1335 1748 2194 2378 2804 2795 2945 3112 3166 2. Others 169 214 222 665 875 1354 1911 2374 3 145 3446 3501 IL Non-financial expenditures 633 1090 1461 2 207 2007 3324 4040 4510 5857 6449 6524 1. Cirrent 577 1037 1381 2045 2549 3156 3898 4370 5569 6157 6396 2. Capitai 56 53 80 162 258 169 142 140 287 291 129 Ill.Others 0 0 0 0 109 1 4 0 0 0 0 IV. Prirnary Blance (-411) 75 57 97 107 371 409 679 659 233 109 143 V. Interest 5 0 0 0 0 1 2 1 1 1 1 VI. Overall Blance 70 57 97 106 371 407 877 658 282 108 141 1. EsSalud 3 65 11 98 211 179 338 531 414 142 -100 92 2. OP 3/ 0 0 0 -126 142 -2 -34 10 -39 65 19 3. FOR 0 0 0 0 0 0 112 72 26 97 53 4. Fonahpu 0 0 0 0 0 0 0 42 42 103 -53 5. Others 4 46 -1 22 50 72 67 119 62 -57 30 1/ bcludes: EsSajd, Oficina do Norretzacl6n Pevisional (CIT), pubic beneficiary societies, Fondo Consoldado do Peservas (F, el Fonda Nacional de Aorro FMbico (Fonahpu), regulated organizatlons and registries. 21 ONPreceives the revenues corresponding to the colections of the Sisterna Necional do Fensiones (SNP). 31 Up to 1993, it shows the operations of PSS In EsSaud, consoldating the provisional pension and health regines. Source: IEF, Sunat, Aduanas, EsSalud, public beneficiary societies and regulated organizations. 202 I STATISTICAL APPENDIX Table A12a. Functional Classification of Central Government Budget Expenditures Millions of Nuevos Soles 1997 1998 1999 2000 2001 2002 1/ Planning and Administration 8,847 6,601 8.246 9,253 8,701 9,514 Pensions and Social Assistance 5,593 5,885 7,149 7,717 7,931 8,196 Pensions 4.041 4,675 5.330 6.127 6.294 6,480 Social Assistance 1,552 1,310 1,819 1,590 1,637 1,716 Education 4,172 4,529 5.230 5,403 5,323 6,067 Pre-primary 303 382 455 448 415 612 Primary 1,518 1,598 1,910 1,957 1,818 2,048 Secondary 1.130 1,134 1,308 1,400 1,424 1.676 Tertiary 743 836 970 955 965 1,039 Others 478 579 587 642 700 692 Defense and National Security 4.523 4.858 4,958 8,366 4,709 4,189 Health and Water 2,250 2,465 2,709 2,914 3,052 3,328 Water and Sewage 283 271 269 333 207 159 Healthcare 1,975 2,194 2,440 2,581 2,845 3,169 TransportatIon 1,709 1,532 1,752 1,383 1,286 1.318 Agriculture 968 1,205 1.289 1.255 1.049 961 Justice 556 621 678 703 733 862 Energy and Natural Resources 269 311 338 266 257 297 Other 21 691 808 1,008 1,109 1,034 1,039 TOTAL 29,586 28,815 33,357 35,369 34,075 35,772 Memorandum /tems: Total Social Expenditure+ Pensions 14,937 15,451 17,667 18.835 18,972 20.286 Total Social Expenditure 10.896 10,876 12,337 12,708 12,678 13,806 Budget social expenditure 31 7,982 8,304 9,758 9,907 10,012 11.111 Essalud 4/ 1,735 2,025 2,489 2,801 2,666 2,695 FonavI 51 1,179 547 90 Exchange Rate 3.49 3.43 3.54 1/Budget 21Includes foreign relations, legislative, housing and urban development, Industry, commerce and services, fishing, communlcations, and employment 31 Includes education, health, and sanitation and other social assistance. 4/Total expenditure by Essalud, the public health Insurance administration 5/ Includes not operations of the national housing fund (Fonavi)-4oan disbursements and amortizations received Source: MEF, IMF Table A12b. Functional Classification of Central Government Budget Expenditures Percent of GDP 1997 1998 1999 2000 2001 2002/1 Planning and Administration 5.6 4.0 4.7 5.0 4.6 4.7 Pensions and Social Assistance 3.6 3.5 4.1 4.1 4.2 4.1 Pensions 2.6 2.7 3.1 3.3 3.3 3.2 Soclal Assistance 1.0 0.8 1.0 0.9 0.9 0.9 Education 2.7 2.7 3.0 2.9 2.8 3.0 Pre-primary 0.2 0.2 0.3 0.2 0.2 0.3 Primary 1.0 1.0 1.1 1.0 1.0 1.0 Secondary 0.7 0.7 0.7 0.7 0.8 0.8 Tertiary 0.5 0.5 0.6 0.5 0.5 0.5 Others 0.3 0.3 0.3 0.3 0.4 0.3 Defense and National Security 2.9 2.9 2.8 2.9 2.5 2.1 Health and Water 1.4 1.5 1.6 1.6 1.6 1.7 Water and Sewage 02 0.2 0.2 0.2 0.1 0.1 Healthcare 1.3 1.3 1.4 1.4 1.5 1.6 Transportation 1.1 0.9 1.0 0.7 0.7 0.7 Agriculture 0.6 0.7 0.7 0.7 0.6 0.5 Justice 0.4 0.4 . 0.4 0.4 0.4 0.4 Energy and Natural Resources 0.2 0.2 0.2 0.1 0.1 0.1 Other 0.4 0.6 0.6 0.6 0.5 0.5 TOTAL 18.8 17.3 19.1 18.9 18.0 17.8 Memorandum Iferns: Total Social Expenditure 6.9 6.5 7.1 6.8 6.7 6.9 Budget social expenditure 5.1 5.0 5.6 5.3 5.3 5.5 Essalud 1.1 1.2 1.4 1.5 1.4 1.3 Fonavi 0.7 0.3 0.1 Footnotes and source: See Table Al 2a 203 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A13. Central Government Spending by Minisby or Insitution Mions of tWevos Sales 1999 2000 2001 Sectors Budgeted Actual Diff% Budgeted Actual 01f % Budgeted Actual 01ff% 01 residency of the Oouncl of Mnisters 268 292 8.8 360 399 10.9 368 279 -24.2 04 JudicalRPwer 413 454 9.9 459 477 38 471 482 2.4 06 Justice 307 269 -12.7 310 392 26.6 381 382 -4.8 07 hiterior Z424 2,891 19.3 2776 3,029 9.1 2.931 2,888 -1.5 08 l5terior Felations 293 346 181 341 382 120 366 374 2.3 09 5conorry and Fance 10,382 10.009 -3.4 12,126 12.533 3.4 13,382 12,741 -4.8 10 Education 2.713 3,040 12.1 2534 2,625 3.6 2478 2.154 -13.1 11 Health 1,623 1,664 2.5 1,828 1,728 -5.5 1,859 1,939 4.3 12LaborandSocialAssistance 126 51 -59.3 25 50 98.9 36 38 5.4 13 Agriculture 648 827 27.6 749 810 8.1 664 787 18.5 14 hIdustry, Tourism Integration and hternational offrercial Negotiations 79 81 2.7 93 95 1.4 103 85 -16.8 15 Transport, Orrrnunications, Housing and Construction 1,817 1,793 -1.4 1,599 1,348 -15.7 1,496 1.196 -20.1 16 energy andm ng 490 325 -33.7 359 259 -27.9 281 254 -9.7 18 Fisheries 90 150 68.6 83 97 16.6 102 110 8.1 19 Generalorptroller 52 49 -6.0 52 48 .7.4 48 49 3.3 20OfficeotheAcole'sOrfbudsperson 12 19 52.1 16 22 33.9 17 21 25.9 21 ltlanafCouncloftheIWgIstrate 3 3 -6.8 3 4 15.9 5 7 44.8 22 Mnistry of the Rifc 156 158 1.2 160 152 -4.8 157 179 13.8 24 ConstitutionalTrbiunal 7 7 3.2 8 8 -0.9 9 8 -2.9 25 Mdstry of the Residency 6,384 6,698 4.9 5.961 6.122 27 6,075 5,802 -4.5 26 Defense 2,773 3,398 22.5 3,228 3,684 14.1 3,486 3,148 -9.7 28 COngress 247 238 -4.4 282 288 2.0 299 304 1.9 31 National Bection Body 12 15 28.0 10 26 157.0 11 41 272.9 32 National Office of the 1lectoral Rocess 9 34 290.0 12 141 1060.3 14 281 1857.1 33 Ntional Pegistry of identification and Ov! State 112 72 -36.0 75 113 49.5 87 79 -9.2 34 Advancement of Worren and HRran Devpent 524 474 -9.5 597 539 -9.8 588 46 -20.7 TOTAL 31,947 33,357 44 34,046 35,388 3.9 35,712 34,075 -46 Source: MEF. SlAFfor 2001 Table A14. Social Spending by Type Mlons of Wtevos Soles 1991 1992 1993 1994 1995 1998 1997 1998 1999 2100 Education 578 1,037 1,944 3,310 4,255 4,287 5,011 5,300 4,827 5,657 Social Security 403 827 1,524 1,933 3,161 3,430 4,712 5,300 4,603 5,305 Nftritin 44 78 155 359 677 587 817 843 776 964 Health 264 459 634 1,037 1,547 1.814 2,200 2372 2.215 2.143 Jlustice 37 70 170 194 194 374 547 498 586 622 Other Sociali vestrant 90 243 815 1,111 1,281 1,480 1,682 2.144 2132 1,783 Total 1,416 2,712 5.242 7.944 11,116 11.971 14,968 16,457 15,139 16,474 Rercent of GDP 5.3% 6.0% 7.6% 8.1% 9.2% 8.7% 9.5% 9.9% 8.7% 8.8% Source: Fvf, notes this table has ninor inconsistencies with the other tables 204 STATISTICAL APPENDIX Table A15. Composition of Social Expenditures (Millions Nuevo Soles) 1999 2000 2001 it 2O0z2Z 1. UNIVERSAL SOCIAL EXPENDITURE 6821 7207 7448 8439 EDUCATION 4552 4812 5042 5797 HEALTH 2268 2395 2405 2643 I1. EXTREME POVERTY PROGRAMS 3257 3043 2979 3531 FONCODES 758 469 350 537 Apoyo Social 232 184 19 Strengthening Social Programs 42 56 32 38 Social investment/ Productive Development 483 228 207 179 Programa de Emergencia Social Productive na na 92 320 EDUCATION 377 315 163 355 Improving primary education 260 219 102 55 Improving education of other levels 116 96 61 271 Literacy 29 HEALTH 126 136 205 291 Lucha contra epidemias 8 39 42 43 Family Planning 32 11 20 9 Seguro Escolar/Matemo Infantil (Prog. Apoyo Reforms Sector Salud -SALUD) 87 86 143 239 AGRICULTURE 317 293 416 276 Agricultural support In Emergency Zones 62 74 257 40 Pronamachs 191 185 143 219 Land Titling 64 34 15 17 PRONAA 237 258 221 216 Food Assistance/Support In Emergency Cases 211 233 201 195 Proyectos de Asistencla Allmentaria 26 25 20 21 PROG.APOYO REPOBLACION Y DESARROLLO ZONA DE EMERG. 22 29 21 17 Programs de Apoyo a la Repoblacl6n 22 29 21 17 ECONOMY AND FINANCE 338 350 357 363 Vaso de Leche (Goblemos Locales) 305 325 332 343 Apoyo Social - PL 480 33 25 25 20 ENERGY AND MINING 77 90 108 Ill Distribution infrastructure for energy 77 96 108 111 TRANSPORT AND COMMUNICATIONS 190 112 59 160 Rural roads 190 112 59 160 MIPRE 102 336 312 329 Prog.Naclonal de Agua Potable y Alcantarlilado 92 63 89 102 Targeted Investment to reduce extreme poverty 10 8 14 20 Prog. Equlpamlento BAsico Municipal na 256 209 207 Water for All na 9 COOPOP 8 6 12 9 Multisectoral Projects 8 6 12 9 INIEDPINFES 77 50 31 50 Education Infrastructure 77 50 31 50 COMISION DE FORMALIZACION DE LA PROP. INFORMAL 86 76 49 61 PROMUDEH 70 84 68 42 Literacy/Educational development 70 84 68 42 BASIC SOCIAL EXPENDITURE 472 433 606 713 Education 39 34 33 59 Health 305 264 255 316 Instituto Naclonal de Salud 83. 93 268 300 Others 45 42 50 38 Ill NON-TARGETED SOCIAL PROGRAMS 9,486 10,526 10,625 10,906 ESSALUD 2489 2801 2666 2695 FonavI 90 Tertiary and Other education 1557 1597 1665 1731 Pensions 5330 6127 6294 6480 TOTAL 19544 20775 21051 22877 Memo Items: Total budgeted social expenditure as (% of GDP) 31 6.7% 6.3% 6.4% 6.8% Total social expenditure and pensions as (% of GDP) 11.2% 11.1% 11.1% 11.4% Total poverty reduction programs (% of GDP) 5.8% 5.5% 5.5% 8.0% Poverty reduction (% of total budgeted social expenditure) 86.6% 86.5% 86.2% 87.4% Non-targeted social programs (% of total social expenditure and pension) 48.4% 50.7% 50.5% 47.7% 1/ Preliminary 2/ Estimated 31 Budgeted social expenditure excludes ESSALUD, Fonavi, and pensions Source: MEF 205 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A16. PESEM: Projected Social Spending by Institution, 2002-2006 2002 2003 2004 2005 2006 Nom Inal Million NS Social Sectors 11898 12554 13224 .14307 15260 Education 6068 6662 7457 8232 9062 Justice 356 386 416 452 491 Haafth 3224 3377 3580 3776 3978 resident 1788 1653 1223 1256 1093 Worrn and Imen Developrnent 462 476 548 592 636 Defense and hIterior 5340 5573 5573 5720 5881 Total 35991 38034 39026 40861 42781 Percent of Budget Social Sectors 33.1 33.0 33.9 35.0 35.7 Education 16.9 17.5 19.1 20.1 21.2 Justice 1.0 1.0 1.1 1.1 1.1 HeaRh 9.0 8.9 92 0.2 9.3 Resident 5.0 4.3 3.1 3.1 2.6 Woman and tman Development 1.3 1.3 1.4 1.4 1.5 Defense and Iterior 14.9 14.7 14.3 14.0 13.7 Total 100.0 100.0 100.0 100.0 100.0 Percent of GDP Social Sectors 5.9 5.8 5.7 5.8 5.7 Education 3.0 3.1 3.2 3.3 3.4 Justice 0.2 0.2 0.2 0.2 0.2 Health 1.6 1.6 1.5 1.5 1.5 President 0.9 0.8 0.5 0.5 0.4 Woren and uman Development 0.2 0.2 0.2 0.2 0.2 Defense and hIterior 2.7 2.6 2.4 2.3 2.2 Total 17.9 17.6 16.8 16.4 16.1 GDP Nominal 201000 216300 232523 248799 266215 Source: KEF and Word Bank staff estimates, based on nufl-annual sectors programs (PES4 Table A17. Pledges at the 2002 Consultative Group for Peru Country Pledges Concesslonal DebtSwaps Loans Total Spain 212.4 9.0 - 271.0 Belgium - 20.5 03nada - . . Denrrark - . Fiand - - - 1.4 France . . - Gernany 22.5 36.1 italy 14.0 127.0 - - Japan - - Luxermburg - Kbrw ay --- Sweden - . . Switzerland . . The Netherlands . . . kited Khgdom Lited States . . . TOTAL 226.4 156.5 - 320.0 Institution IA - - 400.0 Work Food Rog - . . Eutropean Comislon - - - 90.0 WB- 120.0 European Ivest Bank - 20.0 - FAD 25.0 - CAF - - 321.0 - TOTAL - * 886.0 90.0 GRAND TOTAL 226.4 158.5 886.0 419.0 Source: ID 206 STATISTICAL APPENDIX Table A18. Privatizations and Concesslons, 2001-04 Mions of US$ 2001 2002 2003 2004 Sale of participatory stocks 145 100 0 0 Blinor stocks (36%) 80 Ralapasa stocks (32%) 88 Ibcaycruz stocks (25%) 18 Bevensa stocks (38%) 39 Errp. Bectrica lura stocks (40%) 20 Electric Distributors 0 220 160 0 EEeRR Elctricity (JORBSA) 220 Other distrbutors (100%): Puno. Oriente, Sur, etc 160 Bectric Generators 220 0 350 300 Corplete M4antaro (CH 1008 IAN): Concession (up front) 300 BectroAndes (CH183 M) 220 Egasa (CT 154 MN and CH 165 MN) 200 Egesur (CT28.6 Mr and CH 35 Mm) 30 Egerrsa (CH70 1MW) 60 San Gaban (CH 110) 60 Bectric Transmision: Becen and Besur 0 330 0 0 Becen (LT 3000 krr 240 Besur (LT 880 kmr) 90 Petroperu 0 0 35 0 Talara refinery (62 TBD): Concession (up front) 15 Onchan ref Inery (12 TDB) and qultos (10.5 TBD) 20 Concessions 10 0 55 0 Telephone: Rxed Unes 10 Regional ports and from Callao 45 Telephonla Mltitmedios (other bands) 10 Sale of Buildings 0 0 45 0 Eariler sales and other privatisations 85 50 55 s0 TOTAL 460 700 700 350 Marno 1: Revenuesof the Public Tressury and Fonahpu A. Public Treasury 328 600 316 350 Privatization revenues 90 50 201 50 Deducted for the assurption of debt 238 550 115 300 B. Fonahpu 127 100 0 0 CL Revenue FEF 5 0 384 0 TOTAL 480 700 700 350 Memo 2 Deduction for asmming debt bcaycruz (by Mnero Peru) 18 Eectro Andes (by Centrorrin) 220 EE RR Bectricidad: JORBSA (by Bectroperu) 220 Becen and Besur (by Bectroperu) 330 Corrplejo Mntaro (by Bectroperu) 300 Egasa and Egesur (by Bectroperu) 80 Cochen and 4ultos Refinerles (by Bectroperu) 35 TOTAL 21 238 550 115 300 Notes: CT= Central Tdrrrica, CH= Central HIdreca, Mm = Megawatts, LT= Transrrission the, TBD. thous. barrels a i 1/Of this total. $18 rrIn from ocaycuz does not include resources from Fonahpu. 21 The deductions for debt assured fromand capitalization of Bectroperu are for a total of $930 nin. Source: Copri, August 2001 Table A19. Current Revenues by Awarded Concessions Milons of JSS 2001 2002 2003 2004 Jorge Chavez airport 1/ 0 35 40 50 Canisea 0 0 0 90 Conplejo Mntaro (15%) 2 0 0 0 40 Prts (15%) 0 0 0 15 TOTAL 0 35 40 195 Governrent 0 23 40 155 Enterprises 0 12 0 0 Fonahpu 0 0 0 40 TOTAL 0 35 40 195 1/ In 2002, $12 rfn was destined to Copac and $23 rrn to the government 2) Revenues of Fonahpu Source: Copr, August 2001 207 PERU: RESTORING FISCAL DIScIPLInE FOR POVERTY REDUCTION Table A20. Comparative Performance In Social Indicators Life Epectancy Net Primary Enrollment Country-level Divided by Divided by Country-level Divided by Divided by efficiency region avg. world avg efficiency region avg. world avg Latin America/Caribbean 83.5 103.2 76.4 103.1 Argentina 85.3 102.2 105.5 84.8 110.9 114.4 Bahamas, The 85.0 101.8 105.1 Bolivia 73.6 88.1 91.0 80.6 105.5 108.8 Brazil 80.6 96.6 99.7 76.8 100.5 103.7 Chile 89.3 106.9 110.4 72.2 94.5 97.5 Colontia 84.1 100.7 104.0 65.7 85.9 88.6 Costa RIca 91.0 109.0 112.5 72.6 95.0 98.0 Ecuador 82.1 98.4 101.6 77.7 101.7 104.8 9 Salvador 83.6 100.1 103.3 69.8 91.3 94.2 Guatemala 79.1 94.8 97.9 69.6 91.1 93.9 Guyana 76.2 91.3 94.3 Haiti 69.4 83.1 85.8 Honduras 84.7 101.5 104.8 83.5 109.3 112.7 Jamaica 90.3 108.2 111.7 83.2 108.9 112.3 Mexico 86.0 104.0 106.3 85.5 111.9 115.4 Ncaragua 83.1 99.5 102.7 74.1 97.0 100.0 Panama 88.3 105.7 109.2 77.5 101.3 104.5 Paraguay 83.2 99.7 102.9 77.0 100.7 103.9 Peru 81.6 97.7 100.9 77.1 100.9 104.1 Trinidad & Tobago 86.5 103.6 107.0 Uruguay 86.8 104.0 107.4 75.3 98.5 101.6 Venezuela, RB 87.0 104.2 107.5 72.8 95.3 98.2 Source: Wodon (2001) 208 STATISTICAL APPENDIX Table A21. Main Tax Breaks Sector Types of Tax Bresl Estimated Cost (% of GDP) Mining Reinvestment of profits accelerated depreciation;agreements on tax stabilty; 0.4 advanced recovery of IGVIVM; reduced tariffs extended on capital good imports. Hydrocarbons Accelerated depreciation; agreements on tax stability; advance recovery of IGVIVM; 0.1 exemption of IGV/IPM and ISC for imports of Inputs during exploration; exemption of IGV/IPM and ISC on fuel In three regions in the country. Agriculture Exemption of IR for small producers (50 UIT) and reduced tariffs of 15% for the 0.4 sector; reinvestment of profits (up to 20%); accelerated depreciation for Irrigation infrastructure; exoneration from IES; Issuance of documents cancelling taxes for the Imports of certain Inputs and capital goods for agro-Industry IGV of 5% for rice Industry (forest and ExoneratIon from IR for Industrial enterprises; reinvestment of profits; exoneration <0.05 frontier) from IGV/IPM and Peru-Colombla protocol duties; agreements on tax stability; reduced tariffs on capital good imports. Housing Exoneration of IGV/IPM on the first housing sale of less than $35,000; exemption of <0.05 IGV on construction contracts and housing sales In the Amazonia. Tourism and Hotels Accelerated depreciation; exemption from IGVIPM for services borrowed in favor of <0.05 external operators. Public Works and Agreements on tax stability; advanced recovery of IGV/IPM: exemption from IES. <0.05 Concessions Education Reinvestment of profits (up to 50%); accelerated depreciation; exemption from <0.05 IGV/IPM; exemption from IES. CETICOS Exoneration of IR, IGV/IPM, ISC, tariffs; duties of 8% for local commercialization; 0.1 exoneration of IES and other national and municipal taxes and tariffs Amazonia Reduced IR tarif (0, 5, and 10%); reinvestment of profits; exemption from IGV/IPM; 0.9 fiscal credit of 25 and 50% for IGV on sales In the Interior; IES exemption. Other Sectors Agreements on tax stability. < 0.05 TOTAL 2 Source: SUNAT, IMF (2000) Table A22. Alternative Measures for Tax Reform by Sector or Region Proposed Measures Collection Impact Sector or Region whose (% of GDP) privileges will be effected * Elminate exemption of goods in Appendix 1, of the IGV law. 0.25 Agriculture * Eliminate exoneration to financial profits 0.27 Financial Elminate zero-rate tax regime In the Amazonla 0.25 Amazonia * Eiminate exoneration of IGV and ISC to fuels 0.10 Amazonla * Eiminate CETICOS regime 0.13 CETICOS * Unify income tax rates 0.05 Agriculture and Amazonia Source: Schenone (2001) 209 PERU: RESTORING FISCAL DIscIPLINE FOR POVERTY REDUCTION Table A23a. Latin America General Government's Tax Revenuesi 1998 1/ Percent of GDP Lkuguay Brazil Argentina Chile Mexico Colombia Venezuela Ecuador Peru TOTAL 27.8 24.5 23.5 20.4 20.3 18.8 18.4 18.1 14.8 Resources non renewable 21 0 na na 1.1 4.3 1.6 7.6 4.5 0.1 Pest of tax revenues 27.8 24.5 23.5 19.3 16 17.2 10.8 13.6 14.7 Direct taxes 3/ 5.1 4.8 3.4 3.9 4.6 5.0 2.0 1.8 3.5 Social security 6.5 5.1 4 1.5 1.7 3.4 0.7 2.2 1.7 VAT 8.5 8.4 7.1 8.9 3.3 4.5 4.3 4.2 6.1 Special taxes 4 1.8 2.3 2.2 2.3 3.1 1.2 0.6 1.9 Tradetaxes 1.2 0.7 1.0 1.8 0.6 1.2 2 3.1 1A Other revenues 2.5 3.7 5.7 1.0 3.5 0 0.6 1.7 0.1 Memorandum: roductivity of VAT 4/ 0.37 na 0.34 0.49 0.22 0.28 0.26 0.42 0.34 1/ Icludes tax revenues and non-tax revenues of non-renewable resources. For Iixico and Pbru, data is for the year 1999. 21 Excludes VAT, tariffs and special taxes on petrol derivatives. h Chile Includes payrrents not of taxes by OODILO ( the state enterprise that extracts and exports copper) which is 0.4% of GDR the estimated payrrants of the rest of the nining corrpanies is about 0.7% of GDP. 3/hi Pbru includes taxes a Is planlila (LES). 4/ Colection as a percentage of GDP, divided by the legal rate. Source: IvF Table 23b. Latin American Economies* Outline of Fiscal Policy Rules t/ Rule/Country Efective Coverage Basic Rule Escape Additional Statute 4/ Sanction 5f Date 2f 31 Clause 3/ Rule S/ Budget Rule Argentina 2000 NG 6/ 0/DL. CF EL L J Brazil 2001 NG, SG CB WL C L J Colorrbia 2001 SG CB WL L J Ecuador proposal NG OB CF/MY L J Mexico 1917 SG CB C R Mexico proposal NG OB C J Peru 2000 NG OB/L CF EL L J Venezuela 2000 NG CB CFlMY TL C L R Debt Rule Brazil 2001 NG, SG SL C L J Colofbia 1997 SG FL L F Ecuador proposal NG SG FL L J 1/ Ehcopasses perrranent restrictions on aggregate fiscal performance Indicators. Table excludes prohibition or brits on financing from specific sources, notably from the central bank. 2/ General goverrrent (GG), or Irited to nations (central, federal) government (NG) or subnational (including local) govemnment (SG). 3/ Budget rules consist of 4/ (onstitution (C), legal provision (L), or policy guideline (P). 5/ Sanctions for noncornplance: reputational (R), judicial (J), or financial (F). 6/ Ref ers to the national nonfinancial public sector. Source: Kopits, 2002 210 STATISTICAL APPENDIX Table A24. Balicitliesof Social Spending to Total Spending of the Consolidated Public Sector In Peru (1997-2002) l/ Spending Category Average easticity to a Level of Besticity to an Level of Point of Point of share of change in total algnificence Increase In total algnificance elasticity elasticity total public public apending In 1998 in 2001 spending spending (198-2000) Farng andAdrdnitradon 25.7 2.05" 0.05 1.82- 0.05 .76 1.43 Pfnlans and Socle Assistance 21.46 1.37 0.16 1.43- 0.01 020 -0.14 Penlons 16.57 1.32 0.31 1.44 0.17 -0.56 -0.13 Social Assistance 4.89 1.41 0.16 1.36 0.54 248 -0.17 Ekkcatlon 1556 0.92 0.22 0.84 0.1 -0.13 0.61 Re-wpdrary 1.32 1.12 0.55 0.81 0.36 -1.68 1.68 Mbiry 5.5 1.03 0.08 1.02 0.17 0.19 1.64 Secondary 4.08 1.21 0.18 099 - 0.01 0.69 0.05 Tertiary 279 0.56 0.41 0.60 0.46 -0.50 0.16 Ohers 1.86 0.68 0.94 0.15 0.75 -1.26 -1.17 1baldh and Water 8.47 0.74 0.36 0.67 0.13 -0.18 -0.47 Water and Sewags 0.78 -0.24 0.94 0.55 0.71 1.18 8.73 H-ealheare 7.69 0.77 0.51 0.64 - 0.01 -0.386 -1.37 Oher12 2.87 1.38 0.27 1.69" 0.05 -0.90 1.57 Total Social panditure 37.21 0.73" 0.04 0.63 0.19 0.75 0.39 1/ Istirrute 2f ckjdaes forein relaior, legislatve, housing and urban developim industry connwe and services. fishig, comunicatbon, and erployrnt * Idicates signficiance at 10% " Indicates signiicance at 5% h IndIcats algnIliance at 1% Souce Table A12a aid WB Staff eastinme Table A25. Degree of Traruperency In the Publication of Insietuilonal Information Type of Ildrtion Rtshed MEF Oxisoldated Ralc Sector Balance Available frar2002 Flevenues and Bpendiures of the Cetral Govenyrent and Decortralzed requeLts Yes Rans of the Bugetary Law, bdebtedness and Financial fAbrkim Yes Balance and RoIe of the Domestic and Etaeal Rd Det Yes Dsburemnents and Armrtizallons of the Qatic and Sternal Dabt Yes 14jor Fil hwestrrnt Rans of the 1.200 UITs Yes Belance of the Facal Stablizaton Find Available from2002 Bualuaion Faasis of Appicable hdiatora Available frarn2002 FE GOmeaidated Budget Yes General Belance Yes Statennt of Financial AudIts Available from 2002 Rogress hicators to be appled Yes Fesdis of the Bialuation of the Rogress hdialom Avallable from2002 Fneicial Staterrnt of the Fondo nlidado de Reserva revisional (FR audited by 2001 Available from2O02 Financial Statret of the Fondo cioalde Ahorro Rtico (FOA-Rhmxfted by 2001 Avalable from 2002 Fnancial Assets of FaRJand FiChwested Yes CONSUCM0E Mor Acqilsions and Oxwnctng of the 50 UWe Yes Sourme MF Table A26. Published Information of the Budget in 2002 IVnistries CTARS Superintendencies Public Enterprises Original Budget 100% 100% 83% 100% Executed Budget 100% 100% 83% 100% Econonic Iformation on hvestment Plans 100% 100% 33% 100% Info on Contracting and Acquisitions 63% 88% 33% 25% Strategic Plans 100% 0% 83% 100% Source: MEF 211 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A27. Composite Governance Indicators, International Comparsons, 1998 (Global rankings - lower Is rnore postive) Voice and Government Accountablilty Political Instability Efectiveness Rlgulatory Quality Rule of Law Corruption 1998 2001 1998 2001 1998 2001 1998 2001 1998 2001 1998 2001 Peru 125 70 112 95 54 93 37 61. 111 109 74 72 Argentina 59 56 48 51 48 66 38 70 59 66 81 88 BolIvia 63 67 88 122 84 101 22 39 98 100 96 119 Brazil 57 57 98 57 83 89 88 84 88 87 56 70 Chie 54 52 51 30 23 23 18 13 24 24 25 17 Colorrbta 91 112 141 149 74 95 69 93 129 132 103 89 Ecuador 66 92 109 131 111 129 61 95 122 130 125 140 Mexico 90 73 101 81 52 59 42 46 108 101 82 82 Venezuela 68 106 93 100 133 124 92 115 118 136 116 107 No. of CounW,as 173 173 155 161 156 159 168 169 166 170 155 161 LAC Median 64 60 91 69 84 87 46 66 105 98 87 86 PruRankItLAC 25/26 17/26 20/24 17/25 8124 18/25 10/28 13126 15126 16126 10/24 925 Source: Kauffrann, Kraay, and Zodo-Lobiton, January 2002. Table A28. Volatility in Latin America (Standard deviations of grow th rates by decade, percent) GDP PFVATE Country 1980s 1990s 1980s 1990s Argentina 5.6 5.5 5.9 5.7 Bolivia 29 1.0 3.3 0.5 Brazi 4.6 3.0 4.4 4.9 Ohle 6.4 3.5 9.4 3.4 Colorta 1.5 3.3 1.2 1.9 Costa Fica 4.5 2.4 6.1 3.2 Dorrinlcan Fepubic 2.7 4.4 4.5 8.7 Ecuador 4.5 3.4 2.5 0.3 8 Salvador 5.7 1.9 6.0 6.9 Guaternala 2.7 0.8 2.4 1.0 Honduras 2.5 2.5 3.1 1.9 HaltI 2.9 6.4 4.4 Jarraca 4.3 2.2 7.6 16.5 Mexico 4.4 3.6 4.9 5.1 Ncaragua 5.4 2.3 15.8 14.4 Panarm 6.5 2.6 11.3 10.2 Peru 8A 5.2 8.3 4.8 Paraguay 5.3 1.5 11.2 12.5 Trinidad and Tobago 5.7 6.8 14.0 17.6 Uruguay 6.6 2.8 9.4 5.4 Venezuela 4.8 5.0 3.5 4.7 Unw elghted average 4.7 3.3 6.6 6.5 Overall median 4.6 3.0 5.9 5.0 Weighted average 4.6 35 4.9 4.7 Source: WB staff estintes 212 STATISTICAL APPENDIX Table A29. Economic Distribution of General Government Revenue and Expenditure, Selected Countries &olid Argeine Bradi Chl Colombla Costas ike ualaf Mulco Panama Praguy Pwm Veneoaul. LAC (198) (1908) (197) (19W) (1997) (199) (199) (199) (197) (1997) (2001) (199) (85-99) Awmt or Gap TotalRevenue 24.0 19.4 29.2 25.8 274 24,8 11.5 na na 18.7 14.2 3.2 Taxrevenue 19.0 17.8 24.5 19.7 19.7 23.5 8.9 na 19.1 11.8 11.8 10.2 Total Epmxture 27.4 21.9 29.3 234 26.7 26.0 129 22 29.8 20.2 16.8 30.3 20.4 CunmE)peicxtum 21.7 20.2 29.8 19.0 186 25.6 8.1 19.1 23.0 14.8 14.6 21.8 17.6 Inleresl 1.9 1.9 2.5 0. 3.0 5.3 0.9 4.3 3.1 0.5 2.1 2.9 3.6 Other 7.1 1.6 84 8.4 2.1 0.8 1.8 .. .. 2.0 3.7 0.8 3.7 Transfea 4.1 4.4 1.0 53 6.9 34 2.1 6.7 1.7 0.5 3.3 3.2 40 Wagesarwl Saades 8.3 7.3 11.9 4.5 8. 5.8 34 2.6 102 8.5 6.6 4.2 6.3 CapiAlE*pedlhur 60 1.7 2.5 4.4 78 2.4 4.8 3.7 3.6 5.4 22 8.5 3.0 PeNr ortat Total Revenue 100.0 10D.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Tax revenue 79.2 91.8 83.9 77.0 71.8 94.8 77.4 .. .. 63.1 874 31.7 TotlEpendpwm 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 CwrontE)pwndlue 79.2 92.2 91.5 81.2 69.5 91.4 628 83.9 958 73.3 84.5 71.9 86.3 Ilderest 6.9 8.7 8.5 2.6 11.4 1.9 7.0 18.9 11.7 2.5 10.9 9.6 15.6 oIher 25.9 7.2 28.7 35.9 8.1 2.9 14.0 .. .. 9.9 21.5 2. 16.1 Transfer 15.0 20.1 3.4 22.6 25.7 12.1 15.9 29.4 6.3 2.5 13.0 10.6 19.8 WageanmdSelades 30.3 33.3 40.6 19.2 24.7 20.7 29.0 11.6 38.1 42.1 39.1 13.9 33.6 CeaaEVenalre 21.9 7.8 85 18.8 293 8.6 . 37.2 ' 16.1 142 26.7 15.5 28.1 14.3 'Data cornesponds to the central geowent Source: Wod Bank Bila PER. Stall esumewa Table A30. Distribution of Social Investnent by Different Social Programs and by Docile of Poverty Severity Decle Population Investment(Sf.)2001 AmountUS$ Assigned 2001 PRONAA FONCONDEB* VASO DELECHE RURAL ROADS" PESP-URBANO Poorest 2,636,584 21,976.075 93,159,772 39.915.361 53,932,215 0.035 Decie 2 2,616,108 20.682,281 69.530.728 37.123.636 40,256.685 0.018 Decie 3 2,645,999 18,493,705 39,663,491 33,902,357 34,483.305 0.049 Decie 4 2,645,123 12,976,223 16,299,443 31,566,264 21,225,735 0.178 Decile 5 2.631,770 12,823,801 8,013.721 30,507.673 8,081,100 0.190 Decile 6 2,655,524 19,120.704 3,451.029 35,518.612 1,785,900 0.150 Docile 7 2,655,539 18,153,031 1,974.642 38.062.931 240.000 0.143 Decie 8 2,607,142 13,421.899 1,577,795 39,703.391 160.200 0.118 Docile 9 2,616,694 9,699.637 3,983,925 29,057,248 1,528.755 0.085 Least poor 2,636.362 6.675,468 3.025,690 16.889,527 .. 0.035 TOTAL 26,346,845 154.022,822 240.680,236 332,247,000 161,693.895 1.000 Percent Poorest 10 14.27 38.71 12.01 33.35 3.50 Decle 2 10 13.43 28.89 11.17 24.90 1.80 Docile 3 10 12.01 16.48 10.20 21.33 4.85 Decie 4 10 8.42 6.77 9.50 13.13 17.75 Docie 5 10 8.33 3.33 9.18 5.00 19.00 Decile 6 10 12.41 1.43 10.69 1.10 15.00 Docile 7 10 11.79 0.82 11.46 0.15 14.30 Decie 8 10 8.71 0.66 11.95 0.10 11.80 Decie 9 10 6.30 1.66 8.75 0.95 8.50 Least poor 10 4.33 1.26 5.08 .. 3.50 TOTAL 100 100 100 100 100 100 Inckdes the,PESP-RURAL - Accuniated to 2001. Prelminary date. (EstiUation based on US$ 15,000/knf Source: INE 213 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A31a. Average Monthly Per Capita Income of Households, 2000 (US$) Geographic area Total I II III IV V VI ViI VIII Kl x TOTAL 95.e 13.7 25.0 34.7 44.6 53.9 64.3 81.1 103.9 151.6 385A Urbana 129.1 18.6 35.7 48.1 60.7 72.9 86.3 108.4 139.9 203.2 518.2 Rural 38.1 4.5 7.9 11.5 16.4 20.6 25.9 33.3 42.2 61.6 156.8 Source: Diaz (2001) based on ENAHO 2000-4 Table A31b. Household Accessto Social Program aby Income Per Capita Deciles, 2000 Percent of Total Geographic area Total I II Ill IV V VI VII VIII XI X Total 5.632,815 7.4 8.6 8.9 9.1 9.0 9.6 9.8 11.0 12.1 14.5 Beneficiaries 3,330,760 10.6 12.2 11.9 11.1 10.0 10.6 9.9 8.7 8.3 6.8 Non-beneficiaries 2,302,055 2.7 3.4 4.6 6.4 7.5 8.1 9.6 14.3 17.7 25.6 Urban 3,607,764 7.8 8.5 9.0 9.3 8.9 9.5 10.4 11.3 12.2 13.1 Beneficiaries 184,157 12.2 13.9 13.6 11.6 10.8 10.2 10.7 7.3 6.1 3.7 Non-beneficiaries 1,773,607 3.2 2.9 4.1 7.0 7.0 8.7 10.1 15.4 18.6 22.9 Rural 2,025,051 6.7 8.8 8.8 8.8 9.1 9.8 8.6 10.5 12.0 16.8 Beneficiaries 1.496,602 8.7 10.1 9.8 10.4 9.0 11.2 8.9 10.4 10.9 10.5 Non-beneficlarles 528,449 1.0 5.1 6.0 4.3 9.2 6.1 7.8 10.6 14.9 34.8 Source: Diaz (2001) based on eNAHO 2000-1 Table A31c. Household Access to Social Programs by income Per Capita Deciles, 2000 Percent of Sub-total Geographic area Total I II III IV V VI VII VIII XI X Monthly Incomelcapita Us$ 95.8 13.7 25.0 34.7 44.6 53.9 64.3 81.1 103.9 151.6 385.4 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Beneficiaries 59.1 85.2 83.8 79.1 71.6 65.8 65.4 59.8 46.8 40.2 27.6 Non-beneficiaries 40.9 14.8 16.2 20.9 28.4 34.2 34.6 40.2 53.2 59.8 72.4 Urban 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Beneficiaries 50.8 79.9 83.2 77.3 63.3 61.4 54.6 52.2 32.8 25.3 14.3 Non-beneficiaries 49.2 20.1 16.8 22.7 36.7 38.6 45.4 47.8 67.2 74.7 85.7 Rural 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Beneficiaries 73.9 98.0 84.8 82.3 87.2 73.4 83.7 76.2 73.6 67.5 46.1 Non-beneficlaries 26.1 4.0 15.2 17.7 12.8 28.6 16.3 23.8 26.4 32.5 53.9 Source: Diaz (2001) based on ENAHO 2000- 214 STATISTICAL APPENDIX Table A32. Miatargeting by Types of Food Assistance, 2000 Social Program Wde Total Pove Non Beneficiaries Eitreme Not Etreme Total Poor TOTAL Desayuno Escolar 2,972,859 100.0 51.1 18.7 69.8 30.2 Vaso do Leche 2,283,919 100.0 40.1 24.1 642 35.8 Conqdor Popular 746,134 100.0 40.2 14.5 54.7 45.3 URBAND Desayuno Escolar 1,000,349 100.0 21.3 34.5 55.9 44.1 Vaso de Leche 1,104,018 100.0 15.7 34.6 50.2 49.8 Comedor Popular 363,769 100.0 18.3 18.6 36.9 63.1 RURAL Desayuno Escolar 1,972,510 100.0 662 10.7 76.9 23.1 Vaso de Leche 1,179,901 100.0 63.0 14.2 77.2 22.8 Comedor Popular 382,365 100.0 61.1 10.6 71.7 28.3 Source: Diaz (2001) based on ENAHO 2000-1 Table A33. Mie4argeting by Types of Health Programs, 2000 Social Program W do Total Pover Non Beneficiaries Etreme Not Etreme Total Poor TOTAL Regulating chid growth 1,729.899 100.0 33.7 18.4 52.1 47.9 School health Insurance 1,157.912 100.0 36.8 20.6 57.4 42.6 Fanly planning 870,947 100.0 22.1 24.8 46.8 53.2 uAN Regulating chld growth 908,742 100.0 9.2 24.5 33.7 66.3 School health Insurance 636,888 100.0 15.8 31.6 47.4 52.6 Fantly planning 585,964 100.0 11.7 26.0 37.7 62.3 RURAL Regulating chlid growth 821,157 100.0 60.8 11.8 72.5 27.5 School health Insurance 521,024 100.0 62.5 7.1 69.6 30.4 Fantly planning 284,984 100.0 43.4 22.3 65.6 34.4 Source: Diaz (2001) based on ENAHO 2000-1 Table A34. M14argeting by Types of Education Programs, 2000 Social Program W de Total Poverty Non Beneficiaries Eftreme INot Extreme Total Poor TOTAL School text and naterials 2,970.567 100.0 44.4 19.7 64.1 35.9 URBANA School text and rraterials 1,513,755 100.0 22.3 28.7 51.0 49.0 RURAL School text and noterials 1.456,812 100.0 67.3 10.4 77.6 22.4 Source: Diaz (2001) based on EAIAHO 2000-1 Table A35. Evolution of Household Access to Social Programs, 1998-2000 1998 1999 2000 Var % thousands thousands thousands 99198 20009 2000198 Total 5,408 5,490 5,633 1.5 2.6 4.2 Benefillaries 2,848 3,279 3,331 15.1 1.6 17.0 One program 1,004 937 755 -6.7 -19.3 -24.8 Mre than one 1,843 2,342 2,575 27.1 10.0 39.7 Non-beneflclarles 2,561 2,211 2,302 -13.6 4.1 -10.1 Urban 3,546 3,872 3,608 0.7 1.0 1.7 Beneficiaries 1,624 1,915 1,834 17.9 -4.2 12.9 One program 669 633 536 -5.4 -15.3 -19.8 Mdore than one 955 1,282 1,298 34.2 1.2 35.8 Non-beneticlaries 1,921 1,657 1,774 -13.7 7.0 -7.7 Riral 1,663 1,918 2,025 3.0 5.6 8.7 Beneficlarles 1,223 1,364 1,497 11.5 9.7 . 22.3 One program 335 304 219 -9.4 -27.9 -34.7 More than one 888 1,060 1,278 19.4 20.5 43.9 Non-beneflclarles 639 554 528 -133 -4.6 -17.3 Source: laz (2001) based on ENAHO 2000-0 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A36. Annual Private Expenditure on Education by Per Capita Income Decile, 2000 By type and level of education attending, Milons of US$ Declies of Per Capita Income Education Level Total I II fil IV V VI VII VIII XI X TOTAL 4445.4 2.0 2.4 2.5 3.6 5.7 5.5 7.6 10.3 16.8 43.5 niltial 459.7 0.9 1.6 0.9 2.5 3.9 1.1 1.3 7.4 14.6 65.8 Rinary 967.1 4.5 3.5 5.2 4.7 3.6 6.4 7.2 13.4 10.5 41.0 Secondary 762.1 3.3 3.0 4.1 3.7 9.0 11.9 8.7 12.6 15.1 28.5 Bachillerato 6.1 24.8 7.9 5.8 10.2 4.9 11.9 8.4 26.1 0.0 Supplemental Non-tiv. 588.4 1.3 3.0 2.9 6.0 3.6 5.7 19.7 11.3 23.6 23.0 Supplemental Univ. 1476 0.4 1.8 0.7 2.8 4.6 3.6 5.4 8.5 21.1 51.2 Post-graduate 185.9 23.5 3.6 6.1 66.8 PUBLIC 1099 5.3 5.7 7.0 8.1 6.5 10.1 9.5 10.3 17.5 20.0 kintial 53.6 3.5 13.4 4.3 6.9 9.7 5.0 7.0 14.8 18.9 16.6 Primary 298.7 11.0 10.8 11.3 9.2 7.1 9.7 8.3 13.9 8.3 10.4 Secondary 260.8 7.7 6.3 11.3 7.2 10.5 15.2 9.5 10.8 13.2 8.5 Bachillerato 6.1 24.8 7.9 5.8 10.2 4.9 11.9 8.4 26.1 0.0 Supplemental Non-UnIv. 98.6 2.3 2.3 5.1 10.7 8.0 10.2 31.0 15.2 12.6 2.5 Supplemental Univ. 286.8 1.4 2.1 9.8 3.2 10.5 6.9 4.6 37.2 24.3 Post-graduate 94.5 7.0 3.0 90.0 PRIVATE 3346.3 0.9 1.4 1.1 2.2 5.5 4.0 7.0 10.3 16.5 51.2 hitial 406.1 0.6 0.0 0.4 1.9 3.1 0.6 0.6 6.5 14.0 72.3 Primary 668.4 1.6 0.2 2.5 2.7 2.1 4.9 6.7 13.2 11.5 54.7 Secondary 501.3 1.1 1.3 0.4 2.0 8.2 10.1 8.4 13.6 16.0 38.9 Supplemental UnIv. 489.8 1.0 3.1 2.4 5.1 2.7 4.8 17.5 10.4 25.8 27.2 Supplemental Non-Univ. 1189.3 0.5 1.9 0.4 1.1 4.9 1.9 5.1 9.4 17.2 57.7 Post-graduate 91.4 47.9 9.4 42.8 Note: Bcchange rate of SI. 3.50 Source: Diaz (2001) based on ENAHO2000-0 Table A37. Annual Expenditure on Health by Per Capita Income Decile By place of consultation, Mns of US and respective percentage of total by place of consultation Deciles of Per Capita Income Place of Treatment Total I II Ill IV V VI VII VIII XI X Total EKpenditure 742 2.6 5.7 8.8 6.0 7.0 7.6 10.9 11.2 13.2 27.1 Puesto, Central Mnsa 137 5.4 16.5 8.7 11.3 9.6 9.0 11.7 11.4 13.4 3.0 Posta, Fbliclinico IPSS 16 0.2 20.3 5.5 1.9 0.7 7.9 0.4 48.3 9.0 5.9 Hospital MINSA 196 3.4 6.5 21.5 4.7 10.5 8.9 15.9 10.1 4.3 14.3 Hospital ESSALUD 58 0.3 0.9 4.1 5.4 5.2 13.6 5.0 8.2 23.1 34.2 Hospital FF.AA. Pollcia Nac. 27 3.0 0.2 0.7 1.0 0.2 4.4 0.0 15.0 18.7 56.7 Consultorio Medico Part. 148 0.7 8.0 0.5 5.3 5.1 6.1 15.3 13.0 20.2 25.8 Various clinics 146 1.1 0.4 2.6 3.1 3.4 0.6 2.5 9.3 11.1 65.8 Fharrmacy 18 6.0 7.6 14.6 13.1 9.8 8.6 16.8 6.3 7.5 9.6 Home 23 2.7 2.2 8.5 0.1 1.4 9.3 21.5 22.9 18.2 13.1 Indigenous treatment 12 2.9 7.0 7.6 2.4 15.8 9.9 6.6 9.0 34.1 4.7 Others 20 0.7 5.6 1.9 23.4 11.0 13.2 17.5 3.8 8.3 14.7 Self-treatment 38 1.1 7.5 5.6 1.4 17.0 10.7 7.4 4.6 27.6 17.1 Source: Diaz (2001) based on E NAHO 2000-11 216 STATRSTICAL APIPENDX Table A38. Definitions of the Consolidated Public Sector (CPS) Consolidated Public Sector Mnistries, Regional Governments, Public Institutions, Universities and ONP (without funds) * Organized Supervisors of Special General Government Funds (FCR, FE, FONAHPU and Fondo IMvivienda), Beneficiary Societies ESSALUD, Public corpanles, Rest of Publc Sector Regulatory bodies Local government and their companies Local Government Source: ME Table A39. Central Government National and Regional Budget Composition, 2002 National Min MS Regional Mn HS Congress 265 Arazonas 250 CNM 9 Ancash 637 General Comptroller 53 Apurmac 448 Defense 2,226 Arequipa 583 Public Defense 18 Ayacucho 411 Economy and Finance 1,688 Ca|anarca 618 Interior 2,117 Callao 225 JNE 12 Cusco 666 Justice 341 Huancavelica 322 Public IMnistry 228 Hudnuco 381 ONPE 13 Ica 325 Judicial Pow er 467 Junin 526 PCM 260 La Ubertad 530 RENEC 90 Lambayeque 397 Exterior Relations 616 Lima 2.408 Constitutional Tribunal 10 Loreto 520 National Adninistration 1,220 Madre de Dios 100 Moquegua 121 Pasco 160 Plura 739 Puno 713 San Martin 452 Tacna 200 Tumbes 167 Ucayall 256 Total 9633 Total 12,152 Mine US$ Mine US$ Current expenditure 2,468 Current expenditure 2,342 Capital expenditure 253 Capital expenditure 1,091 44% 56% Itf.Stretturflt46tin Mine US i% Debt Service 2,200 Provisional obligations 1,752 Source: MEF 217 PERU. RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A40. Main Revenue Sources for Municipalities, Table A4. Departmental Budget - Education 2002 and Health by Expenditure Category, 2002 NS Percent Mons of NS 1865 100 Department Eucation Hoalth ICurrnt~Weewe 183 9~ Regional Distribution -Percent Vq~s5~ B Llrr 34.4 50.9 Captal 433 Puno 4.9 3.9 Productlon / Consurptlon 41 Pura 4.7 3.8 Others 45 Loreto 4.4 3.8 W-- ", 84 451 La Llbertmd 4.3 3.6 General Adniistratlon 34 Cusco 4.3 3.5 Education 6 Cajarrwca 4.3 2.9 Health 308 Arequipa 4.2 2.8 Housing and Construction 131 Ancash 4.2 2.7 Agroindustry and Mneral Resources 1 Juni 4.1 2.8 Idustry, Mnig and oimnerce 67 Laftayequo 3.0 2.3 Tranport and Corrnunicatons 149 Ayacucho 2.8 2.0 Others 138 Ica 2.8 1.9 no 1. 1 San Marin 2.5 1.9 For Pensions 1 Huanuco 2.5 1.7 For Rblc Works 7 Wayall 1.8 1.6 Others 3 Huancavellca 1.8 1.4 mlotao - 56 ' 30 Apurhmu 1.6 1.4 Agroidustry and Mneral Resources 25 Ammonas 1.4 1.2 Industries 22 Pasco 1.4 1.1 Transport, Canymication and Storage 0.1 Tacna 1.4 0.9 Education, Recreatlon, and Culture 0.1 Turbes 1.3 0.7 Health 2 Kbquegua 0.8 0.6 Others 7 Cabo 0.7 0.5 lst" C6tritlie dre de l 0.4 0.5 Transport, ornmunication and Storage 10 TOTAL 100 100 Education, Pecrealion, and Culture 3 Health 15 By E iture Category Mins NS Others 86 Goods and Services 892.1 1510.0 P0ra hnesrnt 576.6 481.6 Real Estate 90 Personal and Social Obligations 4388.4 1076.7 Anancial 1 Others 210.0 279.8 Others a TOTAL 6007.2 3328.1 Otherses 19 a Pnes 135 By E4moliture Category- Percent Sanctions 54 Goods and Services 14.7 45.4 Others 11 Investrnt 9.5 13.9 Ita vnus - - 0-Personal and Social Obllgations 72.3 32.4 e1 Others 3.5 8.4 Buldings (except land) 7 TOTAL 100.0 10D.0 kban and rural land 15 Vehicles, Equlpnent and Machines (used) 0.3 Others 8 Source: D u n r 2ii STATISTICAL APPENDIX T~.Aoa.M ~ onco7po Tai M 006, 8 ooleisrosw eGuye d Co ilnnin O Rues ena* er 9nea0-0vo1s6seon sete 4oe ne98uro. AuoIal8ur immenipose e0e.a4u0 a le 0n.on 1 Cens8*Sn 806 71.7 72.0 W4 1&4 60A 70.1 0 62 8 o 60 a8A 87. 17.0 0. 43 2 Co0~t a 41.4 50.5 8.7 sej 3.7 19,3 87. C687 02 46.6 492 65.4 06 42 331 ÖD7 3 01 8,p 422 67,4 60.2 3.9 4.2 25.0 06 M, 06. 46.1 44.1 01. 761. 2.4 27.1 508 4 41. 64.4 613 e0.6 41.6 19.4 ca 887 . 1.3 364 542 72.0 44.9 60.0 582 5 Ju~sd PUr 210 88.1 75 3 671 342 57.1 A8.4 D3 57.7 671 36A 622 8k 40.4 25.0 881 66deu.de,4 u 02.0 GO 0 710 746 302 222 85 70, 64.5 18.1 822 882 611. 141 200 37.0 7~N46Pn. 4,4 61.1 64.1 0.1 800 17. 623 74.7 06.1 80.1 361 7.1 79,0 8.4 414 65.1 a MnNMI.IdDD. 07.3 6.7 G24 570 16. 26.0 882 70,4 74.7 8.0 581 64. 71. 5. 3.1 &I .3 9C00rri16ndeFo. 672 76 78,4 719 28.4 00 783 67.0 7.4 90. 8O 79A 76,1 4 130 464 10RgeoP0ed, 61.1 709 7a5 74.7 241 361 7&4 75,4 88.7 70.6 en2 722 7m0 24.1 223 451 Il16.4N. ffi.86 78.1 754 82.8 74.9 133 830 74.7 9. 76.o 64.1 741 701 74.0 10 38 44.0 12M~dsadälinh 535. 67.7 70. 88.2 20. 27.1 71.1 67.0 88S 85.4 4A 62. 60.3 3.7 211 86A 13eden.deEhod. 51.4 50.7 72.1 16 308 12,5 8.7 4,4 e8,7 67.6 74.7 77, 882 19.1 12.5 424 14 P~09rn dW8 Va- 629 Gl2 72.7 11.4 36 20.1 67.7 92 0.4 790 47.1 71.1 73.3 402 301 070 15O00d1p8pdar 854 .4 05 76 30. 33.7 =.1 91.1 770 93.0 402 .1 o.2 2.2 242 001 168 0.7 704 93,3 94. 3, 75.0 81 00.2 704 91.7 701 917 886 33 00 32.7 176UNA7 682 72.1 79.7 89 0.9 738 81 05.6 66,6 88.0 734 64.7 67.8 19.1 99 404 156~ tNa.d.Ad 80 742 79. 884 22. 60.3 79,3 5.2 7&1 630 702 70. 774 22.8 22.2 468 190MPTEL 750 73.4 830 31 400 076 902 730 67. 74.1 D66 531 9 .7 0. 266 20G 1Nam60. 64. 61. 3 2 e6.4 34.7 30.9 du. 85.8 67. 002 801 62. 886 31, 286 50 21 6EAALd. 65.1 71 70.0 701 271 41.1 74. 064 733 78. O6 78 898 t0.7 1,4 40.5 22 ~de Mn. 80.7 705 71.8 741 9 401 6.6 71.7 61.7 614 61.5 88.9 C9.1 202 30.0 482 23 DRE 73.3 70.2 e63 o01 10.0 186 7,3 814 88.7 74.0 853 01.4 610 112 14.1 36.0 24USE/ADE 4901 60.3 77.1 702 412 132 881 88. 0m,9 741 W00 882 708 322 26. B02 250d1o" 5063 64 642 649 35.7 23.1 6.6 700 606 771 450 2.1 862 412 34.1 618 208.de04 4612 .1 s6 735 45,4 100 832 478 552 60.4 322 546 90.6 415 31.3 680 27Un ~rPuj*** 01 862 64.1 09 454 206. 86 71.1 85.1 58 80.1 M2. 70. 296 34. 06 28 edM0.408 13,7 401 487 403 764 60 332 414 49,4 1801 0.6 21.1 90.7 15.2 500 80 2014r89 35.0 50.8 80 5 0 51,6 10.0 .8A 71. 641 88.6 40 43.7 630 30.7 471 88. 30Psts2~ 30,0 57.1 539 404 401 30.0 61 768 1 2 0 1 336 40.1 746 45.5 413 61$4 31Essalud 061 701.3 732 880 442 342 70.9 7.4 61 82 0 63 74.0 75.4 3.9 109.3 42 32Mnde71ab* 882 e L 793 731 732 066 07. 732 060 701 0 01 83 004 226 05 402 33Mn a0~ta 032 62 88 639 300 222 SW. 6.3 57.2 68 456 61 002 25.0 1i 341n6tdeR6.608o 67.9 06 82 67. 33. 37.5 67.9 CO.7 0 470 023 889 701 211 12.5 435 3807NC 466 8.6 63 538 203 33 M 0.3 70. 64 031 007 072 182 250 414 3w004nd8e4 75.3 642 064 60 99 477 91 91.1 772 790 6.1 7.1 79.7 01 12 45.2 37W0. dup 60 . 063 67.4 GLO 867 452 42.1 07.9 74.7 881 2.7 61.1 602 70.6 320 27.1 00g 3B0B n* N.dn~ 702 911 003.6 067 U 42. 01 000 700 70.0 706 08 70.2 24.1 31 44,1 3un.deEn88ga 009 88.2 70.4 70 265 04. 68.4 81.0 9 701 86 81 62.0 20.7 2.2 440 4.0Dg. P~. de Ln*~ 062 75.4 745 0 22 8 1 060 0A 64. 75.1 8583 88 75.9 82.7 209 14.1 s02 40 ~1.P ~d 516 0.5 81 0 807 414 47.9 617 7E.4 70.9 62 741 688 G22 25.1 60 46.2 42C(3*d.9nd 57.9 724 711 681 35.1 417 79.0 82.0 et 62.1 69.5 NA 87. 0 496 31. 053 44 D6daddPueb 633 631 6.9 me 14.1 60 o. 84 .5 716 801 682 75.4 706 l0.t 71 45.3 45Mnleindop8d408 34 881 721 703 25. 38 748 77.7 706 7.4 863 50.8 77.7 200 203 512 4 ~646DE 674 763 79.1 813 2098 70.1 01.4 85 722 986 79 1 882 064 13.0 17.7 48.7 4710lN 6dedc0 ~1.1 N88 066 97.4 106 623 81.7 80 704 4A 74. 80 57.1 357 00 410 40 hnkL d. ~ 769 72,8 06 793 1. 31.9 70 5 7.0 . 0 96 4 0 81. 661 64 4.1 6.1 56.7 49C~rprd WndDea 45.2 6i 74.1 a. 51.5 342 02.7 73,4 637 73,0 54 633 729 30.0 324 0 50#Mn~<§0 nS. 5509 751 6,4 7.1 271 240 72 858 886 71.6 037 630 702 36,1 364 01.6 S61*d.,~ L9 60.9 704 790 73.1 12.6 271 8810 37 61. 78.1 702 701 771 038 0.0 43.1 ~00gi88odeiR1p. 550 563 40. 0672 8.2 371 742 742 60 881 676 64.1 7583 10 1,5 42.7 sBReN6EC 747 067 06,0 84 10a 862 792 06.9 812 871 704 84A 02.1 18. sa 47.6 077Pr~nssah 27.1 61.1 65 400 442 417 71.3 07.7 883 683 45,1 440 67.7 266 482 482 sa84F 503 822 67. a 86 340 403 8.4 724 8. 74A 886 09 776 291 502 52.7 UPRONA 34. 741 044 39.5 652 6.4 746 829 639 64. 67.9 550 771. 292 42.6 535 W 0odAWPoy 402 060 69 '0.3 47 302 860 70 2 07.7 721 04.7 8 60 332 32.9 501. 61 06d=8Nn.0de08o 00.0 703 702 491 MI.7 071 70. 881 88.7 4,4 064 004 88. 34 30 5.0 679 8.1 71.5 886 32 30.0 71.1 758 683 682 0.88 81 73.1 24. 201 47.3 5dp 01 89 704 65.7 34: 386 70,0 78 641 8 07 10. 65.5 792 342 281 61 402 020 66. 63.0 41 310 64 70. a0 881 40 s8 772 31. 283 53.4 e:w.dWa r* 001> Table A43. Corruption In the Judiciary, National and Municipal Agencies Based on public officials responses National Municipal Judiciary Agencies Agencies Variable Man Mean Mean Idex of Corruption 47.3 55.1 53.6 Idex of State Capture 73.1 79.2 77.2 Idex of Corruption In Personnel Adinistration 24.0 34.3 31.9 Idex of Corruption in Budget Adrrtnistration 20.5 28.9 28.3 Idex of Transparency 57.9 51.8 46.2 hIdex of Quality of Rules 68.1 69.9 62.1 Idex of Enforcernent 71.5 71.4 66.9 Extent of Meritocracy 66.6 65.7 63.0 Degree of Pifticization 30.3 34.8 34.6 Wage Satisf action 39.0 36.0 31.6 hIdex of Voice 71.1 70.0 66.4 hIdex of Mssion 75.8 78.8 70.9 (Perceived) Qualty of Service 66.3 64.3 58.3 Service Access 68.2 69.7 52.9 Availabillty of Resources 60.6 50.1 45.0 Existence of Iternal Audit Mechanisms 68.6 65.5 59.4 Source: WBI (2002) 219 g Q casm g -.-. 0 eis es äm ss se si Isa se si rt si si ss i Is I ' sem ss ss is ie el il i l esie esse sess sses isI i e s s e a e i t i m i e semeisesisenesig'ie STATISTICAL APPENDIX Table A45. Expenditure Coefficients by Department, 2001 Poverty rankings by definitions Departments No.ofth Population penditure Coefficient of Coefficient of Poverty litreme MEF INEl IU (Mine of NS) Expenditure erpenditure rate poverty Poverty Extreme by E by Population rate Poverty I Anazonas 12 420,606 215.29 17.9 61.2 74.5 41.1 9 7 10 2 Ancash 27 1.092,662 591.41 21.9 54.1 61.1 33.3 3 15 12 3 Apurime 14 *455.637 298.97 21.4 65.6 78.0 47.4 5 4 5 4 Arequ%)a 19 1.067.469 913.80 48.1 85.6 44.1 14.5 15 19 16 5 Ayacucho 20 541,427 465.12 23.3 85.9 72.6 45.4 4 8 8 6 CajaWerca 20 1,480.690 546.98 27.3 36.9 77.4 50.8 6 5 4 7 Callo 13 774.604 548.98 42.1 70.6 8 Cusco 17 1.194,275 716.58 42.2 60.0 75.3 51.3 11 6 3 9 4jancavelea 11 435,596 215.58 19.6 49.5 88.0 74.4 2 1 1 10 J&inuco 12 800,543 321.97 26.8 402 78.9 61.9 7 2 2 11 Ia 14 676,249 464.60 33.2 68.7 41.7 8.6 20 20 20 12 JAh 20 1.232,343 660.23 33.0 53.6 57.5 24.3 10 16 14 13 La Lbertad 32 1.483,681 741.07 23.2 49.9 52.1 18.3 8 17 17 14 Lartayeque 14 1,110,129 553.80 39.5 49.9 63.0 19.9 17 14 16 15 the 194 7,617,193 23,747.19 122.4 311.8 33.4 3.1 16 22 24 16 Lorelo 18 894.307 511.69 28.4 57.2 70.0 47.2 13 10 6 17 Madre de Cios 8 96,703 79.99 10.0 .82.7 36.7 11.8 23 21 19 18 MOquagua 6 153383 147.72 24.6 96.3 29.6 7.6 21 24 21 19 Pasco 8 259.137 172.66 21.6 66.6 68.1 33.2 18 12 13 20 Flura 16 1.611,573 788.08 49.3 48.9 63.3 21.4 14 13 15 21 Funo 33 1.247,494 651.20 19.7 52.2 78.0 46.1 1 3 7 22 San Marth 23 746.202 445.11 19.4 59.7 66.9 36.2 12 11 11 23 Tacna 9 286W539 22224 24.7 77.6 32.8 5.2 24 23 23 24 Tintes 8 197.605 179.06 22.4 90.6 46.8 7.4 22 18 22 25 Lk:ayal 16 450,693 225.65 14.1 50.1 70.5 44.9 19 9 9 Total 584 26,326K740 34A22.69 58.9 130.8 Totalw/oLkm 390 18.709.547 10,675.50 27.4 57.1 Source: LEF Table A46. Decentralization In LAC Political Functional Fiscal Federal countries Argentina Federal country reverts to type Secondary education and health Minor adjustments in revenue sharing transferred to provinces Brazil Federal country reverts to type, No explicit reallocation of functions Increase in revenue sharing to with new constitutional guarantees municipalities for municipalities Colombia Introduces election of mayors and Transfers primary education, health Increases earmarked transfers for social governors to provinces service expenditures in provinces, general revenue sharing to municipalities Venezuela Introduces election of governors, Introduces optional decoentralization Increase in funding is to follow direct election of mayors of social services, infrastructure to decentralization of functions states M6xico Concedes opposition victories in Transfers education and health to Increases earmarked transfers for social states, municipalities states, water supply, paying, public services in states security to unicipalities Unitary countries Chile Reverts to elected mayors in small Transfers primary education, health Increases earmarked transfers for social municipalities, introduces election to municipalities services to municipalities of mayors in cities Bolivia Introduces election of mayors Transfers primary schools, clinics Increases general revenue sharing to (physical assets only) to municipalities municipalities Guatemala Reverts to elected mayors - Increases general revenue sharing Source: Burli et al., 1999 221 PERU: RESTORING FISCAL DISCIPLINE FOR POVERTY REDUCTION Table A47. Principal Revenue Sources of Subnational Governments Municipal Government Argentina Varying shares of provincial taxes and transfer revenues, fees Brazil Fixed shares of state VAT and central income and excise taxes, property and service taxes Colombia Fixed shares of total central government taxes; property tax; industry and commerce tax, gasoline surcharges Mexico Fixed shares of federal taxes (passed through states), property tax, business licenses Peru Shares of central government revenues, property tax, mining cannon Venezuela Fixed shares of state revenue-sharing receipts, taxes on property, vehicle tax Chile Earmarked capitation grants for education, health, property tax Ecuador Fixed shares of central government oil revenue; taxes on property, business assets, vehicles, business registration Guatemala Fixed shares of total central government revenue, miscellaneous local taxes Bolivia Fixed shares of total central government revenues; taxes on vehicles and property Source: Burki et al., 1999 Table A4f Summary of Mining Taxation Authority for Selected Taxes and Fees Corporate Mineral royalty Dividend ExclaelSales tax on VAT on Property tax Fee based on Stamp tax lCountry Income tax witholding equip and services imported land aea N L N P L N|P L N| P L N P |L N P L N P L N|P L x x x x x x Argentina X X X X 1 1 Bolivia x x X x x Burkina Faso X X X X Canada (Ont) X X X X X Chile x x x x x x China X X X X X x x Ghana X X X x Greenland X X X x Indonesia 2( X X X X 3 Ivory Coast X X X X. X X X Kazakhstan X X X X X X Me)doo X X X 4 4 4 4 4 4 PapuaNGS6 X X X X X x x x Peru X X X X Philippines X X 6 X X X X X Poland X 7 7 X X X South Africa X 8 X X X Sweden X X X X x Tanzania X X X X X x USA X X 9 9 X X X X Uzbekistan X X X X X X West. Australia X X X X X X X x Zimbabwe X X X x N- National government P - Provincial government; L- Local government Notes: Where an X is given, the tax exists although a project may sometimes be exempted where there is a blank the tax does not apply to a typical mine; where a number is given, refer to that attached note with that number. 1/ Bolivia. 70% to national government, 30% to provincial 21 Indonesia. 6th generation contract, taxes under later COWa may differ. 31Indonesia. Status of property tax Is unclear from author's data sources. 4/ Meico. This tax may go to the national, provincial, or local government depending on ownership. 5/ Papua New Guinea. For larger mines specially negotiated revenue sharing agreements between national, provincial, and affected communlties may apply. 61 Philippines. I minerals are located on 'ancestral lands', a special royalty is assessed. 7/ Poland. If a 'basic mineral, paid to the national government; If a common minera (industrial mlneral, paid to local govemment. 81 South Afica. A royalty is paid to national government for mineral in federal lands. Most minerals are privately owned. 9/USA. Royalties are not assessed for most minerals unless they am found In special types of land. Source: Otto (2002) 222 STATISTICAL APPENDIX Table A49. Mining Fiscal Methods and their Amenability to Fiscal Decentralization National Provincial Local Tax type government government government Income or profits based tax Y P N Import duty Y N N Export duty Y N N Royalty (profit based type) Y P N Royalty (ad valorem type) Y Y P Royalty tax (unit type) Y Y Y Royalty tax collected nationally and % distributed Y Y Y Ucensing fees Y Y Y Surface rental or land use fees Y Y Y Withholding taxes on loan Interest, dividends, services Y N N VAT on goods and services Y P N Sales & excise tax Y P P Stamp duty Y Y Y Property tax (on book or assessed value) Y Y Y Payroll based taxes Y P N Surtaxes Y Y Y User fees Y Y Y Source: Otto (2002) Table A50. Total Transfers in Soles (by poverty quantile) Poverty quantile Vaso de Leche Canon Minero Foncomun Poorest 20,740,034 4,506,219 106,413,060 19 19,067,753 5,411,235 103,179,580 18 18,685,593 7,590,344 100,253,079 17 18,545,614 8,705,941 88,758,658 16 18,320,393 10,719,708 99,828,182 15 17,343,102 8,497,593 94,686,509 14 15,464,843 8,564,754 86,423,873 13 14,205,007 3,133,694 70,217,236 12 14,694,335 2,562,790 71,139,439 11 17,086,913 4,805,701 78,724,575 10 15,225,805 1,599,191 58,443,408 9 19,317,850 536.952 55,406,923 8 18,394,082 2,992,030 38,575,661 7 14,362,596 613,003 46,552,702 6 20,370,958 899,337 47,543,907 5 24,174,523 1,432,527 32,616,105 4 13,518,762 2,052,245 44,596,654 3 15,839,307 1,833,122 33,461,090 2 11,384,530 849,128 82,591,765 Richest 5,504,997 3,973,021 28,666,869 Total 332,246,997 81,278,535 1,368,079,275 Source: Apoyo Institute (2002) 223